Period: 69.51 days

	Submissions	Comments
Total	568	12185
Rate (per day)	8.17	172.68
Unique Redditors	216	2083
Combined Score	83314	63532

Top Submitters' Top Submissions

1. 8894 points, 73 submissions: International_Emu947
   
   1. random bakxxodi (380 points, 57 comments)
   2. Respect++++ (284 points, 34 comments)
   3. wtffff (277 points, 17 comments)
   4. neet walo all the best (265 points, 11 comments)
   5. aapas mai hi lad lo pehle (261 points, 29 comments)
   6. fanboy bahar aa jata hai kabhi kabhi :) (261 points, 41 comments)
   7. afternoon:) (257 points, 23 comments)
   8. who is he? (220 points, 22 comments)
   9. W FRAND++++ (196 points, 17 comments)
   10. meme bunch (196 points, 18 comments)
2. 6468 points, 48 submissions: CoolnamePanda
   
   1. Ayo Sus ⁉️ (381 points, 25 comments)
   2. 😌 Me enjoying the comments (326 points, 42 comments)
   3. How To Be As Big As Rapbot In 2 Minutes (298 points, 37 comments)
   4. Master Plan (289 points, 34 comments)
   5. Quality Control (263 points, 36 comments)
   6. Sorry but had to post this , Phir jaa rha hoon underground hone 🥷 (260 points, 22 comments)
   7. Kids Nowadays 😏 (229 points, 20 comments)
   8. Raftaar Paa-G🗿 (229 points, 15 comments)
   9. Haq se (222 points, 28 comments)
   10. The BAWA Anthem (193 points, 21 comments)
3. 3253 points, 29 submissions: Valuable_Currency_27
   
   1. Title Sirf Bantai Records Ko Support Karta Hai [OC] (194 points, 25 comments)
   2. Title Ko Kuch Samjh Nahi Aya (178 points, 18 comments)
   3. Title Ko Hate Mat Kara Karo [OC] (170 points, 22 comments)
   4. Title Ko Saza E Maut Mili Hai (161 points, 23 comments)
   5. Title Ko Bhi Depression Me Daal Diya (156 points, 14 comments)
   6. Title straight hai (156 points, 17 comments)
   7. So, i got the easy process! (145 points, 10 comments)
   8. Title Alag Hi Meme Banata Hai (141 points, 32 comments)
   9. Title Theory Me Bada Yakeen Karta Hai [OC] (141 points, 11 comments)
   10. Title Har Kisi Se Inspire Ho Raha Hai [OC] (123 points, 5 comments)
4. 2806 points, 19 submissions: BiharibabuOp
   
   1. Paji Bol Bsdk (506 points, 60 comments)
   2. SAB KUCH HAI PAAS AUR , PHIR BHI MAI AKELA KYU (399 points, 46 comments)
   3. Hip-hop could have been brought in India in Late 60s itself. However the source is not confirmed so posting on this sub. (392 points, 29 comments)
   4. Nacho BC DeNk ho gaya (312 points, 66 comments)
   5. Sehta nhi Drama main , main naa koi Bhansali (291 points, 20 comments)
   6. Janata hoon Sub dono taraf se , But Free mien karma de toh kyu naa lu - Dilin Nair (188 points, 42 comments)
   7. "Curious case of Rapbot" ft. Dincharya - RapbotVK (123 points, 22 comments)
   8. Jee Bahar Rehti , Teri Aadat Hai Chatna (92 points, 8 comments)
   9. 🥺💦💦 (80 points, 11 comments)
   10. And that's the why you start a Civil War (67 points, 22 comments)
5. 1780 points, 9 submissions: Feisty_Profession119
   
   1. PS:- This song is spinning in my head⚡? (418 points, 41 comments)
   2. Fastest Beef Ever💀💯 (310 points, 40 comments)
   3. BR (285 points, 40 comments)
   4. Kya khayal Awaam😂 (270 points, 24 comments)
   5. A Family that Eats together stays together. (136 points, 36 comments)
   6. Kr$na Bhai Simping on Bhabhi (115 points, 13 comments)
   7. New Beef😃👍 (100 points, 2 comments)
   8. Meme template ka istemal 😂 (95 points, 22 comments)
   9. Ye kya bhai pyaar main darar💔😓 (51 points, 6 comments)
6. 1773 points, 8 submissions: BIG_P0PPA_
   
   1. Beef Labels (571 points, 26 comments)
   2. per dill se Hindi wali pasnd hai (376 points, 56 comments)
   3. Jhoot bolo baar baar jhoot bolo (250 points, 20 comments)
   4. Desi rap ko banadiya hai main ne WrestleMania (219 points, 23 comments)
   5. W moment (152 points, 8 comments)
   6. Title (92 points, 9 comments)
   7. Karma bhai mera joke chori karliye (57 points, 11 comments)
   8. Thug life BAWA! (56 points, 14 comments)
7. 1606 points, 3 submissions: kaatilRandwa
   
   1. Title hierarchy ka matlab dhundh raha hai (899 points, 81 comments)
   2. 50 shades of funny paaji 🥰 (514 points, 36 comments)
   3. Kaatil Randwe ki nayi update (193 points, 37 comments)
8. 1527 points, 9 submissions: Ornery-Reward-2784
   
   1. Eminem blatantly copied anjum sir😨 (386 points, 31 comments)
   2. Title urdu rap god hai (279 points, 21 comments)
   3. YS vs SM part 2 lesgooo (189 points, 23 comments)
   4. Fresh oc hai bawa (175 points, 15 comments)
   5. Oc hai bawe🙄 (174 points, 42 comments)
   6. I tried drawing calm ;) (125 points, 34 comments)
   7. Level : impossible (77 points, 7 comments)
   8. Old beef but fresh mimi (72 points, 13 comments)
   9. Yunus bhai abhi bhi curious hain (50 points, 3 comments)
9. 1505 points, 7 submissions: CatesAreCute
   
   1. Cumback (375 points, 29 comments)
   2. Jaa rha hoon bajne😭👍🏻 (329 points, 47 comments)
   3. We💑🏻 AAWAM👨‍👨‍👦‍👦 new🚼 BTS🍆 ARMY🪖 is🥭 the💜👾🍇🟣☪️ (252 points, 13 comments)
   4. Khanna before he met Raffu bhai (183 points, 7 comments)
   5. Isko👆🏻 Igta👣 hai☑️ mask😷 pehenega🧕🏻 toh🍨 danger☠️ ho🧑🏻‍🔬 jayenga✅ kya🤬 re🏳️‍🌈 bachkane🧒🏻 kuch📉 bade🐯 log👩🏻‍❤️‍💋‍👨🏻 k💋 jaisa👍🏻 baat🗣️ kr👍🏻 (143 points, 22 comments)
   6. This👇🏻 post📬 is🇬🇧 to🇮🇳 break🏹💔 your🫵🏻 no👎🏻 fap💦 (126 points, 10 comments)
   7. Gaali (97 points, 5 comments)
10. 1400 points, 3 submissions: ___sohammmm
    
    1. Throwback to 2009 when Raftaar saw girls doing joints in Delhi (715 points, 76 comments)
    2. Travis Scott from Pune (589 points, 25 comments)
    3. Also Teri Behen Meri (96 points, 17 comments)
11. 1323 points, 4 submissions: Puneet_7164
    
    1. Title gaya Blanko lagane (981 points, 45 comments)
    2. Koi cringe nahi bolega🙏 (145 points, 44 comments)
    3. Not Againnn😭 (108 points, 28 comments)
    4. Kya rappers Rs 1,00,000 kama paa rahe hain ? (89 points, 28 comments)
12. 1277 points, 7 submissions: Level_Ad7440
    
    1. I'm with bawa (266 points, 34 comments)
    2. . (215 points, 34 comments)
    3. Pehle sahi tha sab 🏃 (181 points, 18 comments)
    4. 🕺🕺 (164 points, 31 comments)
    5. Real (159 points, 30 comments)
    6. Seedhe maut please 😥 (157 points, 16 comments)
    7. Bawa (135 points, 25 comments)
13. 1276 points, 8 submissions: Aggressive-Cat3277
    
    1. 🪀🪀🍯🥜 SUPREMACY (277 points, 25 comments)
    2. Kya rappers views la pa rahe hai? Be honest with me (242 points, 54 comments)
    3. Famous tourist taglines (198 points, 19 comments)
    4. RAFTAAR actual full name (191 points, 46 comments)
    5. Lyrically DOPE!!! (130 points, 11 comments)
    6. Jaldi se ek blue hai paani ki tarah MV phek deta hoon, cumback apne AAP hojayega:) (101 points, 14 comments)
    7. Sabko offend kardunga BC!! (100 points, 30 comments)
    8. DHH RAPPERS from their respective regions in 2024 elections (37 points, 18 comments)
14. 1188 points, 6 submissions: BreakLongjumping132
    
    1. Violation (541 points, 61 comments)
    2. Ye YT shorts wale 💀 (237 points, 54 comments)
    3. Yes (164 points, 18 comments)
    4. chini kam (102 points, 23 comments)
    5. Yes (82 points, 15 comments)
    6. menime (62 points, 2 comments)
15. 1015 points, 4 submissions: Aakash_dev00
    
    1. ☠️ (@alpha_mysterio0) • Instagram photos and videos (489 points, 76 comments)
    2. 🤪🤪 (247 points, 27 comments)
    3. sed meme 💔🐼 (212 points, 29 comments)
    4. cause : bhoutt saare bawa memes🤒🤧 (67 points, 9 comments)
16. 1003 points, 5 submissions: WhoimPS
    
    1. Emiway praising young galib for winning this beef (440 points, 28 comments)
    2. Pirated yoyo (214 points, 14 comments)
    3. Who's this ? (175 points, 192 comments)
    4. Pehli nazar me, kesa jaadu kar diya (121 points, 1 comment)
    5. If there is a physical fight , which team will win ? (53 points, 89 comments)
17. 985 points, 3 submissions: GroundbreakingAdi
    
    1. Ain't no way (420 points, 26 comments)
    2. Good luck on the new journey (328 points, 57 comments)
    3. jab tum so rahe the tab mai jaag raha tha (237 points, 65 comments)
18. 949 points, 4 submissions: Avinash_Paryani
    
    1. Name more such songs. (426 points, 56 comments)
    2. Hum toh aise hee hai. (296 points, 9 comments)
    3. Same energy. (166 points, 9 comments)
    4. W for DHH. (61 points, 6 comments)
19. 912 points, 9 submissions: Additional-Peak-1755
    
    1. Additional-peak-1755 (230 points, 22 comments)
    2. I loveeee youuuuu krsna (155 points, 48 comments)
    3. Its additional peak vs everyone {watch the last Kanye part} (154 points, 16 comments)
    4. Take it as a meme 😉 (153 points, 29 comments)
    5. Post p upvotes to kya comments bhi nhi ate upr se koi beef bhi nhi Kara 😭 (70 points, 23 comments)
    6. Inshort okbantaihiphop> dhhmemes> ih3 (65 points, 12 comments)
    7. Sbko shots dene ka mood hai (63 points, 14 comments)
    8. Kaditional analicheeze - why dhinchak pooja is hip hop (17 points, 3 comments)
    9. Winner Announcement of 69k event !!!!! (5 points, 30 comments)
20. 887 points, 4 submissions: GantaiREAPER
    
    1. Itna sach bhi nahi bolna tha Raftaar paaji (300 points, 27 comments)
    2. Ye DHH waale joota flex karte waqt meme kyun bana jaate hain (237 points, 19 comments)
    3. MC STAN and SACHIN TENDULKAR playing cricket together (219 points, 54 comments)
    4. !!69k Celebration!! (131 points, 37 comments)
21. 860 points, 3 submissions: stg_676
    
    1. Mazak lagra kya tereko (471 points, 34 comments)
    2. Me while watching Kapil Sharma show (226 points, 9 comments)
    3. ??? (163 points, 9 comments)
22. 849 points, 4 submissions: Udta_Punjab69
    
    1. when reporter starts pulling your triggers (629 points, 25 comments)
    2. every rapper acts gangsta until real gangster rapper walks in (raftaar stopped posting stories for him after that) (82 points, 18 comments)
    3. raftaar, krsna and khanna when every rapper leaves kalamkaar after getting some fanbase (75 points, 4 comments)
    4. uUuUuUuUuUuUuUuUuUu (63 points, 1 comment)
23. 834 points, 7 submissions: Tyreeed
    
    1. it do be like that (278 points, 42 comments)
    2. Title rap school ja rha hai (171 points, 12 comments)
    3. Bawa new EP ‘Bawa’ teaser (133 points, 7 comments)
    4. AAA MERI JAAN (124 points, 17 comments)
    5. Toh ye hai RAP ka full form 🤯 (73 points, 12 comments)
    6. Shy Bawa (40 points, 12 comments)
    7. Dr Zeus, our very own desi CJ (or Young Maylay) (15 points, 2 comments)
24. 785 points, 2 submissions: Adventurous-Guava981
    
    1. Eminem Becomes Emiway With AI (497 points, 33 comments)
    2. Galib's boobs >>> Khanna's boobs (288 points, 45 comments)
25. 783 points, 2 submissions: belikepac
    
    1. Krsna is a state of mind (660 points, 21 comments)
    2. Yashraj lala mazak lag rela hai tereko (123 points, 14 comments)
26. 721 points, 5 submissions: padh_le_beta
    
    1. Gaykode (231 points, 48 comments)
    2. No more bandi bars now 💔💔😭 (213 points, 42 comments)
    3. pols aa gayi pols 🚨🚨 (131 points, 13 comments)
    4. "Ye dono Bawa pyaar krte ek dusre se" (85 points, 26 comments)
    5. kr$na fan meet alert!! (61 points, 31 comments)
27. 717 points, 2 submissions: Wrong_Ad5174
    
    1. Seedhe Maut k Bina kya jeevan (414 points, 81 comments)
    2. Kendrick should stop copying these DHH gems 😠 (303 points, 57 comments)
28. 705 points, 3 submissions: Ok-Bug7788
    
    1. Guddi sikhran de jatt di Utton jatt dhur’an patt ni (465 points, 35 comments)
    2. Sidhu Moosewala🛐 reincarnated as Khandu Don (135 points, 19 comments)
    3. humpe tho hoo hi naa 😝😛😛 (105 points, 15 comments)
29. 677 points, 2 submissions: Bhutiya_Memer
    
    1. Shemdi 🫡 (470 points, 60 comments)
    2. bawa (207 points, 9 comments)
30. 658 points, 3 submissions: Unbiased-_-
    
    1. Life dedicate (380 points, 104 comments)
    2. ‌ (182 points, 84 comments)
    3. ‌‌ (96 points, 16 comments)
31. 628 points, 1 submission: Dope-Smoke
    
    1. chamdi moti honi chahiye 🗿 (628 points, 34 comments)
32. 622 points, 5 submissions: harjaan_08
    
    1. paji likhna bhul gaya bhay (317 points, 17 comments)
    2. stanny (140 points, 22 comments)
    3. Make comment section look like this guy search history. (66 points, 60 comments)
    4. Ayo! Shorry raff bhai (58 points, 14 comments)
    5. Apna orgasm dene vaala song tell crow. (41 points, 25 comments)
33. 606 points, 3 submissions: SpecialistMagician52
    
    1. Cute vol 1 (329 points, 17 comments)
    2. Parekishoes! (196 points, 22 comments)
    3. Jaat (81 points, 10 comments)
34. 604 points, 3 submissions: 9ighteye
    
    1. lafda posting (339 points, 30 comments)
    2. aw main ta dargi😨 (142 points, 8 comments)
    3. Madam shory (123 points, 13 comments)
35. 577 points, 4 submissions: _UTKRASHT_
    
    1. TITLE GAYA LAFDE DEKHNE (283 points, 40 comments)
    2. FREE LOKA MINTA LEJAO ISKI JAGAH (117 points, 13 comments)
    3. Banatai Ki Buddhi Choti Aur Boobies Badey Rehgye 😹😹 (116 points, 25 comments)
    4. JUST IN- MC STAN met PM Narendra Modi Ji , Home Minister Amit Shah Ji and CM of UP Yogi Adiyanath Ji and talked about LOKA.‼️FREE LOKA‼️ (61 points, 7 comments)
36. 571 points, 2 submissions: Kind-Blackberry-9434
    
    1. Lord from the streets 🛐🙏 (297 points, 12 comments)
    2. खाली स्थान (274 points, 16 comments)
37. 554 points, 1 submission: mohuhkkk_k
    
    1. better than मालूम है ना (554 points, 24 comments)
38. 504 points, 1 submission: Agitated-Ad3717
    
    1. Title gya bjp ko vote dene (504 points, 71 comments)
39. 484 points, 5 submissions: Final-Sample-
    
    1. 🤓 (139 points, 22 comments)
    2. they finally met (135 points, 12 comments)
    3. introvert hu bawa 😔👇 (80 points, 5 comments)
    4. 🍿🍿🍿 (66 points, 35 comments)
    5. Least sane SM fan (64 points, 38 comments)
40. 481 points, 1 submission: aryan_gami
    
    1. Your fav rapper could never ( headphones must ) (481 points, 44 comments)
41. 459 points, 4 submissions: MaroonMafia
    
    1. Aur Raftaar Bhaiya Ko Bahut Pyaar (197 points, 28 comments)
    2. KARMA SELF DISSING?? (119 points, 18 comments)
    3. Let's people be offensive (93 points, 26 comments)
    4. KARMA RIGHT NOW "magar ek din unko rona padhega jisdin bhi meri maiyat uthegi idhar zindagi ka janaza uthega udhar zindagi unki dulhan banegi " ~ Yogi Karma Nath (50 points, 6 comments)
42. 444 points, 1 submission: BRabbit987
    
    1. Rohan Cariappa's brain is finally fucked after decoding KR$NA's lyrics (444 points, 13 comments)
43. 441 points, 1 submission: Terminator_345
    
    1. Raftar 🗿 (441 points, 21 comments)
44. 426 points, 1 submission: MrKayPavlov
    
    1. BRA RAA (426 points, 12 comments)
45. 415 points, 3 submissions: Amogh_12344
    
    1. Kyu maa ki aankh karre ho (251 points, 28 comments)
    2. Aaj Raftaar paneer khayega 😋 (109 points, 20 comments)
    3. Rial heep haap (55 points, 3 comments)
46. 414 points, 1 submission: Minho37
    
    1. Ayoo raftaar...💀 (414 points, 38 comments)
47. 404 points, 2 submissions: DEADPOOL11178105
    
    1. AYYO KRSNA😱 (202 points, 25 comments)
    2. DHANKA DEKHTE HAI🤓👍 (202 points, 26 comments)
48. 403 points, 3 submissions: ishu_reddit
    
    1. Don't cancel Tittle,Tittle wanna exist. (165 points, 12 comments)
    2. Raftaar Bhaiya 🥰 (165 points, 15 comments)
    3. Newton's 3rd law (73 points, 3 comments)
49. 400 points, 1 submission: Darksec_Geek
    
    1. Cu**any goes International (400 points, 17 comments)
50. 395 points, 3 submissions: PuzzleheadedCan4466
    
    1. save nitin plz (150 points, 10 comments)
    2. too late? (143 points, 11 comments)
    3. Ab to dada bhi bol diye (102 points, 1 comment)

Top Commenters

1. CoolnamePanda (2151 points, 407 comments)
2. GaneshRasal (2070 points, 518 comments)
3. TheKingAlchemist (1587 points, 209 comments)
4. Even_Positive_756 (1155 points, 95 comments)
5. coconutanna (974 points, 128 comments)
6. Additional-Peak-1755 (927 points, 251 comments)
7. Aakash_dev00 (836 points, 148 comments)
8. stg_676 (785 points, 54 comments)
9. BIG_P0PPA_ (758 points, 121 comments)
10. Either_Cancel6241 (690 points, 62 comments)
11. Guilty_Surprise5534 (683 points, 44 comments)
12. Aggressive-Cat3277 (682 points, 115 comments)
13. Famous_Amphibian2385 (608 points, 70 comments)
14. International_Emu947 (606 points, 190 comments)
15. Final-Sample- (596 points, 62 comments)
16. Mammoth_Attention_73 (591 points, 35 comments)
17. DankKid2410 (582 points, 42 comments)
18. CheckQuick (566 points, 114 comments)
19. px1618 (558 points, 61 comments)
20. Shedller (527 points, 61 comments)
21. Lit_Extreme05 (481 points, 53 comments)
22. WhoimPS (447 points, 85 comments)
23. JEEASPIRANTOO7 (444 points, 52 comments)
24. NH_hostel (441 points, 51 comments)
25. Angryark (436 points, 51 comments)
26. Ornery-Reward-2784 (431 points, 95 comments)
27. M_A_G_M_A (427 points, 59 comments)
28. AdamQuixote (423 points, 71 comments)
29. DocDiffer (399 points, 94 comments)
30. Ok_Rice4694 (376 points, 119 comments)
31. Adorable_Set7546 (372 points, 32 comments)
32. PARTH-BHARDWAJ (350 points, 39 comments)
33. Valuable_Currency_27 (344 points, 138 comments)
34. letmebeD (336 points, 14 comments)
35. JODDUNMOIST (331 points, 20 comments)
36. Ok_Proposal_9149 (319 points, 35 comments)
37. anujmaral12 (312 points, 15 comments)
38. BrownBomber05 (299 points, 14 comments)
39. organic_mineral (297 points, 12 comments)
40. 9ighteye (292 points, 39 comments)
41. Careless_Landscape45 (282 points, 68 comments)
42. ken_kaneki07 (279 points, 33 comments)
43. en1_gma (267 points, 61 comments)
44. pirupirate07 (266 points, 35 comments)
45. Fit_Neighborhood6332 (258 points, 30 comments)
46. ZealousidealChange11 (248 points, 45 comments)
47. Weeblyweird (247 points, 59 comments)
48. realviivek (245 points, 41 comments)
49. Commercial-Air744 (245 points, 7 comments)
50. GantaiREAPER (243 points, 33 comments)
51. Noeljino (242 points, 18 comments)
52. Lamb_Shiftquantum (241 points, 11 comments)
53. SpecialistMagician52 (239 points, 30 comments)
54. Y0han_liebert (235 points, 19 comments)
55. ShadowLegend444 (234 points, 13 comments)
56. Electrical-Spare-973 (231 points, 23 comments)
57. Avinash_Paryani (225 points, 25 comments)
58. SAYmyNAME041 (223 points, 24 comments)
59. Stunning-Structure-5 (223 points, 24 comments)
60. Chance_Owl_8871 (223 points, 14 comments)
61. WarmRelationship8483 (204 points, 26 comments)
62. Gandi_Aulaad (201 points, 37 comments)
63. lookslikeajobforme (199 points, 27 comments)
64. KRS-100 (198 points, 28 comments)
65. Prior_Asparagus_1922 (194 points, 54 comments)
66. Sad_Firefighter3450 (185 points, 80 comments)
67. Affectionate-Bug875 (183 points, 6 comments)
68. Connect-Duck-8068 (182 points, 23 comments)
69. Fragerop (181 points, 8 comments)
70. retardedkidfr (179 points, 13 comments)
71. Archiet_Tanwar (176 points, 36 comments)
72. BiharibabuOp (171 points, 43 comments)
73. Gamferno (170 points, 7 comments)
74. No_Towel5111 (169 points, 28 comments)
75. Relevant-Pangolin777 (166 points, 30 comments)
76. PrettyIntrovertMe (166 points, 18 comments)
77. InstructionHot9577 (165 points, 34 comments)
78. AITC_worker_77 (163 points, 7 comments)
79. Grizzlyrang (161 points, 30 comments)
80. BRabbit987 (161 points, 27 comments)
81. Bhadwaguy69 (161 points, 23 comments)
82. 3Dips (161 points, 9 comments)
83. Accomplished_Path233 (159 points, 10 comments)
84. Big_a_zoid11 (155 points, 21 comments)
85. Hiyansh (151 points, 21 comments)
86. SH4M4N_OP (151 points, 6 comments)
87. Apprehensive-Tip4504 (150 points, 22 comments)
88. urban_ganwar (149 points, 74 comments)
89. okmishraji (146 points, 16 comments)
90. CharacterNorth1972 (141 points, 34 comments)
91. 0starhunter (141 points, 32 comments)
92. akshroom (140 points, 6 comments)
93. anotherKM7124 (139 points, 15 comments)
94. Anshul1806 (138 points, 43 comments)
95. ApurvX (136 points, 31 comments)
96. padh_le_beta (136 points, 25 comments)
97. Embarrassed_Inside_7 (136 points, 11 comments)
98. Tej_ass135 (132 points, 30 comments)
99. adolf-dumblewhore (130 points, 10 comments)
100. Aarav_Parmar (129 points, 14 comments)
101. HARDY__69 (129 points, 14 comments)
102. Manisheesh (128 points, 18 comments)
103. hackedbyayush (124 points, 18 comments)
104. Wrong_Ad5174 (123 points, 40 comments)
105. yash367 (123 points, 39 comments)
106. just_todi (123 points, 1 comment)
107. assissnator27 (122 points, 10 comments)
108. CatesAreCute (121 points, 10 comments)
109. brokeillionare (120 points, 3 comments)
110. comic_tak (118 points, 15 comments)
111. MysteriousCup1836 (118 points, 12 comments)
112. Beyond_infinity_1729 (118 points, 10 comments)
113. lysosome_destroyer (116 points, 33 comments)
114. fatherless_milf (116 points, 31 comments)
115. GurSubstantial8572 (116 points, 9 comments)
116. Aayosh (115 points, 5 comments)
117. Kind-Blackberry-9434 (114 points, 13 comments)
118. Such-Wolverine-2275 (113 points, 23 comments)
119. harjaan_08 (112 points, 17 comments)
120. shaansmwl77 (112 points, 16 comments)
121. okk_123 (111 points, 19 comments)
122. hridayLejaega (111 points, 9 comments)
123. Prestigious-Coach-81 (110 points, 39 comments)
124. Embarrassed-Cup-2858 (109 points, 1 comment)
125. Puneet_7164 (107 points, 10 comments)
126. BreakLongjumping132 (106 points, 45 comments)
127. PussyMaster1o (106 points, 6 comments)
128. Jaibheem-chhotabheem (105 points, 8 comments)
129. pawan270319999 (104 points, 9 comments)
130. Designer-Stable-3072 (104 points, 6 comments)
131. Interesting-Force347 (102 points, 6 comments)
132. Powerful-Prune-2795 (101 points, 20 comments)
133. rewanshh (101 points, 10 comments)
134. Sparsh_Pandey (100 points, 6 comments)
135. Darknight_hell (99 points, 27 comments)
136. Holiday_Fun_3217 (97 points, 20 comments)
137. DONALD__DRUG (97 points, 10 comments)
138. thanosbutthicc (95 points, 20 comments)
139. LostLaw9417 (95 points, 12 comments)
140. sarvtej_924 (95 points, 7 comments)
141. Scary-Pay-117 (94 points, 19 comments)
142. desislime (94 points, 1 comment)
143. Special-Walrus-8650 (92 points, 44 comments)
144. Tyreeed (92 points, 4 comments)
145. CrewCold9260 (91 points, 11 comments)
146. Educational_Key6764 (89 points, 16 comments)
147. hrshysh (89 points, 4 comments)
148. bucck2929 (87 points, 30 comments)
149. decentmahant69 (86 points, 16 comments)
150. RELUGA_XD (86 points, 9 comments)
151. SkyProfessional9735 (84 points, 20 comments)
152. AlkunPein (84 points, 15 comments)
153. NotKakarott (81 points, 12 comments)
154. shutupmatsuda (81 points, 8 comments)
155. wellwisher54 (80 points, 10 comments)
156. Apart-Big-6120 (80 points, 4 comments)
157. ReeTesHp (80 points, 3 comments)
158. alt94701 (79 points, 12 comments)
159. Dry_Tutor2027 (79 points, 11 comments)
160. harshu2006 (79 points, 7 comments)
161. Agitated_Shadow (78 points, 35 comments)
162. couchfrenchfry (78 points, 13 comments)
163. deba2607 (78 points, 6 comments)
164. theRealMadridGuy (78 points, 6 comments)
165. Ok-Following-6370 (78 points, 3 comments)
166. ignoreddd (78 points, 2 comments)
167. Accomplished-Ad-6007 (77 points, 17 comments)
168. International_Luck56 (77 points, 15 comments)
169. Loan-Fast (77 points, 4 comments)
170. Lexyeusdenini (76 points, 22 comments)
171. danish_bhai (75 points, 14 comments)
172. harshwardhannnn (75 points, 4 comments)
173. 84_years_returns (74 points, 18 comments)
174. Actual_Employment_89 (74 points, 13 comments)
175. Lopsided-Data-2699 (73 points, 20 comments)
176. kaatilRandwa (72 points, 4 comments)
177. WADE_- (71 points, 12 comments)
178. Appropriate-Bird6642 (71 points, 8 comments)
179. Levitillzekealive (71 points, 7 comments)
180. GodKodse (70 points, 24 comments)
181. uknowtfiam (70 points, 18 comments)
182. Awaam_se_hoon (70 points, 17 comments)
183. itzSKIT (70 points, 12 comments)
184. Glum_Feelings (70 points, 5 comments)
185. ravv_idk (70 points, 3 comments)
186. EmiwayBantaiOrganic (69 points, 18 comments)
187. Ansh339 (69 points, 15 comments)
188. DODA05 (69 points, 14 comments)
189. Extreme-Profit-8141 (68 points, 1 comment)
190. DependentImportant51 (67 points, 1 comment)
191. No-Station8968 (66 points, 14 comments)
192. Ill_BLOGGER_633 (65 points, 12 comments)
193. Otherwise_Pace_1133 (65 points, 12 comments)
194. MOURYASAJAL (64 points, 1 comment)
195. PowerFulMan88 (63 points, 23 comments)
196. Impressive_Camera173 (63 points, 2 comments)
197. Regular_Security_179 (62 points, 20 comments)
198. raj-marvel-boy7138 (62 points, 13 comments)
199. Low_Gmer (62 points, 10 comments)
200. allgasm (62 points, 9 comments)

Top Submissions

1. Title gaya Blanko lagane by Puneet_7164 (981 points, 45 comments)
2. Title hierarchy ka matlab dhundh raha hai by kaatilRandwa (899 points, 81 comments)
3. Throwback to 2009 when Raftaar saw girls doing joints in Delhi by ___sohammmm (715 points, 76 comments)
4. Krsna is a state of mind by belikepac (660 points, 21 comments)
5. when reporter starts pulling your triggers by Udta_Punjab69 (629 points, 25 comments)
6. chamdi moti honi chahiye 🗿 by Dope-Smoke (628 points, 34 comments)
7. Ye karke dikhao by deleted (616 points, 33 comments)
8. Travis Scott from Pune by ___sohammmm (589 points, 25 comments)
9. Beef Labels by BIG_P0PPA_ (571 points, 26 comments)
10. better than मालूम है ना by mohuhkkk_k (554 points, 24 comments)

Top Comments

1. 167 points: Lamb_Shiftquantum's comment in Eminem blatantly copied anjum sir😨
2. 147 points: Commercial-Air744's comment in Who's this ?
3. 145 points: ShadowLegend444's comment in Good luck on the new journey
4. 138 points: deleted's comment in No more bandi bars now 💔💔😭
5. 126 points: Gamferno's comment in Seedhe Maut k Bina kya jeevan
6. 125 points: BrownBomber05's comment in Title hierarchy ka matlab dhundh raha hai
7. 123 points: SH4M4N_OP's comment in With a boner
8. 123 points: just_todi's comment in With a boner
9. 118 points: akshroom's comment in Guddi sikhran de jatt di Utton jatt dhur’an patt ni
10. 117 points: Affectionate-Bug875's comment in Most unbiased reaction channel in dhh :)

Generated with BBoe's Subreddit Stats

This report is a detailed and comprehensive analysis of the world market for Commercial Edger, and provides market size (US$ million) and Year-over-Year (YoY) growth, considering 2022 as the base year. This report explores demand trends and competition, as well as details the characteristics of Commercial Edger that contribute to its increasing demand across many markets.
This reports profiles key players in the global Commercial Edger market based on the following parameters – company overview, production, value, price, gross margin, product portfolio, geographical presence, and key developments. Key companies covered as a part of this study include Stihl, RedMax, ECHO, Husqvarna, McLane, EGO, Turf Teq, CRAFTSMAN and Shindaiwa, etc.
Request Sample Report and Full Report TOC:
https://www.globalinforesearch.com/reports/1534955/commercial-edger
Highlights and key features of the study Global Commercial Edger total market, 2018-2029, (USD Million) Global Commercial Edger total market by region & country, CAGR, 2018-2029, (USD Million) U.S. VS China: Commercial Edger total market, key domestic companies and share, (USD Million) Global Commercial Edger revenue by player and market share 2018-2023, (USD Million) Global Commercial Edger total market by Type, CAGR, 2018-2029, (USD Million) Global Commercial Edger total market by Application, CAGR, 2018-2029, (USD Million)
Key Questions Answered 1. How big is the global Commercial Edger market? 2. What is the demand of the global Commercial Edger market? 3. What is the year over year growth of the global Commercial Edger market? 4. What is the total value of the global Commercial Edger market? 5. Who are the major players in the global Commercial Edger market? 6. What are the growth factors driving the market demand?
Global Info Research is a report publisher, a customer, interest-based suppliers. Is in the best interests of our clients, they determine our every move. At the same time, we have great respect for the views of customers. With the improvement of the quality of our research, we develop custom interdisciplinary and comprehensive solution. For further development, we will do better and better. Global Info Research will with excellent professional knowledge and experience to carry out all aspects of our business. At the same time, we will thoroughly look for information, to give a more comprehensive development.
Global Info Research has more than 30000 global well-known customers, covering more than 30 industries, research regions cover China,Japan,Korea,US,EU,Asia,Middle East and Africa,South America,Australia,etc
E-mail: [email protected]
Tel: +86-17665052062 00852-58197708(HK)
Add:FLAT/RM A 9/F SILVERCORP INTERNATIONAL TOWER 707-713 NATHAN ROAD MONGKOK KL HONG KONG
Website: http://www.globalinforesearch.com

Looking for something agile and slightly quick. It'll be my first bike as well so yeah.

The Berlin Major

Presented by ESL

Sponsored by Intel • Predator • DHL • Monster Energy

• Location: Berlin, Germany
• Date: April 26 - May 7, 2023
• Version: 7.33
• ESL Media: Homepage Twitter YouTube Facebook Instagram
• Coverage: DPC Homepage Liquipedia Joindota GosuGamers Dotabuff

Streams

EN: Twitch 1 Twitch 2 Twitch 3 Twitch 4 Youtube RU: Twitch 1 Twitch 2 Twitch 3 Twitch 4 CN: Huya 1 Huya 2 Huya 3 Huya 4 PT: Twitch 1 Twitch 2 Twitch 3 Twitch 4 ES: Twitch 1 Twitch 2 Twitch 3 Twitch 4 YouTube UA: Twitch 1 Twitch 2 Twitch 3 Twitch 4 YouTube

Broadcast Talent

Host:

• Tsunami (Neal Khandheria)
• Sheever (Jorien van der Heijden)
• PyrionFlax (Ted Forsynth)

Sideline Reporter:

• Slacks (Jake Kanner)

CommentatoAnalyst:

• Ephey (Mira Riad)
• Sheepsticked (Alexandra Roberts)
• Jenkins (Andrew Jenkins)
• ODPixel (Owen Davies)
• Fogged (Ioannis Loucas)
• Cap (Austin Walsh)
• SVG (Avery Silverman)
• ppd (Peter Dager)
• BSJ (Brian Canavan)
• syndereN (Troels Lyngholt Nielsen)
• SUNSfan (Shannon Scotten)
• N0tail (Johan Sundstein)

Remote CommentatoAnalyst:

• Gareth (Gareth Bateson)
• Lacoste (Dominik Stipić)
• Ares (Lachlan Williams)
• GoDz (David Parker)
• rkryptic (Ricky Peterson)
• johnxfire (John Nathan Fernandez)
• Lyrical (Gabriel Cruz)
• Fear (Clinton Loomis)
• Trent (Trent MacKenzie)
• Purge (Kevin Godec)
• Dktruman (Tim de Lange)
• Dktruman (Tim de Lange)
• Astini (Filipe Astini)

Observer:

Teams

Team Liquid: miCKe, Nisha, zai, Boxi, iNSaNiA, Coach: Blitz
Tundra Esports: skiter, Nine, 33, Saksa, Sneyking, Coach: Aui_2000
Gaimin Gladiators: DyrachYO, Quinn, Ace, tOfu, Seleri, Coach: Cy-
OG: Yuragi, bzm, MinD_ContRoL¹, Taiga, Kitrak¹
9 Pandas: RAMZES666, kiyotaka, MieRo`, Antares, Solo<sup>2</sup>, Coach: Nofear
Team Spirit: Yatoro, Larl, Collapse, Mira, Miposhka, Coach: Silent
BetBoom Team: Nightfall<sup>3</sup>, gpk~, Resolut1on<sup>3</sup>, Save-, TORONTOTOKYO, Coach: boolk
Xtreme Gaming: Ghost, Paparazi灬, JT-, Kaka, Dy, Coach: LaNm
PSG.LGD: shiro, NothingToSay, zeal, planet, y`, Coach: xiao8
Invictus Gaming: Emo, Dust, 项羽, Pyw, qyqx, Coach: super
Team Aster: Monet, SumaiL⁴, Xxs, BoBoKa, 皮球, Coach: Mad
Talon Esports: 23savage, Mikoto, Jabz, Q, Oli, Coach: SunBhie
Execration: Palos, Bob, Tino, Shanks, Carlo
Team SMG: CDR, MooN, MidOne, ah fu, poloson
TSM: Timado, Bryle, kasane, Ari, Whitemon, Coach: MoonMeander
Shopify Rebellion: Arteezy, Abed, SabeRLight-, Cr1t-, Fly, Coach: BuLba
beastcoast: K1, DarkMago, Sacred, Scofield, StingeR, Coach: Raykill
Evil Geniuses: Pakazs, Chris Luck, Wisper, Matthew, Pandaboo, Coach: Vintage
1: DM, Chu, and Misha are unable to attend the major, MinD_ContRoL and KitraK will stand-in. 2: Solo may be replaced by RodjER during the early days of the tournament. 3: Pure is unable to attend the major, Resolut1on will stand-in as position 3 while Nightfall moves to position 1. 4: Xwy is benched, SumaiL will stand-in..

Format

Group Stage: April 26 - April 30

• Two groups of nine teams each
• Single round-robin
• All matches are Bo2
• Top four teams from each group advance to the upper bracket of the playoffs
• 5th and 6th place teams from each group advance to the lower bracket of the playoffs
• Remaining teams are eliminated

Playoffs: May 01 - May 07

• Double-elimination bracket
• All matches except Grand Final are Bo3
• Grand Final is Bo5

Prize Pool

Place	$ USD	DPC Points
1st	$200,000	500
2nd	$100,000	450
3rd	$75,000	400
4th	$50,000	350
5th-6th	$25,000	300
7th-8th	$12,500	200
9th-12th	-	-
13th-14th	-	-
15th-16th	-	-
17th-18th	-	-

Schedule

Day 1 (April 26) - Group Stage

ID	Team	vs	Team	Result	Cntdwn (CEST)	PST	EST	GMT	SGT	AEST	Frmt
A1	Evil Geniuses	vs	Gaimin Gladiators	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A2	PSG.LGD	vs	Execration	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A3	9 Pandas	vs	TSM	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A4	BetBoom Team	vs	Team Aster	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A5	Team Liquid	vs	Team Aster	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
A6	TSM	vs	BetBoom Team	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
A7	PSG.LGD	vs	Gaimin Gladiators	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
A8	Execration	vs	Evil Geniuses	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
B1	Tundra	vs	Team Spirit	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B2	Talon Esports	vs	beastcoast	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B3	Shopify Rebellion	vs	Invictus Gaming	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B4	Team SMG	vs	OG	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B5	Xtreme Gaming	vs	OG	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2
B6	Team Spirit	vs	Shopify Rebellion	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2
B7	Invictus Gaming	vs	Tundra Esports	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2
B8	beastcoast	vs	Team SMG	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2

Day 2 (April 27) - Group Stage

ID	Team	vs	Team	Result	Cntdwn (CEST)	PST	EST	GMT	SGT	AEST	Frmt
A9	PSG.LGD	vs	Evil Geniuses	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A10	Gaimin Gladiators	vs	BetBoom Team	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A11	Team Liquid	vs	9 Pandas	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A12	Execration	vs	Team Aster	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A13	Team Liquid	vs	BetBoom Team	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
A14	TSM	vs	Evil Geniuses	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
A15	9 Pandas	vs	Team Aster	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
A16	Execration	vs	Gaimin Gladiators	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
B9	Xtreme Gaming	vs	Talon Esports	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B10	Tundra	vs	Shopify Rebellion	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B11	Team Spirit	vs	OG	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B12	Invictus Gaming	vs	Team SMG	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B13	beastcoast	vs	Shopify Rebellion	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2
B14	Talon Esports	vs	OG	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2
B15	Team Spirit	vs	Invictus Gaming	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2
B16	Xtreme Gaming	vs	Team SMG	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2

Day 3 (April 28) - Group Stage

ID	Team	vs	Team	Result	Cntdwn (CEST)	PST	EST	GMT	SGT	AEST	Frmt
A17	9 Pandas	vs	PSG.LGD	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A18	Team Liquid	vs	TSM	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A19	Gaimin Gladiators	vs	Team Aster	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A20	Evil Geniuses	vs	BetBoom Team	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A21	Team Liquid	vs	Gaimin Gladiators	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
A22	9 Pandas	vs	BetBoom Team	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
A23	TSM	vs	Execration	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
A24	PSG.LGD	vs	Team Aster	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
B17	Invictus Gaming	vs	OG	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B18	Talon Esports	vs	Tundra	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B19	Xtreme Gaming	vs	beastcoast	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B20	Shopify Rebellion	vs	Team SMG	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B21	beastcoast	vs	Team Spirit	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2
B22	Tundra Esports	vs	OG	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2
B23	Xtreme Gaming	vs	Invictus Gaming	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2
B24	Talon Esports	vs	Team SMG	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2

Day 4 (April 29) - Group Stage

ID	Team	vs	Team	Result	Cntdwn (CEST)	PST	EST	GMT	SGT	AEST	Frmt
A25	TSM	vs	Team Aster	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A26	9 Pandas	vs	Evil Geniuses	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A27	Team Liquid	vs	PSG.LGD	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A28	Execration	vs	BetBoom Team	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A29	Team Liquid	vs	Execration	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
A30	TSM	vs	PSG.LGD	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
A31	Evil Geniuses	vs	Team Aster	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
A32	9 Pandas	vs	Gaimin Gladiators	0:0	14:30	5:30	8:30	12:30	20:30	22:30	Bo2
B25	Xtreme Gaming	vs	Tundra	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B26	Talon Esports	vs	Shopify Rebellion	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B27	beastcoast	vs	OG	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B28	Team SMG	vs	Team Spirit	0:0	17:15	8:15	11:15	15:15	23:15	1:15	Bo2
B29	Shopify Rebellion	vs	OG	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2
B30	beastcoast	vs	Tundra Esports	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2
B31	Xtreme Gaming	vs	Team Spirit	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2
B32	Talon Esports	vs	Invictus Gaming	0:0	19:45	10:45	13:45	17:45	1:45	3:45	Bo2

Day 5 (April 30) - Group Stage

ID	Team	vs	Team	Result	Cntdwn (CEST)	PST	EST	GMT	SGT	AEST	Frmt
A33	PSG.LGD	vs	BetBoom Team	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A34	TSM	vs	Gaimin Gladiators	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A35	Team Liquid	vs	Evil Geniuses	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
A36	9 Pandas	vs	Execration	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo2
B33	Tundra	vs	Team SMG	0:0	14:45	5:45	8:45	12:45	20:45	22:45	Bo2
B34	Talon Esports	vs	Team Spirit	0:0	14:45	5:45	8:45	12:45	20:45	22:45	Bo2
B35	beastcoast	vs	Invictus Gaming	0:0	14:45	5:45	8:45	12:45	20:45	22:45	Bo2
B36	Xtreme Gaming	vs	Shopify Rebellion	0:0	14:45	5:45	8:45	12:45	20:45	22:45	Bo2

Day 6 (May 01) - Main Event

ID	Team	vs	Team	Result	Cntdwn (CEST)	PST	EST	GMT	SGT	AEST	Frmt
QF1	Gaimin Gladiators	vs	Team Spirit	0:0	11:00	2:00	5:00	9:00	17:00	19:00	Bo3
QF2	OG	vs	Team Liquid	0:0	14:35	5:35	8:35	12:35	20:35	22:35	Bo3
QF3	9 Pandas	vs	Shopify Rebellion	0:0	18:05	9:05	12:05	16:05	0:05	2:05	Bo3
QF4	Tundra Esports	vs	PSG.LGD	0:0	21:40	12:40	15:40	19:40	3:40	5:40	Bo3

Day 7 (May 02) - Main Event

ID	Team	vs	Team	Result	Cntdwn (CEST)	PST	EST	GMT	SGT	AEST	Frmt
LB1-1	Team Spirit	vs	Team Aster	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo3
LB1-2	Team Liquid	vs	Xtreme Gaming	0:0	15:35	6:35	9:35	13:35	21:35	23:35	Bo3
LB1-3	Shopify Rebellion	vs	Evil Geniuses	0:0	19:05	10:05	13:05	17:05	1:05	3:05	Bo3

Day 8 (May 03) - Main Event

ID	Team	vs	Team	Result	Cntdwn (CEST)	PST	EST	GMT	SGT	AEST	Frmt
LB1-4	PSG.LGD	vs	Talon Esports	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo3
SF1	Gaimin Gladiators	vs		0:0	15:35	6:35	9:35	13:35	21:35	23:35	Bo3
SF2	9 Pandas	vs	Tundra Esports	0:0	19:05	10:05	13:05	17:05	1:05	3:05	Bo3

Day 9 (May 05) - Main Event

ID	Team	vs	Team	Result	Cntdwn (CEST)	PST	EST	GMT	SGT	AEST	Frmt
LB2-1	Team Aster	vs	Team Liquid	0:2	12:00	3:00	6:00	10:00	18:00	20:00	Bo3
LB2-2	Evil Geniuses	vs	Talon Esports	2:0	15:35	6:35	9:35	13:35	21:35	23:35	Bo3
LB3-1	Tundra Esports	vs	Team Liquid	0:2	19:05	10:05	13:05	17:05	1:05	3:05	Bo3

Day 10 (May 06) - Main Event

ID	Team	vs	Team	Result	Cntdwn (CEST)	PST	EST	GMT	SGT	AEST	Frmt
LB3-2	OG	vs	Evil Geniuses	0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo3
UBF	Gaimin Gladiators	vs	9 Pandas	0:0	15:35	6:35	9:35	13:35	21:35	23:35	Bo3
LBSF	Team Liquid	vs		0:0	19:05	10:05	13:05	17:05	1:05	3:05	Bo3

Day 11 (May 07) - Main Event

ID	Team	vs	Team	Result	Cntdwn (CEST)	PST	EST	GMT	SGT	AEST	Frmt
LBF		vs		0:0	12:00	3:00	6:00	10:00	18:00	20:00	Bo3
GF		vs		0:0	16:00	7:00	10:00	14:00	22:00	0:00	Bo5

Countdown times are in CEST. All times are subject to change based on the length of matches and delays. Other match discussions: /dota2 on Discord

Hey guys, so I have a small landscaping and lawn service company, and I’m having a bit of an issue. I have a 2 year old Husqvarna 150BT backpack blower and less than a year old, used 4 times, Stihl KM91R.
Both of these units, idle fine and do fine under low to medium throttle, but shut off at high throttle.
I don’t think the current gas is the issue, because my other equipment runs just fine. Any tips or advice? The Husq was used heavily so maybe I should check the spark arrestor but the brand new stihl has me stumped…
Thanks!

Hi,
My father's blower is not starting up. It starts, but then immediately stops as shown in this video:
​
https://reddit.com/link/12jrmt6/video/4jf2ol2uchta1/player
What do you guys suggest I check? Thanks for any help.

I'm not done yet!

Theme - Conviction
Asta grew up as a magicless dweeb in the boonies, who spent his days working out, yelling, and yelling some more. He had a goal in his mind: to become the Wizard King, the Clover Kingdom's number one mage, and defeat his rival and fellow orphan Yuno at said goal. In his adolescence, he received the Grimoire with a Five-Leaf Clover on it, holding the power of a Devil, which revolutionized his chances of becoming number one. Now, he wields his demonic swords to climb the societal ladder and help all who need it to reach his dream burning bright.

• Character profile

KEY
Hover over a feat to see the chapter of occurrence. After Yuno awakened Sylph, Asta's Grimoire gained new text inside of it, and now stores his sword. The Witch Queen strengthened his arms when healing his curse. She also awakened his Demon form. As of fighting Langris, Asta can trigger Demon form twice a day, but it requires him to pose and circulate his ki.
Base < [Black] < [Black]
Also it should be of note that most of the time, Asta's swords...don't really act like swords. They're more like heaps of raw iron.

Strength / Anti-Magic

Lifting

• Does 1000 one-handed upright pushups
• Throws the Demon-Slayer Sword like a spear
• The Demon-Dweller Sword is fairly heavy for Julius
• Hurls the Demon-Dweller Sword into a portal beneath Rades when he can't cross the distance in time
• Tosses a drunkard into a bar's ceiling
• With Yami, escapes from being buried under rubble
• Stops Vetto's arm
• Throws his the Demon-Dweller Sword into Ladros, who is airborne at the time, and Ladros can't remove it due to the weight
• Dug a hole with his sword

Demon-Slayer Sword
Negating Magic (Edge)
Non-Clover Kingdom Wizards

• Cuts the smoke binding Luck
• Slices through the Prison of the Fallen King, but it just reforms
• Chops Mars' Laevateinn in half
  • Does it again
• Destroys a Talos Doll of Mars'
• Destroys some minerals impeding him
• Destroys more minerals
• Cuts through Mars' Titan Armor
• [Black] Shears a path through Ladros' beam
• [Black] Defeats Ladros with Black Meteorite, removing all of his body's mana

Clover Kingdom Wizards

• Easily smashes Sekke's bronze sphere
• Slashes through Noelle's out of control Water Magic and neutralizes it
• Cuts through a magic mist
• Cuts through Heath's gigantic ice projectile
• Cuts through magic mist again
• Defeats Heath, but instantly falls asleep afterwards
• Dispels a bandit's wind cloud
• Defeated a horde of wraiths
• Cancels a spatial magic
• Removes the curse from his wound
• Negates Yuno's wind bird
• Cuts Monster Baro's arm off
• Deletes the barrier protecting two individuals from memory probes
• Cancels Poizot's transparency magic
• Cuts Yuno's wind spell
• Deletes some of Kirsch's cherry petals in a whirlwind
• [Black] Clears a battlefield of both Zora's traps and an AOE cherry blossom illusion magic
• Destroys Langris' space magic
• [Black] Cuts Langris' crystal and ruins his counterattack
• [Black] Rejects an execution squad's blasts

Elves

• Cancels Vetto's spell
• Defeats Vetto
• Cuts a snake tendril abducting a witch
• Slashes through Fana's fireball
• Cancels the fire surrounding Fana
• [Black] Decapitates Lira's four snake creations
• [Black] With Yuno, cuts through Le Chateau du Verre, a giant mobile fortress crafted of ice and stone
• [Black] In another team attack with Yuno, disarms Licht of the Demon-Destroyer Sword

Environment

• Chops through a magic trap's ice
• Cuts the vines holding Noelle
• Cuts through some liquid ghoul monsters
• Cuts the door to a dungeon's treasure hall down
• Removes the curse dulling Marco's mind
  • Does the same for many other children
• Dispels a magic sea current wall
• Cuts through a boulder
• Cancels Raia's suicide bomb
• Cuts through magic elven stone
• Trims some hostile weeds
  • Does it again, this time in a whirlwind

Other

• Wounds Salamander
• [Black] Shifts Zagred's trident back
• [Black] Breaks Zagred's trident

Limits

• Can't break free of Julius' timestop
• [Black] Cannot exorcise elven souls from human bodies
• Can't break the effects of a poison on an abbot

Repelling Magic (Flat)
Wizards

• Hits Magna's fireball back
• Bats aside Heath's icicles
• Hits Magna's fireball into Heath's men from behind
• Protects Luck from a beam of smoke
• Can only block Mars' Harpe blades
• Hits Solid's water bullet back
• Reflects two of Alfred's bolts
• Hits Ladros' beam back
• [Black] Smacks Lumiere away

Elves

• Blocks Raia's beams as they appear from portals
• When Zora doubles five elves combined magic, Asta hits it back into him so he can quadruple it, then launches the quadrupled spell at the elves
• Blocks a stone spear
• [Black] While in Yuno's wind currents, blocks the projectile rays Licht sent from the Demon-Dweller Sword
• Smacks one of Drowa's beams back

Other

• Hits Salamander's fireball back
  • Hits another one into the air
• [Limit] Against a regular sword, it's also just a regular sword

Other (Plain Damage)
Humans

• Launches Revchi and craters him in a wall
• Bludgeons Heath with the point of his sword even after losing his momentum
• Smashes Lotus, breaking his arm and a few ribs
• Sends Mars flying into a pool despite his mineral armor
• Cuts Rades' cheek
• Hits Baro into a cave wall
• Defeated one of the Temple Mages whose magic is meant to turn aside physical attacks

Non-Humans

• Blasts through wraiths
• Launches wraith bodies in a single slash
• Smashes Licht across a cavern
• Slams Vetto into the ground

Black Divider

• [Black] Races through Yuno's wind tunnel to run Zagred through
• [Black] Cuts through Zagred's giant worm snake things
• [Black] Pincers Zagred with Yuno, though his form expires before they can kill him
• [Black] Cleaves Zagred

Demon-Dweller Sword
Ranged

• Destroys Mars' mineral armor and overpowers his flame healing when combined with Noelle's water magic
• Cuts through Monster Baro with a projectile slash of anti-magic light

Negating Magic
Corporeal

• Destroys Mars' Harpes
• Cuts through Alecdora's sand sphere
• As Gauche's Mirror Brigade, dices Monster Baro
• Ruins Yago's snail spell
• Slashes a rock ballista shot down
• Cancels one of Magna's fireballs, and from there, he hits 4/5 of them
• Stops Kirsch's cherry petal blast
• Defeats Kirsch
• Cleaves a magic crystal and the golem it is embedded in
• Cuts through magic elven ice
• [Black] Erases Undine's water knights
• [Black] Cuts through Lolopechka's waterfall curtain

Ethereal

• Disrupts spatial magic to keep Rades from escaping
• Removes the curse from his wound
• When impaled in Vetto, his ability to track Asta's movements fails
• Slashes through Ladros' beam
  • Does it to multiple
• Drains Ladros' magic
• Stops Langris' space magic projectile

Limits

• Can't escape Alecdora's sand golem
• Only partially hits one of Valtos' attacks
• Cuts a brief bit of Vetto's beast magic blast to protect Noelle, the blast easily obliterating three walls behind him
• Can't cut a volcano's ambient heat

Repelling Magic

• Deflects water and mist magics from Solid and Nebra
• Blocks Jimmy's curse barrage
• Blocks one of Alfred's bolts
• Deflects Gauche's beams
• After unlocking his ki sensing fully, reflects Valtos' blast
• Blocks Raia's beams as they appear from portals

Other

• Begins absorbing Julius' magic when he holds it
• Knocks Vetto back, which Vetto says hurts more than some kill spells
• Slices open Vetto's cheek
• Impales Vetto
• Cuts down a volcano rock
• [Black] Rockets up the side of a volcano, one-shots a small magma monster in his path

Demon-Destroyer Sword
Erasure

• Makes it so that a spell never happened

Soul Cleansing

• Removes the elven soul from a magic knight
• [Black] Exorcises Rufel from Luck
• Saves three more mages
• [Black] Exorcises Gauche and Marie's elven souls
• Exorcises Patry from William
• Mass exorcises the elven race when paired with William's world tree

Other

• Removes the poison from Hage's villagers, something the Demon-Slayer Sword could not do
• When shot out of a cannon, dispels a combined effort of three elves' ice, water, and earth magics
• Breaks an immobilization spell on the Black Bulls
• [Black] Whirlwinds to keep Drowa's beams from reaching him
• [Black] With Yuno, creates a chink in Zagred's skin
• [Black] With Yuno backing him, cancels out a dark elven light beam, said to be stronger than the one Yami had trouble with when fighting Patry
• Cuts through Zagred's magic eater gunk stuff
• [Black] Dispels Damnatio's scale magic
• Cancels Nathan's curse oil magic

Other

• This sword is "very sensitive to huge magic"
• Breaks Rufel's lightning spear
• [Black] Cuts Rufel's lightning down
• [Black] Blocks Rufel's kick
• [Black] Tanks a black light barrage that sends him into a wall, in spite of blocking it with his sword and some ambient anti-magic
• [Black] Enters Patry's mind and after giving him a talk, makes him give up his demons

Dual Swords
Slayer & Dweller

• Defeats Jimmy in a tornado of slashes
• Hits back magic blasts on all sides
• Deflects Valtos' attacks for the most part
• Evaporates Salamander
• Rebounds Kirsch's cherry blossom tide and KO's him

Slayer & Destroyer

• [Black] With both the Demon-Slayer and Demon-Destroyer Swords, keeps Licht's attack from obliterating him and Yuno even as it smashes through multiple levels of the megalith they're fighting in
• Takes on Lira's picture magic

Striking (Not Swords)

• Headbutts Gauche to the ground
• Punches a small divot in a cave wall
• Punches Valtos in the face

Other

• Bites Vetto's arm when he can't use his sword
• The Witch Queen expected him to lose the battle of wills against the demon in him
• Briefly resists the Witch Queen's blood magic while unconscious

Durability

Blunt
Magic

• Floored by a Lily's water fist
• Beat up by Revchi's chains
• Hit with the effects of multiple magic traps
• Sent flying by Noelle

Physical

• Rolls down a hill and slams into a castle wall
• Plowed through a wall by Mars' Harpe blades
• Smashed through a wall
• Clubbed by Charmy
• Grazed by Vetto, which still forces him back
• Punched into a cave wall by Vetto
• Kinda sorta blocks Vetto's punches with both of his swords
• Backhanded by Vetto
• Has his arms crushed by Vetto
• "Rock-like" after being healed by the Witch Queen
• [Black] Punched in the face by Mereoleona, who craters him in a volcano's wall, and is basically fine
• [Black] [A kick of Rufel's sends him skidding into rubble
• [Black] Pounded by Zagred's commands
• [Black] Tanks a black light barrage that sends him into a wall, in spite of blocking it with his sword and some ambient anti-magic

Esoteric
Heat

• Blown up
• Launched back by Mars' fire regenerative magic
• Burned by Valtos' attack
• Knocked away by the heat aura surrounding Fana's salamander
• "Would be turned into charcoal" if he kept approaching Fana (which he does, though it's with Mars healing him continuously)
• Cooked by Ladros' beam
• Hikes a volcano
• Blasted by Mereoleona's fireball

Other

• Shot by Ladros
• The Royal Knight robes he gains after being accepted have resistance to magical attacks
• Untouched by magic-leeching organisms because he has no magic

Piercing

• An icicle erupts into him, and he's back on his feet raring to go, multiple times
• One of Mars' Harpe daggers to the gut takes him out
• Impaled on wind needles
• Bit by Sally
• One of Licht's spears goes through his leg
• Cut by some of Kirsch's cherry petals
• [Black] Impaled on a dark elven lightblade that mirrors the Demon-Slayer Sword

Other

• Immune to social pressure
• Could climb a mountain

Speed

Agility

• Rolls out of the way of a thrown boulder
• Uses airborne warlocks as platforms to close the gap on their leader
• Jumps out of the way of Fana's salamander's firebreath
• Spins to stave off Kirsch's tide of cherry petals
• Blocks one of Magna's fireballs he was having trouble sensing the ki of
• Jumps to Kirsch's position in midair
• Charges through Yuno's wind tunnel
• [Black] Slashes some of the "Storm of Blades" down

Dexterity (Arms)

• Faster with the Demon-Dweller Sword
• Cuts down Jimmy's curse barrage
• Blocks Gauche's beams
  • Blocks more
• Slashes Ladros' beams down
• Cuts a ballista shot from 60m down, which moved too fast for Mimosa to see
• Cuts a flurry of Langris' projectiles
• Blocks Raia's beams as they appear from portals
• [Black] Cuts Rufel's lightning down
• Rebounds one of Drowa's beams with his eyes closed
• [Limit] Valtos' attacks from many directions are able to barely break through his defense

Movement (Other Stationary)

• Barely evades Jimmy's curse bolt, and even then, he's still tagged
• Avoids Gauche's beam

Reaction

• Can only partially block Valtos' attack when using his ears to predict the attack path, however, after sensing the ki of the attack, he is able to reflect it
• [Black] Follows Rufel
• [Limit] Cannot see Licht's spears even knowing their path of travel
• [Limit] Tracked Ladros' magic bullet with ki, but couldn't dodge it

Travel

• Moves so fast that the captains wonder if he's using magic to boost his speed
• Gains on a bandit's wind cloud
• Blitzes Valtos through his own portal
• [Black] Rockets up the side of a volcano, one-shotting a small magma monster in his path, but overshoots and almost lands in the volcano's mouth
• [Black] Grazes Lira
• [Black] The immense dashes he makes in this form home in on strong magic, and this can be used against him

Other

Ki

• Finds Valtos hiding in a pile of corpses
  • He did this using ki
• Ki can also sense natural objects and transparency magic users
• Senses Finral's ki as he's stepping out of a portal
• Senses the falling rocks of a volcano
• Hones in on the ki flowing from his sword into his body and uses it to trigger Black form
• Felt Zora's ki shift to disappointment
• Tracks Rufel's ki
• Senses Drowa's oncoming beam
• Has a feeling he's seen Secre before
• Senses Lolopechka behind a waterfall curtain
• [Limit] Magna's disappearing fireballs are too fast for him to sense

Other

• Can feed himself with his feet
• When in Black form, his anti-magic homes in on mana
• His swords are the remnants of the swords once used by the King of the Elves
• Used a vial of the Witch Queen's blood magic to heal Kahono and Kiato's injuries
• Armbars Rades
• Considered to be Arcane level on the Heart Kingdom's magic scale

Post-Timeskip

All life exists to entertain me, and yet…this is no longer fun. What is this? What are you?!! What is this emotion?!!
You don't have to know. You never needed to learn. The highest devils in the first level of the underworld. No one was a threat to you…so of course you can't understand…
It's called FEAR!!!

KEY
Chapter 229 marks a six month timeskip where he trained with the Stage Zero guards and increased his strength. After Dante impales Gauche, Asta unlocks Host form. He increases his strength again after making a Devil's Bargain in his fight with Dante, which lasts 50 seconds. His grimoire gains Yami's katana after defeating Dante. Nacht binds him to his devil, Liebe, allowing them greater power transfer, allowing them 5 minutes of Union form, with a 30 minute recharge time in which Asta can't use any form, not even Black. When Asta and Liebe completely synchronize, they completely shut off the magic of anything that he comes into contact with.
Base < [Black] (unmarked) < [Host] < [Union] < [Union]

Strength / Anti-Magic

Demon-Slayer Sword
Base

• Kills Lucifero

Negating Magic
Devils

• [Union] Cuts Nahamah's arm off
• [Union] Splits Nahamah's flames
• [Union] Whirlwinds to stave off Nahamah's flames and Lilith's ice attacks
• [Union] Cuts Nahamah in half
• [Union] Cuts a path through ice and fire
• [Union] Lucifero cuts his finger on the edge of the blade
• [Union] Severs the tip of Lucifero's horn

Other

• [Host] Obliterates Dante's meteors
• [Host] Forms a whirlwind to repel Dante's three megaswords
• Busts a boulder
• Cuts off Dante's arm
• Slashes through Dante's sword
• Cuts out of Nacht's wrapping
• [Union] Erases a mountain-sized demon's magic ball
• [Union] Cleaves through a colossal demon's forearm long-ways
• [Union] Cuts off the demon's arm, then kills it cutting through its torso
• [Union] Smashes a bit of ground

Repelling Magic

• Deflects Dante's punch
• Blocks a hit from one of Nacht's pack
• [Union] Repels a city-sized magic attack

Other

• [Host] Split the ground trying to kill Dante
• [Host] Extends his blade to try and hit Dante, but only manages to pulverize some ground

Black Divider

• Cuts through a wicker man armor
• Repels three mages' earth, fire, and water spells
• Smashes through a door and the wall surrounding it
• Cuts through an entire fortress
• Slashes a tree trunk
• [Host] Cuts Dante's cheek

Demon-Dweller Sword
Ranged

• Cut through a magic lance and a stone flooring
• Negates Dante's gravity magic with his ranged slashes
• [Union] Ranged slices cut through Lilith's city-sized magic

Other

• With Rill, destroys Charmy's forest-sized sheep
• He can call this to his hand
• [Union] Quadrisects Lilith while also wielding the Demon-Destroyer Sword

Demon-Destroyer Sword

• Expunges the poison in him
• Breaks an immobilization spell trapping his feet
• [Union] Cuts a tendril of fire
• [Union] Quadrisects Lilith while also wielding the Demon-Dweller Sword
• Weakens Megicula's curse on Lolopechka

Demon-Slasher Katana

• Defeats Dante
• Tags Nacht's cloak with a ranged slice
• [Union] Offpanel, performs a slash that cuts through the wall of the dungeon they're training in, dispels Nacht's Dark Prison, and undoes the union between him and his devils on his arm without even hitting him directly
• [Union] Cuts Lucifero's Qliphoth avatar, larger in size than the Giga-Black Bulls Hideout, and saves Yami and William from it
• [Union] Severs the magic Lucifero activated on Nacht's arm
• [Union] Stops Lucifero's punch
• [Union] Cuts off Lucifero's horns and back arms
• [Union] The one of his regular arms
• [Union] Then just cuts him up

Striking

• Knocks Liebe over with a punch
• [Union] Launches Lucifero with a kick

Durability (All Blunt)

VS Lucifero

• Takes the brunt of Lucifero sundering the countryside with gravity magic
• [Union] Punched out of the sky by Lucifero, and has trouble regaining cohesion
• [Union] Stomped by Lucifero into a crater
• [Union] His bones break when Lucifero kicks him
• [Union] No-sells Lucifero's punches
• [Union] Slammed by a rock lifted with Lucifero's gravity magic
• [Union] Tanks another punch from Lucifero

Others

• Hit with six gravity magic-influenced trees, but Rouge undoes the impact
• Gravitated toward the surface from the air
• Cratered again by Dante's gravity magic
• Kicked by Yami
• Smacked by Liebe wielding the Demon-Slayer Sword
• Pressed into stone floor by the effort of blocking Liebe's Demon-Slayer Sword
• Slammed by one of Nacht's pack
• Slapped by a pack member inside of Nacht's Dark Prison
• [Limit] Knocked out (of Host form) after Dante literally beats the hell out of him

Speed

Agility

• Dodges the oncoming Demon-Slayer Sword
• Utilizing his ki, he is able to weave through Liebe's swords and land a hit
• Blocks a hit from one of Nacht's pack after it moved behind him

Dexterity

• Keeps up with Liebe's three telekinetically-moving swords, while limited to one arm
• [Union] Cuts through Lucifero, then cuts him again before the pieces of his body can separate from the first cut

Movement

• [Union] Thrashes Lucifero in five seconds, culminating in an absolute victory

Travel

• [Host] Knew where Dante would appear after he teleported away from Asta's ranged slice, and took him by surprise
• [Union] Shot from the plains of the Spade Kingdom to the inner citadel of its capital
• [Union] Flies away from Nahamah and Lilith's attacks
• [Union] Hounds Lucifero even within his area of crushing gravity influence
• [Limit] [Union] It would "take forever" to go from the Clover Kingdom to the Spade Kingdom (across the continent)
• [Limit] [Union] Even starting before her, has trouble catching up to Lilith when she attempts to flee

Other

• Surfs on his Demon-Slayer Sword to an area of concentrated magic, and he can now return it to his hand at a call
• Senses the ki of Nacht's devil

Post-Post-Timeskip

If you lose somebody, get stronger than that guy. Then never lose to them again. Keep on getting stronger. That's the kind of Wizard King you want to be. Be determined to become…the strongest.

KEY
Chapter 332 marks 1.25 years after the Spade Kingdom War. Asta now can use a Union for 10 minutes when concentrating it into one part of himself. While fighting Yosuga, he's being passively buffed to take less damage and keep full strength. He trains with the Seven Dragons three days before Judgment Day.
Base < [Union] (unmarked)

Strength / Anti-Magic

Demon-Destroyer Sword

• Breaks Lucius' timestop
• Cuts Lucius' time bubbles
• Quenches Ichika's black stars

Demon-Slasher Katana

• Cuts Lucius' time bubbles
• Clashes with Ichika
• Clashes with Yosuga
• Stops the Yami patriarch's slash
• Cuts through Lily's water fist
• Splits Lily's beam

Zetten

• Rots a bamboo shoot by using his anti-magic to affect the life energy in it
• Kills some more bamboo
• Erases the magic inside Yrul
• Cleanses Lily of Lucius' soul remodeling, though to awake her from hibernation requires defeating Lucius
• Carves through the five-headed dragon
• Cuts off all five of the dragon's heads, and breaks the clouds on the horizon

Durability

Blunt

• Lucius karate-chops through his chest
• Sent flying by Ichika's punch, but lands gracefully
• Punted by Ichika, then hurled into a river, blocks her slash that mows down a bamboo forest, and finally takes a kick to the gut that ends the fight
• Thrashed by three other Zetten users
• Kicked by Ichika (serious)
• Slammed into a wall by Yosuga

Esoteric

• Only physical attacks work on him
  • Except when he's teleported to a separate continent while immobilized
• Only a few people could stand up to Ichika's forest cut without blacking out
• Wasn't immediately killed by Yosuga's slash that cut a stone staircase, but did have his shoulder cut
• Conquered self-doubt
• Sacred Devil Control is worthless against him

Speed

• Transforms within Lucius' stopped time
• Clashes with Ichika
• Scrambles away from Yosuga's slash
• Meets Yrul's charge
• Evades Lily's spatial magic
  • Does it some more
• Evades Lily's spatial x water magic lasers

This is a new writeup for Middlemak (renamed since my previous post). Named for the emphasis on the middle finger and use of the middle-finger-upper-row locations for frequent letters
Link to Layout heatmap

QWLDG JFOU: ASRTP YNEIH ZXCVB KM,./ 

Link to Middlemak GitHub download.
Link to GitHub download Middlemak-AH.
Middlemak has also been created. The same writeup has been posted there.
The main design criteria was:
1) Use the middle-finger-upper-row position. The middle finger is a very strong finger, and the middle-finger-upper-row is a very good location for a common letter.
2) To reduce pinballing by putting most vowels on the right hand and putting most of the frequent consonants on the left hand.
3) Reduce the number of Near Finger Bigrams (NFBs).
4) Improve hand balance.
5) Keep SFB down of course.
6) A secondary design criteria was low ring-to-pinky-rolls. I find this to be an awkward movement.
7) Maintain Qwerty similarity to make it easier to learn.
These are all discussed in more detail below. (I’ll be talking in the context of Colemak for some of it. I don’t mean this to be a hit piece on Colemak, it’s just the best way to frame the items.)

1) Middle-Finger-Upper-Row

The main purpose of this layout is to place 2 common letters on the middle-finger-upper-row. The middle finger is very strong and these are excellent locations that are easy to reach.
When you look at the letter frequency, there are 11 common letters before it takes a big drop off. The first eight letters go on the home row under the fingers. The next question is where do the next 3 letters go. Middlemak’s philosophy is to place 2 on the middle-finger-upper-row, and 1 on the index-finger-upper-row. Colemak places 2 on the centre columns, requiring a lateral movement that many find uncomfortable. Workman has 1 middle-finger-upper-row, 1 ring-finger-upper-row, and 1 index-finger-lower-row, a mixed bag.
L pairs well with R for low SFB. Middlemak places the RL column on the middle finger, which puts the common L on the middle-finger-upper-row.
E and O pair very well for low SFB. Placing EO column on the middle finger puts the very common O on the middle-finger-upper-row.
Using the middle finger also unloads the index fingers. The index fingers can get very busy with 6-7 letters compared with the other letters 3, especially when multiple frequent letters are put on them. I’ll give the distance stats in “Excessive amounts of stats” section far below which shows the drastic difference.
(I’ll skip discussing D here, it’s in “Excessive amounts of detail”. I also cover the R position more there.)

2) Pinballing

I wrote about pinballing in length here. The short of it is that there is an incredibly strong interaction between vowels and the consonants, with 75% of all bigrams between vowels and consonants. PInballing can happen when a lot of vowels and consonants arn put on the same hand, the hand movement will pinball back and forth between the vowels and consonants.
I’ll cover Colemak first to outline the issue. Having most of the vowels EOIUY and many frequent consonants of NHLMKJ on the right hand leads to a pinballing problem. I call this the NHLMKJ wall, it’s a wall of consonants that the vowels pinball off of
To reduce pinballing Middlemak does two things:
First Middlemak moves L off the vowel hand and to the consonant hand. This (and moving H) reduces the wall to NMFPJ which is much, much lower frequency for movements to pinball off of. Or if the word starts with NMFPJ, it’s again much lower frequency.
Second it moves H to the pinky. H is a very unique letter because it’s almost always followed by a vowel. Moving H to the pinky means it almost always gives an inside roll. After the vowel, it generally continues to the consonants NMFPJ continuing the inside roll, or it alternates to the other hand. Rarely goes back to H. Generally only after the third letter can it start to pinball back to the vowels or back to H.
Combined, this reduces pinballing and increases alternating. It’s not entirely alternating, Middlemak still has rolls because it has N, H, and MKFJ. For N I find the rolls to be comfortable, likely because N is on the home row.

3) Near Finger Bigrams, NFB

NFB are bigrams that are on adjacent fingers. Too many or an uncomfortable arrangement can be an issue. NFB and pinballing are separate issues: NFB is the smaller pattern of adjacent fingers, pinballing is the much larger pattern of pinballing between all vowels and constants.
I’ll cover Colemak first again to highlight the issue. On Colemak there are extremely common NFBs with HE and LE/EL. (There is also NE/EN, but I find the HE and EL/LE to be awkward because L and H are not on the home row.)
Pretty much all E bigrams are fairly high. I think you want to move common consonants away from E to reduce NFBs. For Middlemak, that means moving them away from the EO column.
Middlemak drastically reduces the NFBs by moving the common letters L and H off the adjacent index finger. The replacements of F and Y have dramatically lower NFBs with E and O. I wanted to keep Colemak’s K there since this doesn’t have many bigrams with E and O.
Looking at other layouts: Workman still has high NFB with L and E. Colemak-DH maintains L and H next to E. Many find the new HE roll to be more comfortable but it’s still a ton of NFB, and LE/EL is still there. MTGAP moved practically all letters away from its E and O and instead put punctuation there, a very good solution.
Middlemak does create some NFB with O placement. Namely OF, ON, and OM. However OF and ON are very comfortable, having the middle finger extended up feels pretty comfortable. OM is ok, and it’s the least common of those 3. The OM scissor-gram is a bit more common than I'd like, but it feels ok. This is about the best location for a scissor-gram too, with the long-middle-finger up and the curl-index-finger down being able to handle it fairly well. (Just to mention Colemak’s EL/LE, not exactly a scissor-gram, I find is bad since the shorter index finger has to reach up and the middle finger is so long it’s already curled just to be on the home row.)
Overall, Middlemak moving common consonants away from E drastically lowers the number of NFBs and takes away a ton of awkward movements. Visually, I think you can see this pretty easily in the heatmaps.
(You can extend this concept to fingers further away, to include Colemaks HI, LI, etc. This can be valid, but for here I’ll focus on adjacent fingers.)

4) Hand balance

Hand balance is a difficult topic. The two most obvious metrics for hand dominance are 1) the frequency of the letters, and 2) the distance the fingers have to travel (which is different from frequency because you don't travel to the home-row). But any pinballing from having too many vowels and consonants on the same hand doesn’t show up in those numbers. Having said all that, I’m going to look at frequency and distance.

Keyboard	Vowel Hand	Dominant Hand	Hand Balance ratio based on Frequency	Hand Balance ratio based on Distance	Hand Balance ratio based on Frequency and Distance	Corrected comparison number for left hand dominant
MTGAP	Left	Right	1.00	1.43	1.43
Workman	Right	Left	0.97	0.82	0.80	1.26
Norman	n/a	Right	0.93	1.27	1.19
Colemak	Right	Right	1.14	1.15	1.32
Middlemak	Right	Right	1.05	1.06	1.11
Semimak	Right	Right	1.15	1.30	1.49
Dvorak	Left	Right	1.23	1.86	2.30
QWERTY	n/a	Left	0.77	1.06	0.82	1.22

Notes: Ratio >1 indicates right hand dominant, <1 indicates left hand dominant. The hand balance ratios are based on each hand's frequency/distance, right hand divided by left hand. E.g. Middlemak: 51.2% right hand frequency divided by 48.8% left hand frequency equals 1.05. Colemak: 53.2% right hand frequency divided by 46.7% left hand frequency equals 1.14. Hand Balance ratio based on Frequency and Distance is a simple multiplication of the two ratios. I did this because having more frequency and more movement on the same hand amplifies the hand dominance. Or if one hand has more frequency and the other hand has more movement, that would mitigate it to some degree. This is by no means the most definitive way to look at it, but it's what I've managed to come up with. The "Corrected comparison number for left hand dominant" is a hard math thing to explain (that I hope I'm right about anyway). E.g. for Workman listing 0.80 gives the wrong impression the hand dominance is equivalent to 1.2 (20%). But it's actually 1.26 (26%), obtained from 1/0.796=1.26.
Middlemak is more balanced on both frequency and distance. When both are considered, it really adds up to being more balanced. Middlemak combined balance ratio is 1.11, compared to Colemak’s 1.32 or Workman’s 1.26.
Ratios can change quite quickly because both the numerator and the denominator (in this case, the left and right hand) can go in opposite directions. Moving L to the other hand does exactly this - changing both the numerator and denominator, and it changes both the frequency and the distance. So that one move really helps to balance out the hands.
(I’ll add the full data in “Excessive amounts of stats” for both frequency and distance, broken down to hands and fingers. The individual finger load is quite interesting too.)

5) SFB

The SFB of this layout is still pretty low, but it does come out a tad higher than Colemak’s. I’ve debated how to show this and decided to go with the full nuclear information. Data is from Mayzner revisited in millions (which only has letter bigrams and doesn’t have punctuation data, that’s what the 0’s are).

Keyboard	Total SFB	Left hand SFB	Right hand SFB	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky
QWERTY	185,270	125,920	59,350	1,105	1,661	54,502	68,651	34,166	3,970	21,214	0
Workman	78,147	29,268	48,878	1,105	3,712	2,806	21,645	27,338	5,037	16,503	0
Dvorak	70,570	28,306	42,264	0	31	0	28,275	12,142	16,265	8,229	5,629
Middlemak	44,698	25,806	18,892	1,105	1,661	2,716	20,323	12,415	3,134	3,343	0
Colemak	39,023	23,336	15,687	1,105	1,267	639	20,323	9,831	5,037	819	0
MTGAP	34,151	14,550	19,601	1,147	339	3,134	9,931	4,629	8,633	5,071	1,267

Notes: This is with index finger pressing C location, how I think most people type. Adding in punctuation the numbers would be higher, I expect they would increase by similar absolute amounts.
For Middlemak: You can see it’s not all that much of an increase in SFB from Colemak. Overall the numbers are low considering it keeps many Qwerty aspects.
On the left hand: The middle finger goes up with R and L. This was to allow more consonants on the left hand and use of the middle-finger-upper-row. I think the middle finger being a strong finger and comfortable reach can handle these quite well. The index finger is the same as Colemak.
On the right hand: The right index goes up a little bit, mostly with NY and MY. They aren’t as high as you may think. But the Y SFBs and adding in NK/KN, NF, etc, they do add up a little. It’s not all that much but it is more. I think it’s still easily manageable. NY is actually not all that common, for example it’s actually less common than EU/UE. (The exact Mayzner numbers are: NY/YN is 2,761 / 373. Compared to EU/UE is 878 / 4,158. MY/YM is 1,753 / 668.) The middle finger actually goes down, EO/OE is very low. The ring finger goes up a bit with UI/IU being more common than Colemak’s IY/YI.
If you want overall percentages (index finger pressing C), Middlemak comes in at 2.078%, vs Colemak’s 1.815%, vs Workmans 3.053%, vs Qwerty’s 6.264%.
Colemak really did a good job finding low SFB pairings of SF, DT, NHL, RW, and IY. So changing 3 of those pairings does come at a cost. But it’s really not much and it’s done so that Middlemak can put more frequent letters to better spots, reduce pinballing, and reduce NFB. I think the SFB is in the same realm as Colemak, it’s still pretty low.

Ortho SFB

Keyboard	Total SFB	Left hand SFB	Right hand SFB	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky
QWERTY	195,687	136,337	59,350	1,105	1,661	86,462	47,108	34,166	3,970	21,214	0
Workman	79,561	30,682	48,878	1,105	3,712	8,206	17,659	27,338	5,037	16,503	0
Dvorak	70,365	28,101	42,264	0	31	1,591	26,480	12,142	16,265	8,229	5,629
Middlemak	42,816	23,924	18,892	1,105	1,661	14,888	6,270	12,415	3,134	3,343	0
Colemak	30,032	14,345	15,687	1,105	1,267	5,702	6,270	9,831	5,037	819	0
MTGAP	34,151	14,550	19,601	1,147	339	3,134	9,931	4,629	8,633	5,071	1,267

Notes: This is with the middle finger pressing C location, “proper” or ortho style. Same data as above.
Middlemak on Ortho actually comes in just a tad lower, but Colemak takes a bigger dip.
This basically moves most of the SFB from the index finger to the middle finger. On Middlemak, ortho takes out CT on the index finger and puts in CR and CL on the middle finger, which is only a tad lower. On Colemak, ortho takes out CT and puts in SC which is much lower.
I think it still works well. The middle finger is a strong finger and can cover most of these well. The one issue is that CL which is a two row jump.
I’ll break down the full Mayzner numbers:
First the index finger pressing CT/TC is 12,998 / 737. This is what ortho removes.
Middlemak Ortho puts in CRC which is 4,214 / 3,423. And puts in CL / LC which is 4,202 / 333. In percentages, CR is 32% of CT, RC is 26% of CT, and CL is 32% of CT. Adding it all up and including all the others (CP, CG, etc), Middlemak Ortho comes in at 87% of CT+TC. Just a tad lower.
Colemak Ortho puts in CS/SC, which is 644 / 4,363. CF/ FC is pretty much nill. In percentages, SC is 34% of CT. Adding it all up and including the others, Colemak Ortho comes in at 36% of CT+TC. A big dip.
If you want overall percentages (middle finger pressing C, Ortho style), Middlemak comes in at 2.16%, vs Colemak’s 1.669%, vs Workman’s 3.147%, vs Qwerty’s 6.575%. This is from Colemak-DH analyzer which says Middlemak ortho increases, as opposed to what I found that Middlemak ortho decreases SFB. I think it comes down to the data set used.

6) Low ring-to-pinky rolls. And the amazing inside roll from H.

A secondary design goal was to decrease the awkward ring-to-pinky roll. Middlemak greatly reduces this roll.
On the left hand: Middlemak’s ring-to-pinky outside roll of SA and WA is actually pretty low. SA is actually the lowest bigram of the common consonants to A. So SA and WA comes out to be a good design. Compared to Colemak, Middlemak’s SA outside roll is far less common than Colemak’s RA. (The exact Mayzner in millions are: Middlemak’s SA is 6,147. Colemak’s RA is 19,333. WA is the same for both at 10,865.)
On the right hand: Middlmak’s ring-to-pinky roll of IH and UH is very low. Vs Colemak: IO is very common and even YO is pretty common. (The exact numbers are Middlemak’s IH is 59, UH is 30. Vs Colemak’s IO 23,542, YO is 4,227.)
This next bit isn’t even a secondary design goal, but it’s worth mentioning: On the right hand, outside rolls from any of the fingers to the pinky H are very low. The common letters that come before H are usually TH, CH, WH, SH, and GH. All of those are on the opposite left hand. Doing the math, 97% of the letters before H are on the left hand. H on the pinky gives an amazing inside roll almost all the time.
Also worth mentioning is I think this also de-emphasizes the pinky a bit. First it limits the ring-to-pinky rolls. Second, the frequency of H is much lower than Colemak’s O. Third, the bigram HH is virtually nill, vs Colemak’s OO which is pretty common. None of these are massive, but I think they add up to de-emphasizing the pinky a bit.

7) Qwerty similarity.

Last but not least, this maintains Qwerty similarity to make it easier to learn.
First, this maintains the bottom row (except N) and maintains Q, W, A, and S. Many letters move on the same finger, including T, G, N, Y, J .
Second, this “keeps” the ASRT order. People learning Colemak commonly say that R and S are reversed. Technically it’s not because R was never on the middle finger to begin with. But it seems to be reversed in the brain. I think it’s two things. First ASRT keeps it in the same order, with R between S and T, even if R changes fingers. Second, I wonder how many people alt-fingered R typing in Qwerty, either all the time or for the RT bigram.
(I've also wondered if, for example, patterns like the very common ION simply being easier to type would make this faster to learn. I’m not certain about that though.)
Comparing layouts:

• Middlemak: 10 letters stay in their original spot and 16 move, with 5 letters moving on the same finger and 11 letters switching fingers.
  
• Colemak: 10 letters stay in their original spot and 16 move, with 4 letters moving on the same finger and 12 letters switching fingers.
  
• Colemak-DH: 5 letters stay in their original spot and 21 move, with 6 letters moving on the same finger and 15 letters switching fingers.
  
• Workman: 6 letters stay in their original spot and 20 move, with 8 letters moving on the same finger and 12 letters switching fingers.
  

*Not accounting for punctuation changes.
Considering letters swapping on the same finger, this has fewer changes than Colemak. And I think “keeping” the ASRT order will make it even easier to adopt than the numbers suggest.
(I hope people don’t think that I’m overly focused on keeping S in the Qwerty location just because. For this layout, it does work better with ASRT, I address it in more detail in excessive amount of details.)

Downsides

1) There is a slight increase in SFB compared to Colemak. But overall I think SFB is still low.
2) There are a handful of NFB with R: BR, GR, PR, etc. Same with L and BL. This isn’t enough to change the home row to ARST. There is more below but ASRT has lower SFB, lower ring-to-pinky rolls, and is easier to learn. Also note these are much, much lower than the extremely high NFB that Colemak has with HE, LE, etc, (to solve those I paid L and R together).
3) I wanted a less common letter in the G location (Qwerty T location). I really tried to put a less common letter there, but it simply didn’t work.
4) I mention this above but to add more: The OM scissor-gram is a bit more common than I'd like. This is a limitation of keeping Qwerty similarity (the M) and wanting to put O on the upper-row-middle-finger. (MTGAP has a good design on this. It paired its EO with a lot of punctuation on the index finger to reduce NFBs.)

Conclusion

I'm really happy with this layout. I think it is a great combination of pretty low SFB, good use of the strong middle-finger-upper-row position, reduced NFB and pinballing, keeping Qwerty similarity, and easy to transition to.
Honestly I’m surprised no one else has stumbled on this arrangement, since many people have taken to trying to fix Colemak’s issues and keep Qwerty similarity.

Some nice rolls and combinations

Although this increases alternating, there are still plenty of rolls left. Some I want to mention:
HE is very easy.
LD feels very good.
OU feels very comfortable.
OF/FO is very common and feels very good.
IO and ION are very nice rolls.
This maintains the nice IEN roll.
IOU is a redirect, but overall feels pretty ok.
YOU has a lateral move, but feels comfortable.
OU+LD together to make OULD feels good.
The very common TH, THE, and THE_ feel very nice not having to move the fingers off the home row.

Vs Colemak and Vs Colemak-DH

I made Middlemak to try to solve the problems I found with Colemak. I think all the items came out above, but as a bit of a summary and to include Colemak-DH.
Colemak:
1) Colemak has an issue with high lateral movement with D and H. You see this especially with H and its NFBs, but D isn’t the best either.
2) Colemak’s right index finger is overworked. It just has too much to do with N, H, and L.
3) Colemak has a pinballing issue due to placing most vowels and many common consonants on the same hand.
4) Colemak has a ton of NFB with EL/LE and HE.
5) Colemak has a hand balance issue. It places a lot of emphasis on the right hand with the vowels and several common consonants. (Hand dominance and pinballing are technically separate items.)
6) Colemak’s awkward ring-to-pinky rolls. Notably IO, YO (especially when typing YOU), and also RA on the right hand.
7) Not exactly a direct issue with Colemak, but Middlemak is easier to learn coming from Qwerty. Many new users want to ‘keep’ the order of ASRT instead of ‘reversing’ S and R.
Middlemak is a pretty low increase in SFBs and it fixes those issues.
Colemak DH:
Colemak-DH solves 1) the lateral movements. But it doesn’t solve 2) the right index finger overwork, 3) the pinballing, 4) the NFBs 5) the hand balance, 6) the awkward ring-to-pinky rolls, and 7) coming from Qwerty it’s pretty much a full change layout, taking out ease of transition. This also means common shortcuts move. They are still accessible, but they move (paste is the same. Undo, cut, and copy move.)

Option: Swap A and H

This gives:

QWLDG JFOU: HSRTP YNEIA ZXCVB KM,./ 

The good: This is good if you want to put all the vowels and most consonants on opposite hands. Overall this increases alternating. This mostly comes from moving A to the right hand because A pairs a lot with the L, R, S, T, etc. This is the biggest appeal if you find those uncomfortable. Moving H actually decreases alternating because there’s a high bigram with TH (you can test your own comfortability on that. Don’t forget capitalized Th is common.) Excessive details: Middlemak A has 41% of bigrams on the opposite hand, swapping gives 65%. Middlemak H has 58% of bigrams on the opposite hand, swapping gives 51% (again, mostly because of TH). Overall, middlemak A+H has 48% on the opposite hand, swapping gives 59%.
The neutral: On the left hand, the ring-to-pinky outside roll of SH is slightly higher, but in a similar range. Data from Mayzner is SH is 8,889, and original is SA is 6,147. Changing WA to WH is practically identical.
The bad: On both the right and left hand, the all-fingers-to-pinky outside roll is higher. On the left hand, it’s because TH is very common. On the right hand, it takes away the remarkable inside roll of H and adds in a moderate number of all-fingers-to-A bigrams with EA, IA, NA, MA, etc, (OA is actually low). Notably, the ring-to-pinky outside roll is higher with IA and UA coming in at 8,072 and 3,844, vs original IH and UH of 59 and 30.
The complicated: This option decreases the pinky-to-all-fingers rolls inside rolls on both hands. It increases the all-fingers-to-pinky rolls outside rolls. The overall sum of inside rolls plus outside rolls is lower. In short: this option lowers inside rolls, increases outside rolls, with an overall sum that is lower.
Basically stick to the original if you want to lower ring-to-pinky movement, lower outside rolls, and easier to learn. Overall I put this as a decent option if someone wants to maximize alternating without going to a full change layout.

Excessive amount of stats

This section has details on distance reductions, hand balance based on frequency, hand balance based on distance, and raw distance numbers. I'm going to give lots of charts to cover all the details. Those that want to know all the details or those that like to analyze will find this interesting, but others don’t have to read.

Distance reduction from Qwerty

Keyboard	Distance	Distance	Distance
	(km)	(miles)	(% reduction from Qwerty)
MTGAP	2.592	1.611	46.3%
Workman	2.613	1.624	45.9%
Norman	2.615	1.625	45.9%
Colemak	2.616	1.625	45.8%
Middlemak	2.710	1.684	43.9%
Semimak	2.739	1.702	43.3%
Dvorak	2.813	1.748	41.8%
QWERTY	4.829	3.001	0%

*Source: 62 page / 31k word compilation. Left index used to press Qwerty C location. Includes punctuation.
Middlemak’s distance reduction is good, most alt keyboards are in the same range. It’s a tad worse on distance than some others because putting O off the home row increases distance, with the O being more frequent. (MTGAP puts O off the home row, but has all the punctuation on the diagonals to decrease distance.)

All Finger Distances

This is the raw distance travelled in metres. Percentages rely on the total distance travelled, which changes between layouts so it's not comparable. Note that distance is different from frequency. The fingers don't move for the home row, the distance travelled is to the other letters.

Keyboard	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky	L Total	R Total	Total
MTGAP	293	116	413	246	667	320	219	318	1068 m	1524 m	2592 m
Workman	123	232	330	753	505	200	146	325	1437 m	1176 m	2613 m
Norman	127	127	223	674	525	377	244	319	1150 m	1465 m	2615 m
Colemak	130	127	116	842	747	200	141	315	1214 m	1402 m	2616 m
Middlemak	134	127	207	847	440	460	189	306	1314 m	1396 m	2710 m
Semimak	238	237	326	390	506	200	450	393	1191 m	1549 m	2739 m
Dvorak	205	54	34	692	753	285	395	396	984 m	1829 m	2813 m
QWERTY	130	127	670	1413	1169	452	450	418	2340 m	2489 m	4829 m

*All distances in metres. Same source as table above.
First, you can see Qwerty’s distances are extremely high. All the alternatives make big reductions.
Notice that the index fingers move way more than the other fingers.
Middlemak balances this out decently and moves more distance to both middle fingers. The left index finger is still a bit high because it has D off the home row (this brings it back to the discussion of where does the 11th letter go). Middlemak’s left index is slightly higher than Colemak’s even though they have the same letters because the row stagger technically puts D on a slight diagonal. A technicality really.
The right index finger is low, a good feature that the index finger next to the vowels doesn’t have to travel much. Compared to Colemak, Middlemak’s movement is 41% lower than Colemaks.

Zoom in on index fingers

Keyboard	Distance	Distance	Total	Index Finger % of total distance
	L Index	R Index	L+R Index	L+R Index
MTGAP	246	667	913 m	35%
Workman	753	505	1259 m	48%
Norman	674	525	1200 m	46%
Colemak	842	747	1589 m	61%
Middlemak	847	440	1287 m	47%
Semimak	390	506	896 m	33%
Dvorak	692	753	1445 m	51%
QWERTY	1413	1169	2582 m	53%

Here you can really see how much work the index fingers do. Middlemak’s sum of the index fingers is nice and low.
You can also look at the percentage of travel with the index finger. Middlemak’s is down at 47% - meaning the index fingers are responsible for 47% of all the distance travelled. On Colemak the index fingers account for 61% the distance travelled, quite a big difference relatively. Workman and Norman are close to Middlemak. MTGAP and Semimak are very low. Dvorak and Qwerty percentages are a bit deceiving though - the higher index finger distances are hidden by the higher total amount of distance.
So even when Middlemak’s left index finger doesn’t have a reduction in distance, the total index finger distance is significantly reduced.

Hand Balance based on Distance

This is the distance in percentages for all the fingers (unlike the raw distances in metres above) so that I can add them up to look at hand balance.

Keyboard	Left hand	Right hand	Hand Balance Ratio	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky
MTGAP	41.2*	58.8	1.43	11.3	4.5	15.9	9.5	25.7	12.3	8.4	12.3
Workman	55.0	45.0*	0.82	4.7	8.9	12.6	28.8	19.3	7.6	5.6	12.5
Norman	44.0	56.0	1.27	4.8	4.8	8.5	25.8	20.1	14.4	9.3	12.2
Colemak	46.4	53.6*	1.15	5.0	4.8	4.4	32.2	28.6	7.6	5.4	12.0
Middlemak	48.5	51.5*	1.06	5.0	4.7	7.6	31.2	16.3	17.0	7.0	11.3
Semimak	43.5	56.5*	1.30	8.7	8.7	11.9	14.2	18.5	7.3	16.4	14.4
Dvorak	35.0*	65.0	1.86	7.3	1.9	1.2	24.6	26.8	10.1	14.0	14.1
QWERTY	48.5	51.5	1.06	2.7	2.6	13.9	29.3	24.2	9.4	9.3	8.7

Notes: * indicates vowel hand (Norman and Qwerty are mixed enough that there is no clear vowel hand). The ratio between the hands can change quite quickly, so I think that highlights the differences between the layouts better. Ratio >1 indicates right hand dominant, <1 indicates left hand dominant.
I think the raw distances are better for comparison between layouts, but some people like percentages. I used this mostly to calculate hand balance.
You can see Middlemak spreads out the work. The pinkies are low, ring fingers are reasonable, middle takes on more, and the indexes have most of the distances.

Hand Balance based on Frequency

Now we’re shifting to frequency. You can see the data for all the fingers, this is enlightening on its own. I also add them up to look at the hand balance.

Keyboard	Left hand	Right hand	Hand Balance Ratio	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky
MTGAP	49.9*	50.0	1.00	10.8	8.4	18.8	11.9	16.3	14.5	9.8	9.4
Workman	50.7	49.3*	0.97	9.3	10.1	11.1	20.2	15.2	14.9	9.7	9.5
Norman	51.7	48.3	0.93	9.3	8.3	15.4	18.7	15.2	13.2	11.8	8.1
Colemak	46.7	53.2*	1.14	9.4	7.7	8.2	21.4	19.4	14.9	9.1	9.8
Middlemak	48.8	51.2*	1.05	9.4	8.3	9.7	21.4	13.7	19.6	9.9	8.0
Semimak	46.6	53.4*	1.15	9.7	10.1	11.8	15.0	12.2	14.9	15.6	10.7
Dvorak	44.8*	55.3	1.23	10.3	8.1	12.2	14.2	17.2	13.4	13.0	11.7
QWERTY	56.5	43.5	0.77	9.4	8.3	15.4	23.4	18.8	8.4	11.8	4.5

* indicates vowel hand (Norman and Qwerty are mixed enough that there is no vowel hand). Ratio of >1 indicates right hand dominant, <1 indicates left hand dominant.
The ratio between the hands can change quite quickly, so I think that highlights the differences between the layouts better than the percentages. I know it doesn't add up to 100%, the analyzer is doing some rounding or something behind the scenes, I'm not going to round any further.
E is so common that it bumps up any finger it’s on.
Middlemak’s left middle finger takes on a bit more work. The left index is still high because of D. The right index is notably lower taking off common letters. The right middle takes on more of the work with O. And the right pinky is lower too with the H there.

Hand Balance Distance + Frequency

This is the same chart that was above, just for a summary.

Keyboard	Dominant Hand	Hand Balance ratio based on Frequency	Hand Balance ratio based on Distance	Hand Balance ratio based on Frequency and Distance	Corrected comparison number for left hand dominant
MTGAP	Right	1.00	1.43	1.43
Workman	Left	0.97	0.82	0.80	1.26
Norman	Right	0.93	1.27	1.19
Colemak	Right	1.14	1.15	1.32
Middlemak	Right	1.05	1.06	1.11
Semimak	Right	1.15	1.30	1.49
Dvorak	Right	1.23	1.86	2.30
QWERTY	Left	0.77	1.06	0.82	1.22

1 indicates right hand dominant, <1 indicates left hand dominant. Hand Balance ratio based on Frequency and Distance is a simple multiplication of the two.

Frequency-off-home-row (F.O.H.R.)

At the risk of too much information, another way to look at this is the frequency that’s off-home-row. I.e. the frequency that is not under the fingers.
Distances weigh the keys differently, while frequency-off-home-row levels that field. This should be read in conjunction with distance. Neither is better than the other, it’s just a different measurement.

Keyboard	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky	L+R Index	Index FOHR as % of All fingers FOHR
MTGAP	4.0	2.0	7.2	4.1	10.8	5.8	3.6	3.8	14.9%	36%
Workman	1.5	3.9	5.6	11.5	8.8	3.3	2.5	2.7	20.3%	51%
Norman	1.5	2.1	3.8	10.0	8.8	6.4	4.6	2.6	18.8%	47%
Colemak	1.6	2.1	2.0	12.7	13.0	3.3	2.3	2.6	25.7%	65%
Middlemak	1.6	2.1	4.1	12.7	7.3	8.0	3.1	2.5	20.0%	48%
Semimak	3.5	4.5	5.4	6.3	8.4	3.3	7.8	3.9	14.7%	34%
Dvorak	2.5	0.9	0.6	11.7	11.7	4.7	6.6	5.5	23.4%	53%
QWERTY	1.6	2.1	11.6	21.4	18.6	7.5	7.8	4.5	40.0%	53%

I did this after noticing, on Colemak, that my right index finger moves off the home row a lot. For Colemak the left index finger has more distance, but the right index finger has more frequency-off-home-row. The long distance to B on the left hand skews the distance travelled.
The effect is similar on most of the layouts: FOHR tones down the left hand dominance and bumps up the right hand dominance. With Colemak the frequency of H and L is enough to actually flip them. This is more noticeable with layouts that keep Qwerty’s B (which is just frequent enough to skew things), layouts that change the letter to a less common one shift less.
Also insightful is “Index FOHR as % of All fingers FOHR”. It shows that on Middlemak the index fingers do 48% of all frequency-off-home-row, quite reasonable. Whereas on Colemak the index fingers have 65% of all frequency-off-home-row, very high. Workman is just a tad higher with the index fingers doing 51% of frequency-off-home-row. Dvorak’s and Qwerty’s 53% for index fingers is a bit deceiving though, it appears low but that's because the overall total FOHR is much higher.
This is similar to “Zoom in on the index fingers”, but this shows an even greater difference between Middlemak and Colemak. This shows Middlemak at 48% and Colemak at 65% based on index finger % of total FOHR. Recall Middlemaks was 47% and Colemaks was 61% based on index finger % of total distance. Same thing, when the distances to the diagonals are levelled out, the frequency stands out.

Excessive amount of detail on letters

You really don’t need to read this, mostly if you have any lingering or specific Q on specific letters.

D position and why not curl?

Back to the discussion of there being 11 common letters: Where to place that 11th letter is difficult, it’s either the ring finger or index finger. Many find the ring finger to not be very dexterous for such a frequent letter. So putting D on the index finger with T works very well for low SFB.
Why not the middle row? First, I think most people find it easier to reach up and down rather than laterally. Second, it works very well with the LD roll.
Why not the bottom row? I want to leave the bottom row the same as Qwerty for easy adoption and to leave the shortcuts of undo=cut-copy-paste the same. I also think C and M (the curl locations) are already near perfect frequencies for those locations. Those are the middle of the pack frequencies, which are perfect for the index fingers - which cover 6 to 7 letters.
I previously said that I’m not a fan of Colemak’s L position, which is the same upper-row-index-finger. But Colemak’s L has a ton of NFBs with the vowels, especially with E but also with all the other vowels. Looking at the whole hand, 62% of Colemak’s L bigrams are on the same hand. I think that’s a big source of discomfort. But with Middlemak’s D, only 22% of bigrams are on the same hand. So Middlemak has far lower same-hand-gymnastics.

The R and L position

I’m sure the R and S location will attract attention, so I’ll discuss it more here.
If you want to skip this, the main factor is that R and L pair well together. Better than S and L. That means the RL column goes on the middle finger. R being more common goes on the home row, and L goes on the middle-finger-upper-row which is a strong location.
[First note that Colemak had to go with ARST because of F on the middle-finger-upper-row. The SFB of RF/FR is more common than SF/FS.]

ASRT, making a RL column

The pros of ASRT is it gives fewer SFB, with LRL being less than half as common as SL/LS. It also gives fewer ring-to-pinky-rolls with SA being far less common than Colemak’s RA. It also makes it much easier to learn from Qwerty.
The cons is that the R position leads to more NFB with the index finger. Most of these feel ok but RD, GR, BR are not so good. These aren’t super common, but they are there. PR gives the commonly disliked lateral NFB, but I think it’s relatively ok for the frequency - it’s not super common. RS creates an outside roll, but I’d rather have the RS on Middlemak’s middle-to-ring fingers than Colemak’s RA on the ring-to-pinky fingers. Colemak’s RA is much more common than Middlemak’s RS too.

ARST, making a SL column

The pros of ARST is that it gives an inside roll of RS. It also separates R from the NFB letters on the index finger, which do add up. However, I’m trying to resolve the NFBs of Colemak’s right hand H and L with E which is a magnitude larger, in addition to the pinballing that it creates which requires moving L.
The cons with ARST is that it gives more SFB. It’s not huge, but it is there. It also gives a lot of ring-to-pinky outside rolls with RA. And finally this makes it much harder to learn coming from Qwerty.

Verdict: ASRT, RL column

None of these are huge deal breakers for one or the other. I went with lower SFB, lower ring-to-pinky rolls, and much easier to learn. At the cost of some NFB on the index finger and inside roll of RS.
(If someone comes from Colemak and wants ARST, they can. It's not a humongous downside to do that. Mostly the more common RA ring-to-pinky roll and slightly more SFB of SD/DS.)
(This layout truly does work better with ASRT. If it didn’t work, I would have changed it. For example, I wanted to keep G on its Qwerty location but it’s better to move it, so I didn’t shy away from changes.)

P and G location

It’s better to move G than keep it in its Qwerty location.
P and G are pretty much the same frequency, so it doesn’t matter in that sense.
P has a SFB with T, and it has common bigrams of PR, SP, and even PA. Those work better with P on the middle row. There is PL, but the ones above outweigh it.
G has a bigram with NG. With N on the opposite hand, that means G can go on the worse diagonal spot. There is GR, but PR is far more common.

O location

O and E together give very low SFB. This is a great opportunity to put them together. The strong middle finger only has 3 keys, so it can handle two very frequent letters.

U, Y and F locations

U works pretty well with I, the SFB are reasonably low. This location also creates a nice OU and YOU roll.
Y goes to the index finger, which actually works pretty well. I broke down the numbers in SFB section.
F goes to the right hand index finger, because putting P or G on the right hand gives too many SFB with M and N. The F location works quite well for both moderate frequency and low SFB. It has a lot of NFB with OF/FO, so placing it next to O makes a comfortable roll.

This is a new writeup for Middlemak (Posted before, but there's a new name for those interested). Named for the emphasis on the middle finger and use of the middle-finger-upper-row locations for frequent letters
Link to Layout heatmap

QWLDG JFOU: ASRTP YNEIH ZXCVB KM,./ 

Link to Middlemak GitHub download.
Link to GitHub download Middlemak-AH.
Middlemak has also been created. The same writeup has been posted there.
The main design criteria was:
1) Use the middle-finger-upper-row position. The middle finger is a very strong finger, and the middle-finger-upper-row is a very good location for a common letter.
2) To reduce pinballing by putting most vowels on the right hand and putting most of the frequent consonants on the left hand.
3) Reduce the number of Near Finger Bigrams (NFBs).
4) Improve hand balance.
5) Keep SFB down of course.
6) A secondary design criteria was low ring-to-pinky-rolls. I find this to be an awkward movement.
7) Maintain Qwerty similarity to make it easier to learn.
These are all discussed in more detail below. (I’ll be talking in the context of Colemak for some of it. I don’t mean this to be a hit piece on Colemak, it’s just the best way to frame the items.)

1) Middle-Finger-Upper-Row

The main purpose of this layout is to place 2 common letters on the middle-finger-upper-row. The middle finger is very strong and these are excellent locations that are easy to reach.
When you look at the letter frequency, there are 11 common letters before it takes a big drop off. The first eight letters go on the home row under the fingers. The next question is where do the next 3 letters go. Middlemak’s philosophy is to place 2 on the middle-finger-upper-row, and 1 on the index-finger-upper-row. Colemak places 2 on the centre columns, requiring a lateral movement that many find uncomfortable. Workman has 1 middle-finger-upper-row, 1 ring-finger-upper-row, and 1 index-finger-lower-row, a mixed bag.
L pairs well with R for low SFB. Middlemak places the RL column on the middle finger, which puts the common L on the middle-finger-upper-row.
E and O pair very well for low SFB. Placing EO column on the middle finger puts the very common O on the middle-finger-upper-row.
Using the middle finger also unloads the index fingers. The index fingers can get very busy with 6-7 letters compared with the other letters 3, especially when multiple frequent letters are put on them. I’ll give the distance stats in “Excessive amounts of stats” section far below which shows the drastic difference.
(I’ll skip discussing D here, it’s in “Excessive amounts of detail”. I also cover the R position more there.)

2) Pinballing

I wrote about pinballing in length here. The short of it is that there is an incredibly strong interaction between vowels and the consonants, with 75% of all bigrams between vowels and consonants. PInballing can happen when a lot of vowels and consonants arn put on the same hand, the hand movement will pinball back and forth between the vowels and consonants.
I’ll cover Colemak first to outline the issue. Having most of the vowels EOIUY and many frequent consonants of NHLMKJ on the right hand leads to a pinballing problem. I call this the NHLMKJ wall, it’s a wall of consonants that the vowels pinball off of
To reduce pinballing Middlemak does two things:
First Middlemak moves L off the vowel hand and to the consonant hand. This (and moving H) reduces the wall to NMFPJ which is much, much lower frequency for movements to pinball off of. Or if the word starts with NMFPJ, it’s again much lower frequency.
Second it moves H to the pinky. H is a very unique letter because it’s almost always followed by a vowel. Moving H to the pinky means it almost always gives an inside roll. After the vowel, it generally continues to the consonants NMFPJ continuing the inside roll, or it alternates to the other hand. Rarely goes back to H. Generally only after the third letter can it start to pinball back to the vowels or back to H.
Combined, this reduces pinballing and increases alternating. It’s not entirely alternating, Middlemak still has rolls because it has N, H, and MKFJ. For N I find the rolls to be comfortable, likely because N is on the home row.

3) Near Finger Bigrams, NFB

NFB are bigrams that are on adjacent fingers. Too many or an uncomfortable arrangement can be an issue. NFB and pinballing are separate issues: NFB is the smaller pattern of adjacent fingers, pinballing is the much larger pattern of pinballing between all vowels and constants.
I’ll cover Colemak first again to highlight the issue. On Colemak there are extremely common NFBs with HE and LE/EL. (There is also NE/EN, but I find the HE and EL/LE to be awkward because L and H are not on the home row.)
Pretty much all E bigrams are fairly high. I think you want to move common consonants away from E to reduce NFBs. For Middlemak, that means moving them away from the EO column.
Middlemak drastically reduces the NFBs by moving the common letters L and H off the adjacent index finger. The replacements of F and Y have dramatically lower NFBs with E and O. I wanted to keep Colemak’s K there since this doesn’t have many bigrams with E and O.
Looking at other layouts: Workman still has high NFB with L and E. Colemak-DH maintains L and H next to E. Many find the new HE roll to be more comfortable but it’s still a ton of NFB, and LE/EL is still there. MTGAP moved practically all letters away from its E and O and instead put punctuation there, a very good solution.
Middlemak does create some NFB with O placement. Namely OF, ON, and OM. However OF and ON are very comfortable, having the middle finger extended up feels pretty comfortable. OM is ok, and it’s the least common of those 3. The OM scissor-gram is a bit more common than I'd like, but it feels ok. This is about the best location for a scissor-gram too, with the long-middle-finger up and the curl-index-finger down being able to handle it fairly well. (Just to mention Colemak’s EL/LE, not exactly a scissor-gram, I find is bad since the shorter index finger has to reach up and the middle finger is so long it’s already curled just to be on the home row.)
Overall, Middlemak moving common consonants away from E drastically lowers the number of NFBs and takes away a ton of awkward movements. Visually, I think you can see this pretty easily in the heatmaps.
(You can extend this concept to fingers further away, to include Colemaks HI, LI, etc. This can be valid, but for here I’ll focus on adjacent fingers.)

4) Hand balance

Hand balance is a difficult topic. The two most obvious metrics for hand dominance are 1) the frequency of the letters, and 2) the distance the fingers have to travel (which is different from frequency because you don't travel to the home-row). But any pinballing from having too many vowels and consonants on the same hand doesn’t show up in those numbers. Having said all that, I’m going to look at frequency and distance.

Keyboard	Vowel Hand	Dominant Hand	Hand Balance ratio based on Frequency	Hand Balance ratio based on Distance	Hand Balance ratio based on Frequency and Distance	Corrected comparison number for left hand dominant
MTGAP	Left	Right	1.00	1.43	1.43
Workman	Right	Left	0.97	0.82	0.80	1.26
Norman	n/a	Right	0.93	1.27	1.19
Colemak	Right	Right	1.14	1.15	1.32
Middlemak	Right	Right	1.05	1.06	1.11
Semimak	Right	Right	1.15	1.30	1.49
Dvorak	Left	Right	1.23	1.86	2.30
QWERTY	n/a	Left	0.77	1.06	0.82	1.22

Notes: Ratio >1 indicates right hand dominant, <1 indicates left hand dominant. The hand balance ratios are based on each hand's frequency/distance, right hand divided by left hand. E.g. Middlemak: 51.2% right hand frequency divided by 48.8% left hand frequency equals 1.05. Colemak: 53.2% right hand frequency divided by 46.7% left hand frequency equals 1.14. Hand Balance ratio based on Frequency and Distance is a simple multiplication of the two ratios. I did this because having more frequency and more movement on the same hand amplifies the hand dominance. Or if one hand has more frequency and the other hand has more movement, that would mitigate it to some degree. This is by no means the most definitive way to look at it, but it's what I've managed to come up with. The "Corrected comparison number for left hand dominant" is a hard math thing to explain (that I hope I'm right about anyway). E.g. for Workman listing 0.80 gives the wrong impression the hand dominance is equivalent to 1.2 (20%). But it's actually 1.26 (26%), obtained from 1/0.796=1.26.
Middlemak is more balanced on both frequency and distance. When both are considered, it really adds up to being more balanced. Middlemak combined balance ratio is 1.11, compared to Colemak’s 1.32 or Workman’s 1.26.
Ratios can change quite quickly because both the numerator and the denominator (in this case, the left and right hand) can go in opposite directions. Moving L to the other hand does exactly this - changing both the numerator and denominator, and it changes both the frequency and the distance. So that one move really helps to balance out the hands.
(I’ll add the full data in “Excessive amounts of stats” for both frequency and distance, broken down to hands and fingers. The individual finger load is quite interesting too.)

5) SFB

The SFB of this layout is still pretty low, but it does come out a tad higher than Colemak’s. I’ve debated how to show this and decided to go with the full nuclear information. Data is from Mayzner revisited in millions (which only has letter bigrams and doesn’t have punctuation data, that’s what the 0’s are).

Keyboard	Total SFB	Left hand SFB	Right hand SFB	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky
QWERTY	185,270	125,920	59,350	1,105	1,661	54,502	68,651	34,166	3,970	21,214	0
Workman	78,147	29,268	48,878	1,105	3,712	2,806	21,645	27,338	5,037	16,503	0
Dvorak	70,570	28,306	42,264	0	31	0	28,275	12,142	16,265	8,229	5,629
Middlemak	44,698	25,806	18,892	1,105	1,661	2,716	20,323	12,415	3,134	3,343	0
Colemak	39,023	23,336	15,687	1,105	1,267	639	20,323	9,831	5,037	819	0
MTGAP	34,151	14,550	19,601	1,147	339	3,134	9,931	4,629	8,633	5,071	1,267

Notes: This is with index finger pressing C location, how I think most people type. Adding in punctuation the numbers would be higher, I expect they would increase by similar absolute amounts.
For Middlemak: You can see it’s not all that much of an increase in SFB from Colemak. Overall the numbers are low considering it keeps many Qwerty aspects.
On the left hand: The middle finger goes up with R and L. This was to allow more consonants on the left hand and use of the middle-finger-upper-row. I think the middle finger being a strong finger and comfortable reach can handle these quite well. The index finger is the same as Colemak.
On the right hand: The right index goes up a little bit, mostly with NY and MY. They aren’t as high as you may think. But the Y SFBs and adding in NK/KN, NF, etc, they do add up a little. It’s not all that much but it is more. I think it’s still easily manageable. NY is actually not all that common, for example it’s actually less common than EU/UE. (The exact Mayzner numbers are: NY/YN is 2,761 / 373. Compared to EU/UE is 878 / 4,158. MY/YM is 1,753 / 668.) The middle finger actually goes down, EO/OE is very low. The ring finger goes up a bit with UI/IU being more common than Colemak’s IY/YI.
If you want overall percentages (index finger pressing C), Middlemak comes in at 2.078%, vs Colemak’s 1.815%, vs Workmans 3.053%, vs Qwerty’s 6.264%.
Colemak really did a good job finding low SFB pairings of SF, DT, NHL, RW, and IY. So changing 3 of those pairings does come at a cost. But it’s really not much and it’s done so that Middlemak can put more frequent letters to better spots, reduce pinballing, and reduce NFB. I think the SFB is in the same realm as Colemak, it’s still pretty low.

Ortho SFB

Keyboard	Total SFB	Left hand SFB	Right hand SFB	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky
QWERTY	195,687	136,337	59,350	1,105	1,661	86,462	47,108	34,166	3,970	21,214	0
Workman	79,561	30,682	48,878	1,105	3,712	8,206	17,659	27,338	5,037	16,503	0
Dvorak	70,365	28,101	42,264	0	31	1,591	26,480	12,142	16,265	8,229	5,629
Middlemak	42,816	23,924	18,892	1,105	1,661	14,888	6,270	12,415	3,134	3,343	0
Colemak	30,032	14,345	15,687	1,105	1,267	5,702	6,270	9,831	5,037	819	0
MTGAP	34,151	14,550	19,601	1,147	339	3,134	9,931	4,629	8,633	5,071	1,267

Notes: This is with the middle finger pressing C location, “proper” or ortho style. Same data as above.
Middlemak on Ortho actually comes in just a tad lower, but Colemak takes a bigger dip.
This basically moves most of the SFB from the index finger to the middle finger. On Middlemak, ortho takes out CT on the index finger and puts in CR and CL on the middle finger, which is only a tad lower. On Colemak, ortho takes out CT and puts in SC which is much lower.
I think it still works well. The middle finger is a strong finger and can cover most of these well. The one issue is that CL which is a two row jump.
I’ll break down the full Mayzner numbers:
First the index finger pressing CT/TC is 12,998 / 737. This is what ortho removes.
Middlemak Ortho puts in CRC which is 4,214 / 3,423. And puts in CL / LC which is 4,202 / 333. In percentages, CR is 32% of CT, RC is 26% of CT, and CL is 32% of CT. Adding it all up and including all the others (CP, CG, etc), Middlemak Ortho comes in at 87% of CT+TC. Just a tad lower.
Colemak Ortho puts in CS/SC, which is 644 / 4,363. CF/ FC is pretty much nill. In percentages, SC is 34% of CT. Adding it all up and including the others, Colemak Ortho comes in at 36% of CT+TC. A big dip.
If you want overall percentages (middle finger pressing C, Ortho style), Middlemak comes in at 2.16%, vs Colemak’s 1.669%, vs Workman’s 3.147%, vs Qwerty’s 6.575%. This is from Colemak-DH analyzer which says Middlemak ortho increases, as opposed to what I found that Middlemak ortho decreases SFB. I think it comes down to the data set used.

6) Low ring-to-pinky rolls. And the amazing inside roll from H.

A secondary design goal was to decrease the awkward ring-to-pinky roll. Middlemak greatly reduces this roll.
On the left hand: Middlemak’s ring-to-pinky outside roll of SA and WA is actually pretty low. SA is actually the lowest bigram of the common consonants to A. So SA and WA comes out to be a good design. Compared to Colemak, Middlemak’s SA outside roll is far less common than Colemak’s RA. (The exact Mayzner in millions are: Middlemak’s SA is 6,147. Colemak’s RA is 19,333. WA is the same for both at 10,865.)
On the right hand: Middlmak’s ring-to-pinky roll of IH and UH is very low. Vs Colemak: IO is very common and even YO is pretty common. (The exact numbers are Middlemak’s IH is 59, UH is 30. Vs Colemak’s IO 23,542, YO is 4,227.)
This next bit isn’t even a secondary design goal, but it’s worth mentioning: On the right hand, outside rolls from any of the fingers to the pinky H are very low. The common letters that come before H are usually TH, CH, WH, SH, and GH. All of those are on the opposite left hand. Doing the math, 97% of the letters before H are on the left hand. H on the pinky gives an amazing inside roll almost all the time.
Also worth mentioning is I think this also de-emphasizes the pinky a bit. First it limits the ring-to-pinky rolls. Second, the frequency of H is much lower than Colemak’s O. Third, the bigram HH is virtually nill, vs Colemak’s OO which is pretty common. None of these are massive, but I think they add up to de-emphasizing the pinky a bit.

7) Qwerty similarity.

Last but not least, this maintains Qwerty similarity to make it easier to learn.
First, this maintains the bottom row (except N) and maintains Q, W, A, and S. Many letters move on the same finger, including T, G, N, Y, J .
Second, this “keeps” the ASRT order. People learning Colemak commonly say that R and S are reversed. Technically it’s not because R was never on the middle finger to begin with. But it seems to be reversed in the brain. I think it’s two things. First ASRT keeps it in the same order, with R between S and T, even if R changes fingers. Second, I wonder how many people alt-fingered R typing in Qwerty, either all the time or for the RT bigram.
(I've also wondered if, for example, patterns like the very common ION simply being easier to type would make this faster to learn. I’m not certain about that though.)
Comparing layouts:

• Middlemak: 10 letters stay in their original spot and 16 move, with 5 letters moving on the same finger and 11 letters switching fingers.
  
• Colemak: 10 letters stay in their original spot and 16 move, with 4 letters moving on the same finger and 12 letters switching fingers.
  
• Colemak-DH: 5 letters stay in their original spot and 21 move, with 6 letters moving on the same finger and 15 letters switching fingers.
  
• Workman: 6 letters stay in their original spot and 20 move, with 8 letters moving on the same finger and 12 letters switching fingers.
  

*Not accounting for punctuation changes.
Considering letters swapping on the same finger, this has fewer changes than Colemak. And I think “keeping” the ASRT order will make it even easier to adopt than the numbers suggest.
(I hope people don’t think that I’m overly focused on keeping S in the Qwerty location just because. For this layout, it does work better with ASRT, I address it in more detail in excessive amount of details.)

Downsides

1) There is a slight increase in SFB compared to Colemak. But overall I think SFB is still low.
2) There are a handful of NFB with R: BR, GR, PR, etc. Same with L and BL. This isn’t enough to change the home row to ARST. There is more below but ASRT has lower SFB, lower ring-to-pinky rolls, and is easier to learn. Also note these are much, much lower than the extremely high NFB that Colemak has with HE, LE, etc, (to solve those I paid L and R together).
3) I wanted a less common letter in the G location (Qwerty T location). I really tried to put a less common letter there, but it simply didn’t work.
4) I mention this above but to add more: The OM scissor-gram is a bit more common than I'd like. This is a limitation of keeping Qwerty similarity (the M) and wanting to put O on the upper-row-middle-finger. (MTGAP has a good design on this. It paired its EO with a lot of punctuation on the index finger to reduce NFBs.)

Conclusion

I'm really happy with this layout. I think it is a great combination of pretty low SFB, good use of the strong middle-finger-upper-row position, reduced NFB and pinballing, keeping Qwerty similarity, and easy to transition to.
Honestly I’m surprised no one else has stumbled on this arrangement, since many people have taken to trying to fix Colemak’s issues and keep Qwerty similarity.

Some nice rolls and combinations

Although this increases alternating, there are still plenty of rolls left. Some I want to mention:
HE is very easy.
LD feels very good.
OU feels very comfortable.
OF/FO is very common and feels very good.
IO and ION are very nice rolls.
This maintains the nice IEN roll.
IOU is a redirect, but overall feels pretty ok.
YOU has a lateral move, but feels comfortable.
OU+LD together to make OULD feels good.
The very common TH, THE, and THE_ feel very nice not having to move the fingers off the home row.

Vs Colemak and Vs Colemak-DH

I made Middlemak to try to solve the problems I found with Colemak. I think all the items came out above, but as a bit of a summary and to include Colemak-DH.
Colemak:
1) Colemak has an issue with high lateral movement with D and H. You see this especially with H and its NFBs, but D isn’t the best either.
2) Colemak’s right index finger is overworked. It just has too much to do with N, H, and L.
3) Colemak has a pinballing issue due to placing most vowels and many common consonants on the same hand.
4) Colemak has a ton of NFB with EL/LE and HE.
5) Colemak has a hand balance issue. It places a lot of emphasis on the right hand with the vowels and several common consonants. (Hand dominance and pinballing are technically separate items.)
6) Colemak’s awkward ring-to-pinky rolls. Notably IO, YO (especially when typing YOU), and also RA on the right hand.
7) Not exactly a direct issue with Colemak, but Middlemak is easier to learn coming from Qwerty. Many new users want to ‘keep’ the order of ASRT instead of ‘reversing’ S and R.
Middlemak is a pretty low increase in SFBs and it fixes those issues.
Colemak DH:
Colemak-DH solves 1) the lateral movements. But it doesn’t solve 2) the right index finger overwork, 3) the pinballing, 4) the NFBs 5) the hand balance, 6) the awkward ring-to-pinky rolls, and 7) coming from Qwerty it’s pretty much a full change layout, taking out ease of transition. This also means common shortcuts move. They are still accessible, but they move (paste is the same. Undo, cut, and copy move.)

Option: Swap A and H

This gives:

QWLDG JFOU: HSRTP YNEIA ZXCVB KM,./ 

The good: This is good if you want to put all the vowels and most consonants on opposite hands. Overall this increases alternating. This mostly comes from moving A to the right hand because A pairs a lot with the L, R, S, T, etc. This is the biggest appeal if you find those uncomfortable. Moving H actually decreases alternating because there’s a high bigram with TH (you can test your own comfortability on that. Don’t forget capitalized Th is common.) Excessive details: Middlemak A has 41% of bigrams on the opposite hand, swapping gives 65%. Middlemak H has 58% of bigrams on the opposite hand, swapping gives 51% (again, mostly because of TH). Overall, middlemak A+H has 48% on the opposite hand, swapping gives 59%.
The neutral: On the left hand, the ring-to-pinky outside roll of SH is slightly higher, but in a similar range. Data from Mayzner is SH is 8,889, and original is SA is 6,147. Changing WA to WH is practically identical.
The bad: On both the right and left hand, the all-fingers-to-pinky outside roll is higher. On the left hand, it’s because TH is very common. On the right hand, it takes away the remarkable inside roll of H and adds in a moderate number of all-fingers-to-A bigrams with EA, IA, NA, MA, etc, (OA is actually low). Notably, the ring-to-pinky outside roll is higher with IA and UA coming in at 8,072 and 3,844, vs original IH and UH of 59 and 30.
The complicated: This option decreases the pinky-to-all-fingers rolls inside rolls on both hands. It increases the all-fingers-to-pinky rolls outside rolls. The overall sum of inside rolls plus outside rolls is lower. In short: this option lowers inside rolls, increases outside rolls, with an overall sum that is lower.
Basically stick to the original if you want to lower ring-to-pinky movement, lower outside rolls, and easier to learn. Overall I put this as a decent option if someone wants to maximize alternating without going to a full change layout.

Excessive amount of stats

This section has details on distance reductions, hand balance based on frequency, hand balance based on distance, and raw distance numbers. I'm going to give lots of charts to cover all the details. Those that want to know all the details or those that like to analyze will find this interesting, but others don’t have to read.

Distance reduction from Qwerty

Keyboard	Distance	Distance	Distance
	(km)	(miles)	(% reduction from Qwerty)
MTGAP	2.592	1.611	46.3%
Workman	2.613	1.624	45.9%
Norman	2.615	1.625	45.9%
Colemak	2.616	1.625	45.8%
Middlemak	2.710	1.684	43.9%
Semimak	2.739	1.702	43.3%
Dvorak	2.813	1.748	41.8%
QWERTY	4.829	3.001	0%

*Source: 62 page / 31k word compilation. Left index used to press Qwerty C location. Includes punctuation.
Middlemak’s distance reduction is good, most alt keyboards are in the same range. It’s a tad worse on distance than some others because putting O off the home row increases distance, with the O being more frequent. (MTGAP puts O off the home row, but has all the punctuation on the diagonals to decrease distance.)

All Finger Distances

This is the raw distance travelled in metres. Percentages rely on the total distance travelled, which changes between layouts so it's not comparable. Note that distance is different from frequency. The fingers don't move for the home row, the distance travelled is to the other letters.

Keyboard	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky	L Total	R Total	Total
MTGAP	293	116	413	246	667	320	219	318	1068 m	1524 m	2592 m
Workman	123	232	330	753	505	200	146	325	1437 m	1176 m	2613 m
Norman	127	127	223	674	525	377	244	319	1150 m	1465 m	2615 m
Colemak	130	127	116	842	747	200	141	315	1214 m	1402 m	2616 m
Middlemak	134	127	207	847	440	460	189	306	1314 m	1396 m	2710 m
Semimak	238	237	326	390	506	200	450	393	1191 m	1549 m	2739 m
Dvorak	205	54	34	692	753	285	395	396	984 m	1829 m	2813 m
QWERTY	130	127	670	1413	1169	452	450	418	2340 m	2489 m	4829 m

*All distances in metres. Same source as table above.
First, you can see Qwerty’s distances are extremely high. All the alternatives make big reductions.
Notice that the index fingers move way more than the other fingers.
Middlemak balances this out decently and moves more distance to both middle fingers. The left index finger is still a bit high because it has D off the home row (this brings it back to the discussion of where does the 11th letter go). Middlemak’s left index is slightly higher than Colemak’s even though they have the same letters because the row stagger technically puts D on a slight diagonal. A technicality really.
The right index finger is low, a good feature that the index finger next to the vowels doesn’t have to travel much. Compared to Colemak, Middlemak’s movement is 41% lower than Colemaks.

Zoom in on index fingers

Keyboard	Distance	Distance	Total	Index Finger % of total distance
	L Index	R Index	L+R Index	L+R Index
MTGAP	246	667	913 m	35%
Workman	753	505	1259 m	48%
Norman	674	525	1200 m	46%
Colemak	842	747	1589 m	61%
Middlemak	847	440	1287 m	47%
Semimak	390	506	896 m	33%
Dvorak	692	753	1445 m	51%
QWERTY	1413	1169	2582 m	53%

Here you can really see how much work the index fingers do. Middlemak’s sum of the index fingers is nice and low.
You can also look at the percentage of travel with the index finger. Middlemak’s is down at 47% - meaning the index fingers are responsible for 47% of all the distance travelled. On Colemak the index fingers account for 61% the distance travelled, quite a big difference relatively. Workman and Norman are close to Middlemak. MTGAP and Semimak are very low. Dvorak and Qwerty percentages are a bit deceiving though - the higher index finger distances are hidden by the higher total amount of distance.
So even when Middlemak’s left index finger doesn’t have a reduction in distance, the total index finger distance is significantly reduced.

Hand Balance based on Distance

This is the distance in percentages for all the fingers (unlike the raw distances in metres above) so that I can add them up to look at hand balance.

Keyboard	Left hand	Right hand	Hand Balance Ratio	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky
MTGAP	41.2*	58.8	1.43	11.3	4.5	15.9	9.5	25.7	12.3	8.4	12.3
Workman	55.0	45.0*	0.82	4.7	8.9	12.6	28.8	19.3	7.6	5.6	12.5
Norman	44.0	56.0	1.27	4.8	4.8	8.5	25.8	20.1	14.4	9.3	12.2
Colemak	46.4	53.6*	1.15	5.0	4.8	4.4	32.2	28.6	7.6	5.4	12.0
Middlemak	48.5	51.5*	1.06	5.0	4.7	7.6	31.2	16.3	17.0	7.0	11.3
Semimak	43.5	56.5*	1.30	8.7	8.7	11.9	14.2	18.5	7.3	16.4	14.4
Dvorak	35.0*	65.0	1.86	7.3	1.9	1.2	24.6	26.8	10.1	14.0	14.1
QWERTY	48.5	51.5	1.06	2.7	2.6	13.9	29.3	24.2	9.4	9.3	8.7

Notes: * indicates vowel hand (Norman and Qwerty are mixed enough that there is no clear vowel hand). The ratio between the hands can change quite quickly, so I think that highlights the differences between the layouts better. Ratio >1 indicates right hand dominant, <1 indicates left hand dominant.
I think the raw distances are better for comparison between layouts, but some people like percentages. I used this mostly to calculate hand balance.
You can see Middlemak spreads out the work. The pinkies are low, ring fingers are reasonable, middle takes on more, and the indexes have most of the distances.

Hand Balance based on Frequency

Now we’re shifting to frequency. You can see the data for all the fingers, this is enlightening on its own. I also add them up to look at the hand balance.

Keyboard	Left hand	Right hand	Hand Balance Ratio	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky
MTGAP	49.9*	50.0	1.00	10.8	8.4	18.8	11.9	16.3	14.5	9.8	9.4
Workman	50.7	49.3*	0.97	9.3	10.1	11.1	20.2	15.2	14.9	9.7	9.5
Norman	51.7	48.3	0.93	9.3	8.3	15.4	18.7	15.2	13.2	11.8	8.1
Colemak	46.7	53.2*	1.14	9.4	7.7	8.2	21.4	19.4	14.9	9.1	9.8
Middlemak	48.8	51.2*	1.05	9.4	8.3	9.7	21.4	13.7	19.6	9.9	8.0
Semimak	46.6	53.4*	1.15	9.7	10.1	11.8	15.0	12.2	14.9	15.6	10.7
Dvorak	44.8*	55.3	1.23	10.3	8.1	12.2	14.2	17.2	13.4	13.0	11.7
QWERTY	56.5	43.5	0.77	9.4	8.3	15.4	23.4	18.8	8.4	11.8	4.5

* indicates vowel hand (Norman and Qwerty are mixed enough that there is no vowel hand). Ratio of >1 indicates right hand dominant, <1 indicates left hand dominant.
The ratio between the hands can change quite quickly, so I think that highlights the differences between the layouts better than the percentages. I know it doesn't add up to 100%, the analyzer is doing some rounding or something behind the scenes, I'm not going to round any further.
E is so common that it bumps up any finger it’s on.
Middlemak’s left middle finger takes on a bit more work. The left index is still high because of D. The right index is notably lower taking off common letters. The right middle takes on more of the work with O. And the right pinky is lower too with the H there.

Hand Balance Distance + Frequency

This is the same chart that was above, just for a summary.

Keyboard	Dominant Hand	Hand Balance ratio based on Frequency	Hand Balance ratio based on Distance	Hand Balance ratio based on Frequency and Distance	Corrected comparison number for left hand dominant
MTGAP	Right	1.00	1.43	1.43
Workman	Left	0.97	0.82	0.80	1.26
Norman	Right	0.93	1.27	1.19
Colemak	Right	1.14	1.15	1.32
Middlemak	Right	1.05	1.06	1.11
Semimak	Right	1.15	1.30	1.49
Dvorak	Right	1.23	1.86	2.30
QWERTY	Left	0.77	1.06	0.82	1.22

1 indicates right hand dominant, <1 indicates left hand dominant. Hand Balance ratio based on Frequency and Distance is a simple multiplication of the two.

Frequency-off-home-row (F.O.H.R.)

At the risk of too much information, another way to look at this is the frequency that’s off-home-row. I.e. the frequency that is not under the fingers.
Distances weigh the keys differently, while frequency-off-home-row levels that field. This should be read in conjunction with distance. Neither is better than the other, it’s just a different measurement.

Keyboard	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky	L+R Index	Index FOHR as % of All fingers FOHR
MTGAP	4.0	2.0	7.2	4.1	10.8	5.8	3.6	3.8	14.9%	36%
Workman	1.5	3.9	5.6	11.5	8.8	3.3	2.5	2.7	20.3%	51%
Norman	1.5	2.1	3.8	10.0	8.8	6.4	4.6	2.6	18.8%	47%
Colemak	1.6	2.1	2.0	12.7	13.0	3.3	2.3	2.6	25.7%	65%
Middlemak	1.6	2.1	4.1	12.7	7.3	8.0	3.1	2.5	20.0%	48%
Semimak	3.5	4.5	5.4	6.3	8.4	3.3	7.8	3.9	14.7%	34%
Dvorak	2.5	0.9	0.6	11.7	11.7	4.7	6.6	5.5	23.4%	53%
QWERTY	1.6	2.1	11.6	21.4	18.6	7.5	7.8	4.5	40.0%	53%

I did this after noticing, on Colemak, that my right index finger moves off the home row a lot. For Colemak the left index finger has more distance, but the right index finger has more frequency-off-home-row. The long distance to B on the left hand skews the distance travelled.
The effect is similar on most of the layouts: FOHR tones down the left hand dominance and bumps up the right hand dominance. With Colemak the frequency of H and L is enough to actually flip them. This is more noticeable with layouts that keep Qwerty’s B (which is just frequent enough to skew things), layouts that change the letter to a less common one shift less.
Also insightful is “Index FOHR as % of All fingers FOHR”. It shows that on Middlemak the index fingers do 48% of all frequency-off-home-row, quite reasonable. Whereas on Colemak the index fingers have 65% of all frequency-off-home-row, very high. Workman is just a tad higher with the index fingers doing 51% of frequency-off-home-row. Dvorak’s and Qwerty’s 53% for index fingers is a bit deceiving though, it appears low but that's because the overall total FOHR is much higher.
This is similar to “Zoom in on the index fingers”, but this shows an even greater difference between Middlemak and Colemak. This shows Middlemak at 48% and Colemak at 65% based on index finger % of total FOHR. Recall Middlemaks was 47% and Colemaks was 61% based on index finger % of total distance. Same thing, when the distances to the diagonals are levelled out, the frequency stands out.

Excessive amount of detail on letters

You really don’t need to read this, mostly if you have any lingering or specific Q on specific letters.

D position and why not curl?

Back to the discussion of there being 11 common letters: Where to place that 11th letter is difficult, it’s either the ring finger or index finger. Many find the ring finger to not be very dexterous for such a frequent letter. So putting D on the index finger with T works very well for low SFB.
Why not the middle row? First, I think most people find it easier to reach up and down rather than laterally. Second, it works very well with the LD roll.
Why not the bottom row? I want to leave the bottom row the same as Qwerty for easy adoption and to leave the shortcuts of undo=cut-copy-paste the same. I also think C and M (the curl locations) are already near perfect frequencies for those locations. Those are the middle of the pack frequencies, which are perfect for the index fingers - which cover 6 to 7 letters.
I previously said that I’m not a fan of Colemak’s L position, which is the same upper-row-index-finger. But Colemak’s L has a ton of NFBs with the vowels, especially with E but also with all the other vowels. Looking at the whole hand, 62% of Colemak’s L bigrams are on the same hand. I think that’s a big source of discomfort. But with Middlemak’s D, only 22% of bigrams are on the same hand. So Middlemak has far lower same-hand-gymnastics.

The R and L position

I’m sure the R and S location will attract attention, so I’ll discuss it more here.
If you want to skip this, the main factor is that R and L pair well together. Better than S and L. That means the RL column goes on the middle finger. R being more common goes on the home row, and L goes on the middle-finger-upper-row which is a strong location.
[First note that Colemak had to go with ARST because of F on the middle-finger-upper-row. The SFB of RF/FR is more common than SF/FS.]

ASRT, making a RL column

The pros of ASRT is it gives fewer SFB, with LRL being less than half as common as SL/LS. It also gives fewer ring-to-pinky-rolls with SA being far less common than Colemak’s RA. It also makes it much easier to learn from Qwerty.
The cons is that the R position leads to more NFB with the index finger. Most of these feel ok but RD, GR, BR are not so good. These aren’t super common, but they are there. PR gives the commonly disliked lateral NFB, but I think it’s relatively ok for the frequency - it’s not super common. RS creates an outside roll, but I’d rather have the RS on Middlemak’s middle-to-ring fingers than Colemak’s RA on the ring-to-pinky fingers. Colemak’s RA is much more common than Middlemak’s RS too.

ARST, making a SL column

The pros of ARST is that it gives an inside roll of RS. It also separates R from the NFB letters on the index finger, which do add up. However, I’m trying to resolve the NFBs of Colemak’s right hand H and L with E which is a magnitude larger, in addition to the pinballing that it creates which requires moving L.
The cons with ARST is that it gives more SFB. It’s not huge, but it is there. It also gives a lot of ring-to-pinky outside rolls with RA. And finally this makes it much harder to learn coming from Qwerty.

Verdict: ASRT, RL column

None of these are huge deal breakers for one or the other. I went with lower SFB, lower ring-to-pinky rolls, and much easier to learn. At the cost of some NFB on the index finger and inside roll of RS.
(If someone comes from Colemak and wants ARST, they can. It's not a humongous downside to do that. Mostly the more common RA ring-to-pinky roll and slightly more SFB of SD/DS.)
(This layout truly does work better with ASRT. If it didn’t work, I would have changed it. For example, I wanted to keep G on its Qwerty location but it’s better to move it, so I didn’t shy away from changes.)

P and G location

It’s better to move G than keep it in its Qwerty location.
P and G are pretty much the same frequency, so it doesn’t matter in that sense.
P has a SFB with T, and it has common bigrams of PR, SP, and even PA. Those work better with P on the middle row. There is PL, but the ones above outweigh it.
G has a bigram with NG. With N on the opposite hand, that means G can go on the worse diagonal spot. There is GR, but PR is far more common.

O location

O and E together give very low SFB. This is a great opportunity to put them together. The strong middle finger only has 3 keys, so it can handle two very frequent letters.

U, Y and F locations

U works pretty well with I, the SFB are reasonably low. This location also creates a nice OU and YOU roll.
Y goes to the index finger, which actually works pretty well. I broke down the numbers in SFB section.
F goes to the right hand index finger, because putting P or G on the right hand gives too many SFB with M and N. The F location works quite well for both moderate frequency and low SFB. It has a lot of NFB with OF/FO, so placing it next to O makes a comfortable roll.

This is a new writeup for Middlemak (renamed since my previous post). Named for the emphasis on the middle finger and use of the middle-finger-upper-row locations for frequent letters
Link to Layout heatmap

QWLDG JFOU: ASRTP YNEIH ZXCVB KM,./ 

The main design criteria was:
1) Use the middle-finger-upper-row position. The middle finger is a very strong finger, and the middle-finger-upper-row is a very good location for a common letter.
2) To reduce pinballing by putting most vowels on the right hand and putting most of the frequent consonants on the left hand.
3) Reduce the number of Near Finger Bigrams (NFBs).
4) Improve hand balance.
5) Keep SFB down of course.
6) A secondary design criteria was low ring-to-pinky-rolls. I find this to be an awkward movement.
7) Maintain Qwerty similarity to make it easier to learn.
These are all discussed in more detail below. (Middlemak was based on Colemak so I’ll be talking in that sense for some of it.)

1) Middle-Finger-Upper-Row

The main purpose of this layout is to place 2 common letters on the middle-finger-upper-row. The middle finger is very strong and these are excellent locations that are easy to reach.
When you look at the letter frequency, there are 11 common letters before it takes a big drop off. The first eight letters go on the home row under the fingers. The next question is where do the next 3 letters go. Middlemak’s philosophy is to place 2 on the middle-finger-upper-row, and 1 on the index-finger-upper-row. Colemak places 2 on the centre columns, requiring a lateral movement that many find uncomfortable. Workman has 1 middle-finger-upper-row, 1 ring-finger-upper-row, and 1 index-finger-lower-row, a mixed bag.
Using the middle finger also unloads the index fingers. The index fingers can get very busy with 6-7 letters compared with the other letters 3. Especially when multiple frequent letters are put on them. I’ll give the distance stats in “Excessive amounts of stats” section far below which shows the drastic difference.

2) Pinballing.

I wrote about pinballing in length here. The short of it is that there is an incredibly strong interaction between vowels and the consonants, with 75% of all bigrams between vowels and consonants.
I’ll cover Colemak first to outline the issue. Having most of the vowels EOIUY and many frequent consonants of NHLMKJ on the right hand leads to a pinballing proble. I call this the NHLMKJ wall, it’s a wall of consonants that the vowels pinball off of
To reduce pinballing Middlemak does two things:
First Middlemak moves L off the vowel hand and to the consonant hand. This (and moving H) reduces the wall to NMFPJ which is much, much lower frequency. Or if the word starts with NMFPJ, it’s again much lower frequency.
Second it moves H to the pinky. H is a very unique letter because it’s almost always followed by a vowel. Moving H to the pinky means it almost always gives an inside roll. After the vowel, it generally continues to the consonants NMFPJ continuing the inside roll, or it alternates to the other hand. Rarely goes back to H. Generally only after the third letter can it start to pinball back to the vowels or back to H.
Combined, this reduces pinballing and increases alternating. It’s not entirely alternating, Middlemak still has rolls because it has N, H, and MKFJ. For N I find the rolls to be comfortable, likely because N is on the home row.

3) Near Finger Bigrams, NFB

NFB are bigrams that are on adjacent fingers. Too many or an uncomfortable arrangement can be an issue. NFB and pinballing are separate issues: NFB is the smaller pattern of adjacent fingers, pinballing is the much larger pattern of pinballing between all vowels and constants.
I’ll cover Colemak first again to highlight the issue. On Colemak there are extremely common NFBs with HE and LE/EL. (There is also NE/EN, but I find the HE and EL/LE to be awkward because L and H are not on the home row.) Pretty much all E bigrams are fairly high. I think you want to move common consonants away from E to reduce NFBs.
Middlemak drastically reduces the NFBs by moving the common letters L and H off the adjacent index finger. The replacement of F and Y have dramatically lower NFBs.
Looking at other layouts: Workman still has high NFB with L and E. Colemak-DH maintains L and H next to E. Many find the new HE roll to be more comfortable but it’s still a ton of NFB, and LE/EL is still there. MTGAP moved practically all letters away from its E and O and instead put punctuation there, a very good solution.
Middlemak does create some NFB with O placement. Namely OF, ON, and OM. However OF and ON are very comfortable, having the middle finger extended up feels pretty comfortable. OM is ok, and it’s the least common of those 3. The OM scissor-gram is a bit more common than I'd like, but it feels ok. This is about the best location for a scissor-gram too, with the long-middle-finger up and the curl-index-finger down being able to handle it fairly well. (Just to mention Colemak’s EL/LE, not exactly a scissor-gram, I find is bad since the shorter index finger has to reach up and the middle finger is so long it’s already curled just to be on the home row.)
Overall, Middlemak moving common consonants away from E drastically lowers the number of NFBs and takes away a ton of awkward movements. Visually, I think you can see this pretty easily in the heatmaps.
(You can extend this concept to fingers further away, to include Colemaks HI, LI, etc. This can be valid, but I’ll focus on adjacent fingers for now.)

4) Hand balance

Hand balance is a difficult topic. The two most obvious metrics for hand dominance are 1) the frequency of the letters, and 2) the distance the fingers have to travel (which is different from frequency because you don't travel to the home-row). But any pinballing from having too many vowels and consonants on the same hand doesn’t show up in those numbers. Having said all that, I’m going to look at frequency and distance.

Keyboard	Vowel Hand	Dominant Hand	Hand Balance ratio based on Frequency	Hand Balance ratio based on Distance	Hand Balance ratio based on Frequency and Distance	Corrected comparison number for left hand dominant
MTGAP	Left	Right	1.00	1.43	1.43
Workman	Right	Left	0.97	0.82	0.80	1.26
Norman	n/a	Right	0.93	1.27	1.19
Colemak	Right	Right	1.14	1.15	1.32
Middlemak	Right	Right	1.05	1.06	1.11
Semimak	Right	Right	1.15	1.30	1.49
Dvorak	Left	Right	1.23	1.86	2.30
QWERTY	n/a	Left	0.77	1.06	0.82	1.22

Notes: Ratio >1 indicates right hand dominant, <1 indicates left hand dominant. The hand balance ratios are based on each hand's frequency/distance, right hand divided by left hand. E.g. Middlemak: 51.2% right hand frequency divided by 48.8% left hand frequency equals 1.05. Colemak: 53.2% right hand frequency divided by 46.7% left hand frequency equals 1.14. Hand Balance ratio based on Frequency and Distance is a simple multiplication of the two ratios. I did this because having more frequency and more movement on the same hand amplifies the hand dominance. Or if one hand has more frequency and the other hand has more movement, that would mitigate it to some degree. This is by no means the most definitive way to look at it, but it's what I've managed to come up with. The "Corrected comparison number for left hand dominant" is a hard math thing to explain (that I hope I'm right about anyway). E.g. for Workman listing 0.80 gives the wrong impression the hand dominance is equivalent to 1.2 (20%). But it's actually 1.26 (26%), obtained from 1/0.796=1.26.
Middlemak is more balanced on both frequency and distance. When both are considered, it really adds up to being more balanced. Middlemak combined balance ratio is 1.11, compared to Colemak’s 1.32 or Workman’s 1.26.
Ratios can change quite quickly because both the numerator and the denominator (in this case, the left and right hand) can go in opposite directions. Moving L to the other hand does exactly this - changing both the numerator and denominator, and it changes both the frequency and the distance. So that one move really helps to balance out the hands.
(I’ll add the full data in “Excessive amounts of stats” for both frequency and distance, broken down to hands and fingers. The individual finger load is quite interesting too.)

5) SFB

The SFB of this layout is still pretty low, but it does come out a tad higher than Colemak’s. I’ve debated how to show this and decided to go with the full nuclear option. Data is from Mayzner revisited in millions (which only has letter bigrams and doesn’t have punctuation data, that’s what the 0’s are).

Keyboard	Total SFB	Left hand SFB	Right hand SFB		L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky
QWERTY	185,270	125,920	59,350		1,105	1,661	54,502	68,651	34,166	3,970	21,214	0
Workman	78,147	29,268	48,878		1,105	3,712	2,806	21,645	27,338	5,037	16,503	0
Dvorak	70,570	28,306	42,264		0	31	0	28,275	12,142	16,265	8,229	5,629
Middlemak	44,698	25,806	18,892		1,105	1,661	2,716	20,323	12,415	3,134	3,343	0
Colemak	39,023	23,336	15,687		1,105	1,267	639	20,323	9,831	5,037	819	0
MTGAP	34,151	14,550	19,601		1,147	339	3,134	9,931	4,629	8,633	5,071	1,267

Notes: This is with index finger pressing C location, how I think most people type. Adding in punctuation the numbers would be higher, I expect they would increase by similar absolute amounts.
For Middlemak: You can see it’s not all that much of an increase in SFB from Colemak. Overall the numbers are low considering it keeps many Qwerty aspects.
On the left hand: The middle finger goes up with R and L. This was to allow more consonants on the left hand and use of the middle-finger-upper-row. I think the middle finger being a strong finger and comfortable reach can handle these quite well. The index finger is the same as Colemak.
On the right hand: The right index goes up a little bit, mostly with NY and MY. They aren’t as high as you may think. But the Y SFB and adding in NK/KN, NF, etc they do add up. It’s not all that much but it is more. I think it’s still easily manageable. (NY is actually not all that common, for example it’s actually less common than EU/UE. The exact Mayzner numbers are: NY/YN is 2,761 / 373. Compared to EU/UE is 878 / 4,158. MY/YM is 1,753 / 668.) The middle finger actually goes down, EO/OE is very low. The ring finger goes up a bit with UI/IU being more common than Colemak’s IY/YI.
If you want overall percentages (index finger pressing C), Middlemak comes in at 2.078%, vs Colemak’s 1.815%, vs Workmans 3.053%, vs Qwerty’s 6.264%.
Colemak really did a good job finding low SFB pairings of SF, DT, NHL, and IY. So changing 3 of those pairings does come at a cost. But it’s really not much and it’s done so that Middlemak can put more frequent letters to better spots, reduce pinballing, and reduce NFB. I think the SFB is in the same realm as Colemak, it’s still pretty low.

Ortho SFB

Keyboard	Total SFB	Left hand SFB	Right hand SFB	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky
QWERTY	195,687	136,337	59,350	1,105	1,661	86,462	47,108	34,166	3,970	21,214	0
Workman	79,561	30,682	48,878	1,105	3,712	8,206	17,659	27,338	5,037	16,503	0
Dvorak	70,365	28,101	42,264	0	31	1,591	26,480	12,142	16,265	8,229	5,629
Middlemak	42,816	23,924	18,892	1,105	1,661	14,888	6,270	12,415	3,134	3,343	0
Colemak	30,032	14,345	15,687	1,105	1,267	5,702	6,270	9,831	5,037	819	0
MTGAP	34,151	14,550	19,601	1,147	339	3,134	9,931	4,629	8,633	5,071	1,267

Notes: This is with the middle finger pressing C location, “proper” or ortho style. Same data as above.
Middlemak on Ortho actually comes in just a tad lower, but Colemak takes a bigger dip.
This basically moves most of the SFB from the index finger to the middle finger. On Middlemak, ortho takes out CT on the index finger and puts in CR and CL on the middle finger, which is only a tad lower. On Colemak, ortho takes out CT and puts in SC which is much lower.
I think it still works well. The middle finger is a strong finger and can cover most of these well. The one issue is that CL which is a two row jump.
I’ll break down the full Mayzner numbers:
First the index finger pressing CT/TC is 12,998 / 737. This is what ortho removes.
Middlemak Ortho puts in CRC which is 4,214 / 3,423. And puts in CL / LC which is 4,202 / 333. In percentages, CR is 32% of CT, RC is 26% of CT, and CL is 32% of CT. Adding it all up and including all the others (CP, CG, etc), Middlemak Ortho comes in at 87% of CT+TC. Just a tad lower.
Colemak Ortho puts in CS/SC, which is 644 / 4,363. CF/ FC is pretty much nill. In percentages, SC is 34% of CT. Adding it all up and including the others, Colemak Ortho comes in at 36% of CT+TC. A big dip.
If you want overall percentages (middle finger pressing C, Ortho style), Middlemak comes in at 2.16%, vs Colemak’s 1.669%, vs Workman’s 3.147%, vs Qwerty’s 6.575%. This is from Colemak-DH analyzer which says Middlemak ortho increases, as opposed to what I found that Middlemak ortho decreases SFB. I think it comes down to the data set used.

6) Low ring-to-pinky rolls. And the amazing inside roll from H.

A secondary design goal was to decrease the awkward ring-to-pinky roll. I find this to be the most awkward movement on the keyboard (for common movements around the home row). Middlemak greatly reduces this roll.
On the left hand: Middlemak’s ring-to-pinky outside roll of SA and WA is actually pretty low. SA is actually the lowest bigram of the common consonants to A. So SA and WA comes out to be a good design. Compared to Colemak, Middlemak’s SA outside roll is far less common than Colemak’s RA. (The exact Mayzner in millions are: Middlemak’s SA is 6,147. Colemak’s RA is 19,333. WA is the same for both at 10,865.)
On the right hand: Middlmak’s ring-to-pinky roll of IH and UH is very low. Vs Colemak: IO is very common and even YO is pretty common. (The exact numbers are Middlemak’s IH is 59, UH is 30. Vs Colemak’s IO 23,542, YO is 4,227.)
This next bit isn’t even a secondary design goal, but it’s worth mentioning. On the right hand, outside rolls from any of the fingers to the pinky H are very low. The common letters that come before H are usually TH, CH, WH, SH, and GH. All of those are on the opposite left hand. Doing the math, 97% of the letters before H are on the left hand. H on the pinky gives an amazing inside roll almost all the time.
Also worth mentioning is I think this also de-emphasizes the pinky a bit. First it limits the ring-to-pinky rolls. Second, the frequency of H is much lower than Colemak’s O. Third, the bigram HH is virtually nill, vs Colemak’s OO which is pretty common. None of these are massive, but I think they add up to de-emphasizing the pinky a bit.

7) Qwerty similarity.

Last but not least, this maintains Qwerty similarity to make it easier to learn.
First, this maintains the bottom row (except N) and maintains Q, W, A, and S.
Second, this “keeps” the ASRT order. People learning Colemak commonly say that R and S are reversed. Technically it’s not because R was never on the middle finger to begin with. But it seems to be reversed in the brain. I think it’s two things. First ASRT keeps it in the same order, with R between S and T, even if R changes fingers. Second, I wonder how many people alt-fingered R typing in Qwerty, either all the time or for the RT bigram.
(I've also wondered if, for example, patterns like the very common ION simply being easier to type would make this faster to learn. I’m not certain about that though.)
Comparing layouts:

• Middlemak: 10 letters stay in their original spot and 16 move, with 5 letters moving on the same finger and 11 letters switching fingers.
  
• Colemak: 10 letters stay in their original spot and 16 move, with 4 letters moving on the same finger and 12 letters switching fingers.
  
• Colemak-DH: 5 letters stay in their original spot and 21 move, with 6 letters moving on the same finger and 15 letters switching fingers.
  
• Workman: 6 letters stay in their original spot and 20 move, with 8 letters moving on the same finger and 12 letters switching fingers.
  

*Not accounting for punctuation changes.
Considering letters swapping on the same finger, this has fewer changes than Colemak. And I think “keeping” the ASRT order will make it even easier to adopt than the numbers suggest.

Downsides

1) There is a slight increase in SFB compared to Colemak. But overall I think SFB is still low.
2) There are a handful of NFB with R: BR, GR, PR, etc. Same with L and BL. This isn’t enough to change the home row to ARST. There is more below but ASRT has lower SFB, lower ring-to-pinky rolls, and is easier to learn. Also note these are much, much lower than the extremely high NFB that Colemak has with HE, LE, etc, (which to solve I paired tL and R).
3) I wanted a less common letter in the G location (Qwerty T location). I really tried to put a less common letter there, but it simply didn’t work.
4) I mention this above but to add more: The OM scissor-gram is a bit more common than I'd like. This is a limitation of keeping Qwerty similarity (the M) and wanting to put O on the upper-row-middle-finger.. (MTGAP has a good design on this. It paired its EO with a lot of punctuation on the index finger to reduce NFBs.)

Conclusion

I'm really happy with this layout. I think it is a great combination of pretty low SFB, good use of the strong middle finger upper room position, reduced NFB and pinballing, keeping Qwerty similarity, and easy to transition to.
Honestly I’m surprised no one else has stumbled on this arrangement, since many people have taken to trying to fix Colemak’s issues and keep Qwerty similarity.

Some nice rolls and combinations

Although this increases alternating, there are still plenty of rolls left. Some I want to mention:
HE is very easy.
LD feels very good.
OU feels very comfortable.
OF/FO is very common and feels very good.
IO and ION are very nice rolls.
This maintains the nice IEN roll.
IOU is a redirect, but overall feels pretty ok.
YOU has a lateral move, but feels comfortable.
OU+LD together to make OULD feels good.
The very common TH, THE, and THE_ feel very nice not having to move the fingers off the home row.

Vs Colemak and Vs Colemak-DH

I made Middlemak to try to solve the problems I found with Colemak. I think all the items came out above, but as a bit of a summary and to compare Colemak-DH. I don’t mean this to be a hit piece on Colemak or Colemak-DH.
Colemak:
1) Colemak has an issue with high lateral movement with D and H. You see this especially with H and its NFBs, but D isn’t the best either.
2) Colemak’s right index finger is overworked. It just has too much to do with N, H, and L.
3) Colemak has a pinballing issue due to placing most vowels and many common consonants on the same hand.
4) Colemak has a ton of NFB with EL/LE and HE.
5) Colemak has a hand balance issue. It places a lot of emphasis on the right hand with the vowels and several common consonants. (Hand dominance and pinballing are technically separate items.)
6) Colemak’s awkward ring-to-pinky rolls. Notably IO, YO (especially when typing YOU), and also RA on the right hand.
7) Not exactly a direct issue with Colemak, but Middlemak is easier to learn coming from Qwerty. Many new users want to ‘keep’ the order of ASRT instead of ‘reversing’ S and R.
Middlemak is a pretty low increase in SFBs and it fixes those issues.
Colemak DH:
Colemak-DH solves 1) the lateral movements. But it doesn’t solve 2) the right index finger overwork, 3) the pinballing, 4) the NFBs 5) the hand balance, 6) the awkward ring-to-pinky rolls, and 7) coming from Qwerty it’s pretty much a full change layout. Shortcuts are accessible, but they move (well paste is the same. Undo, cut, and copy move.)

Option: Swap A and H

This gives:

QWLDG JFOU: HSRTP YNEIA ZXCVB KM,./ 

The good: This is good if you want to put all the vowels and most consonants on opposite hands. Overall this increases alternating. More alternating comes from moving A, it actually decreases alternating with H, mostly because TH is very common. On normal Middlemak, A and H together have 48% of all their bigrams on the opposite hand. With the AH swap, that goes to 59% of bigrams on the opposite hand.
The neutral: On the left hand, the ring-to-pinky outside roll of SH is slightly higher, but in a similar range. Data from Mayzner is SH is 8,889, and original is SA is 6,147.
The bad: On the right hand, the ring-to-pinky outside roll is higher with IA and UA coming in at 8,072 and 3,844, vs original IH and UH of 59 and 30. On both the right and left hand, the all-fingers-to-pinky outside roll is higher.
The complicated: This option decreases the pinky-to-all-fingers rolls inside rolls on both hands. It increases the all-fingers-to-pinky rolls outside rolls. The overall sum of inside rolls plus outside rolls is lower. In short: this lowers inside rolls, increases outside rolls, with an overall sum that is lower.
Basically stick to the original if you want to lower ring-to-pinky movement, lower outside rolls, and easier to learn.
So I can’t recommend this option for most people. I put it as an OK option if someone really wants to increase alternating hands.

Excessive amount of stats

This section has details on distance reductions, hand balance based on frequency, hand balance based on distance, and raw distance numbers. I'm going to give lots of charts. You don’t need to read this all, but these are all the details.

Distance reduction from Qwerty

Keyboard	Distance	Distance	Distance
	(km)	(miles)	(% reduction from Qwerty)
MTGAP	2.592	1.611	46.3%
Workman	2.613	1.624	45.9%
Norman	2.615	1.625	45.9%
Colemak	2.616	1.625	45.8%
Middlemak	2.710	1.684	43.9%
Semimak	2.739	1.702	43.3%
Dvorak	2.813	1.748	41.8%
QWERTY	4.829	3.001	0%

*Source: 62 page / 31k word compilation. Left index used to press Qwerty C location. Includes punctuation.
Middlemak’s distance reduction is good, kost alt keyboards are in the same range. It’s a tad worse on distance than some others because putting O off the home row increases distance, with the O being more frequent. (MTGAP puts O off the home row, but has all the punctuation on the diagonals to decrease distance.)

All Finger Distances

This is the raw distance travelled in metres. Percentages rely on the total distance travelled, which changes between layouts so it's not comparable. Note that distance is different from frequency. The fingers don't move for the home row, the distance travelled is to the other letters.

Keyboard	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky	L Total	R Total	Total
MTGAP	293	116	413	246	667	320	219	318	1068 m	1524 m	2592 m
Workman	123	232	330	753	505	200	146	325	1437 m	1176 m	2613 m
Norman	127	127	223	674	525	377	244	319	1150 m	1465 m	2615 m
Colemak	130	127	116	842	747	200	141	315	1214 m	1402 m	2616 m
Middlemak	134	127	207	847	440	460	189	306	1314 m	1396 m	2710 m
Semimak	238	237	326	390	506	200	450	393	1191 m	1549 m	2739 m
Dvorak	205	54	34	692	753	285	395	396	984 m	1829 m	2813 m
QWERTY	130	127	670	1413	1169	452	450	418	2340 m	2489 m	4829 m

*All distances in metres. Same source as table above.
First, you can see Qwerty’s distances are extremely high. All the alternatives make big reductions.
Notice that the index fingers move way more than the other fingers.
Middlemak balances this out decently and moves more distance to both middle fingers. The left index finger is still a bit high because it has D off the home row (this brings it back to the discussion of where does the 11th letter go). Middlemak’s left index is slightly higher than Colemak’s even though they have the same letters because the row stagger technically puts D on a slight diagonal. A technicality really.
The right index finger is low, a good feature that the index finger next to the vowels doesn’t have to travel much. Compared to Colemak, Middlemak’s movement is 41% lower than Colemaks.

Zoom in on index fingers

Keyboard	Distance	Distance	Total	Index Finger % of total distance
	L Index	R Index	L+R Index	L+R Index
MTGAP	246	667	913 m	35%
Workman	753	505	1259 m	48%
Norman	674	525	1200 m	46%
Colemak	842	747	1589 m	61%
Middlemak	847	440	1287 m	47%
Semimak	390	506	896 m	33%
Dvorak	692	753	1445 m	51%
QWERTY	1413	1169	2582 m	53%

Here you can really see how much work the index fingers do. Middlemak’s sum of the index fingers is nice and low.
You can also look at the percentage of travel with the index finger. Middlemak’s is down at 47% - meaning the index fingers are responsible for 47% of all the distance travelled. On Colemak the index fingers account for 61% the distance travelled, quite a big difference relatively. Workman and Norman are close to Middlemak. MTGAP and Semimak are very low. Dvorak and Qwerty percentages are a bit deceiving though - the higher index finger distances are hidden by the higher total amount of distance.
So even when Middlemak’s left index finger doesn’t have a reduction in distance, the total index finger distance is still significantly reduced.

Hand Balance based on Distance

This is the distance in percentages for all the fingers (unlike the raw distances in metres above) so that I can add them up to look at hand balance.

Keyboard	Left hand	Right hand	Hand Balance Ratio	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky
MTGAP	41.2*	58.8	1.43	11.3	4.5	15.9	9.5	25.7	12.3	8.4	12.3
Workman	55.0	45.0*	0.82	4.7	8.9	12.6	28.8	19.3	7.6	5.6	12.5
Norman	44.0	56.0	1.27	4.8	4.8	8.5	25.8	20.1	14.4	9.3	12.2
Colemak	46.4	53.6*	1.15	5.0	4.8	4.4	32.2	28.6	7.6	5.4	12.0
Middlemak	48.5	51.5*	1.06	5.0	4.7	7.6	31.2	16.3	17.0	7.0	11.3
Semimak	43.5	56.5*	1.30	8.7	8.7	11.9	14.2	18.5	7.3	16.4	14.4
Dvorak	35.0*	65.0	1.86	7.3	1.9	1.2	24.6	26.8	10.1	14.0	14.1
QWERTY	48.5	51.5	1.06	2.7	2.6	13.9	29.3	24.2	9.4	9.3	8.7

Notes: * indicates vowel hand (Norman and Qwerty are mixed enough that there is no clear vowel hand). The ratio between the hands can change quite quickly, so I think that highlights the differences between the layouts better. Ratio >1 indicates right hand dominant, <1 indicates left hand dominant.
I think the raw distances are better for comparison between layouts, but some people like percentages. I used this mostly to calculate hand balance.
You can see Middlemak spreads out the work. The pinkies are low, ring fingers are reasonable, middle takes on more, and the indexes have most of the distances.

Hand Balance based on Frequency

Now we’re shifting to frequency. You can see the data for all the fingers, this is enlightening on its own. I also add them up to look at the hand balance.

Keyboard	Left hand	Right hand	Hand Balance Ratio	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky
MTGAP	49.9*	50.0	1.00	10.8	8.4	18.8	11.9	16.3	14.5	9.8	9.4
Workman	50.7	49.3*	0.97	9.3	10.1	11.1	20.2	15.2	14.9	9.7	9.5
Norman	51.7	48.3	0.93	9.3	8.3	15.4	18.7	15.2	13.2	11.8	8.1
Colemak	46.7	53.2*	1.14	9.4	7.7	8.2	21.4	19.4	14.9	9.1	9.8
Middlemak	48.8	51.2*	1.05	9.4	8.3	9.7	21.4	13.7	19.6	9.9	8.0
Semimak	46.6	53.4*	1.15	9.7	10.1	11.8	15.0	12.2	14.9	15.6	10.7
Dvorak	44.8*	55.3	1.23	10.3	8.1	12.2	14.2	17.2	13.4	13.0	11.7
QWERTY	56.5	43.5	0.77	9.4	8.3	15.4	23.4	18.8	8.4	11.8	4.5

* indicates vowel hand (Norman and Qwerty are mixed enough that there is no vowel hand). Ratio of >1 indicates right hand dominant, <1 indicates left hand dominant.
The ratio between the hands can change quite quickly, so I think that highlights the differences between the layouts better than the percentages. I know it doesn't add up to 100%, the analyzer is doing some rounding or something behind the scenes, I'm not going to round any further.
E is so common that it bumps up any finger it’s on.
Middlemak’s left middle finger takes on a bit more work. The left index is still high because of D. The right index is notably lower taking off common letters. The right middle takes on more of the work with O. And the right pinky is lower too with the H there.

Hand Balance Distance + Frequency

This is the same chart that was above, just for a summary.

Keyboard	Dominant Hand	Hand Balance ratio based on Frequency	Hand Balance ratio based on Distance	Hand Balance ratio based on Frequency and Distance	Corrected comparison number for left hand dominant
MTGAP	Right	1.00	1.43	1.43
Workman	Left	0.97	0.82	0.80	1.26
Norman	Right	0.93	1.27	1.19
Colemak	Right	1.14	1.15	1.32
Middlemak	Right	1.05	1.06	1.11
Semimak	Right	1.15	1.30	1.49
Dvorak	Right	1.23	1.86	2.30
QWERTY	Left	0.77	1.06	0.82	1.22

1 indicates right hand dominant, <1 indicates left hand dominant. Hand Balance ratio based on Frequency and Distance is a simple multiplication of the two.

Frequency-off-home-row (F.O.H.R.)

At the risk of too much information, another way to look at this is the frequency that’s off-home-row. I.e. the frequency that is not under the fingers.
Distances weigh the keys differently, while frequency-off-home-row levels that field. This should be read in conjunction with distance. Neither is better than the other, it’s just a different measurement.

Keyboard	L Pinky	L Ring	L Middle	L Index	R Index	R Middle	R Ring	R Pinky	L+R Index	Index FOHR as % of All fingers FOHR
MTGAP	4.0	2.0	7.2	4.1	10.8	5.8	3.6	3.8	14.9%	36%
Workman	1.5	3.9	5.6	11.5	8.8	3.3	2.5	2.7	20.3%	51%
Norman	1.5	2.1	3.8	10.0	8.8	6.4	4.6	2.6	18.8%	47%
Colemak	1.6	2.1	2.0	12.7	13.0	3.3	2.3	2.6	25.7%	65%
Middlemak	1.6	2.1	4.1	12.7	7.3	8.0	3.1	2.5	20.0%	48%
Semimak	3.5	4.5	5.4	6.3	8.4	3.3	7.8	3.9	14.7%	34%
Dvorak	2.5	0.9	0.6	11.7	11.7	4.7	6.6	5.5	23.4%	53%
QWERTY	1.6	2.1	11.6	21.4	18.6	7.5	7.8	4.5	40.0%	53%

I did this after noticing, on Colemak, that my right index finger moves off the home row a lot. For Colemak the left index finger has more distance, but the right index finger has more frequency-off-home-row. The long distance to B on the left hand skews the distance travelled.
The effect is similar on most of the layouts: FOHR tones down the left hand dominance and bumps up the right hand dominance. With Colemak the frequency of H and L is enough to actually flip them. This is more noticeable with layouts that keep Qwerty’s B (which is just frequent enough to skew things), layouts that change the letter to a less common one shift less.
Also insightful is “Index FOHR as % of All fingers FOHR”. It shows that on Middlemak the index fingers do 48% of all frequency-off-home-row, quite reasonable. Whereas on Colemak the index fingers have 65% of all frequency-off-home-row, very high. Workman is just a tad higher with the index fingers doing 51% of frequency-off-home-row. Dvorak’s and Qwerty’s 53% for index fingers is a bit deceiving though, it appears low but that's because the overall total FOHR is much higher.
This is similar to “Zoom in on the index fingers”, but this shows an even greater difference between Middlemak and Colemak. Recall Middlemaks was 47% and Colemaks was 61% for the index finger % of total distance. Same thing, when the distances to the diagonals are levelled out, the frequency stands out.

Excessive amount of detail on letters

You really don’t need to read this, mostly if you have any lingering or specific Q on specific letters.

D position and why not curl?

Where to place that 11th letter is difficult, it’s either the ring finger or index finger. Many find the ring finger to not be very dexterous for such a frequent letter. So pairing D with T on the index finger works very well for low SFB and to put it on the consonant hand.
Why on the upper-row instead of middle-row? First, I think most people find it easier to reach up and down rather than laterally. Second, it works very well with the LD roll.
Why not the bottom row? I want to leave the bottom row the same as Qwerty for ease of transition and to leave the shortcuts of cut-copy-paste the same. I also think C and M (the curl locations) are already near perfect frequencies for those locations. Those are the middle of the pack frequencies, which are perfect for the index fingers - which cover 6 to 7 letters.
I previously said that I’m not a fan of Colemak’s L position, which is the same upper-row-index-finger. But Colemak’s L has a ton of NFBs with the vowels, especially with E but also with all the other vowels. Looking at the whole hand, 62% of L bigrams are on the same hand. I think that’s the source of discomfort. But with Middlemak’s D, only 22% of bigrams are on the same hand, and 78% are on the opposite hand. So Middlemak has far lower same-hand-gymnastics.

The R and L position

L pairs well with R for low SFB. After finessing with options, the home row can be arranged to ASRT and L can be put above R on the middle-finger-upper-row, a very desirable location.
I’m sure the R and S location will attract attention, so I’ll discuss it more here. Whether to go with ASRT with L above the R, or ARST with L above the S was not the simplest choice.
[Note that Colemak had to go with ARST because of F on the middle-finger-upper-row. RF/FR is more common than SF/FS.]

ASRT with L above R:

The pros of ASRT is it gives fewer SFB, with LRL being less than half as common as SL/LS. It also gives fewer ring-to-pinky-rolls with SA being far less common than Colemak’s RA. It also makes it much easier to learn from Qwerty.
The cons is that the R position leads to more NFB with the index finger. Most of these feel ok but RD, GR, BR are not so good. These aren’t super common, but they are there. PR gives the commonly disliked lateral NFB, but I think it’s relatively ok for the frequency - it’s not super common. RS creates an outside roll, but I’d rather have the RS on Middlemak’s middle-to-ring fingers than Colemak’s RA on the ring-to-pinky fingers. Colemak’s RA is much more common than Middlemak’s RS too.

ARST with L above S

The pros of ARST is that it gives an inside roll of RS. It also separates R from the NFB letters on the index finger, which do add up. However, I’m trying to resolve the NFBs of Colemak’s right hand H and L with E which is a magnitude larger, in addition to the pinballing that it creates which requires moving L.
The cons with ARST is that it gives more SFB. It’s not overly much, but it is there. It also gives a lot of ring-to-pinky outside rolls with RA. And finally this makes it much harder to learn coming from Qwerty.

Verdict: ASRT

None of these are huge deal breakers for one or the other. I went with lower SFB, lower ring-to-pinky rolls, and much easier to learn. At the cost of some NFB on the index finger and inside roll of RS.
(If someone comes from Colemak and wants ARST, they can. It's not a humongous downside to do that. Mostly the more common RA ring-to-pinky roll and slightly more SFB of SD/DS.)
(I hope people don’t think that I’m overly focused on keeping S in the Qwerty location. If it didn’t work, I wouldn’t have kept it. For this layout, it does work better with ASRT. If it didn’t work I would have changed it. For example, I wanted to keep G on its Qwerty location. But it’s better to move it so I didn’t shy away from changing.)

P and G location

It’s better to move G than keep it in its Qwerty location.
P and G are pretty much the same frequency, so it doesn’t matter in that sense.
P has a SFB with T, and it has common bigrams of PR, SP, and even PA. Those work better with P on the middle row. There is PL, but the ones above outweigh it.
G has a bigram with NG. With N on the opposite hand, that means G can go on the worse diagonal spot. There is GR, but PR is far more common.

O location

O and E together give very low SFB. This is a great opportunity to put them together. The strong middle finger only has 3 keys, so it can handle two very frequent letters.

U, Y and F locations

U works pretty well with I, the SFB are reasonably low. This location also creates a nice OU and YOU roll.
Y goes to the index finger, which actually works pretty well. I broke down the numbers in SFB section.
F goes to the right hand index finger, because putting P or G on the right hand gives too many SFB with M and N. The F location works quite well for both moderate frequency and low SFB. It has a lot of NFB with OF/FO, so placing it next to O makes a comfortable roll.