NEO! VOD

Melee Player Rankings

The current top Super Smash Bros. Melee players, rated from every singles bracket set on Liquipedia through November 15, 2026. Cody Schwab is #1, followed by Zain and Hungrybox. 114 players are active enough to be ranked.

Rating history

Each point is the ranking at the start of that month, using only sets played before it. A gap means the player was not active enough to be ranked. Ratings are relative to the top 20 of the time: each month is shifted so its top-20 average matches today's. Earlier years: all-time rating history.

Current ranking

Rating is on the Elo scale: a 200-point gap means the higher-rated player wins about 76% of sets. NEO! score is the player's average chance of beating a current top-50 player in a set, from a second model that allows for player-against-player matchups. Rank range is the 90% interval of the rank. Change is since the last recorded tournament, Port Priority DX (November 14 to 15, 2026).

Show on the chart#PlayerRating90% intervalRank rangeChangeNEO! scoreSets (12 mo)Events (12 mo)
128212766–2876190.4576
227362682–2789284.58511
326642625–2704379.811716
425842527–26414–1072.18011
525762529–26234–971.5639
625612510–26134–1170.0416
725612517–26054–1070.07612
825372489–25855–13+166.98114
925352470–26014–15−166.3325
1025162459–25736–16+363.9467
1125162480–25517–14−164.310617
1224902453–252610–18−161.48213
1324892438–25419–19−160.9498
1424782424–25339–2160.612820
1524742412–25359–2359.9599
1624652420–251012–2258.59315
1724532400–250512–2357.15310
1824492391–250611–2556.75210
1924342383–248415–2654.3255
2024342388–247914–2554.57313
2124212352–249014–3152.7509
2224192345–249314–32+352.2193
2324122372–245218–28−151.8275
2424002348–245217–29−150.1235
2523912349–243420–31−149.06210
2623812217–254610–53new47.4164
2723712320–242322–35−146.95510
2823652234–249612–42−145.9285
2923592296–242322–38−144.6245
3023562302–241023–37−144.15611
3123392274–240325–40new42.2173
3223382295–238127–38−242.0306
3323322287–237628–40−241.0274
3423282254–240125–44−241.0163
3523262275–237728–41−241.3438
3623222268–237527–42−240.0336
3723162238–239325–47−239.1234
3822982225–237128–49−237.1163
3922762196–235731–56−134.5183
4022632208–231935–54−332.9306
4122602199–232235–56−132.5457
4222482177–231836–6030.9469
4322482207–228840–5431.15411
4422472212–228240–5230.58416
4522452140–235027–63−430.6295
4622372165–230836–60−129.75310
4722362158–231535–64−829.7317
4822292154–230537–64+129.1387
4922242099–234932–75−128.3214
5022212147–229438–64−327.9214
5122092135–228340–69−127.2163
5222062124–228936–69−627.0749
5322032147–225943–65new26.5184
5421992043–235532–86−326.2296
5521952059–233131–81−325.9183
5621892096–228141–7925.34410
5721842107–226144–78new24.7235
5821832125–224047–70−524.2388
5921792113–224447–73−424.1175
6021692095–224347–79−623.1317
6121692089–224844–81−423.2264
6221662099–223248–81+422.9245
6321632088–223846–78−422.6215
6421532105–220254–76−421.8153
6521532071–223648–87+421.9337
6621442039–224946–92−420.8255
6721432076–221052–83−920.64410
6821382052–222448–87−320.1539
6921352092–217860–81−820.1449
7021352058–221153–86−320.0236
7121322092–217260–81−719.74710
7221312082–218158–81−919.94510
7321312086–217561–81−319.85011
7421282042–221451–90−619.7266
7521062042–217160–90−417.8388
7621002048–215365–88−317.3216
7721002030–217060–95−517.3246
7820962016–217662–96−316.9338
7920942021–216660–94−316.8214
8020932001–218553–94−616.7535
8120912038–214368–92−316.5369
8220902026–215461–91−516.5214
8320862030–214266–93−116.26415
8420831968–219756–102−416.0358
8520811997–216463–99−615.7245
8620762025–212671–93−515.4399
8720662005–212674–97−414.7368
8820572005–210978–99−414.23712
8920381980–209578–101−413.04711
9020351961–211075–10312.8306
9120341952–211775–107−312.7154
9220301956–210480–105−512.5317
9320301944–211574–107−712.4205
9420231913–213470–110new12.1267
9520201931–211075–107−411.9255
9620181900–213770–111−711.8174
9720171925–210976–109−511.7174
9820111957–206686–106−511.45714
9919971939–205487–107−410.64210
10019931921–206588–110−410.4214
10119861907–206686–111−410.1359
10219851920–205189–110−410.0248
10319841891–207785–112−49.9266
10419831883–208380–112−109.9208
10519621820–210379–114new8.9256
10619591887–203194–112−68.73510
10719591863–205588–113−18.7438
10819591908–200996–111−78.8309
10919461858–203592–113−68.2257
11019451862–202794–113−88.1288
11119381811–206491–114−77.8245
11219201848–1991100–114−57.2268
11318911775–200795–114new6.2197
11418671782–1951105–114−65.4299

How the ranking works

Data. Every singles bracket set on Liquipedia from 2022 on with both players recorded: 113,949 sets among 14,105 players, through November 15, 2026. Names are matched case-insensitively, with a few clear renames merged.

Rating. A Bradley-Terry model: each player has one number, and the chance that player A beats player B in a set depends only on the gap between them. It is fitted to all sets at once, with older sets weighted by 0.5 per 365 days of age, and a Gaussian prior that pulls players with few sets toward the average. The scale is converted to Elo points (1500 + 173.7 × the model's logit). The intervals come from 200 resamples of the sets, each refitted: a 90% interval is the estimate plus or minus 1.645 standard deviations across the resamples, and the rank range runs from the 5th to the 95th percentile of the rank.

NEO! score. The same data fitted with multidimensional Elo (mElo; Balduzzi et al. 2018), which adds 3 pairs of cyclic terms to each player so the model can hold rock-paper-scissors patterns that a single rating cannot: A beats B, B beats C, C beats A. NEO! score is the model's chance that the player beats a player from the current top 50 by Rating, averaged over those 50.

Who is ranked. Players with at least 15 sets at 3 or more tournaments in the last 365 days, and at least 8 of those sets against the provisional top 100: 114 players. Highly rated players outside this rule right now: aMSa (8 sets at 1 events, last played 2025-12); Leffen (0 sets at 0 events, last played 2024-11); iBDW (0 sets at 0 events, last played 2023-10); Mang0 (0 sets at 0 events, last played 2025-06); Plup (0 sets at 0 events, last played 2025-06); Nicki (16 sets at 2 events, last played 2026-04); Zion (0 sets at 0 events, last played 2023-05); Hax$ (0 sets at 0 events, last played 2024-12).

Why these models. Each candidate predicted every month's sets from 2024-01 to 2026-11 using only results from before that month, scored on the 25,438 sets where both players had at least 5 earlier sets (log loss: lower is better), and settings were tuned per model. Bradley-Terry and mElo came out ahead of Elo, Glicko-2 and TrueSkill; Bradley-Terry orders the ranking because it is a single number per player and predicts almost as well as the best mElo. The last column scores only sets between two top-100 players (3,267 sets).

Model comparison (held-out log loss)
ModelSettingsLog lossvs bestTop-100 sets
mElo, 3 cyclic pairs (NEO! score)half-life=365d sigma_r=3 sigma_c=0.50.4748+0.00000.5705
mElo, 2 cyclic pairshalf-life=365d sigma_r=3 sigma_c=0.50.4749+0.00010.5712
mElo, 1 cyclic pairhalf-life=365d sigma_r=3 sigma_c=0.50.4753+0.00050.5732
Bradley-Terry (Rating)half-life=365d sigma_r=30.4755+0.00080.5739
TrueSkillbeta=4.1667 tau=0.40.4837+0.00890.5782
Glicko-2tau=1.2 RD0=350 period=7d0.4944+0.01960.5754
EloK=800.5017+0.02690.5954
coin flip0.6931+0.21840.6931