Virtua Fighter 5 REVO · v1.10 · Character Matchup Winrates
Pai vs Taka
Matches counted574
Win rate47.0%
Widest swing17.6 pts
Survive correction1
Period2026/03/26 – 2026/09/23
Where it was fought
Pai's win rate on each stage, against this matchup everywhere| Stage | Stage effect | Win % | Change | p | Matches |
|---|---|---|---|---|---|
| Breakable half fence | 64.6% | +17.6 | 0.003 | 65 · | |
| Octagon | 62.2% | +15.1 | 0.057 | 37 · | |
| Rectangle | 51.4% | +4.4 | 0.432 | 70 · | |
| Single wall | 51.2% | +4.2 | 0.578 | 41 · | |
| Low fence | 44.2% | -2.9 | 0.697 | 43 · | |
| Breakable full fence | 43.5% | -3.6 | 0.528 | 69 · | |
| Open | 43.0% | -4.0 | 0.443 | 79 · | |
| Half fence | 41.6% | -5.5 | 0.301 | 77 · | |
| Full fence | 32.6% | -14.4 | 0.041 | 46 · | |
| Fence and open | 32.0% | -15.0 | 0.123 | 25 · |
p < 0.05 on its ownStill there after correcting for how many were tested
Stage against stage
Pai's win rate on the row's stage, minus the same on the column'sAll ten shapes at once: chi-square 20.32 on 9 df, p 0.016
| Stage | Full | Break. full | Half | Break. half | Low | Wall | Mixed | Open | Rect | Oct |
|---|---|---|---|---|---|---|---|---|---|---|
| Full fence | 32.6%46 | -10.90.242 | -8.90.323 | -32.0<0.001 | -11.60.261 | -18.60.079 | +0.60.958 | -10.40.249 | -18.80.046 | -29.60.007 |
| Breakable full fence | +10.90.242 | 43.5%69 | +1.90.815 | -21.10.014 | -0.70.941 | -7.70.431 | +11.50.316 | +0.40.957 | -8.00.348 | -18.70.067 |
| Half fence | +8.90.323 | -1.90.815 | 41.6%77 | -23.10.006 | -2.60.780 | -9.70.315 | +9.60.395 | -1.50.852 | -9.90.231 | -20.60.039 |
| Breakable half fence | +32.0<0.001 | +21.10.014 | +23.10.006 | 64.6%65 | +20.40.036 | +13.40.171 | +32.60.005 | +21.60.010 | +13.20.121 | +2.50.804 |
| Low fence | +11.60.261 | +0.70.941 | +2.60.780 | -20.40.036 | 44.2%43 | -7.00.519 | +12.20.322 | +1.10.903 | -7.20.455 | -18.00.108 |
| Single wall | +18.60.079 | +7.70.431 | +9.70.315 | -13.40.171 | +7.00.519 | 51.2%41 | +19.20.127 | +8.20.394 | -0.20.983 | -10.90.330 |
| Fence and open | -0.60.958 | -11.50.316 | -9.60.395 | -32.60.005 | -12.20.322 | -19.20.127 | 32.0%25 | -11.00.327 | -19.40.094 | -30.20.020 |
| Open | +10.40.249 | -0.40.957 | +1.50.852 | -21.60.010 | -1.10.903 | -8.20.394 | +11.00.327 | 43.0%79 | -8.40.306 | -19.10.055 |
| Rectangle | +18.80.046 | +8.00.348 | +9.90.231 | -13.20.121 | +7.20.455 | +0.20.983 | +19.40.094 | +8.40.306 | 51.4%70 | -10.70.288 |
| Octagon | +29.60.007 | +18.70.067 | +20.60.039 | -2.50.804 | +18.00.108 | +10.90.330 | +30.20.020 | +19.10.055 | +10.70.288 | 62.2%37 |
Better on the row's stage — deeper at 5 and 10 pointsBetter on the column's stagep < 0.05 on its ownStill there after correcting for how many were tested· fewer than 80 decided matches on one side of the comparison
How significance is decided
- Each gap is two win rates compared with a pooled two-proportion z test, over decided matches only.
- The 95% interval is Newcombe's hybrid score method, which keeps its shape near 0% and 100% where a symmetric interval does not.
- q is p corrected for how many tests were asked at once, by Benjamini–Hochberg at a 5% false discovery rate. A figure is called real on q, never on p alone.
- Standard errors are widened by a measured design effect, because one player appears in many matches and those matches are not independent of each other.
- Stages are assigned at random, so a gap between them is the stage's own effect and not a difference in who chose to play where.
Matches counted479
Win rate47.0%
Widest swing19.7 pts
Survive correction0
Period2026/03/26 – 2026/09/23
Where it was fought
Pai's win rate on each stage, against this matchup everywhere| Stage | Stage effect | Win % | Change | p | Matches |
|---|---|---|---|---|---|
| Octagon | 66.7% | +19.7 | 0.026 | 30 · | |
| Breakable half fence | 64.7% | +17.7 | 0.007 | 51 · | |
| Single wall | 50.0% | +3.0 | 0.714 | 34 · | |
| Rectangle | 48.1% | +1.2 | 0.854 | 54 · | |
| Breakable full fence | 47.4% | +0.4 | 0.949 | 57 · | |
| Open | 44.9% | -2.0 | 0.713 | 69 · | |
| Low fence | 40.0% | -7.0 | 0.391 | 35 · | |
| Half fence | 39.1% | -7.9 | 0.173 | 64 · | |
| Full fence | 35.0% | -12.0 | 0.113 | 40 · | |
| Fence and open | 34.8% | -12.2 | 0.230 | 23 · |
p < 0.05 on its ownStill there after correcting for how many were tested
Stage against stage
Pai's win rate on the row's stage, minus the same on the column'sAll ten shapes at once: chi-square 17.35 on 9 df, p 0.044
| Stage | Full | Break. full | Half | Break. half | Low | Wall | Mixed | Open | Rect | Oct |
|---|---|---|---|---|---|---|---|---|---|---|
| Full fence | 35.0%40 | -12.40.225 | -4.10.677 | -29.70.005 | -5.00.655 | -15.00.192 | +0.20.986 | -9.90.310 | -13.20.202 | -31.70.009 |
| Breakable full fence | +12.40.225 | 47.4%57 | +8.30.357 | -17.30.070 | +7.40.490 | -2.60.808 | +12.60.304 | +2.40.784 | -0.80.934 | -19.30.086 |
| Half fence | +4.10.677 | -8.30.357 | 39.1%64 | -25.60.006 | -0.90.927 | -10.90.298 | +4.30.717 | -5.90.494 | -9.10.321 | -27.60.013 |
| Breakable half fence | +29.70.005 | +17.30.070 | +25.60.006 | 64.7%51 | +24.70.024 | +14.70.177 | +29.90.017 | +19.80.032 | +16.60.087 | -2.00.858 |
| Low fence | +5.00.655 | -7.40.490 | +0.90.927 | -24.70.024 | 40.0%35 | -10.00.404 | +5.20.689 | -4.90.632 | -8.20.450 | -26.70.032 |
| Single wall | +15.00.192 | +2.60.808 | +10.90.298 | -14.70.177 | +10.00.404 | 50.0%34 | +15.20.256 | +5.10.627 | +1.90.866 | -16.70.178 |
| Fence and open | -0.20.986 | -12.60.304 | -4.30.717 | -29.90.017 | -5.20.689 | -15.20.256 | 34.8%23 | -10.20.394 | -13.40.280 | -31.90.021 |
| Open | +9.90.310 | -2.40.784 | +5.90.494 | -19.80.032 | +4.90.632 | -5.10.627 | +10.20.394 | 44.9%69 | -3.20.722 | -21.70.047 |
| Rectangle | +13.20.202 | +0.80.934 | +9.10.321 | -16.60.087 | +8.20.450 | -1.90.866 | +13.40.280 | +3.20.722 | 48.1%54 | -18.50.102 |
| Octagon | +31.70.009 | +19.30.086 | +27.60.013 | +2.00.858 | +26.70.032 | +16.70.178 | +31.90.021 | +21.70.047 | +18.50.102 | 66.7%30 |
Better on the row's stage — deeper at 5 and 10 pointsBetter on the column's stagep < 0.05 on its ownStill there after correcting for how many were tested· fewer than 80 decided matches on one side of the comparison
How significance is decided
- Each gap is two win rates compared with a pooled two-proportion z test, over decided matches only.
- The 95% interval is Newcombe's hybrid score method, which keeps its shape near 0% and 100% where a symmetric interval does not.
- q is p corrected for how many tests were asked at once, by Benjamini–Hochberg at a 5% false discovery rate. A figure is called real on q, never on p alone.
- Standard errors are widened by a measured design effect, because one player appears in many matches and those matches are not independent of each other.
- Stages are assigned at random, so a gap between them is the stage's own effect and not a difference in who chose to play where.
Matches counted95
Win rate47.4%
Widest swing-47.4 pts
Survive correction0
Period2026/03/26 – 2026/09/23
Where it was fought
Pai's win rate on each stage, against this matchup everywhere| Stage | Stage effect | Win % | Change | p | Matches |
|---|---|---|---|---|---|
| Breakable half fence | 64.3% | +16.9 | 0.170 | 14 · | |
| Low fence | 62.5% | +15.1 | 0.370 | 8 · | |
| Rectangle | 62.5% | +15.1 | 0.184 | 16 · | |
| Single wall | 57.1% | +9.8 | 0.590 | 7 · | |
| Half fence | 53.9% | +6.5 | 0.615 | 13 · | |
| Octagon | 42.9% | -4.5 | 0.804 | 7 · | |
| Open | 30.0% | -17.4 | 0.245 | 10 · | |
| Breakable full fence | 25.0% | -22.4 | 0.097 | 12 · | |
| Full fence | 16.7% | -30.7 | 0.120 | 6 · | |
| Fence and open | 0.0% | -47.4 | 0.175 | 2 · |
p < 0.05 on its ownStill there after correcting for how many were tested
Stage against stage
Pai's win rate on the row's stage, minus the same on the column'sAll ten shapes at once: chi-square 12.04 on 9 df, p 0.211
| Stage | Full | Break. full | Half | Break. half | Low | Wall | Mixed | Open | Rect | Oct |
|---|---|---|---|---|---|---|---|---|---|---|
| Full fence | 16.7%6 | -8.30.688 | -37.20.127 | -47.60.051 | -45.80.086 | -40.50.135 | +16.70.537 | -13.30.551 | -45.80.056 | -26.20.308 |
| Breakable full fence | +8.30.688 | 25.0%12 | -28.90.141 | -39.30.045 | -37.50.094 | -32.10.161 | +25.00.425 | -5.00.793 | -37.50.049 | -17.90.419 |
| Half fence | +37.20.127 | +28.90.141 | 53.9%13 | -10.40.581 | -8.70.697 | -3.30.888 | +53.90.155 | +23.90.253 | -8.70.638 | +11.00.639 |
| Breakable half fence | +47.60.051 | +39.30.045 | +10.40.581 | 64.3%14 | +1.80.933 | +7.20.751 | +64.30.086 | +34.30.098 | +1.80.919 | +21.40.350 |
| Low fence | +45.80.086 | +37.50.094 | +8.70.697 | -1.80.933 | 62.5%8 | +5.40.833 | +62.50.114 | +32.50.168 | +0.01.000 | +19.60.447 |
| Single wall | +40.50.135 | +32.10.161 | +3.30.888 | -7.20.751 | -5.40.833 | 57.1%7 | +57.10.151 | +27.10.263 | -5.40.809 | +14.30.593 |
| Fence and open | -16.70.537 | -25.00.425 | -53.90.155 | -64.30.086 | -62.50.114 | -57.10.151 | 0.0%2 | -30.00.371 | -62.50.094 | -42.90.257 |
| Open | +13.30.551 | +5.00.793 | -23.90.253 | -34.30.098 | -32.50.168 | -27.10.263 | +30.00.371 | 30.0%10 | -32.50.107 | -12.90.585 |
| Rectangle | +45.80.056 | +37.50.049 | +8.70.638 | -1.80.919 | -0.01.000 | +5.40.809 | +62.50.094 | +32.50.107 | 62.5%16 | +19.60.382 |
| Octagon | +26.20.308 | +17.90.419 | -11.00.639 | -21.40.350 | -19.60.447 | -14.30.593 | +42.90.257 | +12.90.585 | -19.60.382 | 42.9%7 |
Better on the row's stage — deeper at 5 and 10 pointsBetter on the column's stagep < 0.05 on its ownStill there after correcting for how many were tested· fewer than 80 decided matches on one side of the comparison
How significance is decided
- Each gap is two win rates compared with a pooled two-proportion z test, over decided matches only.
- The 95% interval is Newcombe's hybrid score method, which keeps its shape near 0% and 100% where a symmetric interval does not.
- q is p corrected for how many tests were asked at once, by Benjamini–Hochberg at a 5% false discovery rate. A figure is called real on q, never on p alone.
- Standard errors are widened by a measured design effect, because one player appears in many matches and those matches are not independent of each other.
- Stages are assigned at random, so a gap between them is the stage's own effect and not a difference in who chose to play where.