Virtua Fighter 5 REVO · v1.10 · Character Matchup Winrates
Brad vs Lei Fei
Matches counted1,462
Win rate48.1%
Widest swing-8.4 pts
Survive correction0
Period2026/03/26 – 2026/09/23
Where it was fought
Brad's win rate on each stage, against this matchup everywhere| Stage | Stage effect | Win % | Change | p | Matches |
|---|---|---|---|---|---|
| Single wall | 56.3% | +8.2 | 0.040 | 142 | |
| Fence and open | 54.0% | +5.9 | 0.259 | 87 | |
| Rectangle | 52.8% | +4.6 | 0.277 | 127 | |
| Half fence | 51.1% | +3.0 | 0.471 | 133 | |
| Breakable full fence | 50.9% | +2.8 | 0.452 | 165 | |
| Octagon | 46.1% | -2.0 | 0.608 | 141 | |
| Low fence | 46.0% | -2.1 | 0.584 | 152 | |
| Open | 44.9% | -3.2 | 0.405 | 147 | |
| Breakable half fence | 43.7% | -4.4 | 0.248 | 151 | |
| Full fence | 39.7% | -8.4 | 0.029 | 151 |
p < 0.05 on its ownStill there after correcting for how many were tested
Stage against stage
Brad's win rate on the row's stage, minus the same on the column'sAll ten shapes at once: chi-square 13.67 on 9 df, p 0.135
| Stage | Full | Break. full | Half | Break. half | Low | Wall | Mixed | Open | Rect | Oct |
|---|---|---|---|---|---|---|---|---|---|---|
| Full fence | 39.7%151 | -11.20.046 | -11.40.054 | -4.00.484 | -6.30.267 | -16.60.004 | -14.30.033 | -5.20.367 | -13.00.030 | -6.40.272 |
| Breakable full fence | +11.20.046 | 50.9%165 | -0.20.970 | +7.20.200 | +4.90.387 | -5.40.342 | -3.10.638 | +6.00.289 | -1.90.754 | +4.80.401 |
| Half fence | +11.40.054 | +0.20.970 | 51.1%133 | +7.40.211 | +5.10.392 | -5.20.386 | -2.90.674 | +6.20.297 | -1.60.793 | +5.00.405 |
| Breakable half fence | +4.00.484 | -7.20.200 | -7.40.211 | 43.7%151 | -2.30.682 | -12.60.031 | -10.30.125 | -1.20.836 | -9.10.133 | -2.40.681 |
| Low fence | +6.30.267 | -4.90.387 | -5.10.392 | +2.30.682 | 46.0%152 | -10.30.078 | -8.00.236 | +1.10.841 | -6.70.265 | -0.10.994 |
| Single wall | +16.60.004 | +5.40.342 | +5.20.386 | +12.60.031 | +10.30.078 | 56.3%142 | +2.30.732 | +11.40.052 | +3.60.556 | +10.20.085 |
| Fence and open | +14.30.033 | +3.10.638 | +2.90.674 | +10.30.125 | +8.00.236 | -2.30.732 | 54.0%87 | +9.10.177 | +1.30.855 | +7.90.245 |
| Open | +5.20.367 | -6.00.289 | -6.20.297 | +1.20.836 | -1.10.841 | -11.40.052 | -9.10.177 | 44.9%147 | -7.90.194 | -1.20.838 |
| Rectangle | +13.00.030 | +1.90.754 | +1.60.793 | +9.10.133 | +6.70.265 | -3.60.556 | -1.30.855 | +7.90.194 | 52.8%127 | +6.70.276 |
| Octagon | +6.40.272 | -4.80.401 | -5.00.405 | +2.40.681 | +0.10.994 | -10.20.085 | -7.90.245 | +1.20.838 | -6.70.276 | 46.1%141 |
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 counted1,160
Win rate48.8%
Widest swing9.6 pts
Survive correction0
Period2026/03/26 – 2026/09/23
Where it was fought
Brad's win rate on each stage, against this matchup everywhere| Stage | Stage effect | Win % | Change | p | Matches |
|---|---|---|---|---|---|
| Single wall | 58.4% | +9.6 | 0.023 | 125 | |
| Half fence | 57.9% | +9.2 | 0.047 | 107 | |
| Fence and open | 55.2% | +6.4 | 0.278 | 67 · | |
| Rectangle | 51.0% | +2.2 | 0.644 | 102 | |
| Breakable full fence | 49.6% | +0.8 | 0.854 | 121 | |
| Low fence | 47.0% | -1.8 | 0.684 | 117 | |
| Octagon | 44.2% | -4.5 | 0.309 | 113 | |
| Breakable half fence | 44.1% | -4.7 | 0.303 | 111 | |
| Full fence | 41.9% | -6.9 | 0.095 | 129 | |
| Open | 40.7% | -8.1 | 0.070 | 113 |
p < 0.05 on its ownStill there after correcting for how many were tested
Stage against stage
Brad's win rate on the row's stage, minus the same on the column'sAll ten shapes at once: chi-square 17.02 on 9 df, p 0.049
| Stage | Full | Break. full | Half | Break. half | Low | Wall | Mixed | Open | Rect | Oct |
|---|---|---|---|---|---|---|---|---|---|---|
| Full fence | 41.9%129 | -7.70.220 | -16.10.014 | -2.30.722 | -5.20.417 | -16.50.008 | -13.40.075 | +1.10.856 | -9.10.167 | -2.40.708 |
| Breakable full fence | +7.70.220 | 49.6%121 | -8.30.207 | +5.50.407 | +2.60.691 | -8.80.166 | -5.60.459 | +8.90.173 | -1.40.836 | +5.30.414 |
| Half fence | +16.10.014 | +8.30.207 | 57.9%107 | +13.80.042 | +10.90.102 | -0.50.944 | +2.70.724 | +17.20.011 | +7.00.312 | +13.70.042 |
| Breakable half fence | +2.30.722 | -5.50.407 | -13.80.042 | 44.1%111 | -2.90.664 | -14.30.029 | -11.10.152 | +3.40.603 | -6.80.318 | -0.10.988 |
| Low fence | +5.20.417 | -2.60.691 | -10.90.102 | +2.90.664 | 47.0%117 | -11.40.076 | -8.20.284 | +6.30.336 | -4.00.558 | +2.80.674 |
| Single wall | +16.50.008 | +8.80.166 | +0.50.944 | +14.30.029 | +11.40.076 | 58.4%125 | +3.20.672 | +17.70.006 | +7.40.264 | +14.20.029 |
| Fence and open | +13.40.075 | +5.60.459 | -2.70.724 | +11.10.152 | +8.20.284 | -3.20.672 | 55.2%67 | +14.50.059 | +4.20.589 | +11.00.154 |
| Open | -1.10.856 | -8.90.173 | -17.20.011 | -3.40.603 | -6.30.336 | -17.70.006 | -14.50.059 | 40.7%113 | -10.30.131 | -3.50.590 |
| Rectangle | +9.10.167 | +1.40.836 | -7.00.312 | +6.80.318 | +4.00.558 | -7.40.264 | -4.20.589 | +10.30.131 | 51.0%102 | +6.70.324 |
| Octagon | +2.40.708 | -5.30.414 | -13.70.042 | +0.10.988 | -2.80.674 | -14.20.029 | -11.00.154 | +3.50.590 | -6.70.324 | 44.2%113 |
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 counted302
Win rate45.7%
Widest swing-22.6 pts
Survive correction0
Period2026/03/26 – 2026/09/23
Where it was fought
Brad's win rate on each stage, against this matchup everywhere| Stage | Stage effect | Win % | Change | p | Matches |
|---|---|---|---|---|---|
| Rectangle | 60.0% | +14.3 | 0.134 | 25 · | |
| Open | 58.8% | +13.1 | 0.103 | 34 · | |
| Breakable full fence | 54.5% | +8.8 | 0.202 | 44 · | |
| Octagon | 53.6% | +7.9 | 0.380 | 28 · | |
| Fence and open | 50.0% | +4.3 | 0.689 | 20 · | |
| Low fence | 42.9% | -2.8 | 0.720 | 35 · | |
| Breakable half fence | 42.5% | -3.2 | 0.663 | 40 · | |
| Single wall | 41.2% | -4.5 | 0.700 | 17 · | |
| Full fence | 27.3% | -18.4 | 0.072 | 22 · | |
| Half fence | 23.1% | -22.6 | 0.015 | 26 · |
p < 0.05 on its ownStill there after correcting for how many were tested
Stage against stage
Brad's win rate on the row's stage, minus the same on the column'sAll ten shapes at once: chi-square 15.36 on 9 df, p 0.082
| Stage | Full | Break. full | Half | Break. half | Low | Wall | Mixed | Open | Rect | Oct |
|---|---|---|---|---|---|---|---|---|---|---|
| Full fence | 27.3%22 | -27.30.036 | +4.20.738 | -15.20.235 | -15.60.235 | -13.90.361 | -22.70.130 | -31.60.021 | -32.70.024 | -26.30.061 |
| Breakable full fence | +27.30.036 | 54.5%44 | +31.50.010 | +12.10.270 | +11.70.302 | +13.40.349 | +4.50.736 | -4.30.706 | -5.50.660 | +1.00.936 |
| Half fence | -4.20.738 | -31.50.010 | 23.1%26 | -19.40.106 | -19.80.108 | -18.10.206 | -26.90.057 | -35.70.006 | -36.90.007 | -30.50.022 |
| Breakable half fence | +15.20.235 | -12.10.270 | +19.40.106 | 42.5%40 | -0.40.975 | +1.30.926 | -7.50.582 | -16.30.162 | -17.50.170 | -11.10.368 |
| Low fence | +15.60.235 | -11.70.302 | +19.80.108 | +0.40.975 | 42.9%35 | +1.70.908 | -7.10.609 | -16.00.185 | -17.10.190 | -10.70.397 |
| Single wall | +13.90.361 | -13.40.349 | +18.10.206 | -1.30.926 | -1.70.908 | 41.2%17 | -8.80.591 | -17.60.234 | -18.80.231 | -12.40.420 |
| Fence and open | +22.70.130 | -4.50.736 | +26.90.057 | +7.50.582 | +7.10.609 | +8.80.591 | 50.0%20 | -8.80.529 | -10.00.502 | -3.60.807 |
| Open | +31.60.021 | +4.30.706 | +35.70.006 | +16.30.162 | +16.00.185 | +17.60.234 | +8.80.529 | 58.8%34 | -1.20.928 | +5.20.678 |
| Rectangle | +32.70.024 | +5.50.660 | +36.90.007 | +17.50.170 | +17.10.190 | +18.80.231 | +10.00.502 | +1.20.928 | 60.0%25 | +6.40.637 |
| Octagon | +26.30.061 | -1.00.936 | +30.50.022 | +11.10.368 | +10.70.397 | +12.40.420 | +3.60.807 | -5.20.678 | -6.40.637 | 53.6%28 |
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.