Virtua Fighter 5 REVO · v1.10 · Character Matchup Winrates
Aoi vs Lei Fei
Matches counted1,328
Win rate51.7%
Widest swing-8.2 pts
Survive correction0
Period2026/03/26 – 2026/09/23
Where it was fought
Aoi's win rate on each stage, against this matchup everywhere| Stage | Stage effect | Win % | Change | p | Matches |
|---|---|---|---|---|---|
| Octagon | 56.8% | +5.1 | 0.233 | 125 | |
| Rectangle | 54.7% | +3.0 | 0.501 | 117 | |
| Open | 54.5% | +2.7 | 0.523 | 123 | |
| Breakable half fence | 54.2% | +2.5 | 0.529 | 142 | |
| Single wall | 54.0% | +2.3 | 0.572 | 137 | |
| Breakable full fence | 52.1% | +0.4 | 0.923 | 142 | |
| Half fence | 51.6% | -0.1 | 0.973 | 126 | |
| Full fence | 48.3% | -3.5 | 0.436 | 116 | |
| Low fence | 46.4% | -5.3 | 0.141 | 166 | |
| Fence and open | 43.5% | -8.2 | 0.159 | 69 · |
p < 0.05 on its ownStill there after correcting for how many were tested
Stage against stage
Aoi's win rate on the row's stage, minus the same on the column'sAll ten shapes at once: chi-square 7.05 on 9 df, p 0.632
| Stage | Full | Break. full | Half | Break. half | Low | Wall | Mixed | Open | Rect | Oct |
|---|---|---|---|---|---|---|---|---|---|---|
| Full fence | 48.3%116 | -3.80.540 | -3.30.607 | -6.00.341 | +1.90.754 | -5.70.363 | +4.80.527 | -6.20.338 | -6.40.327 | -8.50.185 |
| Breakable full fence | +3.80.540 | 52.1%142 | +0.50.932 | -2.10.721 | +5.70.316 | -1.90.750 | +8.60.239 | -2.40.701 | -2.60.678 | -4.70.443 |
| Half fence | +3.30.607 | -0.50.932 | 51.6%126 | -2.60.666 | +5.20.378 | -2.40.694 | +8.10.279 | -2.90.648 | -3.10.627 | -5.20.407 |
| Breakable half fence | +6.00.341 | +2.10.721 | +2.60.666 | 54.2%142 | +7.80.170 | +0.20.972 | +10.80.143 | -0.20.968 | -0.50.939 | -2.60.673 |
| Low fence | -1.90.754 | -5.70.316 | -5.20.378 | -7.80.170 | 46.4%166 | -7.60.186 | +2.90.684 | -8.10.174 | -8.30.168 | -10.40.079 |
| Single wall | +5.70.363 | +1.90.750 | +2.40.694 | -0.20.972 | +7.60.186 | 54.0%137 | +10.50.153 | -0.50.941 | -0.70.913 | -2.80.651 |
| Fence and open | -4.80.527 | -8.60.239 | -8.10.279 | -10.80.143 | -2.90.684 | -10.50.153 | 43.5%69 | -11.00.144 | -11.20.139 | -13.30.075 |
| Open | +6.20.338 | +2.40.701 | +2.90.648 | +0.20.968 | +8.10.174 | +0.50.941 | +11.00.144 | 54.5%123 | -0.20.972 | -2.30.712 |
| Rectangle | +6.40.327 | +2.60.678 | +3.10.627 | +0.50.939 | +8.30.168 | +0.70.913 | +11.20.139 | +0.20.972 | 54.7%117 | -2.10.742 |
| Octagon | +8.50.185 | +4.70.443 | +5.20.407 | +2.60.673 | +10.40.079 | +2.80.651 | +13.30.075 | +2.30.712 | +2.10.742 | 56.8%125 |
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 counted1,169
Win rate52.5%
Widest swing-13.2 pts
Survive correction0
Period2026/03/26 – 2026/09/23
Where it was fought
Aoi's win rate on each stage, against this matchup everywhere| Stage | Stage effect | Win % | Change | p | Matches |
|---|---|---|---|---|---|
| Octagon | 58.2% | +5.7 | 0.212 | 110 | |
| Single wall | 55.8% | +3.3 | 0.443 | 120 | |
| Open | 55.6% | +3.1 | 0.479 | 115 | |
| Rectangle | 55.3% | +2.8 | 0.549 | 103 | |
| Breakable half fence | 54.1% | +1.6 | 0.701 | 135 | |
| Half fence | 52.4% | -0.1 | 0.976 | 105 | |
| Breakable full fence | 51.6% | -0.9 | 0.830 | 124 | |
| Full fence | 50.0% | -2.5 | 0.589 | 104 | |
| Low fence | 48.2% | -4.3 | 0.275 | 143 | |
| Fence and open | 39.3% | -13.2 | 0.042 | 56 · |
p < 0.05 on its ownStill there after correcting for how many were tested
Stage against stage
Aoi's win rate on the row's stage, minus the same on the column'sAll ten shapes at once: chi-square 8.13 on 9 df, p 0.521
| Stage | Full | Break. full | Half | Break. half | Low | Wall | Mixed | Open | Rect | Oct |
|---|---|---|---|---|---|---|---|---|---|---|
| Full fence | 50.0%104 | -1.60.808 | -2.40.731 | -4.10.532 | +1.80.786 | -5.80.383 | +10.70.195 | -5.70.403 | -5.30.442 | -8.20.230 |
| Breakable full fence | +1.60.808 | 51.6%124 | -0.80.908 | -2.50.692 | +3.40.584 | -4.20.509 | +12.30.125 | -4.00.532 | -3.70.575 | -6.60.314 |
| Half fence | +2.40.731 | +0.80.908 | 52.4%105 | -1.70.794 | +4.10.520 | -3.50.604 | +13.10.113 | -3.30.627 | -3.00.669 | -5.80.392 |
| Breakable half fence | +4.10.532 | +2.50.692 | +1.70.794 | 54.1%135 | +5.80.332 | -1.80.778 | +14.80.063 | -1.60.803 | -1.30.846 | -4.10.519 |
| Low fence | -1.80.786 | -3.40.584 | -4.10.520 | -5.80.332 | 48.2%143 | -7.60.220 | +9.00.254 | -7.40.237 | -7.10.273 | -9.90.117 |
| Single wall | +5.80.383 | +4.20.509 | +3.50.604 | +1.80.778 | +7.60.220 | 55.8%120 | +16.50.041 | +0.20.978 | +0.50.941 | -2.40.719 |
| Fence and open | -10.70.195 | -12.30.125 | -13.10.113 | -14.80.063 | -9.00.254 | -16.50.041 | 39.3%56 | -16.40.045 | -16.10.053 | -18.90.021 |
| Open | +5.70.403 | +4.00.532 | +3.30.627 | +1.60.803 | +7.40.237 | -0.20.978 | +16.40.045 | 55.6%115 | +0.30.963 | -2.50.702 |
| Rectangle | +5.30.442 | +3.70.575 | +3.00.669 | +1.30.846 | +7.10.273 | -0.50.941 | +16.10.053 | -0.30.963 | 55.3%103 | -2.80.676 |
| Octagon | +8.20.230 | +6.60.314 | +5.80.392 | +4.10.519 | +9.90.117 | +2.40.719 | +18.90.021 | +2.50.702 | +2.80.676 | 58.2%110 |
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 counted159
Win rate45.9%
Widest swing15.6 pts
Survive correction0
Period2026/03/26 – 2026/09/23
Where it was fought
Aoi's win rate on each stage, against this matchup everywhere| Stage | Stage effect | Win % | Change | p | Matches |
|---|---|---|---|---|---|
| Fence and open | 61.5% | +15.6 | 0.238 | 13 · | |
| Breakable half fence | 57.1% | +11.2 | 0.542 | 7 · | |
| Breakable full fence | 55.6% | +9.7 | 0.383 | 18 · | |
| Rectangle | 50.0% | +4.1 | 0.748 | 14 · | |
| Half fence | 47.6% | +1.7 | 0.866 | 21 · | |
| Octagon | 46.7% | +0.8 | 0.951 | 15 · | |
| Single wall | 41.2% | -4.7 | 0.678 | 17 · | |
| Open | 37.5% | -8.4 | 0.624 | 8 · | |
| Low fence | 34.8% | -11.1 | 0.247 | 23 · | |
| Full fence | 33.3% | -12.6 | 0.363 | 12 · |
p < 0.05 on its ownStill there after correcting for how many were tested
Stage against stage
Aoi's win rate on the row's stage, minus the same on the column'sAll ten shapes at once: chi-square 4.72 on 9 df, p 0.858
| Stage | Full | Break. full | Half | Break. half | Low | Wall | Mixed | Open | Rect | Oct |
|---|---|---|---|---|---|---|---|---|---|---|
| Full fence | 33.3%12 | -22.20.232 | -14.30.424 | -23.80.311 | -1.40.932 | -7.80.668 | -28.20.158 | -4.20.848 | -16.70.391 | -13.30.484 |
| Breakable full fence | +22.20.232 | 55.6%18 | +7.90.621 | -1.60.943 | +20.80.183 | +14.40.395 | -6.00.739 | +18.10.395 | +5.60.755 | +8.90.611 |
| Half fence | +14.30.424 | -7.90.621 | 47.6%21 | -9.50.663 | +12.80.387 | +6.40.691 | -13.90.429 | +10.10.624 | -2.40.890 | +0.90.955 |
| Breakable half fence | +23.80.311 | +1.60.943 | +9.50.663 | 57.1%7 | +22.40.290 | +16.00.476 | -4.40.848 | +19.60.447 | +7.10.757 | +10.50.647 |
| Low fence | +1.40.932 | -20.80.183 | -12.80.387 | -22.40.290 | 34.8%23 | -6.40.680 | -26.80.121 | -2.70.890 | -15.20.361 | -11.90.464 |
| Single wall | +7.80.668 | -14.40.395 | -6.40.691 | -16.00.476 | +6.40.680 | 41.2%17 | -20.40.269 | +3.70.861 | -8.80.623 | -5.50.755 |
| Fence and open | +28.20.158 | +6.00.739 | +13.90.429 | +4.40.848 | +26.80.121 | +20.40.269 | 61.5%13 | +24.00.284 | +11.50.547 | +14.90.431 |
| Open | +4.20.848 | -18.10.395 | -10.10.624 | -19.60.447 | +2.70.890 | -3.70.861 | -24.00.284 | 37.5%8 | -12.50.571 | -9.20.673 |
| Rectangle | +16.70.391 | -5.60.755 | +2.40.890 | -7.10.757 | +15.20.361 | +8.80.623 | -11.50.547 | +12.50.571 | 50.0%14 | +3.30.858 |
| Octagon | +13.30.484 | -8.90.611 | -0.90.955 | -10.50.647 | +11.90.464 | +5.50.755 | -14.90.431 | +9.20.673 | -3.30.858 | 46.7%15 |
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.