Virtua Fighter 5 REVO · v1.10 · Character Matchup Winrates
Lei Fei vs Taka
Matches counted572
Win rate51.4%
Widest swing-8.0 pts
Survive correction0
Period2026/03/26 – 2026/09/23
Where it was fought
Lei Fei's win rate on each stage, against this matchup everywhere| Stage | Stage effect | Win % | Change | p | Matches |
|---|---|---|---|---|---|
| Breakable full fence | 59.0% | +7.6 | 0.208 | 61 · | |
| Full fence | 59.0% | +7.6 | 0.208 | 61 · | |
| Rectangle | 57.9% | +6.5 | 0.301 | 57 · | |
| Octagon | 56.2% | +4.8 | 0.349 | 80 | |
| Open | 47.1% | -4.3 | 0.516 | 51 · | |
| Fence and open | 46.7% | -4.7 | 0.710 | 15 · | |
| Half fence | 45.6% | -5.8 | 0.357 | 57 · | |
| Low fence | 44.4% | -7.0 | 0.242 | 63 · | |
| Single wall | 43.5% | -7.9 | 0.262 | 46 · | |
| Breakable half fence | 43.4% | -8.0 | 0.221 | 53 · |
p < 0.05 on its ownStill there after correcting for how many were tested
Stage against stage
Lei Fei's win rate on the row's stage, minus the same on the column'sAll ten shapes at once: chi-square 9.54 on 9 df, p 0.389
| Stage | Full | Break. full | Half | Break. half | Low | Wall | Mixed | Open | Rect | Oct |
|---|---|---|---|---|---|---|---|---|---|---|
| Full fence | 59.0%61 | +0.01.000 | +13.40.145 | +15.60.096 | +14.60.105 | +15.50.111 | +12.30.387 | +12.00.206 | +1.10.902 | +2.80.742 |
| Breakable full fence | -0.01.000 | 59.0%61 | +13.40.145 | +15.60.096 | +14.60.105 | +15.50.111 | +12.30.387 | +12.00.206 | +1.10.902 | +2.80.742 |
| Half fence | -13.40.145 | -13.40.145 | 45.6%57 | +2.20.815 | +1.20.898 | +2.10.828 | -1.10.942 | -1.40.881 | -12.30.190 | -10.60.219 |
| Breakable half fence | -15.60.096 | -15.60.096 | -2.20.815 | 43.4%53 | -1.00.910 | -0.10.993 | -3.30.822 | -3.70.708 | -14.50.129 | -12.80.147 |
| Low fence | -14.60.105 | -14.60.105 | -1.20.898 | +1.00.910 | 44.4%63 | +1.00.920 | -2.20.876 | -2.60.781 | -13.40.141 | -11.80.161 |
| Single wall | -15.50.111 | -15.50.111 | -2.10.828 | +0.10.993 | -1.00.920 | 43.5%46 | -3.20.829 | -3.60.724 | -14.40.146 | -12.80.167 |
| Fence and open | -12.30.387 | -12.30.387 | +1.10.942 | +3.30.822 | +2.20.876 | +3.20.829 | 46.7%15 | -0.40.979 | -11.20.436 | -9.60.494 |
| Open | -12.00.206 | -12.00.206 | +1.40.881 | +3.70.708 | +2.60.781 | +3.60.724 | +0.40.979 | 47.1%51 | -10.80.260 | -9.20.304 |
| Rectangle | -1.10.902 | -1.10.902 | +12.30.190 | +14.50.129 | +13.40.141 | +14.40.146 | +11.20.436 | +10.80.260 | 57.9%57 | +1.60.848 |
| Octagon | -2.80.742 | -2.80.742 | +10.60.219 | +12.80.147 | +11.80.161 | +12.80.167 | +9.60.494 | +9.20.304 | -1.60.848 | 56.2%80 |
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 counted403
Win rate52.1%
Widest swing-12.5 pts
Survive correction0
Period2026/03/26 – 2026/09/23
Where it was fought
Lei Fei's win rate on each stage, against this matchup everywhere| Stage | Stage effect | Win % | Change | p | Matches |
|---|---|---|---|---|---|
| Breakable full fence | 64.6% | +12.5 | 0.065 | 48 · | |
| Full fence | 61.1% | +9.0 | 0.257 | 36 · | |
| Octagon | 58.5% | +6.4 | 0.318 | 53 · | |
| Rectangle | 57.6% | +5.5 | 0.512 | 33 · | |
| Single wall | 50.0% | -2.1 | 0.791 | 36 · | |
| Open | 47.4% | -4.7 | 0.539 | 38 · | |
| Breakable half fence | 46.0% | -6.2 | 0.431 | 37 · | |
| Low fence | 44.2% | -7.9 | 0.271 | 43 · | |
| Fence and open | 41.7% | -10.4 | 0.462 | 12 · | |
| Half fence | 39.6% | -12.5 | 0.064 | 48 · |
p < 0.05 on its ownStill there after correcting for how many were tested
Stage against stage
Lei Fei's win rate on the row's stage, minus the same on the column'sAll ten shapes at once: chi-square 11.00 on 9 df, p 0.276
| Stage | Full | Break. full | Half | Break. half | Low | Wall | Mixed | Open | Rect | Oct |
|---|---|---|---|---|---|---|---|---|---|---|
| Full fence | 61.1%36 | -3.50.744 | +21.50.051 | +15.20.194 | +16.90.134 | +11.10.343 | +19.40.240 | +13.70.236 | +3.50.765 | +2.60.805 |
| Breakable full fence | +3.50.744 | 64.6%48 | +25.00.014 | +18.60.086 | +20.40.051 | +14.60.180 | +22.90.147 | +17.20.109 | +7.00.524 | +6.10.530 |
| Half fence | -21.50.051 | -25.00.014 | 39.6%48 | -6.40.556 | -4.60.657 | -10.40.341 | -2.10.895 | -7.80.469 | -18.00.111 | -18.90.058 |
| Breakable half fence | -15.20.194 | -18.60.086 | +6.40.556 | 46.0%37 | +1.80.875 | -4.00.729 | +4.30.796 | -1.40.902 | -11.60.331 | -12.50.240 |
| Low fence | -16.90.134 | -20.40.051 | +4.60.657 | -1.80.875 | 44.2%43 | -5.80.606 | +2.50.876 | -3.20.774 | -13.40.247 | -14.30.163 |
| Single wall | -11.10.343 | -14.60.180 | +10.40.341 | +4.00.729 | +5.80.606 | 50.0%36 | +8.30.617 | +2.60.821 | -7.60.528 | -8.50.429 |
| Fence and open | -19.40.240 | -22.90.147 | +2.10.895 | -4.30.796 | -2.50.876 | -8.30.617 | 41.7%12 | -5.70.730 | -15.90.344 | -16.80.290 |
| Open | -13.70.236 | -17.20.109 | +7.80.469 | +1.40.902 | +3.20.774 | -2.60.821 | +5.70.730 | 47.4%38 | -10.20.390 | -11.10.294 |
| Rectangle | -3.50.765 | -7.00.524 | +18.00.111 | +11.60.331 | +13.40.247 | +7.60.528 | +15.90.344 | +10.20.390 | 57.6%33 | -0.90.933 |
| Octagon | -2.60.805 | -6.10.530 | +18.90.058 | +12.50.240 | +14.30.163 | +8.50.429 | +16.80.290 | +11.10.294 | +0.90.933 | 58.5%53 |
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 counted169
Win rate49.7%
Widest swing-29.7 pts
Survive correction0
Period2026/03/26 – 2026/09/23
Where it was fought
Lei Fei's win rate on each stage, against this matchup everywhere| Stage | Stage effect | Win % | Change | p | Matches |
|---|---|---|---|---|---|
| Half fence | 77.8% | +28.1 | 0.083 | 9 · | |
| Fence and open | 66.7% | +17.0 | 0.553 | 3 · | |
| Rectangle | 58.3% | +8.6 | 0.361 | 24 · | |
| Full fence | 56.0% | +6.3 | 0.495 | 25 · | |
| Octagon | 51.9% | +2.1 | 0.808 | 27 · | |
| Open | 46.1% | -3.5 | 0.790 | 13 · | |
| Low fence | 45.0% | -4.7 | 0.654 | 20 · | |
| Breakable full fence | 38.5% | -11.2 | 0.399 | 13 · | |
| Breakable half fence | 37.5% | -12.2 | 0.305 | 16 · | |
| Single wall | 20.0% | -29.7 | 0.053 | 10 · |
p < 0.05 on its ownStill there after correcting for how many were tested
Stage against stage
Lei Fei's win rate on the row's stage, minus the same on the column'sAll ten shapes at once: chi-square 9.72 on 9 df, p 0.374
| Stage | Full | Break. full | Half | Break. half | Low | Wall | Mixed | Open | Rect | Oct |
|---|---|---|---|---|---|---|---|---|---|---|
| Full fence | 56.0%25 | +17.50.305 | -21.80.249 | +18.50.248 | +11.00.463 | +36.00.053 | -10.70.724 | +9.80.564 | -2.30.869 | +4.20.764 |
| Breakable full fence | -17.50.305 | 38.5%13 | -39.30.069 | +1.00.958 | -6.50.710 | +18.50.340 | -28.20.375 | -7.70.691 | -19.90.248 | -13.40.427 |
| Half fence | +21.80.249 | +39.30.069 | 77.8%9 | +40.30.053 | +32.80.101 | +57.80.012 | +11.10.700 | +31.60.138 | +19.40.301 | +25.90.172 |
| Breakable half fence | -18.50.248 | -1.00.958 | -40.30.053 | 37.5%16 | -7.50.650 | +17.50.347 | -29.20.348 | -8.70.638 | -20.80.197 | -14.30.362 |
| Low fence | -11.00.463 | +6.50.710 | -32.80.101 | +7.50.650 | 45.0%20 | +25.00.180 | -21.70.484 | -1.10.948 | -13.30.378 | -6.80.642 |
| Single wall | -36.00.053 | -18.50.340 | -57.80.012 | -17.50.347 | -25.00.180 | 20.0%10 | -46.70.125 | -26.10.192 | -38.30.041 | -31.90.082 |
| Fence and open | +10.70.724 | +28.20.375 | -11.10.700 | +29.20.348 | +21.70.484 | +46.70.125 | 66.7%3 | +20.50.522 | +8.30.782 | +14.80.626 |
| Open | -9.80.564 | +7.70.691 | -31.60.138 | +8.70.638 | +1.10.948 | +26.10.192 | -20.50.522 | 46.1%13 | -12.20.478 | -5.70.736 |
| Rectangle | +2.30.869 | +19.90.248 | -19.40.301 | +20.80.197 | +13.30.378 | +38.30.041 | -8.30.782 | +12.20.478 | 58.3%24 | +6.50.642 |
| Octagon | -4.20.764 | +13.40.427 | -25.90.172 | +14.30.362 | +6.80.642 | +31.90.082 | -14.80.626 | +5.70.736 | -6.50.642 | 51.9%27 |
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.