| Abû Seïf vs Kazuya | 21–41 | 33.87% |
| Abû Seïf vs Law | 17–35 | 32.69% |
| Abû Seïf vs Hwoarang | 20–32 | 38.46% |
| Abû Seïf vs King | 26–25 | 50.98% |
| Abû Seïf vs Jin | 15–34 | 30.61% |
| Abû Seïf vs Reina | 24–23 | 51.06% |
| Abû Seïf vs Lili | 20–17 | 54.05% |
| Abû Seïf vs Steve | 17–19 | 47.22% |
| Abû Seïf vs Eddy | 14–19 | 42.42% |
| Abû Seïf vs Dragunov | 12–13 | 48.00% |
| Abû Seïf vs Jack-8 | 7–13 | 35.00% |
| Abû Seïf vs Victor | 7–13 | 35.00% |
| Abû Seïf vs Nina | 9–10 | 47.37% |
| Abû Seïf vs Bryan | 6–12 | 33.33% |
| Abû Seïf vs Lee | 7–11 | 38.89% |
| Abû Seïf vs Azucena | 9–9 | 50.00% |
| Abû Seïf vs Lidia | 9–9 | 50.00% |
| Abû Seïf vs Feng | 7–10 | 41.18% |
| Abû Seïf vs Yoshimitsu | 5–11 | 31.25% |
| Abû Seïf vs Asuka | 6–9 | 40.00% |
| Abû Seïf vs Jun | 7–8 | 46.67% |
| Abû Seïf vs Alisa | 7–7 | 50.00% |
| Abû Seïf vs Leo | 3–10 | 23.08% |
| Abû Seïf vs Devil Jin | 4–8 | 33.33% |
| Abû Seïf vs Xiaoyu | 4–7 | 36.36% |
| Abû Seïf vs Lars | 6–5 | 54.55% |
| Abû Seïf vs Kuma | 7–2 | 77.78% |
| Abû Seïf vs Paul | 1–7 | 12.50% |
| Abû Seïf vs Raven | 4–4 | 50.00% |
| Abû Seïf vs Leroy | 4–3 | 57.14% |
| Abû Seïf vs Clive | 1–6 | 14.29% |
| Abû Seïf vs Claudio | 2–4 | 33.33% |
| Abû Seïf vs Shaheen | 1–5 | 16.67% |
| Abû Seïf vs Heihachi | 1–5 | 16.67% |
| Abû Seïf vs Zafina | 1–1 | 50.00% |
| Abû Seïf vs Panda | 0–1 | 0.00% |
Limitations
This data is often requested to give insight into which characters you have more trouble with than others, but it is not particularly helpful for that. The main issue is that it is heavily skewed by how strong the opponents you play are.
For example, this data suggests my worst matchup is clearly vs Reina, but that's just because most of those games are vs Yagami.
There is a way to account for this being worked on. The central idea is to assign each matchup a rating vs you which adjusts based on the result, much like the regular rating but also based on the rating of each player. With this, it would give a better summary of how well you perform vs each character.
In the meantime, this page is here to present the data as requested.