EPA = expected points added per play. Compared with every FBS RB who qualifies
2 seasons since 2014 · box-score totals, snap share estimated from play-by-play
| Season | Team | G | Snaps |
|---|---|---|---|
| 2026 so far | UCLA#16 · 6-0 | 1 | 2% |
| 2025 | UCLA#89 · 3-9 | 1 | 1% |
Past seasons are summed from ESPN box scores; success rates and snap shares come from the play-by-play. Karson Cox's team rank is that season's final TDC rank. A dash means the stat wasn't recorded.
Karson Cox's box score, game by game, from ESPN, with a 0–10 game score for each (hover it for how it works). Click a result for the full game
FBS rank among qualifying players
| Rushing | 2026 | FBS rank |
|---|---|---|
| Carries | 1 | — |
| Yards | 10 | 1037th / 1458 |
| Yards / carry | 10 | — |
| Touchdowns | 0 | — |
| Long | 10 | — |
| Fumbles | 2026 | FBS rank |
|---|---|---|
| Fumbles | 1 | — |
| Lost | 1 | — |
| Recovered | 0 | — |
Estimated snaps from a model trained on real NFL snap counts. Ranges in brackets are where the real number lands 80% of the time: wide for a single game, tighter over the season. How it works
| Game | Team plays | Est. snaps | Plays involved | Snap share |
|---|---|---|---|---|
| Wk 4 · Maryland | 63 | ~10 (2–19) | 1 | 16% |
| Season | 404 | 10 (2–19) | 1 | 2% |
6 RBs on the UCLA roster, ordered by car
| # | Player | Pos | Class | Car | Yds | YPC | TD | Rec | EPA/play |
|---|---|---|---|---|---|---|---|---|---|
| 3 | Wayne Knight | RB | SR | 84 | 571 | 6.8 | 11 | 14 | +0.38 |
| Anthony Woods | RB | 53 | 347 | 6.5 | 5 | 5 | +0.31 | ||
| Jaivian Thomas | RB | 41 | 200 | 4.9 | 3 | 2 | +0.17 | ||
| 32 | Troy Leigber | RB | SR | 5 | 83 | 16.6 | 1 | — | +1.31 |
| 33 | Karson Cox | RB | SO | 1 | 10 | 10 | 0 | — | -4.50 |
| 22 | Dylan Lee | RB | JR | — | — | — | — | — | — |
Stats, EPA and usage come from CollegeFootballData.com and refresh daily. FBS ranks only count players with enough volume for the stat to mean something. Qualifying for EPA takes 6+ plays per team game. UCLA depth chart →