Fantasy Football Regression Analysis: How to Use Regression to the Mean in Fantasy Football
How regression to the mean works in fantasy football: when a player performs far above or far below their expected level — based on their role, usage, and historical baselines — their performance tends to return toward average over subsequent weeks. This is not because the player changes; it is because extreme outcomes involve some element of luck (touchdowns on limited targets, unusually high yards per carry, fumble-free weeks in a fumble-prone player) and luck does not persist at the same rate indefinitely. A WR who scored 3 touchdowns on 4 targets in week 5 is unlikely to repeat that rate in week 6 — their underlying target volume does not support a 75% touchdown rate, and the week-5 result was inflated by scoring luck.
Regression Analysis Quick Reference
| Stat category | Regression speed | What to buy/sell | Mistake to avoid |
|---|---|---|---|
| Touchdowns on low targets | Fast — TDs on fewer than 5 targets per game are the most volatile fantasy stat | Sell high on a WR with 3 TDs in a week they only had 5 targets; buy low on a WR who has 0 TDs with 9 targets/week | Holding a low-target TD scorer as a reliable starter — the production is not sustainable without target volume supporting it |
| High yards per carry (RB) | Moderate — elite RBs sustain above-average YPC; average RBs regress to the mean | Sell high on an average RB at 6.2 YPC in weeks 1–4; buy low on an elite RB at 3.8 YPC in a rough start | Treating early-season YPC as a stable signal for average-talent RBs — regression will pull them toward their true mean |
| Fumble streaks (good or bad) | Fast — fumble rate is highly variable week to week | Buy low on a fumble-prone player who has lost 3 fumbles in 3 weeks; sell before a fumble-free streak ends | Rostering a fumble-prone player assuming the fumbles will stop — sell while others are buying the fumble-free weeks |
| Field goal percentage (kickers) | Moderate — elite kickers sustain 85%+; average kickers regress from early hot streaks | Sell a kicker who is 95% through 4 weeks; buy a kicker who is 55% despite consistent opportunities | Dropping a kicker hitting 55% in week 5 — they may regress positively to their true 80% mean by week 10 |
| Catch rate above expectation | Moderate — elite WRs sustain high catch rates; average WRs regress from inflated early rates | Sell a 90% catch rate WR on 6 targets/game — unsustainable; normal WR2 catch rate is 65–72% | Treating a 90% early-season catch rate as a real skill signal for a non-elite WR |
| QB completion percentage spikes | Moderate — elite QBs sustain 68%+; average QBs regress from above-75% early-season rates | Be cautious rostering players in a game-manager QB’s offense after a 78% completion week — production was inflated | Trading major assets for a non-elite QB’s receivers after a 78% completion, 4-TD game |
Stats That Regress Most in Fantasy Football
The most regression-prone fantasy statistics. High-regression stats in order: (1) touchdowns per target (WR/TE) — the most volatile stat; a WR scoring touchdowns on 30% of their targets will regress; normal touchdown rate is 8–12% of targets; (2) yards per carry (RB) — explosive plays inflate season-early YPC; players with fewer than 100 carries in a sample have highly unstable YPC; (3) red zone touchdown percentage — teams and players who convert red zone opportunities at 50%+ early in the season will regress toward the 30–35% historical average; and (4) yards per reception for players with low routes run — explosive plays on limited routes produce inflated per-reception averages that will normalize as the target volume grows.
Stats that are more stable. Low-regression fantasy stats: (1) target share — how much of the passing attack flows through a specific WR or TE is the most stable week-to-week metric; (2) snap percentage — players on the field more consistently produce more consistently; (3) routes run per game — a consistent route tree indicates a consistent offensive role; and (4) carries per game — a featured back’s carry load is more stable than efficiency metrics.
Using Regression for Trade and Waiver Decisions
How to act on regression signals. Practical regression applications: (1) buy low on a player whose underlying metrics (targets, snaps, routes run) are strong but whose fantasy scores have been suppressed by touchdown drought — these players are likely to score more as the season progresses; (2) sell high on a player whose fantasy scores are inflated by unsustainable touchdowns on limited target volume — their high output has inflated their trade value above their true worth; and (3) wait out a slump for proven players — a WR with an established 22% target share who posts two mediocre weeks has not changed; the short sample variance will normalize.
For how regression analysis connects to overall fantasy strategy, see: Fantasy Football Waiver Wire Strategy: How to Win the Waiver Wire in Fantasy Football.