Tuesday, September 22, 2026
Full Seasons Make 2025 Fantasy Scoring More Useful
Weighting last season’s scoring by games played separates durable evidence from averages built on small samples.
Bobby's AI Desk·written from our own draft-market and 2025 production data·how this works
Takeaways
- → Discount nine- and ten-game scoring samples until the ranges fit your roster risk.
- → Give 16- and 17-game evidence more weight, but keep the full range visible.
- → Treat rookie ADP as projection, not NFL evidence.
In this piece
Last season’s fantasy averages are evidence, but they are not all equally useful. To weight them honestly, take each player’s 2025 per-game scoring, regress it toward his positional average based on games played, and keep a range around the result. This is not a 2026 projection or forecast; it is 2025 evidence adjusted for how much of it we actually saw.
Small samples need the hardest pull
Omarion Hampton averaged 15.08 PPR points across 9 games last season. That is a strong average from a limited sample, so the weighted center falls well below 15.08 and carries a 6.06-18.57 range. His 2.09 ADP — average draft position, or where a player is typically selected — therefore requires more projection than the raw average suggests.
Christian Watson presents a similar problem. He scored 13.24 points per game over 10 games, producing a weighted evidence range of 4.86-16.99. His weekly results also show boom/bust scoring, meaning frequent high and low outcomes: 30% of his games reached 20 points, while 40% finished under eight. The range matters more than treating 13.24 as a stable expectation.
Full seasons preserve more evidence
A full season does not make an average repeatable, but it gives the method more information to retain. Caleb Williams played 17 games and averaged 18.75 points, leaving a weighted range of 10.68-24.83. Compare that with Hampton: the method gives Williams’ average more weight because it was observed nearly twice as often.
The same principle applies at the top of the board. Jaxon Smith-Njigba averaged 21.17 points over 17 games, with an 11.35-24.22 range. Trey McBride produced 18.58 across 17 games, with an 8.39-22.78 range. Their full seasons do not guarantee another one, but they provide firmer evidence than a nine- or ten-game average. Both belong in the Draft Kit as ranges, not fixed scoring promises.
Price can compensate for uncertainty
Value means paying a 2026 draft price below what 2025 production supports. Stefon Diggs, at 8.02, and Deebo Samuel Sr., at 9.12, are each being drafted 7 picks later than their 2025 production justifies.
That discount does not erase variance. Diggs averaged 12.37 points over 17 games, but his weighted range remains 3.67-18.31; he scored 20-plus in 18% of games and fewer than eight in 41%. Deebo averaged 11.76 over 16 games, with a 3.06-17.86 range; his corresponding rates were 25% and 44%. Those prices can be reasonable while the weekly experience remains uneven. Use the Usage Report to check whether later evidence supports the original draft case.
Rookies are a separate evidence class
Jeremiyah Love and Jadarian Price each logged 0 NFL games in 2025. Their respective prices of 3.04 and 6.05 are projection prices, not conclusions supported by an NFL scoring sample. Love’s playoff matchup score is 67, but schedule context cannot replace missing production evidence.
How to use this: Draft from ranges rather than last season’s headline averages. Favor larger samples when prices are close, demand discounts for volatile profiles, and treat rookies separately. Use the Anytime TD Board only as another decision input, not a substitute for sample-aware scoring evidence.
Get the ADP movers by email
One email a week through draft season: who's climbing, who's falling, and which players the market is pricing furthest from what they actually produced last year. Free — no account needed.
Weekly through Week 1. No spam, unsubscribe any time.