Browse every forecast field · Fantasy research & full exports

Scheduled games272

Tested decisive games569

Backtest winner accuracy66.4%

Elo 63.4% / Market 68.4%

Verified market matches3 / bounded capture

WeekFind a team

16 games

WEEK 1 · HOME FIELD

NE at SEA

Sep 9, 8:20 PM ET

NE31.7%

WIN

SEA68.3%

Conditional on a decisive result; ties are excluded.

Expected scoreNE 19.1
SEA 24.5

Home margin+5.4

Total points43.7

80% forecast ranges

Home scoring margin-12.6 to 23.5

-40Estimate 5.440

Combined points27.3 to 60.0

0Estimate 43.790

Ranges use held-out 2023 errors. Historical coverage appears in Backtests; future coverage can differ.

What goes into this forecast

Earlier completed weeks inform team strength and scoring form. These are input values, not causal explanations.

Elo difference (home − away): 45.5
Rest days (home / away): 21 / 21
Scoring form difference: +4.4

Injuries, starting quarterbacks and roster changes are not yet included.

MARKET OBSERVATION

No verified market match in this capture

The bounded sample is not a complete market inventory. Missing observations do not imply zero probability or liquidity.

Sources, methodology and limitations

Existing Sleeper, ESPN and Yahoo connections preserved; this frozen game experiment does not yet use their live fields.

Schedule captured Sep 4, 9:24 PM ET from a pinned nflverse source. Historical rows were reconstructed from later-retrieved final results.

Initial training: 2000-2017
Development: 2018-2022 rolling annual folds
Selected probability model: logistic:5
Evaluation: 2024-2025 annual refit using earlier seasons only; 2024 results enter 2025 training without reselection

No claim of best-in-market performance or public model promotion.
Historical results and odds may have been corrected after the historical cutoff; capture hashes do not establish historical availability.
Market odds are retrospective benchmarks only and never enter model features.
Same-week results cannot inform other games in that week, including later kickoff games.
No injury, starting quarterback, player availability, roster turnover or live prediction-market features in this game model.
Intervals transfer 2023 residuals to later seasons and do not guarantee conditional or future coverage.
Season-week bootstrap assumes blocks independent; between-week and between-season dependence can understate uncertainty.
Season simulations assume independent decisive outcomes and fixed team strength; no playoff or tie-breaker simulation.
All weeks in this preseason snapshot use decayed 2025 form and zero current-season completed-game counts. Late-season feature combinations differ from in-season backtests; they are preseason scenarios until refreshed from newly completed games.
Retrospective research: 2019-2025 outcomes were already inspected before this experiment; no pristine confirmation set.
Source snapshots are later-retrieved nflverse history, with assumed season-complete feature timestamps, not archived publication-as-of receipts.
Original football is an unchanged historical comparator containing recent_team scoring attribution; the nested candidate excludes that index and constant age.
Full-PPR season totals train directly on explicit dataset totals; missing labels are excluded, never filled with zero. Existing dataset zero-target semantics are inherited.
Prior-season eligible-player cohort excludes rookies and other players without qualifying prior production; results do not cover every draftable player.
No historical Sleeper, ESPN, Yahoo consensus or prediction-market probabilities are synthesized or used in these historical folds.
Earlier outer residuals give out-of-sample temporal conformal-style intervals; dependence and distribution shift mean nominal coverage is not guaranteed.
Player-block bootstrap preserves repeated players but not all shared season/team shocks; intervals describe conditional retrospective uncertainty.
Public ranking influence and promotion are false; this experiment cannot establish market leadership.

Game report: a1276c623e927ab248e61c7fb73cc0b1a1d919dee602d8bd468188fd18fd3699
Player report: 9fb75caa6a26e48973092102a4b3760fada59eb211151b9153ccf62da2edf340
Source: 08447bfca7733c18e969b2675e0f2220f3d76ee1cb29d8597b4eb004a6a10735

Loading forecast evidence…

NE at SEA

Sep 9, 8:20 PM ET

NE31.7%

WIN

SEA68.3%

Conditional on a decisive result; ties are excluded.

Expected scoreNE 19.1
SEA 24.5

Home margin+5.4

Total points43.7

80% forecast ranges

Home scoring margin-12.6 to 23.5

-40Estimate 5.440

Combined points27.3 to 60.0

0Estimate 43.790

Ranges use held-out 2023 errors. Historical coverage appears in Backtests; future coverage can differ.

What goes into this forecast

Earlier completed weeks inform team strength and scoring form. These are input values, not causal explanations.

Elo difference (home − away): 45.5
Rest days (home / away): 21 / 21
Scoring form difference: +4.4

Injuries, starting quarterbacks and roster changes are not yet included.

MARKET OBSERVATION

No verified market match in this capture

The bounded sample is not a complete market inventory. Missing observations do not imply zero probability or liquidity.

NFL Forecast Lab2026

NE at SEA

80% forecast ranges

No verified market match in this capture

NFL Forecast Lab2026

NE at SEA

80% forecast ranges

No verified market match in this capture