Skip to main content
Educational analyticsNot betting advice · research use only.
GameTime Picks
Today · Oct 30 · 10 eventsLatest results· Oct 29

NFL

Week 8 · regular seasonSun, Nov 1 – Mon, Nov 2times in ET · updated Oct 30, 11:53 AM ET
Official Forecast · frozen before kickoff · graded after

The authorized GameTimePicks output for a game or player, frozen before the start and never edited afterwards. Each one is graded against the official result once the game is final, and results are published as games are graded. Player results grade the official player board's own pre-kickoff projections, each under the model and version that made it — the same projections the game pages and Weekly leaders show. The supporting simulations some game pages show are not graded in these results, so none of these results describe them.

Upcoming · Sun, Nov 1

Games

13 scheduled · 13 with a report · 13 with a supported read

Win chances add up to less than 100%: about 2–3% chance of a tie after regulation is not assigned to either team.

Model Performance. Our win chances come from a team rating that weights how decisively teams win. Tested on 4,281 past games it had never seen (2006–2021), it scored 0.629 on log loss (lower is better) — better than our previous rating's 0.642 and a coin flip's 0.693, but short of the sportsbooks' own odds at 0.610. It has NOT been shown to beat the sportsbook market. Where our win and margin ratings disagree on the favourite, the previous rating is used for that game. The 2026 season evaluation still runs under bars frozen in August, so every forecast stays experimental. These ratings are built from results, not from who is playing: they do not know that a starting quarterback has been ruled out. Our player projections do remove players who cannot play; the team win chance beside them does not. We tested whether adding that would help, using perfect hindsight about whether each team's main quarterback actually played — on 855 games from 2023 to 2025 it improved the win rating on average but was WORSE in one of the three seasons, so it did not clear the bar we hold these ratings to and we left the rating alone. Official Forecasts do not enter paper products (Bank Builder, Moonshot, Parlays) until the model clears its preregistered bar. Model Performance →

Forecast details for every game this weekThe full slate table: each game's simulation state, score range and report link.
This week

Week 8 · regular season — 13 games

The full Week 8 · regular season: 13 of 13 carry a published simulation.

OFFICIAL FORECAST
Kickoff (ET)MatchupLikely winnerProjected scoreOur total (range)Sportsbook totalWeatherStatus
Sun 1:00 PMBAL atBUFBUF 70.8%BAL 22 — 30 BUF52 (35–69)—63°F · 3 mph wind from the S · 0% chance of rain · partly sunnySimulated · 10,000 runsView game →
Sun 1:00 PMTEN atCINCIN 81.8%TEN 18 — 28 CIN46 (29–63)—68°F · 3 mph wind from the S · 40% chance of rain · chance light rainSimulated · 10,000 runsView game →
Sun 1:00 PMARI atDALDAL 65.5%ARI 24 — 31 DAL55 (38–72)—78°F · 5 mph wind from the SE · 1% chance of rain · partly sunny — the schedule carries no roof for this stadium, so this is the open-air forecastSimulated · 10,000 runsView game →
Sun 1:00 PMMIN atDETDET 56.9%MIN 20 — 24 DET44 (27–62)—Indoors — roofed stadium, no weather to reportSimulated · 10,000 runsView game →
Sun 1:00 PMLV atNYJNYJ 55.7%LV 20 — 24 NYJ44 (27–61)—62°F · 3 mph wind from the SW · 0% chance of rain · partly sunnySimulated · 10,000 runsView game →
Sun 1:00 PMCLE atPITPIT 65.5%CLE 17 — 23 PIT40 (24–57)—67°F · 3 mph wind from the S · 0% chance of rain · mostly cloudySimulated · 10,000 runsView game →
Sun 1:00 PMATL atTBTB 51.3%ATL 22 — 24 TB46 (29–63)—80°F · 13 mph wind from the E · 69% chance of rain · chance showers and thunderstormsSimulated · 10,000 runsView game →
Sun 1:00 PMIND atJAXJAX 76.5%IND 20 — 27 JAX47 (30–64)—75°F · 17 mph wind from the E · 36% chance of rain · chance showers and thunderstormsSimulated · 10,000 runsView game →
Sun 4:05 PMLAC atLARLAR 74.3%LAC 19 — 26 LAR45 (28–62)—Indoors — roofed stadium, no weather to reportSimulated · 10,000 runsView game →
Sun 4:25 PMKC atDENDEN 61.9%KC 18 — 22 DEN40 (23–57)—61°F · 6 mph wind from the ENE · 8% chance of rain · partly sunnySimulated · 10,000 runsView game →
Sun 4:25 PMNE atMIANE 70.9%NE 24 — 20 MIA44 (27–61)—79°F · 14 mph wind from the E · 89% chance of rain · showers and thunderstormsSimulated · 10,000 runsView game →
Sun 8:20 PMPHI atWSHPHI 50.8%PHI 24 — 23 WSH47 (30–64)—55°F · 2 mph wind from the SE · 0% chance of rain · partly cloudySimulated · 10,000 runsView game →
Mon 8:15 PMCHI atSEASEA 65.5%CHI 20 — 26 SEA46 (29–63)—51°F · 1 mph wind from the N · 12% chance of rain · mostly cloudySimulated · 10,000 runsView game →

How to read this table. Likely winner is the team our model gives the higher win chance, and that chance. It comes from the model's team rating, not from counting wins in the 10,000 simulated games. Projected score is the middle of our simulated outcomes, not a call on the exact final; it adds up to our total and always leans toward the favourite, even by one point. Our total is the points we expect both teams to score combined, with the range where 8 in 10 simulations landed. Nothing here is a bet recommendation.

Before it was allowed on this page, our totals model was tested on 5,878 past NFL games (2000–2021) it had never seen. Its projected total missed the real combined score by 10.9 points on average; the sportsbooks' closing totals missed by 10.7. Game totals are hard to predict for everyone, so read any single total as the middle of a wide range.

Ours next to the sportsbooks’: Across these 13 games our projected totals run from 40 to 55 points. Real NFL scoring swings far more than that from game to game, so treat small differences between games as small.

Conditions are the forecast nearest kickoff, captured Oct 29, 7:55 PM ET. The model does not use them — no weather term enters the total or the win chance, and these are shown beside the forecast so you can see what it is not accounting for. Forecast: US National Weather Service (api.weather.gov), public domain. Schedule: nflverse (CC BY 4.0). Coordinates: © OpenStreetMap contributors (ODbL).

Updated Fri 11:53 AM ET · frozen pre-kickoff · Model details · Week permalink

This week · 13 games

Top boards

The official player board's top five in passing yards, rushing yards, receiving yards, receptions and anytime touchdown — the same numbers as each game page. Players listed as Questionable, Doubtful or Out are left out; a statistic marked Estimate has not yet met its calibration standard and is labelled so.

Estimate · calibration standard not yet met · The chance of going over a line is not yet well calibrated for this family — read the range, not a precise over/under chance. Evaluation: not cleared on a second review — failed bar: threshold calibration (ECE 0.055 vs 0.05); beats both baselines on typical miss.

  1. 1Jared Goff
    Jared GoffMatchupDET vs MIN · StartSun 1:00 PM ET
    264yds80%: 151–424
  2. 2Matthew Stafford
    Matthew StaffordMatchupLAR vs LAC · StartSun 4:05 PM ET
    251yds80%: 146–397
  3. 3Joe Burrow
    Joe BurrowMatchupCIN vs TEN · StartSun 1:00 PM ET
    249yds80%: 144–396
  4. 4Drake Maye
    Drake MayeMatchupNE vs MIA · StartSun 4:25 PM ET
    244yds80%: 133–404
  5. 5Dak Prescott
    Dak PrescottMatchupDAL vs ARI · StartSun 1:00 PM ET
    237yds80%: 136–378

Projected = the official player board's median passing yards; 80% of outcomes fall in the range shown · Official player board

  • Passing yards: Ranked by projected median passing yards (the middle of the model's range). Not a chance to go over any line.
  • Rushing yards: Ranked by projected median rushing yards (the middle of the model's range). Not a chance to go over any line.
  • Receiving yards: Ranked by projected median receiving yards (the middle of the model's range). Not a chance to go over any line.
  • Receptions: Ranked by projected median receptions (the middle of the model's range). Not a chance to go over any line.
  • Anytime touchdown: Ranked by the official touchdown model's chance that the player scores at least one touchdown.

Board stamp

Every forecast we publishedThe twelve most recent game reports, each marked with where that game now stands.
Per-game reports

Every forecast we published

Each one shows what the model said before kickoff, and is labelled with where that game now stands.

Schedule

The rest of Week 9

NFL schedule as of Oct 30, 11:50 AM ET. Simulations publish inside each game's own window before kickoff, not weeks ahead.

Players

Endzone Vault

Who our model thinks is most likely to score

WATCHLIST
  1. Jahmyr GibbsJahmyr GibbsMatchupDET vs MINStartSun 1:00 PM ETexpected to playMarketNot checkedModelGameTimePicks78.9%Game →
  2. James Cook IIIJames Cook IIIMatchupBUF vs BALStartSun 1:00 PM ETexpected to playMarketNot checkedModelGameTimePicks76.6%Game →
  3. Javonte WilliamsJavonte WilliamsMatchupDAL vs ARIStartSun 1:00 PM ETexpected to playMarketNot checkedModelGameTimePicks73.4%Game →
  4. Jonathan TaylorJonathan TaylorMatchupIND vs JAXStartSun 1:00 PM ETexpected to playMarketNot checkedModelGameTimePicks71.0%Game →
  5. Kenneth Walker IIIKenneth Walker IIIMatchupKC vs DENStartSun 4:25 PM ETexpected to playMarketNot checkedModelGameTimePicks69.1%Game →
  6. D'Andre SwiftD'Andre SwiftMatchupCHI vs SEAStartMon 8:15 PM ETexpected to playMarketNot checkedModelGameTimePicks66.9%Game →
  7. Derrick HenryDerrick HenryMatchupBAL vs BUFStartSun 1:00 PM ETexpected to playMarketNot checkedModelGameTimePicks62.1%Game →
  8. Kyren WilliamsKyren WilliamsMatchupLAR vs LACStartSun 4:05 PM ETexpected to playMarketNot checkedModelGameTimePicks61.5%Game →

We have not checked the sportsbooks for this market, so we are not showing a price we do not hold.

Sportsbook prices for this slate: none current — last capture Oct 11 ET.

Recent final scores15 official finals from the scoreboard.
Results

Recent finals

Official final scores. A result we cannot match to a game we listed before kickoff is held back and reported, never graded.

  • NE 24 — 30 CHIfinal · Thu, Oct 22
  • PIT 16 — 10 NOfinal · Sun, Oct 25
  • SF 20 — 29 ATLfinal · Sun, Oct 25
  • CLE 27 — 24 TENfinal · Sun, Oct 25
  • IND 10 — 16 MINfinal · Sun, Oct 25
  • MIA 28 — 25 NYJfinal · Sun, Oct 25
  • TB 43 — 13 CARfinal · Sun, Oct 25
  • CIN 35 — 13 BALfinal · Sun, Oct 25
  • NYG 27 — 21 HOUfinal · Sun, Oct 25
  • DEN 30 — 16 ARIfinal · Sun, Oct 25
  • GB 17 — 20 DETfinal · Sun, Oct 25
  • LAR 35 — 21 LVfinal · Sun, Oct 25
  • KC 3 — 20 SEAfinal · Sun, Oct 25
  • DAL 38 — 21 PHIfinal · Mon, Oct 26
  • CAR 19 — 27 GBfinal · Thu, Oct 29
Picks vs outcomes · 147 graded

How the model's picks actually turned out

Game forecasts: the model-favored team (the higher frozen win probability), graded against the official final. This is historical model-favored winner accuracy, not a record of published picks.

Full record →
Historical model-favored winner accuracy: 61 hit · 44 missedSides frozen before kickoff: 24 hit · 18 missed3 void — a condition that did not hold is never scored as a miss

147 graded picks is still a small sample. A hit rate over this many outcomes has a wide enough margin of error to be consistent with almost any true accuracy.

EXPERIMENTAL. The NFL team model has not cleared a preregistered bar and is not promoted into any paper product. A tie is recorded as void, not a miss.

DateEventModel saidWhat happenedResult
2026-10-26DAL @ PHIHOMEAWAYMISS
2026-10-25PIT @ NOAWAYAWAYHIT
2026-10-25IND @ MINHOMEHOMEHIT
2026-10-25TB @ CARHOMEAWAYMISS
2026-10-25CIN @ BALHOMEAWAYMISS
2026-10-25CLE @ TENAWAYAWAYHIT

Showing the 6 most recent of 147 graded.

Technical detailsCalibration, coverage by market, playing time, the pregame audit and product evaluations.

Were the intervals right?

Every game page prints a margin and a total with an 80% interval. This is how often the real result actually landed inside one, graded against official final scores. Seasons are never pooled.

preseason41 settled
winner 53.7% · margin off by 12.4 avg · total off by 8.2 avg
Margin65.8%of 41 landed inside a 80% intervaltoo narrow
Total87.8%of 41 landed inside a 80% intervalmatches the label
regular-season107 settled
winner 59.8% · margin off by 9.7 avg · total off by 10.9 avg
Margin81.3%of 107 landed inside a 80% intervalmatches the label
Total83.2%of 107 landed inside a 80% intervalmatches the label

an interval printed as 80% contained the result only 65.9% of the time over 41 settled forecasts (z -2.26) — the interval is too narrow and the label overstates what the model knows The correction is not applied automatically — changing a model because of a result it has already seen is how a track record stops meaning anything. It needs a preregistered bar and a held-out season first.

Status

Coverage, market by market

Readiness is stated per layer — a page section is never filled merely because it exists.

LayerStatusWhat that means
Schedule & identitiesLIVEdaily capture · 15 events in window, captured Oct 30, 11:50 AM ET
ResultsLIVEofficial finals join by durable event id; results without pre-event schedule lineage are quarantined and say so
Team game simulationOfficial ForecastWe publish a simulated score range and win chance for every game, published as the Official Forecast (13 games forecast today under nfl-regular-season-public-v1). The win chance comes from a team rating that weights margin of victory. Tested on 4,281 past games it had never seen (2006–2021), it scored 0.629 on log loss (lower is better) against 0.642 for our previous rating, 0.693 for a coin flip and 0.610 for the sportsbooks' own odds — so it makes no claim to beat the sportsbook market. Every forecast is frozen before kickoff and settled against the official result. Next: A validated pick additionally needs the frozen 2026 walk-forward evaluation to pass its own bars — 64 decisive games and 4 completed weeks before any reading counts.
Moneyline / spread / total pricesArchived pricesThe most recent authorized capture (2026-10-11) priced 14 game(s) whose kickoffs have all passed. Those prices live with their games in the archive; no current price is shown and none is invented. Next: The next authorized market capture.
Passing yardsEstimatenot cleared on a second review — failed bar: threshold calibration (ECE 0.055 vs 0.05); beats both baselines on typical miss — the chance of going over a line is not yet well calibrated for this family — read the range, not a precise over/under chance. Next: Current role evidence, a current offered line, and settlement support.
Rushing yardsEstimateeligible for a private test — failed bar(s): calOk — its uncertainty calibration failed the preregistered bar — the range shown may be mis-sized even where the middle is reasonable. Next: Current role evidence, a current offered line, and settlement support.
Receiving yardsPublishedEvery game on this week's slate publishes receiving yards from the same evaluated model. Each game page states its own range.
ReceptionsPublishedShare-level model: cleared every bar when re-tested on 2014–2021 (27,134 player-games) — a second look at seasons examined once before, not a blind test. Each week's forecast is frozen before the first kickoff and graded; its blind 2026 record is holding every bar (1129 graded player-games).
Anytime touchdown · Endzone VaultPublishedEvery game on this week's boards publishes touchdown chances from the calibrated scoring model; they assume the player plays (a player who does not play settles void). Endzone Vault makes no selection: current role evidence for these games has not cleared its own bar. Next: An offered anytime-touchdown market plus current role evidence.
SettlementDEPLOYEDteam and scorer results are graded automatically from official finals; each pre-kickoff forecast settles exactly once when its result exists
Paper products · Bank Builder, Moonshot, Card builderEvaluated · none qualify13 pre-kickoff NFL games evaluated; none qualified — the reason is under "Why NFL is not in the paper products" below.
Paper products · Endzone VaultGatedthe Vault reports watchlist only on its own evaluation: 166 model candidates · no current comparable touchdown price is available
Playing time · 13 games

Playing time refreshes until kickoff.

Game-day actives are published about 90 minutes before kickoff; until then availability comes from the injury report.

What we do instead

Each player carries a range of his team's work, not a single number, and the share nobody on the list accounts for is published as unallocated.

Why it matters

A projection is only as good as the playing-time behind it.

Graded · Thu, Aug 13

We scored all 6 of our 2026-08-13 NFL forecasts against the official finals.

Every forecast on this slate was a baseline: the model had no measured read separating the two teams, so these numbers grade a league-average preseason prior, not a game-specific prediction. 6 games. A record this small says almost nothing about a model; it is published because grading every forecast is the only way the numbers stay honest, not because it is evidence of skill.

GameWe saidResultMargin off byTotal off byInside our range
DET @ CIN19-1914-1627yes
GB @ PIT19-189-28200no
IND @ NE · tie19-1813-13112yes
ARI @ LV19-1827-14134yes
LAC @ HOU19-1827-7193no
TEN @ SF19-1819-1345yes
Winners called
3 of 5 (1 tie excluded)
Score off by, on average
6.25 points
Result inside our range
margin 67% · total 100% (aiming for 80%)

Our winner probabilities scored WORSE than the sportsbook consensus (0.2391 vs 0.2237, lower is better) across 5 decisive games.

What this model knows

Every part of this model reacts to the specific teams playing.

Win probabilities come from an evaluated strength rating; each game's total comes from that matchup's own scoring ratings.

Who wins, and by how much

This part uses each team's own strength, so it differs from game to game — across this slate our win percentages range from 26.2% to 81.8%.

How many points are scored

This part reads the two teams: each game's total comes from that matchup's own scoring ratings and from how efficiently both offences and defences have played, play by play, updated after every game. Across this slate the published totals span 40 to 55 points. It replaced the earlier totals model after doing better on 5,878 past games (2000–2021) it had never been tested on, under rules written down before the test.

How we tested this

What we found and fixed: An earlier version of this page applied the team-strength input anyway. Because the measured effect pointed slightly the wrong way, it was quietly favouring the WEAKER side in every game. We caught it, tested the input properly, and switched it off.

Every part of this model is re-tested before it changes: a new totals or win model has to do better than the current one on games it has never seen, under rules written down before the test.

Products · last run Oct 30, 11:53 AM ET · next Oct 30, 5:00 PM ET

All four NFL lanes ran

All four NFL lanes ran and none produced a card.

Per-product detail
ProductResultCandidates looked atWhy none qualified
Bank BuilderNo qualified card1313 × the team model has not cleared its preregistered bar
MoonshotNo qualified card1313 × the team model has not cleared its preregistered bar
Card builderNo qualified card1313 × the team model has not cleared its preregistered bar
End Zone VaultNo qualified card166166 × no NFL player-market price was captured for these games — the market was not checked in this window

What would change this: the team model is baseline only (the P181 challenger was rejected on bars frozen before it was built); all four player families were rejected on their own frozen bars (P183 Release B); no NFL player-market price was captured for these games — the market was not checked in this window (games probed in this window: 0 of 13).

Evaluated Oct 30, 11:53 AM ET

Why NFL is not in the paper products

13 pre-kickoff NFL games were evaluated and none qualified. What would change that:

  • an NFL model version that meets its own preregistered promotion bar on held-out data
  • enough settled Official Forecasts to measure whether it is calibrated
  • an explicit validated block on the forecast — the classifier requires one and the public-beta engine never emits it

This evaluation runs on every NFL event window and is recorded, so the answer for any past day stays recoverable. The same gate is enforced in the money path itself — an ineligible sport is refused there too, not merely omitted from the pool.

Where to go next
MethodHow this worksWhat the model does — and cannot do.
RecordTracked resultsEvery forecast graded against the official result.
CoverageAll sportsWhat is live and what is deliberately not.
MLBMLB hubThe settled, longer-running side of the site.

Official Forecasts and Game Simulations on this page are never presented as picks. Player markets and the Endzone Vault stay gated on their own separate evidence — team readiness never auto-publishes a player market. See methodology and all sports coverage.