Data status
We check the data behind every number each hour: are the ratings current, does every game have a forecast before it starts, are the scores and closing lines in. This page shows the newest check.
Checked Nov 17, 1:46 AM PT.
College basketball
| Check | Status | Detail |
|---|---|---|
| Export is current | OK | The site export: last update 2026-11-17 09:45 UTC (0.0 h ago; limit 3 h). |
| Ratings are current | OK | The team ratings: last update 2026-11-16 17:05 UTC (16.7 h ago; limit 30 h). |
| Every game has a forecast before it starts | OK | All 0 games that start in the next 2 hours have a forecast. |
| Final scores are loaded | OK | Every finished game has a final score. |
| Closing lines are recorded | OK | 114 of 114 graded games of the last 3 days have a closing line (100%). |
| Published forecasts are never changed | OK | 914 published forecasts; earlier entries unchanged. |
| Daily forecast issue is current | OK | Newest rc2_availability forecast issue: last update 2026-11-16 17:05 UTC (16.7 h ago; limit 26 h). |
| Every finished game had a pregame forecast | Watch | 245 of 638 finished D-I games have no pregame forecast (earlier outage): 2026-11-01 (1), 2026-11-05 (43), 2026-11-06 (59), 2026-11-07 (48), 2026-11-08 (29), 2026-11-09 (34), 2026-11-10 (31). |
| Coverage numbers match the source | Problem | The export says 415 of 660; the source gives 393 of 638. |
| The site lists every scheduled game | OK | All 434 D-I games of the next 10 days are in the export. |
| Team records are correct | Problem | 44 team record(s) differ. |
| Home, away and neutral sites agree | OK | All 476 forecast games agree with the schedule; 7 neutral-site games only swap the home label. |
| Team identities are unique and complete | OK | 365 teams, each id, name and slug once; every game team is a D-I member with a rating. |
| Forecasts follow the rating rule | OK | All 914 issued forecasts reproduce: margin = rating gap x pace / 100 + the home edge (3.5 at home sites; 3.5 + the home-court v2 shift in the 0 v10 rows; 0 at neutral sites); totals follow the offense and defense ratings; win chances fit the margins. |
| Site forecasts are the last pregame issue | OK | All 475 site forecasts equal the last same-day issue before the start, with the correct home side. |
| Each game says what the forecast includes | OK | All 730 scheduled and recent games say that the forecast does not include injuries. |
NBA
| Check | Status | Detail |
|---|---|---|
| Export is current | OK | The site export: last update 2026-11-17 09:45 UTC (0.0 h ago; limit 3 h). |
| Ratings are current | OK | The team ratings: last update 2026-11-16 23:02 UTC (10.7 h ago; limit 30 h). |
| Every game has a forecast before it starts | OK | All 0 games that start in the next 2 hours have a forecast. |
| Final scores are loaded | Watch | 8 games started more than 5 hours ago and have no final score. |
| Closing lines are recorded | OK | 14 of 14 graded games of the last 3 days have a closing line (100%). |
| Published forecasts are never changed | OK | 218 published forecasts; earlier entries unchanged. |
| NBA export is fresh | OK | NBA export: last update 2026-11-17 09:45 UTC (0.0 h ago; limit 5 h). |
| NBA results are complete | Problem | 7 games before today have no box score in the game tables (build 11-16 06:05 ET). Results, records and ratings miss them. |
| Every NBA game has a forecast | OK | 200 of 205 started games have a forecast before tip. Known gap: 5 games of 2026-10-23 (Mac asleep); CHI @ MIL 10-23 came after a moved tip. |
| Games in the next 3 hours have a forecast | OK | 0 games start in the next 3 hours; 0 are past their cutoff and all have an issued forecast. |
| NBA scores and records match the NBA | OK | 197 box scores: scores, period sums and all 30 records match the export. |
| Forecast numbers follow the rule | OK | 200 logged forecasts: margin = 0.98 x net gap + 2.0 at home, nets = sum of the players, issued between C and tip; 17 previews follow the Tonight gap rule. |
| NBA Strength table is sane | OK | 30 teams; full strength mean 0, SD 4.80, range -8.4 to +8.5; offense + defense = strength; Tonight matches the forecast log. Largest full - tonight gap: Golden State Warriors +2.8. |
| Closing lines are right | Watch | 27 of 37 final games since 2026-11-11 have a close (14 checked against the raw snapshots). Missing closes come from Mac sleep before tip. |
| Injury status on the site is complete | OK | 163 games have the official report status; 80 games show a status change after our forecast. |
| 'If he sits' line moves are sane | OK | 255 players in 25 games of the next 2 days have a line move; each Questionable or Doubtful player has a sentence. Games that no report lists yet have none. |
College football
| Check | Status | Detail |
|---|---|---|
| Export is current | OK | The site export: last update 2026-11-17 09:45 UTC (0.0 h ago; limit 3 h). |
| Ratings are current | OK | The team ratings: last update 2026-11-16 16:01 UTC (17.7 h ago; limit 192 h). |
| Every game has a forecast before it starts | OK | All 0 games that start in the next 2 hours have a forecast. |
| Final scores are loaded | OK | Every finished game has a final score. |
| Closing lines are recorded | OK | 56 of 56 graded games of the last 3 days have a closing line (100%). |
| Published forecasts are never changed | OK | 836 published forecasts; earlier entries unchanged. |
| The weekly rating issue is current | OK | The newest weekly issue is from Mon 11-16 08:01 PT (0.7 days ago). |
| The weekly issue covers every FBS game of its week | OK | The issue of Mon 11-16 08:01 PT has forecasts for all 70 FBS games of its week, and its CFBD capture has every final score up to Sun 11-15 16:00 PT. |
| The forecast uses only allowed inputs | OK | C02 reads only CFBD scores, sites, talent and returning production. All lines come from The Odds API. |
| CFBD schedule and scores are current | Problem | The CFBD kickoff-check capture: last update 2026-11-17 02:49 UTC (7.0 h ago; limit 3 h). |
| Every finished FBS game has a forecast logged before kickoff | Watch | 484 of 492 finished FBS games since 2026-09-21 have a forecast logged before kickoff. The 8 missing games are known gaps (late weekly issues). |
| Final scores agree across sources | OK | 484 exported results match CFBD; 1410 final games agree between the weekly capture and the newest kickoff check. |
| Team records match the final scores | OK | All 138 records match CFBD; every team has 9 or more finished games. |
| Home, away and neutral sites agree | OK | All 984 logged forecasts name the CFBD home and away teams; 4 neutral sites agree. |
| Line signs agree with our forecasts | OK | Our margin and the market margin correlate at 0.96 over 300 games; 1 games differ by more than 10 points. Each one is confirmed from the raw snapshot. |
| Each forecast equals the rating edge plus the home edge | OK | 488 forecasts equal the rating edge plus 2.38 points (0 at neutral sites) within 0.02; 71 forecasts of the first issue have no saved ratings. |
| Team ratings are in a sane range | OK | 138 teams, mean +0.00, SD 13.0, from -27.8 to +29.9. No earlier weekly edition to compare. |
| Closing lines follow the rule | OK | 110 closing lines are from the last 30 minutes before kickoff, from The Odds API. |
| Closing lines are recorded each week | Watch | Week 11: 58 of 61 finished games have a closing line. Weeks with no closes: 10. |
| The CLV number uses no SpotOdds line | OK | The CLV row counts only the 18 entries with The Odds API lines. |
| Game pages show the official availability reports | OK | 199 games have the block (not included in our forecast); 62 have official lists, with 980 players listed Out, Doubtful or Questionable. |
NFL
| Check | Status | Detail |
|---|---|---|
| Export is current | OK | The site export: last update 2026-11-17 09:45 UTC (0.0 h ago; limit 3 h). |
| Ratings are current | OK | The team ratings: last update 2026-11-16 14:12 UTC (19.6 h ago; limit 30 h). |
| Every game has a forecast before it starts | OK | All 0 games that start in the next 2 hours have a forecast. |
| Final scores are loaded | Watch | 1 games started more than 8 hours ago and have no final score. |
| Closing lines are recorded | OK | 12 of 12 graded games of the last 3 days have a closing line (100%). |
| Published forecasts are never changed | OK | 772 published forecasts; earlier entries unchanged. |
| Daily ratings file is current | OK | The newest ratings file 2026_wk09_20261116 was written 2026-11-16 06:12 PT (19.6 h ago; the build runs daily at 06:10 PT, limit 26 h). |
| Ratings include the last full week | OK | The ratings (house model, file 2026_wk09_20261116) use games through week 9; week 10 is not complete yet. |
| Every game has a forecast from before kickoff | OK | 56 of 56 final games since the house model went live (2026-10-20 06:13 PT) have a logged forecast from before kickoff. Of 93 earlier games, 44 keep the forecast of the earlier rating (team results only) and 49 have none on the site. |
| Every game of the next 10 days is on the site | OK | 17 of 17 scheduled games of the next 10 days have a forecast on the site. |
| Final scores agree across sources | OK | 149 final games: 149 checked against the play-by-play and 85 against ESPN; 0 disagree. |
| Team records are correct | OK | 32 of 32 records match the schedule (ties shown as a third number). |
| Neutral sites are marked | OK | 9 neutral-site games this season (2026_01_SF_LA, 2026_03_BAL_DAL, 2026_04_IND_WAS, 2026_05_PHI_JAX, 2026_06_HOU_JAX, 2026_07_PIT_NO, 2026_09_CIN_ATL, 2026_10_NE_DET, 2026_11_MIN_SF). The site differs on 0. 5 logged rows used another neutral flag than the schedule now has. |
| Home and away teams agree | OK | 272 games: the game ids, the forecast log, the site and 86 ESPN games agree on the home and away team; 32 teams with 17 games each. |
| Spread signs are right | OK | The site's home spread is the negative of the nflverse spread_line for every line (0 errors). ESPN's early-week line agrees within 1.5 points on 72 of 86 games. 3 games changed favorite after ESPN's line; 3 of them had a new starting QB. |
| Forecast = home rating - away rating + home advantage | OK | 118 forecasts: margin (with the home term of the ratings file that made it), win chance (normal, 13.5-point spread) and issue time follow the rule; 0 do not. |
| Site ratings are the newest ratings file | OK | 32 of 32 'this week' values match model_only_margin in 2026_wk09_20261116; 0 ranks do not follow full strength (a tie of the shown values may take either rank). |
| Ratings are in a sane range | OK | Full strength: mean -0.25, spread (SD) 4.83, range -9.48 to +7.33. Largest injury-report gaps (full - this week): SEA +7.33, GB +1.60, TEN +1.50. This week's value from week 8 to 9: mean change 1.52, largest 6.66. |
| Closing and latest lines are present | OK | 149 of 149 final games have the nflverse closing line; the site misses 0. 1 of 1 games of the next 72 h have a latest line. The latest line updates once a day (06:10 PT). |
| Game pages say what the number includes | OK | 46 of 46 scheduled and recent games have the availability block with included true; 17 of 17 upcoming games have the injury effect and 17 have 'if he sits' line moves. |