Scale of Sultry — Week 10 Results: November 3–7, 2026

Week 10 outcomes: The revised August 15 campus-baseline model correctly selected 29 winners in 54 eligible pregame matchups (29–25, 53.7%). The eight featured selections finished 4–4. The original published November 3 forecasting model, tracked separately, finished 33–30 over 63 games and calculates to a season record of 359–324 (52.6%) through Week 10, including the Week 0 pilot. The original archive publication cutoff remains subject to confirmation.

Published November 6, 2026 at 6:47 PM CT. All 63 matchups are included below. Games whose scheduled kickoff has passed are labeled retrospective and excluded from this revised model’s pregame prediction record.

Eight featured games and results · All 54 pregame predictions and outcomes · Nine retrospective calculations and outcomes · Week 9 results archive

How this week’s model works

Each school keeps its fixed August 15, 2026 campus SOS. The higher unrounded baseline supplies the winner selection; exact ties receive no selection. Stadium conditions near kickoff provide the context for testing whether the baseline advantage matters more in sultry weather. These are model selections, not confidence ratings or betting recommendations.

The modified SOS uses the saved daily high temperature and daily average relative humidity. Those values need not occur at the same time. A heat-index-based proxy is mapped to a bounded 1–10 scale, using 70°F = 1 and 160°F = 10. It uses temperature and humidity only; solar exposure, wind and fabric factors are not included. It does not establish actual training exposure or acclimatization.

Weather timing: all kickoff temperatures and humidity values below are the forecasts saved November 3, 2026. They have not been refreshed for this publication and are not observations. Modified game-site SOS values use the same temperature/humidity mapping as the August baselines, so they differ from the older site’s SOS scale.

Eight featured games — final results (4–4)

Displayed baselines are rounded to one decimal. Texas A&M and South Carolina both display 7.9, but Texas A&M leads before rounding. The same rule selects North Texas over Rice, although both display 6.6.

What Week 10 showed

The revised model correctly selected Houston over Cincinnati, Alabama over LSU, Arizona State over Colorado, and Texas over Missouri among its eight featured games; it missed Florida–Oklahoma, Texas A&M–South Carolina, Missouri State–FIU, and BYU–Utah. Across all 54 eligible pregame selections, it finished 29–25 (53.7%). A higher campus summer-weather baseline remains a geographic proxy, not a direct measurement of conditioning; the results do not establish that heat or humidity caused any victory. The stadium temperatures and humidity percentages below are saved November 3 forecasts, not observed game conditions.

All 54 pregame selections — final outcomes (29–25)

Every game in this table has a final score and is marked Correct or Incorrect against the revised August 15 campus-baseline selection published before kickoff. The eight featured matchups also appear here and are not counted twice. The nine retrospective calculations are listed separately below and excluded from the revised pregame total.

Earlier games — retrospective calculations

Excluded from the revised model’s pregame record. These selections were calculated after their scheduled kickoff. They do not replace or rewrite the original November 3 published predictions.

Sources and record keeping

Campus baselines: August_15_2026_D1_Football_Weather_Fresh.xlsx, Modified SOS tab, version dated October 23, using Iowa Environmental Mesonet station summaries. Weather data definitions · Heat-index method. Schedule and saved weather: the November 3 Week 10 archive, sourced from the ESPN Week 10 scoreboard and Open-Meteo.

Season results through Week 10: The original published Scale of Sultry forecast series stood at 326–294 through Week 9 and finished Week 10 at 33–30 over 63 games. That produces 359–324 across 683 games (52.6%), including the Week 0 pilot, subject to final confirmation that the November 3 original forecast archive was frozen before the first kickoff. The revised August 15 campus-baseline research model finished Week 10 at 29–25 across 54 eligible pregame selections (53.7%); its nine retrospective calculations are excluded. The models have different rules and denominators, so their records must not be combined. The original Week 10 archive remains available for audit: original November 3 forecasts and picks.

Week 9 results archive

The archived content below describes the original published model and its results through Week 9. The revised Week 10 model, scores and outcomes appear above. The original model’s Week 10 record is 33–30, bringing its calculated 2026 season total to 359–324 (52.6%), including the Week 0 pilot. This new total supersedes the older Week 9-only season figures shown inside the historical archive below.