Market vs Reality [Football Season]

← All reports
2026 Season · Week 9

Week 9: Favorites Rule the Weekend—Except in New England

Week 9 delivered 15 games and a remarkably lopsided favorite record: favorites went 14-1. The prediction market’s average Brier score was 0.1166, better than the 0.2500 baseline. There were no ties.

That strong overall mark came with one glaring miss: Green Bay beating New England 45-27 despite the market making the Patriots the favorite.

Market Accuracy by Week

FOOTBALLBLOGGER / NFL MARKET ACCURACY
===============================================
Average Brier score (LOWER IS BETTER)
Each # represents approximately 0.01 Brier points

WEEK  BRIER   GAMES  CHART
-----------------------------------------------
   3  0.2517     16  #########################
   4  0.2172     16  ######################
   5  0.2236     14  ######################
   6  0.2612     14  ##########################
   7  0.2272     14  #######################
   8  0.1788     14  ##################
   9  0.1166     15  ############ <-- CURRENT
LATEST VS PREVIOUS: IMPROVED (-0.0622)
-----------------------------------------------
50/50 BASELINE: 0.2500
===============================================

Best Call

The cleanest prediction came in Kansas City. The market gave the Chiefs an 80.5% chance against the Jets, and Kansas City handled business with a 35-12 victory.

That was the week’s lowest Brier score. Jacksonville’s 33-10 win at Baltimore was another strong call, with the Jaguars carrying a 77.5% market probability and producing a 0.0506 Brier score.

Biggest Miss

The market’s worst result came in Green Bay’s visit to New England. The Patriots were priced as a 61.5% favorite, but the Packers won convincingly, 45-27.

It was a surprising result, but one upset does not prove the original probability was unreasonable. A 61.5% forecast still leaves plenty of room for the other team to win.

Quarter Kelly

The original fictional Quarter Kelly strategy stayed on the sidelines this week. It started week 9 with $153.44 and finished at $153.44.

All 15 games had usable 24-hour snapshots, but every game fell below the strategy’s five-percentage-point entry threshold. That suggests the fixed threshold was cautious this week, though more wagers would not automatically have been better.

Since the experiment began with $160.00, the original strategy now has:

The walk-forward model used only earlier completed weeks and 24-hour snapshots, with historical calibration adjustments, quarter Kelly sizing, and a 10% bankroll cap. Weeks 3 and 4 were training only; Week 5 was the first test week. No real money was wagered.

Quarter Kelly 2.5% (Retrospective Comparison)

The 2.5% version is a RETROSPECTIVE historical comparison introduced later, not a prospectively tracked strategy. It uses the same walk-forward model and quarter Kelly sizing but keeps a separate bankroll.

This lower-threshold version found one eligible wager: Carolina against Denver. The model’s estimated probability for Carolina was 45.63%, compared with an assumed 42.5% price. Carolina lost, while Denver won 26-21.

From the experiment’s original $160.00 bankroll:

This one-week comparison does not establish that either entry threshold is better.

Bet the House

The fictional Bet the House benchmark divides the carried-forward bankroll equally among eligible favorites, using the final market midpoint as the assumed price. It began with $10 per game in Week 3.

The lone loss was the Patriots against Green Bay, but the other 14 favorites won, producing a huge week.

Since the original $160.00 bankroll:

What We Learned This Week

All betting described here is fictional. Midpoint prices may not be executable, fees are excluded, and this is not betting advice.

← Week 8 All Reports Week 10 →