AI vs Market Scoreboard

Every hourly AI forecast is stored with the market price captured at the same moment, then scored when the market resolves. Append-only receipts, deterministic math, losses published. Lower Brier is better. Read the full research writeup → Submit your own forecasts →

Forecasts scored
6442
13080 awaiting resolution
AI Brier (mean)
0.182
sample: 6442
Market Brier (mean)
0.131
same timestamps
AI accuracy
74%
market: 81%
Across 6442 scored forecasts, VoxOdds AI (Brier 0.182) trailed the market price (Brier 0.131) taken at the same moment. A sweep over every AI/market mixture weight (n=6442) finds no weight of AI improves on the price — the market alone is optimal so far.

Recently scored forecasts

MarketAI saidResolved AI BrierMarket BrierVerdict
Miami Marlins vs. Chicago Cubs Chicago Cubs @ 90% Chicago Cubs 0.010 0.009 Market closer
Miami Marlins vs. Chicago Cubs Chicago Cubs @ 89% Chicago Cubs 0.013 0.009 Market closer
Miami Marlins vs. Chicago Cubs Chicago Cubs @ 86% Chicago Cubs 0.020 0.009 Market closer
Miami Marlins vs. Chicago Cubs Chicago Cubs @ 81% Chicago Cubs 0.035 0.009 Market closer
Miami Marlins vs. Chicago Cubs Chicago Cubs @ 72% Chicago Cubs 0.078 0.009 Market closer
Counter-Strike: Betclic Apogee Esports vs Eternal Fire (BO3) - Stake Pulse Beat II Group D Betclic Apogee Esports @ 96% Betclic Apogee Esports 0.002 0.002 AI closer
Counter-Strike: Betclic Apogee Esports vs Eternal Fire (BO3) - Stake Pulse Beat II Group D Betclic Apogee Esports @ 95% Betclic Apogee Esports 0.002 0.002 Market closer
Counter-Strike: Betclic Apogee Esports vs Eternal Fire (BO3) - Stake Pulse Beat II Group D Betclic Apogee Esports @ 95% Betclic Apogee Esports 0.002 0.002 Market closer
Counter-Strike: Betclic Apogee Esports vs Eternal Fire (BO3) - Stake Pulse Beat II Group D Betclic Apogee Esports @ 95% Betclic Apogee Esports 0.003 0.002 Market closer
Counter-Strike: Betclic Apogee Esports vs Eternal Fire (BO3) - Stake Pulse Beat II Group D Betclic Apogee Esports @ 94% Betclic Apogee Esports 0.004 0.002 Market closer
LoL: Guinea Pink vs Kawaii Kiwis (BO5) - LES Promotion Playoffs Guinea Pink @ 94% Kawaii Kiwis 0.884 0.893 AI closer
LoL: Guinea Pink vs Kawaii Kiwis (BO5) - LES Promotion Playoffs Guinea Pink @ 94% Kawaii Kiwis 0.890 0.893 AI closer
LoL: Guinea Pink vs Kawaii Kiwis (BO5) - LES Promotion Playoffs Guinea Pink @ 94% Kawaii Kiwis 0.886 0.893 AI closer
LoL: Guinea Pink vs Kawaii Kiwis (BO5) - LES Promotion Playoffs Guinea Pink @ 94% Kawaii Kiwis 0.879 0.893 AI closer
LoL: Guinea Pink vs Kawaii Kiwis (BO5) - LES Promotion Playoffs Guinea Pink @ 93% Kawaii Kiwis 0.865 0.893 AI closer
Counter-Strike: Phantom vs Iberian Soul (BO3) - Stake Ranked Episode 5: Closed Qualifier Playoffs Iberian Soul @ 96% Iberian Soul 0.002 0.001 Market closer
Counter-Strike: Phantom vs Iberian Soul (BO3) - Stake Ranked Episode 5: Closed Qualifier Playoffs Iberian Soul @ 96% Iberian Soul 0.001 0.001 Market closer
Counter-Strike: Phantom vs Iberian Soul (BO3) - Stake Ranked Episode 5: Closed Qualifier Playoffs Iberian Soul @ 96% Iberian Soul 0.002 0.001 Market closer
Counter-Strike: Phantom vs Iberian Soul (BO3) - Stake Ranked Episode 5: Closed Qualifier Playoffs Iberian Soul @ 95% Iberian Soul 0.003 0.001 Market closer
Counter-Strike: Phantom vs Iberian Soul (BO3) - Stake Ranked Episode 5: Closed Qualifier Playoffs Iberian Soul @ 93% Iberian Soul 0.005 0.001 Market closer
Chicago White Sox vs. Kansas City Royals Chicago White Sox @ 98% Chicago White Sox 0.000 0.000 Market closer
Chicago White Sox vs. Kansas City Royals Chicago White Sox @ 98% Chicago White Sox 0.000 0.000 Market closer
Chicago White Sox vs. Kansas City Royals Chicago White Sox @ 98% Chicago White Sox 0.000 0.000 Market closer
Chicago White Sox vs. Kansas City Royals Chicago White Sox @ 98% Chicago White Sox 0.000 0.000 Market closer
Chicago White Sox vs. Kansas City Royals Chicago White Sox @ 98% Chicago White Sox 0.000 0.000 Market closer

Model variants — the self-correction experiment

ModelForecastsBrier Market BrierBeats marketAccuracy
blend-mkt50-v1 10577 0.129 0.121 51% 80%
blend-mkt75-v1 9895 0.123 0.121 51% 82%
blend-mkt90-v1 9736 0.121 0.120 50% 83%
gemini-2.5-flash 6442 0.182 0.131 54% 74%
gemini-2.5-flash-cal2 3874 0.136 0.125 45% 79%
gemini-3.7-flash 6391 0.134 0.126 50% 81%
gemini-3.7-flash-cal2 6380 0.126 0.126 49% 82%

blend-mkt50-v1 = the raw AI probability blended 50/50 with the market price at forecast time — a deterministic self-correction deployed after the calibration analysis showed mid-range overconfidence. If the blend's Brier beats the market's, the AI adds information beyond price.

Calibration — when the AI says X%, how often does it happen?

AI confidenceForecastsRealized AI BrierMarket Brier
0-10% 8 100% 0.941 0.000
20-30% 1 100% 0.640 0.922
50-60% 419 57% 0.252 0.173
60-70% 1078 44% 0.265 0.212
70-80% 1225 59% 0.232 0.189
80-90% 981 60% 0.231 0.218
90-100% 2730 86% 0.097 0.036

Red = overconfident (events happened less often than the AI's stated probability).

AI vs market, by market line

Market line (winner's price at forecast time)Forecasts AI BrierMarket BrierAI edge
mkt 95%+ 1691 0.082 0.001 -0.081
mkt 80-95% 1632 0.079 0.016 -0.062
mkt 60-80% 1222 0.114 0.094 -0.020
mkt 40-60% 1084 0.276 0.230 -0.046
mkt <40% (upset) 813 0.578 0.559 -0.019

Methodology

Hourly AI forecasts are stored with the contemporaneous market price (append-only receipts). At resolution every receipt is scored: Brier = (1 - probability assigned to the eventual winner)^2, for both the AI and the market price captured at the same timestamp. Lower is better. Brier is computed only for 2-outcome markets; a winner requires one outcome priced >=0.95 at resolution. No receipt is dropped silently.

Unscorable receipts: 38323 ({'no_probability': 1, 'market_expired_unresolved': 38239, 'ambiguous_resolution': 15, 'unmatched_outcome_text': 68}) — counted, never silently dropped.

Raw JSON: /api/v1/scoreboard · Updated 2026-09-24T23:30:02.695903+00:00 · Market-implied odds, not financial advice.

Get the edge — weekly

Top divergences, biggest movers, and survey insights. No spam.