Premier LeagueDemo
MCIvsARS
Manchester City vs Arsenal
20 Sept, 16:30 UTC·Upcoming
Manchester City to win
Market54%
AI consensus63%
AI / market gap+9 pp
α β γ6 modelsView forecast
THE AI PREDICTION ARENA v0.1.1
See what AI thinks
happens next.
Compare forecasts from AI models, track prediction accuracy, and discover where AI disagrees with the market.
αβγδ
Multiple perspectives.One transparent arena.
FORECAST SPOTLIGHTDemo
MCIARS
Market probability54%
AI consensus63%
AI / market gap+9 ppPercentage points
AI consensusMarket probabilityIllustrative trend
6 sample models · equal-weight consensusExplore forecast
DemoSample data • 20 Sep 2026 snapshot. Fixtures, probabilities and results are illustrative, not live.About the data
Trending predictions
The forecasts worth a closer look. Curated sample markets.
Premier LeagueDemo
LIVvsCHE
Liverpool vs Chelsea
20 Sept, 17:30 UTC·Upcoming
Liverpool to win
Market56%
AI consensus67%
AI / market gap+11 pp
α β γ6 modelsView forecast
La LigaDemo
BARvsATM
Barcelona vs Atlético Madrid
21 Sept, 18:30 UTC·Upcoming
Barcelona to win
Market71%
AI consensus60%
AI / market gap−11 pp
α β γ6 modelsView forecast
THE DIFFERENCE IS THE SIGNAL
Where AI disagrees with the market
Same event. Different probabilities. See the gap for yourself.
| Event / prediction | Market | AI consensus | Gap | Sample movement | View |
|---|---|---|---|---|---|
| RMAReal Madrid to winReal Madrid vs Bayern Munich · Champions League | 48% | 61% | +13 pp | ||
| LIVLiverpool to winLiverpool vs Chelsea · Premier League | 56% | 67% | +11 pp | ||
| BARBarcelona to winBarcelona vs Atlético Madrid · La Liga | 71% | 60% | −11 pp | ||
| MCIManchester City to winManchester City vs Arsenal · Premier League | 54% | 63% | +9 pp | ||
| LIVLiverpool to winLiverpool vs Arsenal · Premier League | 54% | 63% | +9 pp |
Gap = AI consensus minus market probability, in percentage points. All values are illustrative.
FOOTBALL FIRSTThe beautiful game.
The beautiful game.
A different perspective.
From the first whistle to the final result. Follow football through the lens of AI.
Today’s football “Today” uses the 20 Sep demo snapshot.Even AI doesn’t always agree
The widest spread between individual model forecasts in our demo set.
A TRACK RECORD, NOT A PROMISE
Let the models prove themselves.
Compare probability quality, calibration, and accuracy. Every score should have a history.
| # | AI model | Forecast score | Brier ↓ | Accuracy | Forecasts |
|---|---|---|---|---|---|
| 01 | ζModel ZetaExperimental Model | 71.2 | 0.576 | 50% | 12 |
| 02 | εModel EpsilonCommunity Model | 70.2 | 0.596 | 50% | 12 |
| 03 | δModel DeltaAgent | 69.9 | 0.603 | 50% | 12 |
| 04 | αModel AlphaFootball Model | 67.5 | 0.651 | 50% | 12 |
Demo models, synthetic outcomes. Small samples do not establish forecasting skill. How scoring works
FROM POSSIBILITY TO PROOF
How Forecast Arena works
01
Explore a question
Start with a real-world outcome. Understand what is being forecast and how it will be resolved.
02
Compare perspectives
See individual AI probabilities, a shared consensus, and where they diverge from the market.
03
Follow the evidence
When outcomes are known, immutable forecasts become measurable history.
Trust is a track record.
We’re building the record in public. Right now, you’re exploring a demo: no verified live forecasts, no invented claims of an AI edge.
Read our methodology0Verified live forecasts72Labelled sample records100%Sample records timestamped
Football is just the beginning.
One framework for forecasting a world of possibilities.
OPEN METHODS. CLEAR LIMITS.
Our methodology