Understand the model
Transparent proof and a clear introduction to ContextXI.
£0forever
- Full model calibration, every band
- How accuracy was measured
- Selected match intelligence
- Core probability explanations
Football Intelligence, Explained
Stats tell you what happened.ContextXI tells you what may happen next.132 calculations. Per player. Per match.
Premier League and La Liga player probabilities, built from real form and matchup data. One number, no jargon.
Premier League and La Liga player probabilities, built from every signal that matters. One number, no jargon.
What the model considersRecent formPositionOpponentHome or away
Probabilities, not promises.See why probabilities change.Every published forecast stays on record.
Explore ContextXI →Free to explore — no card requiredor start a 5-day free trial · card required →
ARSArsenal
PREMIER LEAGUE · 20:00vsEmirates Stadium
MCIMan City
BS
PLAYER INTELLIGENCE
STRONG POSITIVEBukayo Saka
ARS · Right wing · Expected 88 mins81%
1+ ShotsChance of this event
74% → 81%+7 percentage points
SOT67%
Shots 3+39%
Fouled 1+74%
PROBABILITY MOVEMENT+7Percentage points
MODEL UPDATED2 minutes ago
ARS LIVEMCI
BS
PLAYER INTELLIGENCE
STRONG POSITIVEBukayo Saka
ARS · RW · Expected 88 mins81%
1+ ShotsChance of this event
74% → 81%+7 percentage points
ARS LIVEMCI
EH
PLAYER INTELLIGENCE
STRONG POSITIVEErling Haaland
MCI · ST · Expected 90 mins35%
2+ Shots On TargetChance of this event
31% → 35%+4 percentage points
EXAMPLE Illustrative players and figures — not live predictions.
SEE CONTEXTXI IN ACTIONPlayer intelligence,
Player intelligence,
without the black box.
Start with the probability. See how it moved. Open the reasoning only when you want the deeper context.
EXAMPLE Illustrative players and figures, shown to demonstrate the interface — not live predictions.
BS
Bukayo Saka
ARS · RW · Expected 88 mins1+ Shots
81%
Movement+7 percentage points
SOT 67%Shots 39%Fouled 74%
WHY?
Confirmed start · advanced role
EH
Erling Haaland
MCI · ST · Expected 90 mins2+ Shots On Target
35%
Movement+4 percentage points
SOT 61%Shots 55%Fouled 46%
WHY?
Box volume · opponent profile
MO
Martin Ødegaard
ARS · AM · Expected 86 mins1+ Shots On Target
56%
Movement+9 percentage points
SOT 31%Shots 24%Fouled 52%
WHY?
Set-piece role · central overload
OPEN ANY PLAYER — SEE WHY THE NUMBER MOVED
BS
UPDATED 2M AGOBukayo Saka1+ Shots
EARLIER74%
→NOW81%
+7 percentage points
What changed
✓
Confirmed in starting XIMinutes confidence increased from 76 to 88.
↗
More advanced roleExpected touch position moved closer to goal.
◎
Favourable matchupOpponent concedes more chances down this side.
HOW THE NUMBERS GET MADENot a guess.
Not a guess.
A calculation.
Every percentage on ContextXI comes from real match data, not a hunch.
Matches analysed2,410+6 completed Premier League and La Liga seasons, growing every matchweek
Players tracked1,060+Every registered Premier League and La Liga squad
Sync frequency15 minHow often live match data refreshes
Predictions validated125,609Scored against a held-out season, no band off by more than 1.6pp
Not a guess.
A calculation.
Every percentage comes from real match data — not a hunch.
Matches analysed2,410+6 completed Premier League and La Liga seasons, growing every matchweek
Players tracked1,060+Every registered Premier League and La Liga squad
Sync frequency15 minHow often live match data refreshes
Predictions validated125,609Scored against a held-out season, no band off by more than 1.6pp
When we say 50-60%, it happens 54.5% of the time.
See how that was measured →MODEL VALIDATION, FREE TO EVERYONETested on a season
Tested on a season
it had never seen.
We publish how the model scored against completed matches it was never developed on — every probability band, including the thin ones. No band removed for looking bad.
HELD-OUT BACKTEST50-60% BAND
Average predicted54.2%across 9,046 predictions
Actually occurred54.5%+0.3pp vs predicted
i
Backtest validation, not a live track record. ContextXI has only just begun publishing live predictions; none of those fixtures has been played yet.
FROM HINDSIGHT TO FORESIGHTWhy a stat sheet
Why a stat sheet
isn’t a forecast.
TRADITIONAL STATS
The record
- Goals and assists
- Shots and touches
- Past averages
- Final outcomes
Useful history, but the match has already happened.
→
CONTEXTXI
The changing picture
- Player probabilities
- Role and matchup context
- Live probability movement
- Explainable reasons
A forward-looking view that changes when the football context does.
TRADITIONAL STATS
- Goals, shots, touches
- Past averages
- Final outcomes
CONTEXTXI
- Live probabilities
- Matchup context
- Explainable reasons
The record shows what happened. ContextXI shows what may happen next — and updates as the picture changes.
CHOOSE YOUR VIEWProof is free.
Proof is free.
Current intelligence is Pro.
Explore how the model has performed before deciding whether today’s complete player intelligence is for you.
See today before it unfolds
Complete current intelligence across every player and match.
£2.99/monthafter your 5-day free trial
- Every projected player visible
- Current player probabilities
- Probability movement tracking
- Full “Why?” explanations
- Matchup and team-news context
ACCOUNTABILITY BY DESIGNIntelligence you can
Intelligence you can
inspect, not just accept.
ContextXI is built around explainability. The original probability, the movement, the reasons and the final outcome remain distinct and visible.
Probabilities stay probabilities
We communicate uncertainty instead of disguising it as certainty.
Signals are explained
“Why?” connects the number to the football context behind it.
Outcomes stay separate
What happened never overwrites what the model originally expected.
The full record remains visible
The model’s complete calibration record is free, without a Pro subscription.
FREQUENTLY ASKEDClear questions.
Clear questions.
Straight answers.
ContextXI should be as understandable as the intelligence it presents.
Is ContextXI a betting tips service?+
No. ContextXI is a football intelligence product. It presents modelled probabilities and the context behind them, not guaranteed outcomes or financial advice.
What does a player probability mean?+
It is the model’s estimate of how likely a specific player event is. It starts from that player’s own record in that market, weighted toward more recent matches and balanced against the typical rate for their position, then adjusts for the opponent’s record against that position and whether the match is at home or away.
Why do probabilities move?+
They update as completed matches enter the model. Each new result changes that player’s own record, the average for their position, and the opponent’s record against it. A player who becomes unavailable through injury or suspension is removed from the list rather than shown a lower number.
What is available for free?+
Free access includes the model’s full calibration record — how its published probability bands scored against a season it was never developed on — plus selected match intelligence. Pro adds current intelligence, full player coverage and live movement explanations.
Does a 70% probability count as a win if it happens?+
We avoid that framing. A probability is judged through calibration across many predictions, not one result. Our accuracy page shows every probability band and how often events in it actually occurred.
See the game’s changing
context before it unfolds.
Explore today’s matches, open any player and understand the probability — and the reasons behind it.