The bar is the model's probability; the caret on the same scale is the market's, with the margin removed. The distance between them is the disagreement.
OutcomeModel fill · market caretProbFairΔ mkt
MUN
26.2%3.81−4.7 pp
Draw
24.8%4.04+0.8 pp
MCI
49.0%2.04+3.9 pp
Ch 02 · Book
27 markets
dixon_coles xi001-20260903
Model fair prices, computed as 1 / probability from a Dixon-Coles score matrix. No bookmaker margin is included, and no bookmaker is quoting these. A real price will be shorter than every number here.
Match result
Man Utd26.2%3.81
Draw24.8%4.04
Man City49.0%2.04
Double chance
Man Utd or draw51.0%1.96
Man Utd or Man City75.2%1.33
Draw or Man City73.8%1.36
Draw no bet
Man Utd34.9%2.87
Man City65.1%1.53
A draw returns the stake, so these prices exclude it.
Win to nil
Man Utd11.6%8.59
Man City23.6%4.25
Clean sheet
Man Utd17.3%5.76
Man City29.3%3.42
Ch 03 · Reasoning
What is driving this
computed, not generated
Computed arithmetic, not generated text. Each driver is a stored quantity: the Elo difference replayed from results, points per game over the last five matches, rolling expected goals, and each absent player's share of team expected goals.
Underlying expected goals+0.66 net xG per match
Favours MUN. MUN +1.34, MCI +0.69 across the form window — results are noisier than the chances behind them
Team rating gap-110 Elo
Favours MCI. MUN 1617 against MCI 1728, computed from stored results; an Elo-driven model scores RPS around 0.2015 on this league against roughly 0.194 for de-vigged closing odds
Home advantage+60 Elo equivalent
Favours MUN. the football default is around 100 Elo points, but Premier League home advantage fell during the crowdless 2020/21 season and settled lower afterwards, so the ratings carry a smaller term
Recent form2.00 against 2.00 points per game
last 5 and 5 completed matches: WLWWD against WWLDW
Unavailable players0% home against 0% away xG missing
5 out and 1 doubtful for MUN, 1 out and 1 doubtful for MCI; measured as the share of recent expected goals those players contributed
Ch 04 · Comparison
Both sides
MUN
WLWWD
9 May – 30 Aug
MCI
WDLWW
13 May – 28 Aug
2.00Points per game2.00
11Goals for11
6Goals against5
13.79Expected goals9.65
1,617Elo1,728
Head to head over 10 meetings in the seasons held (Premier League from 2021-22, Championship from 2024-25): MUN 3, drawn 1, MCI 6. Cup ties and earlier seasons are not included, so a full-history site will show more.
Ch 05 · Market
Bookmaker prices
Real bookmaker prices, collected from published market data.
Market
Selection
Best
Book
Implied
N
h2h
away
2.23
1xBet
44.8%
31
h2h
draw
4.20
Betfair
23.8%
31
h2h
home
3.17
1xBet
31.5%
31
Ch 06 · Availability
Man Utd
0.0% xG out
Share of attacking output is what moves the number, so it is shown rather than implied.
AmadMIDinjured0%
Unspecified injury - Unknown return date
BalebaMIDinjured0%
Ankle injury - Expected back 19 Sep
De LigtDEFinjured0%
Back injury - Unknown return date
HeatonGKPinjured0%
Hamstring injury - Unknown return date
UgarteMIDinjured0%
Knee injury - Unknown return date
MountMIDdoubtful75%
Foot injury - 75% chance of playing
Ch 06 · Availability
Man City
0.0% xG out
Share of attacking output is what moves the number, so it is shown rather than implied.
DokuMIDinjured0%
Calf injury - Expected back 5 Sep
Matheus N.DEFdoubtful75%
Unspecified injury - 75% chance of playing
Ch 07 · Simulation
Re-price without a player
Removing a player scales that side's expected goals by their share of recent attacking output, then rebuilds the score matrix. Every market moves together.
Ch 09 · Divergence
Where the model differs from a price
Shown with the model's own record, because that is what decides how much these are worth.
h2h awaymodel 49.0% vs 2.23 (1xBet)+9.3 ppstake 1.9%medium confidence
h2h drawmodel 24.8% vs 4.20 (Betfair)+4.0 ppstake 0.3%low confidence
Ch 10 · Calibration
How this model actually scores
Dixon-Coles bivariate Poisson fitted to Premier League results with exponential time decay. Statistical model, no language model involved.
loses to closing market2,281 out-of-sample fixtures
Metric
Model
Market
Δ
Ranked probability score
0.2047
0.1965
+0.0082
Log loss
0.9882
0.9623
+0.0259
Brier
0.5883
0.5711
+0.0172
Accuracy
52.5%
54.8%
−2.3 pp
Lower is better for the first three. These probabilities are for reading a fixture, not for expecting profit against closing prices. What these numbers mean