Fixture 26scheduledGW 3

Man City v Coventry City

Sat 5 Sept 14:00Zmarket prices held

Expected goals2.14 — 0.86

Expected points2.20 — 0.60

Over 2.557.9%fair 1.73

Both score51.7%fair 1.93

Most likely2-011.4%

Markets priced27

Ch 01 · Outcome

Match result

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

MCI

66.8%1.50−16.3 pp

Draw

19.9%5.03+8.6 pp

COV

13.3%7.51+7.7 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.

Thin sample

Coventry City has only 2 completed matches in the training window, so its fitted strength is imprecise. Treat the long prices in this book as arithmetic rather than information: a real bookmaker would quote a much shorter price on the outsider.

Matches behind each fit: Man City 149 · Coventry City 2

Match result

Man City66.8%1.50

Draw19.9%5.03

Coventry City13.3%7.51

Double chance

Man City or draw86.7%1.15

Man City or Coventry City80.1%1.25

Draw or Coventry City33.2%3.01

Draw no bet

Man City83.4%1.20

Coventry City16.6%6.01

A draw returns the stake, so these prices exclude it.

Win to nil

Man City37.2%2.69

Coventry City6.1%16.32

Clean sheet

Man City42.2%2.37

Coventry City11.1%9.01

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.

Team rating gap+242 Elo

Favours MCI. MCI 1728 against COV 1486, 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

Underlying expected goals+0.80 net xG per match

Favours MCI. MCI +0.69, COV -0.11 across the form window — results are noisier than the chances behind them

Home advantage+60 Elo equivalent

Favours MCI. 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 1.80 points per game

Favours MCI. last 5 and 5 completed matches: WWLDW against LLWWW

Unavailable players0% home against 0% away xG missing

1 out and 1 doubtful for MCI, 3 out and 0 doubtful for COV; measured as the share of recent expected goals those players contributed

Ch 04 · Comparison

Both sides

MCI

WWLDW

13 May – 28 Aug

COV · incl. Championship

WWWLL

21 Apr – 29 Aug

2.00Points per game1.80

11Goals for12

5Goals against6

9.65Expected goals (2 of 5)1.74

1,728Elo1,486

Ch 05 · Market

Bookmaker prices

Real bookmaker prices, collected from published market data.

Market	Selection	Best	Book	Implied	N
h2h	away	18.00	Betfair	5.6%	43
h2h	draw	9.00	Smarkets	11.1%	43
h2h	home	1.21	GTbets	82.6%	43

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 06 · Availability

Coventry City

0.0% xG out

Share of attacking output is what moves the number, so it is shown rather than implied.

Kesler-HaydenDEFinjured0%

Hamstring injury - Unknown return date

WoolfendenDEFinjured0%

Knee injury - Unknown return date

WrightFWDinjured0%

Thigh injury - Unknown return date

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 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