How accurate is the Poisson football analysis calculator?

The Poisson model historically calibrates within roughly ±2% of actual outcome frequencies on settled matches. No statistical model predicts individual football matches with certainty — the sport has high variance. The calculator's value comes from identifying bets where bookmaker odds consistently diverge from mathematically correct prices, not from predicting individual results. Past performance does not guarantee future results.

What does the Poisson model calculate?

The Poisson model calculates attack strength and defensive record for each team relative to league averages, then derives two lambda values — the expected goals for home and away sides. The formula P(k) = (lambda^k x e^-lambda) / k! generates a full 8x8 scoreline probability matrix. Win, draw, loss, BTTS, Over/Under, and Asian handicap probabilities are all derived from this matrix.

Free tool

AI Football Analysis Calculator — Interactive Poisson Distribution Calculator

Enter expected goals below for instant win/draw/loss probabilities, correct scores, Over/Under, BTTS, fair odds & value bets versus 40+ bookmakers

This calculator uses the Poisson distribution model to generate statistically grounded match probabilities for 10+ European leagues. Every number is derived from historical attack and defensive ratings — not gut feel, not trends. The model is transparent: the full methodology is published at footballinteligence.com/methodology.

18+ | Informational tool only | Not financial advice. Football Intelligence does not accept or facilitate bets. Always gamble responsibly.

Expected goals — home team (λH)

Expected goals — away team (λA)

Not sure what to enter? A practical starting point: each team's goals-scored-per-game this season, nudged down against a strong defence (or up against a weak one), then adjusted roughly +10–20% for home advantage. Most of Europe's top-five leagues average 2.6–3.0 combined goals per game; recent form (last 5–6 matches) is usually a better guide than a full-season average.

Home win44.1%

Draw25.5%

Away win30.4%

Over 2.5 goals50.6%

Under 2.5 goals49.4%

BTTS — Yes54.3%

Top 5 most likely correct scores

Calculation: P(k) = (λ^k × e^−λ) / k!, evaluated for 0–9 goals on each side (a 100-cell scoreline matrix) assuming home and away goals are independent. Runs entirely in your browser — nothing is sent to our server. See how the formula works for the full derivation.

Check a bookmaker price against the model

Market

Bookmaker decimal odds

Model probability44.1%

Implied probability47.6%

Value edge-7.3%

Value edge = model probability × bookmaker odds − 1. A positive edge is not a guaranteed win — individual match results are random regardless of the maths. Edge only pays off over a large sample of similarly-priced bets, and only if your λ inputs are genuinely accurate. Once you have an edge, size the stake with the Kelly Criterion calculator, or check the raw value math on the dedicated value bet calculator.

Use the live calculator now

Open the dashboard to access all upcoming matches across 10+ leagues. The AI analysis calculator runs automatically on every match.

Open AI Football Calculator

Free plan · No credit card · 18+

What the AI football analysis calculator does

Football Intelligence is a statistical engine that treats football as a Poisson process. For any upcoming match, it computes:

Lambda values — expected goals for home and away teams, derived from season-long attack strength and defensive record
Win / draw / loss probabilities — percentage likelihoods summed from the scoreline matrix
Fair decimal odds — the mathematically correct price for each outcome (before bookmaker margin)
Value edge — the percentage difference between fair odds and live bookmaker odds from 40+ sportsbooks
Kelly Criterion stake — the fraction of bankroll to stake to maximise long-term growth given the edge
AI narrative — a plain-English match analysis generated by Claude Sonnet 4.6 contextualising the numbers

The calculator covers Premier League, La Liga, Serie A, Bundesliga, Ligue 1, Champions League, Europa League, Eredivisie, Primeira Liga, and select cup competitions. New leagues are added regularly.

How the Poisson formula works

The Poisson distribution models the probability of a discrete number of events occurring in a fixed interval — in football, goals in 90 minutes. Given an expected number of goals (lambda), the probability of exactly k goals is:

P(k) = (λ^k × e^−λ) / k!

Where:

λ (lambda) = expected goals, calculated from team attack strength × opponent defensive record × league average × home advantage multiplier
k = the exact number of goals (0, 1, 2 ... 7+)
e = Euler's number (2.71828...)

Calculating this for k = 0 through 7 for both home and away teams produces an 8×8 scoreline probability matrix. Summing the cells above / below / on the diagonal gives win, draw, and loss probabilities respectively. BTTS (both teams to score) is 1 minus the probability that either team scores zero goals.

Football Intelligence adds Bayesian regularisation to prevent small sample sizes from producing extreme lambda values at the start of a season, and applies a home advantage multiplier calibrated from multi-season league data. The full derivation is at the methodology page.

How to use the calculator

Open the dashboard and select a match
Sign in free (no card required) and go to the Matches tab. Click any upcoming match. The AI analysis calculator loads automatically for every match across the covered leagues.
Read the probability output
The match detail panel shows home win / draw / away win percentages, lambda values (expected goals), and fair odds. Compare the fair odds against the live bookmaker odds shown beneath — a green edge indicator means the bookmaker is offering more than the fair price.
Size your stake with Kelly Criterion
If there is a positive value edge, the Kelly Criterion calculator shows the optimal stake as a percentage of your bankroll. We recommend Half-Kelly (50% of the full Kelly stake) to reduce variance from model uncertainty. Never bet more than the Kelly suggests — overbetting leads to bankroll ruin even with a genuine edge.

The Claude AI analysis layer

On top of the Poisson numbers, Football Intelligence runs an AI narrative layer powered by Claude Sonnet 4.6 (Anthropic). When you open a match, the AI reads the model output and generates a 150-200 word match analysis covering:

Which team the model favours and by how much
The most statistically likely score range
Which markets (BTTS, Over/Under 2.5, Asian handicap) have the strongest edge
Any contextual factors the model cannot capture (cup competitions, rotation risk)

You can also ask SportGPT (the full conversational interface) any football analysis question in plain English. Free accounts get 2 queries per day; PRO gets 15, Elite gets 40.

Example: how the calculator produces odds

Consider a hypothetical Premier League match. The Poisson model derives:

Home team lambda: 1.52 (expected goals, based on recent attack strength vs opponent's defensive record)
Away team lambda: 1.18 (expected goals)

Feeding these into the Poisson formula for k = 0 to 7 goals each produces a scoreline matrix. Summing the relevant cells:

Outcome	Model probability	Model fair odds
Home win	44.8%	2.23
Draw	25.6%	3.91
Away win	29.6%	3.38

If a bookmaker is offering the home win at 2.40 (implying 41.7%), the value edge is (44.8% − 41.7%) / 41.7% = +7.4%. With a £1000 bankroll, full Kelly would suggest a £33 stake; Half-Kelly would be £16.50.

This is exactly the calculation Football Intelligence performs live for every upcoming match across 40+ bookmakers simultaneously.

Finding value bets with the AI calculator

The calculator automatically flags every match where the model's fair odds are better than at least one live bookmaker price. These are the value bets. The AI Picks tab collects only the highest-confidence value bets — those meeting Football Intelligence's data-driven risk gate built from 380+ settled historical bets.

Value bets are not guaranteed wins. They are bets where the mathematical expected value is positive if your probability estimate is accurate — over a large sample such bets should net positive even though individual bets lose. It is the disciplined approach professional bettors use, but it is not a guarantee of profit: the edge is only real when your model genuinely beats the bookmaker's price.

For a deeper explanation of how to identify and act on value bets, see the Football Intelligence value betting guide.

Common questions

Is the AI football analysis calculator free?

Yes. The free plan includes 1 featured prediction daily, league standings, live scores, and a €1000 DEMO bankroll — no credit card required. The PRO plan ($4.99/month) unlocks unlimited match analysis across all leagues and the full live Value Calculator comparing your probabilities against 40+ bookmakers in real time.

How accurate is the Poisson football calculator?

The Poisson model historically calibrates within roughly ±2% of actual outcome frequencies on settled matches. No statistical model predicts individual football matches with certainty — football has high variance. The calculator's value is in finding bets where bookmaker odds are consistently worse than the mathematical fair price, not in predicting individual results.

Past performance does not guarantee future results. Never bet more than you can afford to lose.

What is a value bet in football?

A value bet exists when the bookmaker's implied probability is lower than the model's estimated probability. If the calculator gives a team a 55% win chance and the bookmaker implies 47.6% (odds of 2.10), there is a +15.6% value edge. Over many such bets, positive expected value generates long-term returns. See the full value betting guide for the complete explanation.

What does the Poisson model actually calculate?

For each team it calculates an attack strength score and defensive record score, normalised to the league average. Combined with a home advantage multiplier, these produce two lambda values (expected goals). The formula P(k) = (λ^k × e^−λ) / k! generates probabilities for 0–7 goals for each side, building an 8×8 matrix from which all market probabilities are derived. The full methodology explains every step.

How do I use the Poisson calculator on this page?

Enter expected goals (lambda) for the home and away team in the calculator at the top of this page — a rough starting point is each team's goals-scored-per-game this season, adjusted for opponent strength and home advantage. The calculator instantly computes win/draw/loss probabilities, Over/Under 2.5, BTTS, and the 5 most likely correct scores, all in your browser. Optionally enter a bookmaker's decimal odds for any of those markets to see the implied probability and the model's value edge. For stake sizing once you have an edge, use the Kelly Criterion calculator.

Try the AI football calculator now

Free plan, no credit card. Open any upcoming match and see the Poisson probability breakdown, fair odds, and live value edges.