Poisson Distribution Calculator
Model match scoring with Poisson probabilities for totals and props.
Enter an expected average (lambda) for the number of events—goals, corners, cards—and see probabilities for exact counts and over/under cutoffs.
Poisson Calculator
Adjust inputs to see updated results.
Table of Contents
How to Use This Calculator
Set lambda (λ) to your expected average count of events in the match or period you are modeling (goals, corners, cards, etc.). Set the line to the total you want to test against—for example 2.5 for a classic goals total. The tool reports P(over line), the Poisson probability that the realized count is strictly above that number.
Pick a defensible λ
Use team xG trends, league averages, or your own projection. λ is not “odds”; it is the mean of the distribution. If you change λ, the whole curve shifts.
Match the line to the market
Enter the same half-point total the book posts (2.5, 1.5, 10.5 corners, etc.). Half points avoid push logic in this simplified model.
Read P(over) as a benchmark
Compare the model probability to the implied probability from decimal odds (1/price) to see whether the line looks cheap or expensive—before you bet.
Cross-check extremes
Very large λ or odd lines can still be computed, but the model assumes the process is Poisson. Use common sense on injury time, red cards, and sample size.
What Is the Poisson Model?
The Poisson distribution describes the probability of seeing exactly k events when events occur independently at a constant average rate. In betting, people use it for goal totals, corner counts, booking points, and other count markets.
Each integer count k has probability P(X = k) = e−λ λk / k!. Summing those from the line upward gives P(over). That is what the calculator implements for the “over” side of a total.
How the Math Maps to Betting
For a posted total T (e.g. 2.5), “over” means the realized count is 3 or more. Under our half-point convention, P(over) = Σk > T P(X = k).
P(X = k) = e−λ · λk / k! → P(over T) = Σk>T P(X = k)
Compare P(over) to the implied probability from odds to frame value. Pair with implied probability and value tools on the site.
Worked Examples
Example 1 — Goals: λ = 2.5, line 2.5
Enter Lambda 2.5 and Line 2.5. The model sums mass on counts 3, 4, … For λ = 2.5 this gives P(over 2.5) ≈ 45.62%. If a book prices over 2.5 near even money (50% implied), the raw Poisson view says the over is slightly expensive—unless you believe λ should be higher than 2.5.
Example 2 — Low total: λ = 1.2, line 0.5
λ = 1.2, line = 0.5. Then P(over 0.5) = 1 − P(0) = 1 − e−1.2 ≈ 69.9%. Useful when you model a low-scoring match but still want the chance of at least one event (e.g. “anytime” style thinking).
Example 3 — Sanity-check against the market
If decimal odds on over 2.5 are 2.00, implied probability = 50%. Your model at λ = 2.5 yields ~45.6% for over—so you would need a reason (team news, weather) to lift λ above 2.5 before liking the over. The calculator is the first pass; the narrative closes the case.
Limits and Caveats
Real football is not perfectly Poisson: scoring rates change with game state, and corners may cluster. Use this as a transparent baseline, not a live trading model. Always verify rules for voids, Asian lines, and pushes at your book.
Frequently Asked Questions
λ is the expected average number of events in the scope you are modeling—often one match for goals. Raise λ if you expect more scoring; lower it for a cagey game.
This Poisson tool does not input odds—it outputs a probability. Convert that to implied fair odds as 1/P if you want to compare to decimal prices.
The calculator focuses on the cumulative probability above your line, which matches how many total markets are quoted. Exact goal probabilities are the individual terms P(X = k).
It is a standard first approximation. Bivariate or time-varying models can be more accurate but need more inputs. Start simple, then adjust λ for context.
Yes, if you are comfortable estimating λ for those counts. Corner and card processes deviate more from ideal Poisson, so treat outputs as directional.
See our expected value and implied probability calculators, and the related calculators block below.
