Sharkline

← All articles

How to Estimate the Probability of a Football Result Yourself

·5 min read

You estimate a football result by working out how many goals each team is likely to score, then turning those two averages into home win, draw and away win percentages. The standard tool is a Poisson grid, and you can build one in a spreadsheet in ten minutes.

Here is the method with real numbers, and where it breaks.

Step 1. Rate how many goals each team scores

Take the last 10 to 15 league matches. Count goals scored and conceded, home and away separately. Then compare each team with the league average. A side that scores 1.5 goals a game in a league that averages 1.3 attacks 15% better than normal.

For the match you are looking at, multiply three things: the home team’s attack, the away team’s defence, and the league’s average home goals. Do the same for the away side. You end up with an expected number of goals for each team. Call them λ.

Goals scored are noisy. Expected goals (xG) are steadier, so if you have them, use them instead. The idea is in what is xG in football.

Step 2. Turn each average into a goal distribution

Say the home side’s λ is 1.5 and the away side’s is 1.0. The Poisson formula gives the chance of exactly k goals: e^(−λ) × λ^k / k!.

Goals Home (λ = 1.5) Away (λ = 1.0)
0 22% 37%
1 33% 37%
2 25% 18%
3 13% 6%
4+ 7% 2%

Step 3. Build the scoreline grid and add it up

Multiply the home and away chances for every scoreline. A 1–1 draw is 33% × 37%, about 12%. A 0–0 is 22% × 37%, about 8%. Then add:

  • every cell where home goals beat away goals: home win, about 49%;
  • the diagonal (0–0, 1–1, 2–2 and so on): draw, about 26%;
  • every cell where away goals beat home goals: away win, about 25%.

That is your estimate. Notice what it says: the favourite is a favourite, and still fails more often than it wins outright. A 49% home win is a coin flip with a small tilt. For why even a clear edge guarantees nothing, read why nobody can guarantee a sports result.

Step 4. Adjust for what the numbers cannot see

The grid knows nothing about line-ups, injuries, suspensions, a manager’s change or a dead-rubber fixture. Check the news and move the numbers by hand, in small steps. Losing a first-choice striker might cut the home λ from 1.5 to 1.3. The grid then gives you new percentages. Write down why you moved it. The full pre-match routine is in how to analyse a football match before kick-off.

Where the home-made model breaks

  • It assumes goals are independent. They are not: a team that scores often changes how the other plays.
  • Form windows are short, so one 5–0 can distort a whole average.
  • It ignores home advantage unless you build it in. See home advantage in football.
  • You must keep it updated by hand, match after match.
Match analysisAI ANALYSIS
Manchester City

Manchester City — Liverpool

Premier League

Liverpool

Win probability

1 · City

44%

X

28%

2 · Liverpool

28%

Read

Goals at both ends likely

Both attacks are in form, and their meetings usually run high.

74%

Confidence

★★★data quality

The home side have scored two or more in eight home games running.

Risk: rotation ahead of a midweek European tie.

The same output, done automatically: three probabilities, confidence and the main risk. Figures are illustrative.

When a tool does it for you

A spreadsheet gives you one number per outcome. An AI analysis does the same job with more inputs. Sharkline covers 9 sports, combines 20+ data sources, returns the probability of each outcome, a confidence figure in percent and the main risk, and skips a match when data is thin. See the AI sports analysis page for what it covers, how Sharkline works, and how AI predicts a match result for the mechanics. Whichever you use, check that the percentages hold up over many matches: that is covered in can you trust an AI match analysis.

FAQ

How do I calculate the probability of a football match result? Estimate each team’s expected goals, apply a Poisson distribution to both, multiply into a scoreline grid and add up the home win, draw and away win cells.

What is a Poisson model in football? A formula that gives the chance of exactly 0, 1, 2 or more goals when you know a team’s average. It treats goals as independent events.

How accurate is a home-made probability estimate? It is a rough guide. It ignores line-ups and injuries until you adjust for them by hand. Judge it over many matches, not one.

Should I use goals or xG? xG if you have it. It is steadier than goals over a short run of matches.

Is an estimate the same as a prediction? No. An estimate shows how outcomes are spread. A prediction picks one. See AI analysis or prediction channels.

The matches are read. The call is yours.

Open in Telegram