StatFairEspañolPractical guide ·
How to calculate football statistics: a worked example
To calculate football statistics, define your sample, collect complete records and use a formula that someone else can check. Start with one question: a team's home performance is not the same thing as the total goals in all its matches.
1. Define the sample before calculating
Choose a team, competition, date range and venue filter: home, away or both. Keep that definition unchanged throughout the calculation. For a home-team study, collect its recent home matches rather than quietly replacing missing games with away fixtures. Keep the date, opponent and source beside each record.
The following results are fictional. Each score is written from team A's perspective, with its goals first. These are calculation examples, not real fixtures or a StatFair forecast.
| Match | Goals for | Goals against | Points | Total goals |
|---|---|---|---|---|
| 1 | 2 | 1 | 3 | 3 |
| 2 | 0 | 0 | 1 | 0 |
| 3 | 1 | 2 | 0 | 3 |
| 4 | 3 | 1 | 3 | 4 |
| 5 | 1 | 1 | 1 | 2 |
2. Calculate averages and points per game
Team A scores 7 goals and concedes 5 across five matches. Goals for per game are 7 ÷ 5 = 1.40; goals against per game are 5 ÷ 5 = 1.00. The matches contain 12 goals altogether, so total goals per game are 12 ÷ 5 = 2.40. A team average and a match-total average answer different questions.
Using three points for a win and one for a draw, A collects 8 points: 3 + 1 + 0 + 3 + 1. Points per game are 8 ÷ 5 = 1.60. State the scoring system if you are analysing a competition with a different format.
3. Count frequencies separately
Matches 1, 3 and 4 have over 2.5 goals: 3 out of 5, or 60%. Both teams score in four matches: 80%. Neither percentage can be recovered from the 2.40 average alone. You need the individual records to count how often a condition occurred.
A 2.5-goal line separates totals of two or fewer from totals of three or more; it does not imply fractional goals. With five matches, each result contributes twenty percentage points. Display the count alongside the percentage so readers can see how small the sample is.
4. Treat missing values differently from zero
A confirmed 0-0 is a valid zero-goal result. An unknown score is not a 0-0. Apply the same rule to corners, shots and cards: use only valid observations for each metric and publish its denominator. Five complete scores may contain only four valid corner records.
5. Compare like with like
Show the average, frequency, sample size and venue definition together. Sort records by date to distinguish recent form from a season average. Repeat the calculation with ten matches when available, but do not mix competitions or venue filters without explaining the change.
Use the public football statistics to explore recorded data. A historical percentage describes the sample; it is not a guarantee for the next match. Advanced interactive tools and public reading pages have different access conditions, which are explained on the pricing page.