Team Form: How to Evaluate Recent Matches

Team form is how a team has played in its recent matches: how many points it took, how many goals it scored and conceded, how convincingly it created chances. To assess team form properly, a string like "W-W-D-L-W" is not enough. You need to split matches by venue, account for the strength of the opponents and check whether the results are backed by the quality of play. Here is how to do that.

Why the last 5 matches are the usual sample

Five matches are a compromise between freshness and reliability. Take fewer and a single fluky win or loss skews the picture. Take many more and the sample includes games played with a different squad, under a different coach or in a different part of the season.

The xG.Football model also uses a team's last 5 matches. For every match the form snapshot is fixed before kick-off, so nothing that happened later ever gets into it. Only finished matches with available statistics are included.

Keep in mind that five matches are a guideline, not a law. Three straight wins over bottom sides and three wins over teams from the top half mean different things, even though they look the same in a form line.

How to assess team form: five slices

A single form line mixes very different matches. It is more useful to look at several slices separately:

1. All matches. The overall picture: is the team picking up points or not.
2. Home matches only, for the home side. Many teams are noticeably stronger at home.
3. Away matches only, for the away side. The other side of the same effect.
4. Head-to-head. The history of matches between this exact pair.
5. Head-to-head at the home side's ground. The narrowest slice: same teams, same stadium.

These are exactly the five groups of matches the xG.Football model stores for every game. For a match result prediction the model compares the teams on overall form, home and away form, goal threat, goal difference and head-to-head. For totals (corners, shots on target, cards) it checks the share of matches in each slice where the stat went over the line.

Form stats to look at beyond points

Points show the result but do not explain it. Form stats become useful when you look at the reasons:

• Goals scored and conceded. A team winning 1-0 and a team winning 3-2 are in different situations.
• xG. If a team is collecting points but is consistently outplayed on chance quality, the run will probably end. Conversely, a team without wins but with high xG often starts winning soon.
• Shots on target and corners. They show how much pressure a team applies even when it does not score.
• Opponent strength. Check where in the table the teams it took points from are sitting.

Common mistakes when reading form

• Mixing home and away. "Five wins in a row" may be five home games, while the next one is away.
• Missing a change of coach or squad. After a coaching change, old matches say little about the new team.
• Overrating streaks. Three or four matches in a row are too few to call it a pattern.
• Reading results without context. A win with ten men or a defeat after an early red card are special cases.
• Confusing form with class. A strong team in poor form can still be stronger than a weak team in good form.

Where to see form on xG.Football

Recent matches and stats for every club are on the team and competition pages: results, goals and season figures. On a match page, both teams' form is broken down into the same slices the model uses: all matches, home, away and head-to-head. Click a sample to open the list of matches behind the calculation.

Ready-made estimates in which form is one of the key signals are collected in predictions. Form explains a lot but not everything: a prediction is still a probability, and even a team in excellent form can lose.

The short version

Team form is a useful but easily distorted signal. Assess it over the last 5 matches, split home and away, look at goals and chance quality rather than points alone, and account for opponent strength. Then form becomes support for a prediction rather than a trap.