How to Analyse a Football Match Before Kick-Off: 7 Steps

To analyse a football match before kick-off, ask seven questions, always in the same order: what is at stake for each side, who starts, how they have really played lately, how they do at home and away, what the head-to-head says, how the fixture list looks, and how much all of it is worth in percentages. It takes about fifteen minutes. You don’t need a scout’s eye.

Here is each step with an example. At the end, how to put the answers into one estimate.

Step 1. Check what is at stake

The same match is worth different things to each side. Matchday one is not a relegation six-pointer.

Ask whether either team has already hit its target. Whether a cup tie is three days away. Whether it’s a derby for one of them. A side with nothing to play for often drops the intensity. A side fighting the drop goes all out.

You won’t see this in the table. That is why many previews skip it.

Step 2. Line-ups and absences

The injury list is only the start. What matters is what that player did.

Losing the twelve-goal striker hurts. Losing the holding midfielder who won the ball in front of the back line can cost more: the whole midfield comes apart. So don’t count goals. Count the role.

Check the confirmed line-ups an hour before kick-off. “Predicted XIs” published two days earlier are often wrong.

Step 3. Form, but with quality

Five wins in a row sounds good. The question is against whom, and how.

See whether the team was creating chances or won on one penalty. Look at the opponents in the run. Five wins against the bottom of the Premier League are not five wins against the top. How form is calculated is in the article on team form.

If you have xG, set it next to the goals. When goals run well above xG, the streak rests on finishing luck, and that rarely lasts. The numbers on a match sheet are explained in how to read football match stats.

Step 4. Home and away, separately

The overall table hides gaps. A team can sit fourth and still be poor on the road.

Take the home side’s last five to eight games at its own ground and the visitor’s last five to eight away. Don’t blend them. How much home advantage is really worth depends on the league and the team: we cover it in home advantage in football.

Step 5. The head-to-head, in moderation

Previous meetings are the most quoted number and the least reliable. Three games from four years ago, with other managers and other squads, are an anecdote.

Look at the latest meetings, ideally from the last two or three seasons. Look for patterns of play, not scorelines. Does one side always struggle against the other’s press? Useful. “They won 5-1 in 2019”? Not. The detail is in what head-to-head means in football.

Step 6. Fixtures and fatigue

Three games in eight days, with a European trip in the middle. It shows, mostly in the second half.

Count each side’s rest days and note any long journey. See whether the manager could rotate. Two extra days of rest don’t decide anything alone, but together with the absences they can move your estimate.

Step 7. Turn it into a probability

This is where people get lost. Instead of asking “who wins?”, ask what percentage each outcome has. Three numbers that add up to a hundred.

1. Start from a baseline. In the big leagues the home side wins roughly one match in two, and draws sit near a quarter. Use 45 / 27 / 28 as a neutral starting point.
2. Move in small steps. Better form, and of real quality: a few percentage points. A key centre-back missing: a few points the other way. Never twenty at once.
3. Judge how much confidence the estimate deserves. Little data, a new manager, many changes in the XI: low confidence, even if the numbers look tidy.
4. Name a single risk. What could flip the match? A red card, rotation, the captain’s absence. One thing, not a list.

The result, for example: 52 / 26 / 22, medium confidence, main risk rotation before the cup tie. That is an analysis. It promises no score. It says how the outcomes are split and how far to trust it.

Example of an analysis card. The numbers are illustrative.

And if you don’t want to do it by hand

You can run all seven steps yourself, and for one match it’s a good exercise. With ten matches a day, patience runs out.

Sharkline gathers this data from more than 20 sources and returns what step 7 describes: the probability of each outcome, confidence in percent and the main risk. How it works is on the AI sports analysis page, and the full method in how Sharkline works. What an AI cannot know about a match: what AI does not know about a match.

FAQ

Where do I start when analysing a match? With what’s at stake for both teams. Context first, then line-ups, form and stats. The same run of results means something different in a relegation fight and in mid-table.

How many recent matches should I look at? Five to eight for form, with home and away split. Fewer is noise. More mixes different eras, managers and squads.

What stats should I check before a match? Shots and their quality (xG), goals conceded, form split by home and away, and the line-up. Possession describes a style, not strength.

How do I estimate the probability of a result myself? Start from the league’s typical split and shift it a few points for each relevant factor. The three outcomes must add up to 100%. No formula is free of error: it is an estimate, not a certainty.

What is the difference between analysis and prediction? Analysis shows how the outcomes are split and what drives them. A prediction tries to point at one result. A good analysis admits the score keeps its share of chance. More in AI analysis or a predictions channel.

Can AI analyse a match for me? With the data, yes: it collects stats, form and history, compares sources and gives probabilities. It doesn’t know what never enters the data, like the mood in the dressing room.