What are data analysts in Premier League football clubs?

Football data analysts work at the point where raw match events become decisions. In the Premier League, the rules give them a fixed data cycle: 20 clubs play 38 matches each, home and away, for 380 games a season, and every result changes the table. If you think their job is just counting goals, you are missing most of the work.

The mechanism

Start with the league table. The Premier League has 20 clubs, each playing 38 matches, once at home and once away. That creates 380 matches per season. Three points for a win, one for a draw, zero for a loss. The bottom three are relegated to the Championship, and three clubs come up, two automatically and one through the play-off.

Now add tiebreakers. After points, position is decided by goal difference, then goals scored, then head-to-head record. If still level and the position decides the title, relegation, or European qualification, a play-off at a neutral venue may be used. An analyst cannot just report that a team won. They need to know how many goals were scored and conceded, because goal difference is a tiebreaker. Every extra goal in a win is a tiebreaker asset.

Then add the squad rules. A club's 25-player squad list for players over 21 must include at least 8 homegrown players, meaning players trained in England or Wales for three years before turning 21. Under-21 players do not count against the 25. This is a data problem as much as a scouting problem, because a club must know which players qualify and when they qualify.

Substitutions are also a data constraint. A club can make five substitutions in a match, in a maximum of three windows plus half-time. An analyst's job is to help a manager use those windows well: who comes on, at what minute, and with which players still eligible to be used.

European qualification is another layer. The top four go to the Champions League league phase. Fifth and the FA Cup winner go to the Europa League. The League Cup winner goes to the Conference League play-off. Additional Champions League places can be earned through UEFA coefficient performance. If a cup winner has already qualified through league position, the place passes down. This gives analysts a second table to model.

History/evolution

The Premier League was founded on 20 February 1992, when First Division clubs broke away to negotiate their own broadcast deals. The first season, 1992/93, had 22 clubs. The league reduced to 20 clubs in 1995/96 and settled into the format analysts work with today. That historical shift matters because it changed the size of the dataset. A 22-club league creates a different match calendar and a different points race. The current 20-club format, with its 38 matches per club and 380 matches per season, is the base every analyst works from.

Edge cases

Relegation is the most visible edge case. The bottom three go down, while three clubs come up from the Championship, two automatically and one via play-off. An analyst working with a promoted club has to plan a squad that meets Premier League rules: a 25-player list with at least 8 homegrown players, and under-21 players outside that list.

Tiebreakers also produce edge cases. If two clubs are level after points, goal difference, goals scored, and head-to-head, and the position is for the title, relegation, or European qualification, a play-off at a neutral venue may be used. That means a season can end without the final table being settled.

Homegrown status is an edge case hidden in the calendar. A player counts as homegrown if trained in England or Wales for three years before turning 21. An analyst must track exact dates, because a player can move from under-21 status to counting against the 25-player squad list at a specific point.

Substitutions are an edge case within every match. Five substitutes can be made, but only in a maximum of three windows plus half-time. That means a manager cannot simply make five separate changes whenever they want; the changes have to be grouped. The analyst has to keep track of windows as well as players.

Even the scoring system has edge cases. In FPL scoring, a goal is 6 for a goalkeeper or defender, 5 for a midfielder, 4 for a forward. An assist is 3. A clean sheet is 4 for a goalkeeper or defender, 1 for a midfielder. Own goal minus 2, red card minus 3, penalty miss minus 2. A goalkeeper gets 5 for a penalty save and 1 for every 3 saves. Every two goals conceded costs a goalkeeper or defender minus 1. This is the kind of event-level model that shows why one point can be won or lost in a single action.

FAQ

What do data analysts in football do? They study league structure, tiebreakers, squad rules and substitution limits. In the Premier League, that means tracking 38 matches per club, goal difference, the 25-player squad list, the 8-homegrown-player rule, and the five-substitute limit.

Why is goal difference important for data analysts? Because the Premier League uses goal difference as the first tiebreaker after points. If clubs finish level, goal difference decides, then goals scored, then head-to-head. A late goal can move a team up or down the table.

How do data analysts handle the homegrown player rule? They track which players qualify as homegrown. A player counts if trained in England or Wales for three years before turning 21, and at least 8 homegrown players must be in the 25-player squad. Under-21 players do not count against the 25.

Do data analysts use Fantasy Premier League scoring? FPL scoring is a useful model, not the job itself. It gives 6 points for a goal by a defender, 5 for a midfielder, 4 for a forward, and adds bonus points and deductions. Analysts use similar event scoring to compare players.

Key takeaways

• For data analysts, the Premier League's fixed structure is the starting point: 20 clubs, 38 matches each, 380 matches per season.
• Tiebreakers matter after points: goal difference, then goals scored, then head-to-head, so every goal counts.
• The homegrown rule forces analysts to track training history: 8 of the 25 players over 21 must be homegrown, and under-21 players sit outside the 25.
• Substitution limits create in-game decisions: five substitutes, maximum three windows plus half-time.
• For data analysts, European qualification is a live model: top four to the Champions League, fifth and the FA Cup winner to the Europa League, League Cup winner to the Conference League play-off.