When do player ratings actually work in football?

Player ratings work when they measure what they can see: goals, assists, saves, tackles won. They fail when they try to measure what they cannot: defensive positioning, movement off the ball, the pass that breaks a press before the assist. The gap between a 6.8 and a 7.2 is often noise, not signal.

For anyone who has ever stared at a match rating and thought "that does not match what I watched", the instinct is correct. Ratings are a blunt instrument. But they are not useless. The trick is knowing when the number means something and when it is just a number.

The mechanism: what ratings actually capture

Every player rating system -- whether from WhoScored, Sofascore, FotMob, or the old-school newspaper marks out of 10 -- follows the same logic. They assign a baseline score, then add or subtract for events that happen during a match.

A goal is +X. An assist is +Y. A completed dribble is +Z. A yellow card is -A. A defensive error leading to a shot is -B. The final score is the sum of those adjustments.

The problem is obvious: not everything that matters is an event. A centre-back who reads danger and steps up three metres to catch a striker offside does not get a positive event. The striker who makes that run but does not receive the pass gets nothing. The midfielder who shifts five yards left to block a passing lane -- invisible.

Ratings work best for positions where the game is event-dense. Forwards and attacking midfielders produce shots, chances, dribbles, and key passes. Goalkeepers make saves and claim crosses. These are countable. A goalkeeper who makes eight saves and keeps a clean sheet will rate highly, and that rating will match what you saw.

The problem lives in central defence and central midfield, where the best work is often the work that prevents something from happening.

History: how we got here

Player ratings in English football did not start with algorithms. They started with a man in the press box with a notebook.

Newspapers have published match ratings since at least the 1960s. A reporter watched the game and gave each player a mark out of 10. The system was simple: 6 was average, 7 was good, 8 was very good, 9 was exceptional, 10 was perfect. A 5 meant a poor game. A 4 meant dreadful. The ratings were subjective, but they came from someone who watched the full 90 minutes and understood the context.

The problem was consistency. One reporter might give a 7 for a performance another would call a 6. There was no standard. A player could get a 7 from the Guardian and a 5 from the Daily Mail for the same match.

Then came the data era. Companies like Opta began collecting every event in a match. Passes, tackles, shots, interceptions, clearances. By the late 2000s, rating systems shifted from subjective opinion to weighted algorithms. The number was no longer what one journalist thought. It was a calculation based on thousands of data points.

That made ratings more consistent. It did not make them more accurate. The algorithm replaced one form of bias with another: the bias of counting only what can be counted.

In 2019, the Premier League introduced its own official player ratings through a partnership with Opta and EA Sports. The system uses a complex algorithm weighing 35+ metrics. It is the most sophisticated rating system in football. It still gets things wrong every weekend.

Edge cases: when ratings break

There are three situations where player ratings consistently fail.

The defensive midfielder who does not tackle. A holding midfielder who reads the game well and intercepts passes before they reach the defence often has low tackle numbers. The algorithm sees few tackles and assigns a lower score. What it misses is that the player did not need to tackle because he was always in the right position. Rodri at Manchester City routinely produced matches with low tackle counts and high positional intelligence. His ratings were good but rarely reflected his actual influence.

The centre-back in a low block. A defender who makes 15 clearances and 10 blocks will rate highly. A defender who plays in a high line, steps up, and prevents the ball from arriving in the box at all gets no clearance events. The second defender is often the better one. The algorithm rewards the first.

The forward who does not score. A striker who occupies two centre-backs, creates space for teammates, and makes intelligent runs but does not get on the scoresheet will rate poorly. The system values goals and assists above all else. If a forward plays a 9/10 performance in terms of movement and pressing but does not score, the rating will say 6.5. This is the most common complaint about ratings among fans who watch their team every week.

When to trust the number

Ratings become useful when you aggregate them. A single match rating is noisy. A season average across 30+ matches smooths out the noise. If a player averages 7.4 over a full Premier League season, they are almost certainly performing at a high level. The outliers balance out.

Ratings also work for comparison within the same position. Two centre-forwards playing in similar systems can be compared on their season averages because they face the same event bias. The number is not perfect, but it is directional.

For a single match, ignore the rating. Watch the game. Your eyes are still better than the algorithm.

Key takeaways

• Player ratings measure countable events: goals, assists, tackles, saves. They miss defensive positioning, off-ball movement, and intelligent pressing.
• Single-match ratings are unreliable. Season averages across 30+ matches are more meaningful.
• Forwards and goalkeepers rate more accurately than centre-backs and defensive midfielders because their work produces more countable events.
• The newspaper 10-point system was subjective but contextual. Modern algorithms are consistent but context-blind.
• When a player's rating disagrees with what you saw, trust your eyes for that match. Trust the aggregate over a season.

FAQ

Do player ratings matter for Fantasy Premier League?

FPL uses its own bonus points system based on the Bonus Points System (BPS), not match ratings. The BPS rewards clear actions: goals, assists, clean sheets, saves, dribbles completed. It is event-based and predictable. A player who scores and keeps a clean sheet will almost always get bonus points.

Which rating system is most accurate?

No system is consistently accurate for every position. The Opta-powered official Premier League ratings use the most data points. Sofascore and WhoScored are similar. The best approach is to use season averages from any major provider and treat single-match ratings with skepticism.

Why do defenders sometimes get low ratings in clean sheets?

A defender in a team that dominates possession may have little defending to do. The algorithm sees few tackles, few clearances, and few interceptions. The rating drops even though the defender did their job by preventing chances before they happened.

Can ratings predict future performance?

Partially. A player with a high season average in a stable system is likely to maintain that level. A player with a high rating from one match is not. Regression to the mean is strong in rating data. One 9.0 performance does not make a world-class player.