Rising Transfers · Methodology
Per 90 Stats in Football Explained: Why They Matter More Than Raw Numbers
Per-90 statistics measure a footballer's output per 90 minutes of actual playing time, rather than per game or per season. This normalisation makes players genuinely comparable regardless of how many games they started, how long they were substituted on, or which league they play in. A striker with 12 goals in 3,000 minutes is producing 0.36 goals per 90 — a player with 8 goals in 1,400 minutes is producing 0.51 goals per 90. Without per-90 normalisation, the first player looks better. With it, the second player is clearly the sharper finisher. Per-90 is now the standard measurement unit used by professional clubs, analytics platforms, and AI scouting systems to evaluate and compare players.
Why Raw Stats Mislead You
Comparing players on total goals, assists, or tackles is intuitive but systematically misleading. A central midfielder who plays 90 minutes every game for a high-possession team will accumulate more passing stats than an equally skilled player used as a 60-minute impact substitute. A striker who plays 2,000 minutes will have more goals than a superior finisher who played 1,200 minutes through injury. Raw totals reward availability as much as they reward quality.
The same problem affects seasonal comparisons. A player who played 34 games in a season cannot be fairly compared on totals to one who played 22. Leagues have different match lengths in terms of effective playing time, and managers use squad rotation differently. Every one of these variables corrupts a raw total as a measure of individual ability.
Per-90 stats do not solve every problem — a player in a dominant team will have easier chances, and a player in a bottom-half team may be exposed to more defensive pressure — but they remove the single biggest source of noise in player comparison: playing time differences.
How Per-90 Statistics Work
The calculation is straightforward: divide any raw statistic by total minutes played, then multiply by 90. The result is what a player would produce if they played exactly one full match at their observed rate.
The Basic Calculation
Goals per 90 = (Total goals ÷ Total minutes played) × 90. The same formula applies to any countable stat: assists, shots, key passes, tackles, interceptions, progressive carries. A player with 6 goals in 1,620 minutes produces 0.33 G/90. A player with 4 goals in 720 minutes produces 0.50 G/90. The second player is a more prolific finisher, despite having fewer goals.
Which Stats Benefit Most From Per-90 Normalisation
Attacking output metrics — goals, shots, shots on target, expected goals (xG), big chances — benefit most from per-90 normalisation because they are the most sensitive to playing time variation. Creative metrics — key passes, chance creation, assists, expected assists (xA) — benefit similarly. Defensive metrics like tackles and interceptions are more context-dependent (a team that concedes less possession will have fewer tackle opportunities), but per-90 still removes the playing time distortion.
Minimum Minutes Threshold
Per-90 stats require a minimum playing time threshold to be meaningful. A player who came on for 12 minutes and scored a penalty has 7.5 goals per 90 — technically accurate but statistically meaningless. Most analytics platforms apply a minimum of 900 minutes (roughly 10 full games) before treating per-90 numbers as representative. Rising Transfers flags profiles with limited samples explicitly, so you know when a number is a small-sample signal versus a stable performance indicator.
Position-Specific Interpretation
The same per-90 number means different things in different positions. 0.5 G/90 is elite output for a central midfielder; it is ordinary for a centre-forward in a top league. 0.1 G/90 is unremarkable for a striker and exceptional for a centre-back. Per-90 stats should always be read relative to position norms and league context — which is why the most useful scouting tools present per-90 data with peer comparisons, not in isolation.
Per-90 vs Per-Game
Per-game stats (averaging over matches played) treat a 45-minute substitute appearance the same as a 90-minute start. Per-90 treats them correctly — a player who scores in their only 30 minutes on the pitch contributed exactly one-third of a game's worth of output. Per-90 is strictly more accurate than per-game for measuring individual output rate, which is why it has become the industry standard.
Per-90 in Practice: Reading a Player Profile
Consider two Premier League wingers. Player A finished the season with 9 goals and 7 assists — a strong return. Player B finished with 6 goals and 4 assists — on paper, clearly worse. But Player A played 3,150 minutes across 35 appearances; Player B played 1,800 minutes across 26 appearances (many as a substitute).
On a per-90 basis: Player A produced 0.26 G/90 and 0.20 A/90. Player B produced 0.30 G/90 and 0.20 A/90. Player B is marginally the more efficient contributor — and had he played the same minutes as Player A, the projection is 10.5 goals rather than 9.
This is the insight that changes transfer valuations. A club looking to replace Player A with a player of similar output would be making a mistake if they overlooked Player B based on raw stats. Per-90 normalisation surfaces the correct comparison.
Rising Transfers applies per-90 normalisation across 6,000+ players as the foundation for DNA style matching and player comparison. When you see a similarity score between two players on the alternatives page, it is per-90 performance vectors that are being compared — not totals.
Frequently Asked Questions
What does per 90 mean in football?
Per 90 means per 90 minutes of playing time — one full match. A per-90 stat is calculated by dividing a player's total for a given metric (goals, assists, tackles, etc.) by their total minutes played, then multiplying by 90. This normalises performance to a standard unit of playing time, making players comparable regardless of how many games they started or how many minutes they actually played.
Why are per 90 stats better than total stats?
Total stats reward playing time as much as they reward quality. A player who plays 3,000 minutes will accumulate more goals and assists than an equally good player who played 1,500 minutes, simply because they had more opportunity. Per-90 stats remove this distortion by measuring what a player produces per unit of playing time, making it possible to compare a regular starter with a rotation player, or a player returning from injury with one who stayed fit all season.
What is a good goals per 90 in football?
Context matters significantly. For a centre-forward in a top European league, 0.5+ G/90 is elite (only the very best strikers sustain this). 0.3–0.5 G/90 is above average. Below 0.2 G/90 suggests a striker who is not primarily a goalscorer. For a central midfielder, 0.2+ G/90 from open play is excellent. For a defender, 0.1 G/90 is notable. The benchmark always depends on position and role.
How many minutes do you need for per 90 stats to be reliable?
Most analysts apply a minimum threshold of 900 minutes (roughly 10 full matches) before treating per-90 numbers as a stable indicator rather than a small-sample noise signal. Below 900 minutes, a player's per-90 stats can be heavily distorted by a single good or bad performance. Rising Transfers flags profiles with limited samples so you can judge reliability accordingly.
Do football clubs actually use per 90 stats?
Yes. Per-90 statistics are now the standard unit of player evaluation in professional football analytics. Most Premier League clubs, and elite clubs across Europe, use per-90 metrics as the foundation of their data scouting. Companies like StatsBomb, Opta, and Wyscout provide per-90 data as a core product. The shift happened progressively between 2010 and 2020, accelerated by the Moneyball-style analytics movement reaching football.
See per-90 stats for any player in the Rising Transfers database.