Eight Data Points Scouts Track Beyond Goals and Assists

Data scouting is the use of match statistics to narrow a list of transfer targets before anyone travels to watch them. Goals and assists are where every shortlist starts, and they are almost never where a recruitment analyst stops. The reason is specific: those two numbers travel badly between leagues, teams and roles.

The metrics below are the ones that survive the journey. Each is chosen for the same reason — it isolates something about the player rather than something about the situation the player happened to be in.

The organising question

Every item on this list answers one question: if this player were moved into a different squad, in a different competition, playing a slightly different job, what part of his output would come with him?

Goals travel poorly because they depend on service, on how many chances a team creates, and on a finishing rate that fluctuates heavily over a single season. Assists travel worse still, because an assist requires a teammate to convert. A recruitment model built on those two columns is largely modelling the selling club.

A metric earns its place on a scouting dashboard when it satisfies most of the following:

• It describes an action the player initiates rather than one a teammate completes
• It is normalised for the volume of opportunity the player's team provides
• It is compared against players doing a genuinely similar job
• It is stable enough across a season that a good month cannot manufacture it
• It has a known relationship to the standard of the competition it was recorded in

Very few raw counting statistics pass all five. That is the whole reason recruitment departments build derived measures instead of sorting a spreadsheet by goals.

Availability and minutes load

The first number checked is usually the least analytical: how much football has this player actually played.

Availability is a trait, not an accident. A player with a long record of missing blocks of matches carries risk that no attacking metric offsets, and a player who has played heavy minutes for several consecutive seasons has demonstrated something durable. It also functions as a manager's verdict — sustained selection by successive coaches is a form of assessment produced by people who see every training session.

Analysts look at total minutes, the share of available minutes played, the pattern of absences, and whether minutes came in meaningful fixtures or in dead rubbers.

Age measured against the level played at

Raw age is close to useless on its own. Age relative to the standard of competition is not.

A player holding down regular minutes in a strong league at twenty is a different proposition from one doing the same in a weaker division at twenty-four, even if their output columns look identical. Recruitment models routinely apply an age curve, which assumes output improves toward a peak band and declines afterwards, with the peak arriving earlier for roles that depend on acceleration and later for roles that depend on positioning and passing.

The practical use is projection. A scout is rarely buying current output. They are buying the intersection of current output and the slope the player is on.

Output adjusted for possession, not just per ninety

Per-ninety normalisation fixes the problem of unequal playing time. It does nothing about unequal team context.

A midfielder in a side that dominates the ball will accumulate more passes, more touches and more final-third entries than an equally good midfielder in a side that spends matches defending. Comparing them per ninety compares their teams.

The correction is to normalise by possession — output per one hundred touches, per one hundred passes, or per unit of team possession. This is one of the most consistently informative adjustments in recruitment analysis, because it converts a team-level statistic into something closer to a player-level rate.

Comparison against the right positional peer group

A number means nothing until it has a reference group, and choosing that group is a decision with consequences.

Comparing a wing-back against all defenders makes him look extraordinary going forward. Comparing him against other wing-backs in similar systems tells you whether he is actually unusual. The peer group needs to match on position, on the tactical role within that position, and ideally on the style of the team.

This is why role classification sits underneath most modern scouting databases. Before a player can be ranked, the system has to decide what job he does — and the same nominal position can conceal several distinct jobs.

Set-piece dependency, stripped out

Attacking output that comes overwhelmingly from dead-ball situations behaves differently from output generated in open play.

Set-piece contribution is real and valuable, but it is more dependent on the delivery a team provides, on the routines it drills, and on being nominated to take them. A forward whose numbers rest heavily on headers from corners may find those numbers do not follow him to a club with different set-piece designers.

Splitting expected goals into open-play and set-piece components is standard practice for this reason. The open-play share is the part more likely to transfer.

Progressive actions relative to involvement

Progressive passes and progressive carries measure whether a player moves the ball meaningfully toward the opposition goal. Their scouting value increases sharply when expressed as a share of involvement rather than a raw count.

A defender attempting many progressive passes may be brave or may simply have the ball constantly. The informative version is the rate: of the passes he attempts, how many advance play, and at what completion rate. That combination separates players who genuinely break lines from those who merely have possession in a system that funnels the ball through them.

The same logic applies to carries. Volume tells you about opportunity, and rate tells you about the player.

Defensive numbers read against team shape

Defensive statistics are the most frequently misread family in scouting, because they are inverted by team style.

A player in a deep-defending side accumulates tackles, interceptions and clearances because his team concedes territory. A player in a dominant, high-pressing side records fewer of those events not because he is worse defensively but because opponents rarely reach him with the ball. Ranking defenders on raw defensive volume systematically rewards playing for weaker teams.

The usable versions are adjusted: defensive actions per opponent possession, the height up the pitch at which they occur, and success rate rather than attempt count. Where the action happens is often more informative than how often it happens.

The distribution, not the season average

A season average conceals whether output arrived steadily or in two extraordinary months.

Recruitment analysts increasingly examine the shape of a player's contribution across matches — the floor as much as the ceiling. Two players with identical season totals can present very different risk profiles if one delivered consistently and the other produced almost everything in a short hot streak against weaker opposition.

Match-level breakdowns also allow the quality of opposition to be layered on. Output concentrated against the bottom of a table is a different asset from output spread evenly across it.

The distribution also reveals dependence. A creative midfielder whose best matches all coincide with one particular striker being available is telling you something about the partnership rather than about himself, and that is precisely the sort of dependency a transfer breaks.

The adjustment that sits over all eight

None of these metrics is comparable across competitions until league strength is accounted for.

Recruitment models apply a difficulty adjustment, estimated from how players historically perform after moving between specific leagues. It is why a strong season in a mid-tier European division is discounted rather than taken at face value, and why the discount differs depending on which two competitions are involved. Without it, cross-league shortlists are close to meaningless.

Building that adjustment requires the same data structure for every competition being compared — consistent event definitions, consistent minutes records, consistent season boundaries. Multi-competition platforms such as RubiScore matter here for that reason rather than for any individual statistic they publish.

What none of it can see

The honest limit of data scouting is worth stating plainly. These eight measures narrow a list. They do not decide anything.

They cannot assess how a player handles instruction, whether he suits a dressing room, how he responds to being dropped, or whether his body will hold up under a heavier schedule. They cannot see decision-making without the ball in phases where nothing is recorded as an event. They are a filter placed in front of human judgement, not a replacement for it.

The workflow that works treats data as the tool that decides who gets watched. Scouts still decide who gets signed. Season-long player, club and competition data for building that first filter is published on rubiscore.com.