The Advanced Metrics Every Soccer Analytics Platform Should Track

Transition Analysis

Modern soccer is often decided in the six or seven seconds after possession changes hands. Transition analysis is where a lot of analytics platforms fall short, because it demands tracking speed, structure, and decision-making in a phase of play that lasts only moments.

PPDA (Passes Allowed per Defensive Action) is the standard entry point here. It measures how many passes an opponent is allowed to string together before a team makes a tackle, interception, or challenge, and a lower number signals a more aggressive, higher-intensity press. PPDA is as much a transition metric as a defensive one, since a team that presses well typically wins the ball back in advanced areas and can transition into attack almost immediately.

Recoveries in opponent half works alongside PPDA to show where the ball is actually being won back, which matters because a high press that never recovers the ball high up the pitch is a lot of running for very little return.

Average passing chain length after a turnover is a metric worth building as it measures how many passes a team strings together in the moments right after regaining possession, before the opponent's defence resets. A short, direct chain suggests a team is exploiting disorganisation quickly; a longer chain might mean they are choosing control over speed, or that they lack the passing options to break forward at pace. Neither is inherently better, but the pattern tells a coaching staff a lot about a team's transitional identity.

Pressure success rate rounds this out by measuring how often pressing actions actually disrupt the opponent rather than just chasing the ball. A team can apply plenty of pressure and still leak transitions if that pressure rarely wins the ball back.

xG and xA Analysis

A team taking twenty shots from distance is not creating the same threat as a team taking eight shots inside the box, and this is exactly the gap xG analysis was built to close. But xG on its own only tells half the story, the finishing half. Pairing it with expected assists and a genuine measure of chance creation gives a fuller picture of where a team's attacking threat actually comes from and whether it is being converted efficiently.

xG (expected goals) assigns a probability to every shot based on factors like location, angle, and the type of chance, then sums those probabilities to estimate how many goals a team or player "should" have scored given the chances created. It is the single most useful corrective to results that get skewed by finishing variance or goalkeeping heroics over a small sample.

xA (expected assists) does the same work on the creation side, assigning a probability to the shot that follows a key pass and crediting the passer with that shot's expected value rather than waiting on whether it was actually scored. This separates a player consistently setting up high-quality chances from one who is just racking up assists because a teammate is finishing well, or being denied credit because a teammate is finishing poorly. Read alongside chances created and key passes, xA shows whether a team's creative output is repeatable or dependent on a handful of moments.

Goals minus xG (and assists minus xA) is where finishing enters the picture directly. A player or team consistently scoring above their xG is either finishing at an elite level or benefiting from variance that tends to regress, and a platform that surfaces this gap over a rolling sample gives a technical director an early signal on which one it is. Consistently underperforming xG, by contrast, can point to a genuine finishing issue worth coaching, or simply a small sample still catching up to the underlying numbers.

xG per shot matters just as much as total xG, arguably more for evaluating chance creation quality rather than shot volume. A team generating a high xG per shot is consistently getting into dangerous positions before pulling the trigger, while a team with high total xG but low xG per shot may just be shooting often from moderate positions.

Non-penalty xG strips out penalty conversions, which follow a different probability model entirely and can distort a player's or team's underlying open-play threat if left in the raw number.

It is worth pairing xG with deep completions, passes completed into the zone within about 20 meters of goal, since this shows whether a team's chance creation is being generated through sustained buildup into dangerous areas or through more isolated moments. A team with strong xG but weak deep completion numbers may be relying heavily on individual moments rather than a repeatable attacking process, which is a useful distinction for a technical director assessing sustainability.

Defensive Efficiency

Defensive efficiency is the hardest layer to quantify well because so much of good defending is about preventing situations from developing in the first place, not making a visible action to stop them.

Total defensive actions and, more specifically, crucial defensive actions (tackles, interceptions, and clearances classified as stopping a genuine goal-scoring opportunity) give a much better read on defensive value than raw tackle counts, which can reward players who are constantly out of position and forced into recovery challenges.

Duel win rate, split between ground and aerial duels, is one of the clearest per-player defensive efficiency indicators, since it normalises for how often a player is actually engaged in a contest rather than just how many contests they have.

PPDA belongs here too, viewed from the defensive efficiency side rather than the transition side. It is genuinely a dual-purpose metric, useful both for describing pressing intensity and for evaluating how well a team's structure prevents the opponent from building comfortably.

Why Per-90 Values Matter for Soccer KPIs

None of these soccer metrics means much without normalization, and per-90 values are the standard fix. A player who completes 15 progressive passes in 45 minutes off the bench is doing something very different from one completing 15 across a full match, and raw totals flatten that distinction completely.

Per-90 values matter even more for team and squad-level soccer KPIs used in recruitment or squad planning, because they let a technical director compare a rotation player's output fairly against a nailed-on starter, or benchmark a trialist against existing squad members without match-minutes skewing the picture. They also matter across age groups and competition levels, where match length and squad rotation patterns vary widely. Any platform serious about being used for recruitment or performance review needs per-90 as a default option.

What Clubs Should Look For in an Analytics Platform

Not every platform that offers "advanced metrics" is built the same way, and technical directors or decision-makers at a club evaluating options should look past the size of the data catalogue to how the numbers are organised. A platform that exposes hundreds of raw data points without structure puts the interpretation work back on the club, which defeats the purpose of paying for analytics in the first place.

The stronger question to ask a vendor is whether their metrics actually answer something. Can the platform tell you if your possession is purposeful, using field tilt, progressive actions, and box entries rather than a bare possession percentage? Can it show whether your team wins the moments that matter, through PPDA, transition passing chains, and pressure success, or does it stop at raw tackle and pass counts? Does it separate chance quality from chance volume through the xG family and deep completions, rather than reporting shots and goals as if they carry equal weight? And does it distinguish defensive efficiency from defensive busyness, using crucial actions and duel win rates instead of counting every challenge a player makes regardless of outcome?

Per-90 normalisation should be a default across the platform, not a manual toggle buried in a filter menu, since that is what makes comparisons across rotation players, trialists, and different competition levels usable in a recruitment meeting or a Monday debrief. A club evaluating an analytics platform is really evaluating whether it can answer these questions quickly under time pressure, not whether it has the longest glossary.