Advanced Football Statistics Beyond xG — The Metrics That Actually Predict Match Outcomes

June 22, 2026 · 11 min read

Expected goals changed how we watch football. But xG alone tells less than half the story. The best analysts and clubs now use a suite of advanced metrics — from pressing intensity to territorial dominance to ball progression — that capture what xG misses. Here are the statistics that separate surface-level analysis from genuine predictive insight.

Why xG Is Not Enough

Expected goals measures shot quality — the probability that a given shot results in a goal based on distance, angle, body part, and assist type. It revolutionized football analysis by showing that a team winning 1-0 despite being outshot 20-3 was probably lucky, not good.

But xG only captures what happens at the moment of shooting. It says nothing about how a team builds attacks, how aggressively they press, whether they dominate dangerous territory, or how effectively they progress the ball into scoring positions. Two teams can post identical xG numbers while playing completely different styles — and one might be far more sustainable than the other.

The metrics below fill those gaps. Each measures a different phase of play, and together they paint a complete picture of team and player quality.

PPDA — Measuring Pressing Intensity

Passes Per Defensive Action (PPDA) quantifies how aggressively a team presses. It counts how many passes the opponent completes in their own defensive 60% of the pitch before the pressing team makes a defensive action — a tackle, interception, foul, or aerial duel.

The formula is straightforward: divide the opponent's completed passes in their defensive 60% by the team's defensive actions in that same zone. A lower PPDA means more aggressive pressing.

• Elite pressing (6-9 PPDA): Jürgen Klopp's Liverpool consistently posted PPDA values around 7-8 during their 2018-19 Champions League and Premier League title runs. Marcelo Bielsa's Leeds averaged roughly 8.0 in their first Premier League season.
• Average pressing (10-12 PPDA): Most mid-table Premier League teams sit in this range — pressing in moments rather than as a system.
• Passive defending (15+ PPDA): Sean Dyche's Burnley regularly hit 15-18 PPDA, deliberately sitting deep in a compact low block and letting opponents have the ball in harmless areas.

The metric was popularized by StatsBomb founder Ted Knutson in the early 2010s and became widely available on FBref. Brentford — owned by Smartodds founder Matthew Benham — used PPDA extensively in their recruitment model, identifying pressing-capable players from lower leagues and European markets before their 2021 Premier League promotion.

Liverpool's 2018-19 PPDA of ~7.5 was among the lowest (most aggressive) ever recorded in a top-five league. They won the Champions League and finished with 97 Premier League points.

Why PPDA Matters for Predictions

Pressing intensity correlates strongly with chance creation. Teams that press high force turnovers in dangerous areas, leading to high-quality scoring opportunities that xG models reward. A team with a suddenly high PPDA (pressing less intensely) may be fatigued, tactically adjusting, or missing key pressing triggers — all predictive signals.

Field Tilt — Territorial Dominance in the Final Third

Possession percentage is one of football's most misleading stats. A team can have 60% possession while passing endlessly in their own half. Field tilt solves this by measuring only final-third activity: what percentage of total completed passes in the attacking third belongs to each team.

• Dominant teams (60-75%): Pep Guardiola's Manchester City consistently posts field tilt values of 65-72%, suffocating opponents in their own defensive third.
• Balanced matches (45-55%): Most evenly-matched Premier League games fall in this range.
• Counter-attacking teams (30-40%): Diego Simeone's Atlético Madrid deliberately cedes territory, often recording field tilt around 35-42% while remaining extremely dangerous on the break.

The predictive value of field tilt lies in its ability to separate sustainable dominance from possession illusion. A team winning games with 40% field tilt is likely relying on counter-attacking efficiency and may struggle against deep-defending opponents. A team with 65%+ field tilt that isn't winning is probably underperforming their chances — a regression candidate in either direction.

Progressive Passes and Carries — Ball Progression Metrics

Progressive actions measure how effectively a player or team moves the ball into dangerous areas. FBref defines a progressive pass as one that advances the ball at least 10 yards toward the opponent's goal when starting in the team's own 40% of the pitch, or at least 5 yards when starting in the middle 40%, or any pass into the penalty area. Progressive carries follow similar thresholds for dribbles and runs.

These metrics cut through the noise of raw pass completion. A center-back who completes 95% of his passes mostly sideways adds little attacking value. A midfielder who completes 80% but plays 10 progressive passes per 90 is driving the team forward constantly.

Toni Kroos averaged roughly 10 progressive passes per 90 in his final Real Madrid seasons — one of the highest rates in European football history. Virgil van Dijk consistently ranks among the top center-backs for progressive passing (7-8 per 90), reflecting Liverpool's build-from-the-back philosophy.

xAG — The Metric That Separates Creators from Assist-Dependent Players

Traditional assists are binary — a pass either leads to a goal or it doesn't. This creates enormous noise. A perfect through ball that puts a striker one-on-one with the goalkeeper gets zero credit if the striker skies it over the bar. Expected Assisted Goals (xAG) fixes this by measuring the xG value of shots assisted by a player's passes.

The difference between xAG and expected assists (xA) is subtle but important. xA measures the probability that a completed pass becomes an assist based on pass characteristics (type, distance, angle). xAG measures the quality of the resulting shot — summing up the xG of all shots that came from a player's key passes.

Why does this matter? A through ball that creates a 0.5 xG chance gets high xAG credit but might get lower xA if through balls are generally unlikely to become assists on average. xAG rewards players who create genuinely dangerous chances, regardless of whether their teammates finish them.

• Elite creators (0.30-0.50 xAG per 90): Kevin De Bruyne consistently posts the highest xAG in the Premier League, often exceeding 0.40 per 90 during Manchester City's treble-winning 2022-23 season.
• Good creators (0.15-0.25 per 90): Bruno Fernandes typically falls in this range, with xAG often exceeding his actual assist totals — evidence that Manchester United's finishing has let him down.
• Average players (0.05-0.10 per 90): Most non-creative midfielders and defenders.

Shot-Creating and Goal-Creating Actions

If xAG measures the quality of chances created through passing, Shot-Creating Actions (SCA) and Goal-Creating Actions (GCA) measure all forms of chance creation. SCA captures the two offensive actions directly leading to any shot — passes, dribbles, fouls drawn, shots that rebound, or interceptions that start an attack. GCA does the same for goals.

This broader lens matters because it credits players who contribute to attacks in ways traditional stats miss. A winger who draws a foul on the edge of the box gets SCA credit. A midfielder whose intercepted pass launches a counter-attack gets GCA credit if it leads to a goal. Assists alone miss these contributions entirely.

• Elite SCA (5-8 per 90): Lionel Messi regularly posted 7-10 SCA per 90 during his peak Barcelona years, combining dribbling, passing, and shooting into a single creative force.
• Elite GCA (0.5-1.0 per 90): Much rarer than SCA since goals are inherently scarce. Only the most decisive creators sustain this rate over a full season.

The Packing Rate — Germany's Secret Weapon

Developed by German analytics company Impect around 2014-2016, the packing rate measures how many opponents a pass or dribble takes out of the game. An opponent is "packed" if they end up behind the ball after the action completes. A sideways pass in midfield might pack one opponent. A progressive through ball might pack five.

The German national team under Joachim Löw used packing data during Euro 2016, and Bundesliga clubs including Borussia Dortmund and Bayern Munich have adopted the metric for player evaluation. Toni Kroos was historically one of the highest-rated midfielders by packing rate — his progressive passes routinely took multiple opponents out of the game.

The packing rate addresses a fundamental limitation of traditional passing statistics: not all completed passes are equal. A center-back who completes 50 sideways passes and a playmaker who completes 30 progressive passes through the lines have very different impacts, but traditional stats treat them identically.

Defensive Metrics — Beyond Tackles and Interceptions

Modern defensive analysis goes far beyond counting tackles. The best defensive metrics capture a player's total defensive contribution: pressures applied, balls recovered, and the zones where these actions occur.

Pressures count how many times a player challenges an opponent receiving, carrying, or releasing the ball. The key distinction is between raw pressures and successful pressures — those that actually result in winning possession. Roberto Firmino averaged roughly 25 pressures per 90 during Liverpool's 2018-19 title challenge, with an elite success rate that made him the fulcrum of Klopp's counter-pressing system.

Combined tackles and interceptions (Tkl+Int) remain a solid composite metric. N'Golo Kanté's ~7.5 Tkl+Int per 90 during Leicester's miraculous 2015-16 title win remains one of the highest single-season defensive rates in Premier League history. Rodri balances ~5-6 Tkl+Int per 90 with elite passing — a combination that made him indispensable to Manchester City's treble.

How These Metrics Work Together

No single metric tells the full story. The real power comes from combining them into a multi-dimensional profile:

1. Team identity: PPDA + field tilt + possession reveal style. Low PPDA + high field tilt = dominant high-pressing team. High PPDA + low field tilt = organized counter-attackers.
2. Sustainability check: Compare xG with field tilt and progressive actions. A team winning with high xG but low field tilt and few progressive passes is likely riding unsustainable finishing.
3. Player evaluation: Progressive passes + xAG + SCA identify elite creators. Tkl+Int + pressures + progressive carries identify complete midfielders.
4. Prediction models: Feed PPDA, field tilt, progressive pass volume, and xAG into a model alongside xG, and prediction accuracy improves significantly compared to xG alone.

Key Takeaways

• PPDA measures pressing intensity: Lower values mean more aggressive pressing. Elite teams like Klopp's Liverpool posted values around 7-8, while deep-defending teams sit at 15+.
• Field tilt captures territorial dominance: Unlike possession percentage, it only measures activity in the attacking third, separating genuine dominance from sterile ball circulation.
• Progressive actions reveal ball progression quality: Progressive passes and carries identify players who move the ball into dangerous areas, not just those who complete safe passes.
• xAG decouples creation from finishing: It measures the quality of chances a player creates regardless of whether teammates convert them, making it far more predictive than raw assist counts.
• Combine metrics for predictive power: Using PPDA, field tilt, progressive actions, and xAG alongside xG produces significantly more accurate predictions than any single metric alone.