The Ultimate Guide to Football Analytics: Player Stats, Team Performance, and Goalkeeper Save Percentage

So you’ve heard people throw around terms like “xG,” “PPDA,” and “save percentage” during match analysis, but you’re not entirely sure how they fit together—or which ones actually matter. You’re not alone. Football analytics has exploded over the last decade, but the sheer volume of metrics can feel overwhelming. This guide cuts through the noise. We’ll walk you through the essential player stats, team performance indicators, and—most importantly—how to interpret goalkeeper save percentage in context. By the end, you’ll have a practical checklist to evaluate any match or player performance without falling for misleading numbers.

Why Save Percentage Alone Is a Trap

Let’s start with the most misunderstood stat in football: goalkeeper save percentage. At first glance, it seems straightforward—saves divided by shots on target. But here’s the catch: a keeper facing ten low-quality shots from distance and saving nine has a 90% save percentage, while one facing five high-quality chances from inside the box and saving three has only 60%. Which one performed better? The second keeper likely prevented more dangerous goals, but the raw percentage says otherwise.

The key insight: Save percentage is heavily influenced by the quality of shots faced. That’s where Expected Goals (xG) comes in. Instead of counting all shots equally, xG assigns a probability value to each attempt based on distance, angle, assist type, and body part used. A shot from six yards out might have an xG of 0.6, while a long-range effort might be 0.03. When you compare a keeper’s actual goals conceded to the total xG they faced, you get a much fairer measure of performance.

How to use it: Look for “Goals Prevented” or “PSxG-GA” (Post-Shot Expected Goals minus Goals Allowed) on sites like FBref. A positive number means the keeper saved more than expected; negative means they underperformed. This adjusts for the defense in front of them.

Step 1: Start with Team-Level Metrics

Before diving into individual players, assess the team’s overall performance. This gives context to any player stat you’ll examine later.

Your checklist:

• Possession percentage: Not a quality metric on its own—a team can have 70% possession and create few chances. Pair it with passes in the final third.
• Expected Goals (xG) for and against: This tells you the quality of chances created and conceded. A team with high xG for but low actual goals might be finishing poorly (or facing a hot keeper).
• PPDA (Passes Per Defensive Action): Measures pressing intensity. A low PPDA (e.g., 8-10) means the team presses high and aggressively. A high PPDA (e.g., 15+) suggests a deeper block. Compare it to the opponent’s build-up style.
• Shots on target ratio: If a team takes 20 shots but only 3 are on target, their shot selection is poor. Conversely, a team with 5 shots and 4 on target is creating high-quality chances.

Example table (hypothetical match data):

Notice how Team A dominated possession and shots but only marginally out-xG’d Team B. That suggests Team B’s chances were higher quality—and their low PPDA indicates they sat deep and hit on the counter. The final score might not reflect the actual threat.

Step 2: Evaluate Individual Outfield Players

Once you understand the team context, assess players using stats that align with their role. Avoid judging a defensive midfielder by goals scored or a winger by tackles won.

For attackers:

• Non-penalty xG per 90: Removes penalty bias. Compare to actual goals to see if they’re over- or underperforming.
• Key passes and expected assists (xA): A key pass is a pass leading to a shot; xA measures the quality of that pass. A player with high key passes but low xA is creating low-quality chances.
• Dribbles completed and progressive carries: Shows ability to advance the ball into dangerous areas.

For midfielders:

• Pass completion percentage in the final third: High completion here means they’re breaking lines. Avoid the trap of overall pass completion, which can be inflated by sideways passes.
• Tackles and interceptions per 90: Context matters—high numbers in a low-block team are different from a high-press team.
• Progressive passes: Passes that move the ball toward the opponent’s goal by a significant distance.

For defenders:

• Clearances and blocks: Useful for center-backs in deep defenses, but less relevant for high-line teams. Pair with /clearances-and-blocks-stats for deeper analysis.
• Aerial duel win rate: Important for set-piece defense and long-ball teams.
• Passes into the final third: Modern center-backs are judged by their build-up play, not just defending.

Step 3: The Goalkeeper Deep Dive

Now you’re ready to evaluate keepers properly. Here’s your checklist:

• Save percentage: Write it down, but don’t trust it alone.
• PSxG-GA (Post-Shot Expected Goals minus Goals Allowed): The gold standard. A positive number means they saved more than expected given the shot quality. Available on FBref and Opta-powered sites.
• Crosses claimed percentage: High numbers (above 8-10% of crosses) indicate command of the box. Low numbers suggest reliance on defense to clear.
• Sweeper actions: How often does the keeper leave the box to clear through balls? Pair with /goalkeeper-sweeper-keeper-tactics to see if their style fits the team’s defensive line.
• Clean sheet percentage: Highly dependent on team defense. A keeper on a top-four team will have more clean sheets than one on a relegation-threatened side, regardless of individual quality.

The trap to avoid: Never compare save percentages across different leagues. The Premier League has higher shot quality on average than Ligue 1 or the Bundesliga, so raw numbers aren’t transferable. Use PSxG-GA for cross-league comparisons.

Step 4: Team Performance in Context

Now combine everything. A team’s defensive performance isn’t just about the keeper—it’s the whole system.

Key relationships:

• High PPDA + low xG against: The press is working—opponents can’t build quality chances.
• Low PPDA + low xG against: The team sits deep and limits space effectively (e.g., Atletico Madrid under Simeone).
• High xG against + high save percentage: The keeper is bailing out a poor defense. This is unsustainable—if the keeper’s form drops, the team will concede more.
• Low xG against + low save percentage: The defense is solid, but the keeper is underperforming. This might be a fixable problem with a change in goal.

Example scenario: A team concedes 1.5 xG per game but has a keeper with a 75% save percentage. They’re conceding about 0.38 goals per game from xG (1.5 * 0.25). If they switch to a keeper with a 65% save percentage (league average), they’d concede 0.53 goals per game from xG—a significant increase. The defense might look good on paper, but the keeper is masking issues.

Step 5: Use Public Data Sources

Don’t guess. Use these free resources to pull real stats:

• FBref.com: Comprehensive stats for all major leagues, including xG, PSxG, and advanced metrics. Export to CSV for deeper analysis.
• WhoScored.com: Player ratings based on weighted stats. Good for quick comparisons.
• Transfermarkt.com: For market values, contract expiry dates, and release clauses—useful for scouting potential transfers. Note that Transfermarkt valuations are estimates, not actual transfer fees.
• Understat.com: Focuses on xG and xA for European leagues. Clean interface.

How to use them: When evaluating a keeper, pull their PSxG-GA over the last two seasons. A consistent positive value indicates real skill. A one-season spike might be luck. Compare to the team’s xG against to see if the defense is helping or hurting.

Step 6: Avoid Common Pitfalls

Here’s what not to do:

• Don’t use total saves as a quality metric. A keeper facing 30 shots per game will have more saves than one facing 10, but that doesn’t mean they’re better.
• Don’t ignore sample size. A keeper with 5 games and a 90% save percentage might regress to the mean. Look for at least 20 games of data.
• Don’t confuse correlation with causation. A team with a high save percentage might also have a strong defense—but the stats can’t tell you which caused which.
• Don’t rely on clean sheets alone. A clean sheet against a weak opponent is less impressive than one against a top attack. Check the opponent’s xG in that match.

Step 7: Build Your Own Match Report

After a match, run through this checklist:

1. Team xG for and against: Who created better chances?
2. Keeper PSxG-GA: Did the keeper outperform or underperform expectations?
3. PPDA comparison: Which team pressed harder? Did it work?
4. Key player stats: Did the star striker underperform xG? Did the midfielder complete progressive passes?
5. Context: Was it a cup game with rotated squads? Was the pitch weather-affected?

Write your conclusions, but remember: stats describe what happened, not why. Use them to ask better questions, not to find definitive answers.

Football analytics isn’t about finding a single number that tells you everything. It’s about building a framework to evaluate performance fairly. Save percentage is a starting point, but it’s meaningless without shot quality context. Team metrics like xG and PPDA give you the big picture, while individual stats like PSxG-GA and progressive passes add the detail.

The next time you watch a match or read a scouting report, run through this checklist. You’ll spot the keepers who are carrying their teams, the defenders who are being exposed, and the attackers who are due for regression. And when you see a stat that seems too good to be true, dig deeper—it probably is.

Remember: Betting on football carries financial risk. Use analytics to inform your understanding, not to guarantee outcomes. No model predicts the future—it only describes the past more accurately.