Is a positive delta the same as good finishing?

It is one half of it. The delta measures placement and power, not whether the shot went in. A player can strike the ball superbly and still be denied by the goalkeeper.

How large a sample does it need?

More than most stats. Shot placement varies enormously from match to match, so treat anything under about fifty shots as indicative rather than settled.

Which way round is the subtraction?

Post-shot xG minus pre-shot xG, actual striking value minus chance quality, so positive means the shots were placed and struck better than their starting quality suggested. This follows the same actual-minus-expected direction as the player deltas elsewhere on the site.

Why does this need more shots than goals minus xG?

Because it is measuring a finer-grained thing. Goals minus xG only asks whether the ball crossed the line. This asks how close to the corner every shot travelled, on target or not, which is a noisier quantity that takes longer to settle into a reliable average.

Does it credit the same player as goals minus xG?

Yes, and only that player. It has nothing to do with a passer or a teammate, unlike assists minus expected assists, which depends on someone else's finishing entirely. This is the cleanest one-player read among the site's over and under performance deltas.

Can a low xG shot still post a high post-shot value?

Yes, and that is the mechanism behind a large positive delta. A tight-angle effort with little underlying chance, placed perfectly into the far corner, generates far more post-shot value than its pre-shot xG implied, even if the goalkeeper gets a hand to it.

Post-Shot xG Delta

PSxG Δ

Post-shot expected goals delta is post-shot xG minus xG, so a positive figure means a player's ball striking adds value to the chances they get.

Why it matters. It separates clean strikers of a ball from players who reach the same positions and waste them with poor placement.

Actual striking value, minus the chance it started as

Take post-shot expected goals and subtract ordinary xG from it. The result strips out where the shot was taken from and leaves only what the shooter did with it once it left their foot or head.

Positive. Shots are travelling into more dangerous areas of the goal than their starting position deserved. Clean, well-placed striking.
Negative. Shots are travelling into less dangerous areas than their starting position deserved. That includes anything that missed the target entirely, which scores zero on post-shot xG regardless of the chance it came from.
Near zero. Placement is matching chance quality, the typical outcome across a large sample of shots.

Pre-shot xG prices in distance, angle and pressure before the shot is struck. Post-shot xG re-prices the same shot using where it actually travelled. The gap between the two is entirely down to execution: power, placement, and how well the shooter beat the goalkeeper's position.

This follows the same convention as the other player deltas on the site, actual minus expected, positive meaning overperformance. It is still the reverse of the league table deltas, where xPts and xGD run expected minus actual. A positive figure there flags a team due an improvement, not one already ahead.

Three shooters, three verdicts

A forward on +0.35 per shot across a hot fortnight. Every effort has found the top corner or the far post. Verdict: enjoy the highlight reel, but expect placement to regress towards the shooter's normal level over the following weeks. A short run of well-struck shots is closer to variance than a rebuilt technique.

A striker on -0.20 per shot across a full season, with goals roughly matching ordinary xG. Their shots are landing centrally, close to the goalkeeper, more often than the model expects, yet enough are still going in. Verdict: there may be a genuine placement habit worth coaching, since this has held over a large sample rather than a handful of matches.

A winger holding close to zero across two full seasons of shot volume. Verdict: this is close to their real striking quality, whatever level that lands at. A stable reading over that much data is the strongest kind of signal this metric produces.

Why fifty shots is the floor, not the target

Shot placement swings harder from game to game than shot outcome does. A single dipping strike into the top corner can move a low-volume player's average by a wide margin on its own.

Under 20 shots: description only. It tells you what happened in those shots, nothing about the level to expect going forward.
50 to 100 shots: the point where a genuine placement tendency starts to separate from a run of lucky or unlucky strikes.
A full season of regular shot volume: the number a rating decision should actually rest on.

Compare it against goals minus expected goals over the same stretch. If the two move together, placement is driving the finishing story. If they diverge, the goalkeeper or the woodwork is doing more of the work than the shooter is.

Where this is cleaner than the other deltas, and where it is not

It is the purest one-player read on the page. Goals minus xG can be nudged by a deflection or a goalkeeping error after the shot is struck. Assists minus expected assists depends entirely on a teammate's finishing. Post-shot xG delta measures only where the ball was sent, which is under the shooter's control alone.

"A positive delta proves this is a natural finisher." It proves the placement was good across this sample of shots. Whether it is a repeatable skill or a hot run only becomes clear once the sample runs past fifty or so shots. Even strong shooters see this number drift season to season.

Common questions

Is a positive delta the same as good finishing?: It is one half of it. The delta measures placement and power, not whether the shot went in. A player can strike the ball superbly and still be denied by the goalkeeper.
How large a sample does it need?: More than most stats. Shot placement varies enormously from match to match, so treat anything under about fifty shots as indicative rather than settled.
Which way round is the subtraction?: Post-shot xG minus pre-shot xG, actual striking value minus chance quality, so positive means the shots were placed and struck better than their starting quality suggested. This follows the same actual-minus-expected direction as the player deltas elsewhere on the site.
Why does this need more shots than goals minus xG?: Because it is measuring a finer-grained thing. Goals minus xG only asks whether the ball crossed the line. This asks how close to the corner every shot travelled, on target or not, which is a noisier quantity that takes longer to settle into a reliable average.
Does it credit the same player as goals minus xG?: Yes, and only that player. It has nothing to do with a passer or a teammate, unlike assists minus expected assists, which depends on someone else's finishing entirely. This is the cleanest one-player read among the site's over and under performance deltas.
Can a low xG shot still post a high post-shot value?: Yes, and that is the mechanism behind a large positive delta. A tight-angle effort with little underlying chance, placed perfectly into the far corner, generates far more post-shot value than its pre-shot xG implied, even if the goalkeeper gets a hand to it.

Related terms

See PSxG Δ across the Premier League

Every player, team and match measured on the chances created rather than the scoreline.

Last updated 6 August 2026