Expected Goals on Target

Expected goals rates the chance before the ball is kicked. Expected goals on target rates the finish: once a shot is heading for goal, how often do shots aimed at that spot go in? A shot into the top corner scores far more often than one straight at the goalkeeper.

About a 6-minute read.

See every topic

Start with one shot

Once it’s struck, where is it going?

Expected goals is set before the ball is kicked, from where and how the shot was taken. Expected goals on target waits for the strike. For a shot that’s on target, it asks: out of 100 shots like this, struck to the same part of the goal, how many go in?

Take a shot from close in. Shots like it go in about 45 times out of 100: 0.45 expected goals. Struck low and straight at the goalkeeper, shots like it go in only about 22 times out of 100: 0.22 expected goals on target. Struck into a top corner, about 70 times out of 100: 0.70.

0.45 expected goals before the strike

A goal split into six areas. One strike heads for the top right corner: 0.70. The other goes low and straight at the goalkeeper: 0.22.

Same chance, two strikes. The dashed lines split the goal into the six areas this page uses. Numbers use the Riverfield League’s expected-goals-on-target model, rounded.

Try it

Pick the chance, then aim the shot

Choose where the shot was taken from, then tap the part of the goal it’s heading for. The readout compares the number before the strike with the number after it.

Tap one of the six areas, or use the arrow keys and Enter.Tap one of the six areas. The goal is drawn to scale: 7.32 meters wide and 2.44 meters high (24 by 8 feet). Numbers are estimates, rounded.

Before the strike

45 out of 100

That’s 0.45 expected goals.

After the strike, heading for the top right

70 out of 100

That’s 0.70 expected goals on target.

The strike’s difference

+0.25

0.70 − 0.45 = +0.25. Struck there, shots like this go in more often than the chance alone suggests.

Why the number moves

  • The corners

    Shots heading for a corner are the hardest to reach. Wherever the shot came from, the corners have the biggest numbers.

    Try it: tap a top corner, then a bottom corner.

  • Straight at the goalkeeper

    The goalkeeper usually starts in the middle, so strikes there are the easiest to stop, even from close in.

    Try it: tap the low middle area.

  • High or low

    In the middle, a high shot is harder to stop than a low one. In the corners, high and low are both hard to reach.

    Try it: tap low middle, then high middle.

  • The chance still counts

    A better chance raises the number in every part of the goal. From close in, the goalkeeper has less time to react.

    Try it: switch between Close in and Long range.

Real models use the exact spot where the ball crosses the goal line, how fast it’s moving, and more. This page splits the goal into six areas and uses the Riverfield League’s model for each area. A shot that misses the target gets no expected goals on target at all.

What it doesn’t tell you

A great strike can still be saved

Expected goals on target says how often strikes like this go in. It can’t say whether this one goes in. When a goalkeeper stops a shot with a high number, that’s excellent goalkeeping, not bad luck for the shooter.

It rates the finish, not the chance. A poor chance can be struck beautifully, and a great one can be scuffed.

Shots that miss the target don’t get expected goals on target at all, so a team’s total counts only its shots on target.

What counts as on target
★ Challenge Chance or Finish?

On Matchday 6 of the Riverfield League, our fictional league, an Owls shot from 17 meters out and 10 meters to the side had 0.05 expected goals. It flew toward a top corner with 0.67 expected goals on target, and the Geese’s goalkeeper saved it. Was it a good chance? Was it a good finish?

Not a good chance: shots like it go in about 5 times out of 100. But a great finish: struck like that, about 67 out of 100 go in. The save was excellent.

A whole match

Put the two numbers side by side

Add up a team’s expected goals, then its expected goals on target, and compare both with its goals. Each gap tells you something different.

The Robins’ finishing didn’t match their chances

On Matchday 1, the Bison beat the Robins 2–0. The Robins’ 7 shots added up to 1.58 expected goals, including a penalty kick. Only 1 was on target, so their strikes added up to just 0.21 expected goals on target. The penalty went wide, and a miss gets no expected goals on target at all.

The Bison went the other way. Their 13 shots on target added up to 3.80 expected goals on target, more than their 2.46 expected goals, and they scored 2. The Robins’ goalkeeper made 11 saves.

Bison FC 2–0 Sporting Robins

Riverfield League · Matchday 1

Bison FC · 18 shots, 13 on target13 of 18 on target

Expected goals

2.46

Expected goals on target

3.80

Goals

2

3.80 - 2 = 1.80 kept out by the keeper

Sporting Robins · 7 shots, 1 on target1 of 7 on target

Expected goals

1.58

Expected goals on target

0.21

Goals

0

-1.37 lost in the strikes

All bars use one scale, from 0 to 4 goals. Expected goals rates the chances, expected goals on target rates the strikes, and goals are what went in.

A minus sign means the strikes went in less often than chances like theirs usually do; a plus sign means more. The Robins lost 1.37 in the strikes, and the Bison added 1.34 (3.80 − 2.46).

Comparing goals with expected goals on target shows the goalkeeper. Analysts call it goals prevented: the Robins’ goalkeeper kept out about 1.80. Over the season, the Turtles’ goalkeeper led the league at +3.03. The Geese’s was −1.53, meaning 1.53 more goals went in than strikes like those usually produce.

Over the season, the Robins had 17.42 expected goals, 11.50 expected goals on target, and 11 goals. Most of their gap, 5.92 goals, came in the strikes, not from the goalkeepers they faced.