Does winning the corner count mean you're on top?
"They've had eight corners, they're all over them." Over 5,877 Scottish Premiership matches, the side with more corners won only 44% of the time, and once you know the shots, corners add nothing.
Beginner
Contents
The claim
"They've had eight corners. They're all over them."
The corner count sits on every scoreboard graphic next to possession and shots, and commentators read it as a measure of who's on top. Win lots of corners, the thinking goes, and you're the side doing the pressing, so the goals will come.
Why people believe it
Corners come from attacking. You can't win one without getting the ball into the opposition's end, and a run of corners looks and sounds like pressure: the crowd rises, the big centre-halves go forward, the keeper gets crowded. It feels like a goal is coming.
The data
Every Scottish Premiership match from 2000/01 to 2025/26 that records corners and shots: 5,877 matches, from football-data.co.uk.
The method
- In a single match: when one side wins more corners, how often does it win, draw or lose?
- Corners against shots: if you already know how many shots each side had, does the corner count tell you anything more about the goals they scored?
The evidence
More corners, only just more wins
When one side won more corners, it won 44.2% of those matches and lost 32.3%. That's better than a coin toss, but not by much. Even with eight or more extra corners, a side still lost one match in four.
Compare that with shots. A side with ten or more extra shots won 72% and lost fewer than one in ten, as the shots myth shows. The corner count is a much weaker sign of who's on top.
Corners without shots lose
The telling case is a side that wins more corners but has fewer shots than the opposition. It happened 1,313 times, and that side lost 56.5% and won only 20%. Corners that don't turn into shots are territory, not threat.
Once you know the shots, corners add nothing
Fit a simple model of each side's goals using its shots on target, its other shots, and its corners. Each shot on target is worth about 0.26 goals. Each corner is worth about −0.04: slightly less than nothing.
In plain football
- The model asks: for two sides with exactly the same shots, does the one with more corners score more?
- The answer is no. If anything, it scores a fraction less.
- That doesn't mean corners never produce goals. It means that when they do, it's through the shots they create, and those are already counted.
Over a whole season, teams that win lots of corners do score lots of goals (a correlation of 0.75). But that's because good teams win corners, just as they win shots, possession and matches. It isn't the corners doing the work.
Verdict
Not supported. Winning the corner count is only a faint sign that a side is on top: it wins 44% of those matches and loses nearly a third. Corners matter when they produce shots. On their own they tell you where the ball has been, not who is going to win.
Caveats
- The files don't say which goals came from corners. This tests whether the corner count reflects dominance, not how well any team takes its corners. A side with a great set-piece routine gets its reward through the shots it creates, which this counts as shots.
- Game state matters here too. A side chasing the game pushes forward and wins corners; a side protecting a lead tends to concede them. That pulls the corner count towards whichever side is behind.
- Corner quality varies. A short corner played back to the halfway line counts the same as one whipped onto the penalty spot.
- The model is simple. It's a straight-line fit on individual matches, and doesn't allow for the opponent's strength.
Reproduce the analysis
The results files are published by football-data.co.uk. Download the Premiership file (SC0) for each season and save each under its own name, such as SC0_2425.csv; they aren't rehosted on this site. Then:
import csv, glob
from collections import Counter
more_corners, corners_no_shots = Counter(), Counter()
for path in glob.glob("SC0_*.csv"):
with open(path, encoding="latin-1") as f:
for r in csv.DictReader(f):
if not all(r.get(c) for c in ("FTHG", "FTAG", "HC", "AC", "HS", "AS")):
continue
hg, ag, hc, ac, hs, as_ = (int(r[c]) for c in ("FTHG", "FTAG", "HC", "AC", "HS", "AS"))
for gf, ga, cf, ca, sf, sa in ((hg, ag, hc, ac, hs, as_), (ag, hg, ac, hc, as_, hs)):
result = "won" if gf > ga else "drew" if gf == ga else "lost"
if cf > ca:
more_corners[result] += 1
if sf < sa:
corners_no_shots[result] += 1
for name, c in (("more corners", more_corners), ("more corners, fewer shots", corners_no_shots)):
n = sum(c.values())
print(name, n, {k: f"{v / n:.1%}" for k, v in c.items()})
Spotted a flaw in the method, or have a football claim you'd like tested? Suggest it on LinkedIn, where discussion of these articles happens.