- The 9.5 line is close to a coin flip across all 16 leagues (50.9%), but it swings a lot from league to league: from 38.6% to 59.4%.
- Confirmed: favorites and home teams win more corners, but false that more goals bring more corners (correlation −0.01).
- Limit: this data has no corner odds, so we cannot say which market pays, only how often each line lands.
What did we measure, and with what data?
This article does not give opinions: it counts.
We used the public files from football-data.co.uk.
That is 128 files: 16 European leagues and 8 seasons (2018/19, 2019/20, and 2021/22 through 2026/27).
The 2026/27 season is still in progress, and 2020/21 is not in the sample.
In total there are 39,856 matches with corners recorded.
They cover the Premier League and Championship, the Scottish league, the Bundesliga and Bundesliga 2, LaLiga and LaLiga 2.
Also Serie A and Serie B, Ligue 1 and Ligue 2.
Plus the leagues of the Netherlands, Belgium, Portugal, Turkey and Greece.
We ran the calculations ourselves with a Python script.
Every number in this article comes from that script.
If you want to redo it, download the same files and count.
Honest limit: these files do not include corner odds.
That is why we cannot say which market pays.
We can say how often each line lands, and that alone already dismantles several myths.
This is not betting advice and is for adults 18 and over only.
How many corners per match, and how many go over 9.5?
First the basics: total corners per match.
The average across the 16 leagues is 9.80 corners.
The home team adds 5.39 and the away team 4.41.
In other words, a typical home side wins almost one more corner than its opponent.
Now, how often the best-known lines are beaten.
The table uses all matches together.
| Corner line | Matches | Frequency | Reading |
|---|---|---|---|
| Over 8.5 | 39,856 | 62.7% | Frequent |
| Over 9.5 | 39,856 | 50.9% | Coin flip |
| Over 10.5 | 39,856 | 39.4% | Less often |
| Over 11.5 | 39,856 | 29.1% | Rare |
The 9.5 line is beaten in 50.9% of matches.
That is why it is the most common one at bookmakers: it is the most balanced.
However, this overall average is no use for your match.
Does corner frequency change by league?
A lot.
It is the biggest difference we found.
The table lists the 16 leagues in alphabetical order of their file code.
| League | Matches | Average corners | Over 9.5 | Over 10.5 |
|---|---|---|---|---|
| Belgium (B1) | 2,093 | 10.05 | 54.0% | 42.7% |
| Bundesliga (D1) | 2,204 | 9.86 | 52.0% | 40.3% |
| Bundesliga 2 (D2) | 2,223 | 10.21 | 54.6% | 43.4% |
| Premier League (E0) | 2,740 | 10.35 | 57.7% | 46.2% |
| Championship (E1) | 4,007 | 10.29 | 56.0% | 44.6% |
| Ligue 1 (F1) | 2,409 | 9.51 | 46.7% | 35.0% |
| Ligue 2 (F2) | 2,498 | 9.15 | 43.2% | 32.3% |
| Greece (G1) | 1,722 | 8.73 | 38.6% | 28.0% |
| Serie A (I1) | 2,750 | 9.72 | 50.5% | 38.7% |
| Serie B (I2) | 2,711 | 9.65 | 49.7% | 37.6% |
| Netherlands (N1) | 2,158 | 10.30 | 56.8% | 45.6% |
| Portugal (P1) | 2,223 | 10.01 | 52.9% | 41.4% |
| Scotland (SC0) | 1,613 | 10.52 | 59.4% | 47.6% |
| LaLiga (SP1) | 2,759 | 9.49 | 48.0% | 36.4% |
| LaLiga 2 (SP2) | 3,334 | 9.31 | 45.0% | 32.7% |
| Turkey (T1) | 2,412 | 9.65 | 49.3% | 37.7% |
The Scottish league goes over 9.5 in 59.4% of matches.
The Greek league does so in only 38.6%.
That is a gap of almost 21 percentage points.
The same line, two leagues, two very different realities.
This answers a common question: "is 9.5 an over or an under?"
It depends on where they play.
A loose corner figure with no league attached is of no use to you.
Myth 1: does the favorite win more corners?
To measure it, we defined the favorite as the team with the lowest average 1X2 odds.
We used the average of the pre-match odds in the file.
That gives 39,721 matches with valid odds.
| Group | Matches | Favorite's corners | Opponent's corners | Favorite wins more |
|---|---|---|---|---|
| All matches | 39,721 | 5.67 | 4.13 | 59.3% |
| Strong favorite (odds below 1.60) | 8,408 | 6.79 | 3.32 | 75.6% |
The favorite wins more corners in 59.3% of matches.
In 9.5% the count is tied, and in 31.1% the opponent wins more.
When the favorite is very clear-cut, the figure rises to 75.6%.
Verdict: confirmed.
The logic is simple: the side that attacks more earns more corners.
But be careful: the odds already know this, so this fact alone gives you no edge.
Myth 2: does the home team win more corners than the away team?
Across the 16 leagues, the home team wins 5.39 corners and the away team 4.41.
The home team wins more in 54.0% of matches.
The away team wins more in 36.4%, and 9.5% are tied.
The advantage shows up in every league in the sample.
In LaLiga, for example, the home side averages 5.35 and the away side 4.13.
In Portugal the gap is the smallest: 5.41 against 4.60.
Verdict: confirmed.
It is a stable pattern.
It is also a pattern the market knows very well.
Myth 3: do more goals mean more corners?
It sounds logical: open game, more attacks, more corners.
We tested it by grouping matches by total goals.
| Goals in the match | Matches | Corners (average) | Over 9.5 |
|---|---|---|---|
| 0 or 1 goal | 9,898 | 9.82 | 51.3% |
| 2 goals | 9,601 | 9.87 | 51.9% |
| 3 goals | 8,818 | 9.79 | 50.4% |
| 4 goals | 6,004 | 9.78 | 51.0% |
| 5 or more | 5,535 | 9.69 | 49.2% |
Matches with 0 or 1 goal average 9.82 corners.
Matches with 5 goals or more average 9.69.
There is almost no difference.
The correlation between goals and corners is −0.01, which is to say, zero.
Verdict: false.
A corner is not a small goal.
A team can press for 90 minutes without scoring, or score three on the break without winning a single corner.
Who wins does not matter much either.
If the home team wins, the average is 9.69 corners.
If it is a draw, 9.95.
If the away team wins, 9.84.
Myth 4: are there more corners in the second half?
It is the most repeated line in the market.
To answer it we would need the corners for each half.
The football-data.co.uk files include first-half goals, but not corners by half.
Verdict: it cannot be known with this data.
We would rather say so than invent a percentage for you.
If an AI gives you one, ask for the source, league, season and number of matches.
When an AI gives you a corners-by-half percentage with no source, treat it as a rumor.
See how to compare AIs for betting and how to always ask them for the source of every figure.
Myth 5: does an "unusual" season change everything?
Another common doubt is whether one season's numbers carry over to the next.
We looked at total corners by season.
| Season | Matches | Corners (average) | Over 9.5 |
|---|---|---|---|
| 2018/19 | 5,470 | 9.96 | 52.9% |
| 2019/20 | 5,200 | 9.90 | 51.4% |
| 2021/22 | 5,668 | 9.73 | 50.5% |
| 2022/23 | 5,602 | 9.71 | 49.5% |
| 2023/24 | 5,601 | 9.90 | 52.0% |
| 2024/25 | 5,482 | 9.71 | 50.0% |
| 2025/26 | 5,450 | 9.69 | 50.1% |
| 2026/27 (in progress) | 1,383 | 9.85 | 50.9% |
The average moves between 9.69 and 9.96.
Over 9.5 ranges from 49.5% to 52.9%.
That is a narrow range: three percentage points across eight seasons.
Verdict: partly true.
The picture is fairly stable, though there is always some season-to-season noise, and the league weighs far more.
What can't we say with this data?
- Which market pays: we have no corner odds from any bookmaker.
- What happens by half: there are no corners by half in the files.
- What happens inside a live match: the data is from full time.
- Which specific team wins more corners: here we measured entire leagues, not clubs.
Any article that tells you to "bet the corners over" based on this data alone is overreaching.
A frequency without a price is not a bet.
For there to be value, the real frequency has to be higher than the one the odds imply.
How do you turn frequency into a decision?
What if I want to automate this?
Counting corners by hand is slow.
If you want to go further, you can build your own rules and test them.
We have a guide on how to create your own AI for betting with your own parameters.
There is also the corner bot, built for people who want to follow this market in an organized way.
If you are more interested in live action, read when the next corner comes.
It is a different approach, not an exact prediction.
None of this is a promise of profit.
What to read next
If corners left you wanting more data, move on to cards.
The article Card stats: myths and data uses the same 39,856 matches.
After that, how the referee influences cards shows how much the game changes depending on who officiates it.
And to see how we measure our own picks, check the Betting Journal.
The full context is in the guide to AI for betting.
Betting carries risk and is for adults 18 and over only.
If you feel you are losing control, ask for help.
Past data guarantees nothing going forward.
Frequently asked questions
How many corners are there per match on average?
9.80 in our sample.
They come from 39,856 matches in 16 European leagues, between 2018/19 and 2026/27.
The home team wins 5.39 and the away team 4.41.
Each league drifts away from that average, so check the table for your league.
Does over 9.5 corners land more than half the time?
Almost exactly half: 50.9% overall.
But it depends on the league.
In the Scottish league it landed 59.4% of the time and in the Greek league 38.6%.
A global average hides big differences.
Are there more corners in the second half?
This data cannot tell us.
The files give the full-match total, not corners by half.
Anyone who gives you a percentage without showing a half-by-half source is improvising.
Can I use this data to tell if corner betting is profitable?
No.
The public files we used do not include corner odds.
We measured how often each line lands, not how much it pays.
To find value you need each bookmaker's odds, compared against the frequency.
Does the favorite always win more corners?
Not always, but most of the time.
The favorite wins more corners in 59.3% of matches.
For strong favorites (average odds below 1.60) that rises to 75.6%.
In 9.5% of matches the corner count is tied.
Why can an AI get corner stats wrong?
Because chat AIs often invent percentages that sound good.
Always ask for the source, the league, the seasons and the number of matches.
If it cannot give you those, treat the figure as an opinion.