Corners

Corner predictions explained: ranges, lines and “today”

· 9 min read

In 30 seconds

  • Manchester City's last 10 away games averaged 12.40 corners and Nottingham Forest's last 10 at home averaged 8.70. The range between the two averages caught the real total: 10.
  • In other words, a range tells you the zone a game tends to fall into — and the next day Sunderland v Chelsea showed that it sometimes misses badly.
  • "Over 9.5 corners" means 10 or more, both teams added together. This season 47% of Premier League games got there.

Corner predictions explained: ranges, lines and "today"

Corner predictions tell you how many corners a game tends to produce, not how many it will produce. A range is a zone, a line is a yes-or-no question, and "today" means the numbers are rebuilt every morning with the fixtures of the day.

Two Premier League games from November 2026 show both sides of that. Nottingham Forest v Manchester City pointed at roughly 9 to 12 corners and finished with 10. The next afternoon, Sunderland v Chelsea pointed at the same 9 to 12 and finished with 4.

What a corner prediction is (and what it isn't)

A corner prediction is a probability built from history: how many corners each side's recent games produced, where they were played, and how the two sides add up in this fixture.

That makes it useful for three things and useless for a fourth:

  • Where the game sits compared with a normal game in that league.
  • Which side of a line the total is more likely to land on.
  • How much the two teams disagree, which is often the most interesting part.
  • What the final number will be. No prediction does this, and any site that claims it does is selling certainty it cannot have.

So treat a corner prediction the way you would treat a weather forecast. It narrows the possibilities and it is still wrong sometimes.

The range: what "9–12" (or "0–4", "0–8") means

A corner range is the band the total tends to fall into. Before you read one, check what it counts, because the same-looking range can mean three different things.

RangeWhat it usually coversHow to read it
9–12Both teams, full 90 minutes plus stoppage timeA normal-to-busy game: the total tends to land in this zone
0–4One team only, or one halfA low band like this almost never means a full match total
0–8The low end of a full-match band, or the "under" side of a line8 or fewer corners in the game, which is a quiet match

The reason a full-match range almost never starts at 0 is simple: the Premier League averages 9.49 corners per game this season, so a total of 0 to 4 is an outlier, not a band. In other words, if a range you are reading starts at zero, it is probably counting one team, one half, or a short window, and comparing it with a match total will send you the wrong way.

Short windows have their own numbers, and they are much lower by nature. A band for the first ten minutes of a game lives on a completely different scale from the full-match band.

The line: over 8.5, 9.5 and 10.5 in half a sentence

A line turns the range into a yes-or-no question. "Over 9.5 corners" means 10 or more in the game, counting both teams; with 9 it is under. The half corner exists so the total can never tie with the line.

LeagueAvg corners per gameGames with 10+ (over 9.5)
Bundesliga 2026/2710.4259%
Serie A 2026/279.7545%
LaLiga 2026/279.7150%
Premier League 2026/279.4947%
Ligue 1 2026/279.0844%
Liga Portugal 2026/278.9043%

CornerPro data, finished games to 16/11/2026: Premier League 100, LaLiga 119, Serie A 110, Bundesliga 81, Ligue 1 90, Liga Portugal 98. Second source: ESPN.

The Premier League line is the one to keep in mind here, because both examples in this guide are Premier League games. Fewer than half of them, 47%, reached 10 corners this season. In other words, "over 9.5" is close to a coin flip in England, while in the Bundesliga it comes up in nearly six games out of ten.

This also explains why a 9.49 average does not mean most games go over 9.5. An average is pulled up by the handful of games with 15 or 16 corners, and those games do not make the quiet ones any busier.

Where a range comes from: build one by hand

You can build a usable range with two averages and no maths beyond adding up and dividing. Take the home side's last 10 home games in the league, take the away side's last 10 away games, and the zone between the two averages is your range.

  1. Home team, last 10 at home. Total corners in each game, both sides added together, then the average.
  2. Away team, last 10 away. Same thing.
  3. The range runs from the lower average to the higher one.
  4. The midpoint is the single number to compare with the line.

Nottingham Forest v Manchester City, on 7 November 2026, works as the worked example:

StepNumber
Nottingham Forest, last 10 home games in the Premier League8.70 corners per game
Manchester City, last 10 away games in the Premier League12.40
Range between the two averages~9 to 12
Midpoint, the number to hold against the line10.55

CornerPro's own model weighs more than these two averages — corners for and against, the pace of the league — and its range for an upcoming game is a Premium feature. The hand-built version above is not a copy of it, just the quick version anyone can do with the free stats, and it is the one used in both examples below.

A real example that worked

Forest v City landed inside the hand-built range. The total was 10 corners, against a range of roughly 9 to 12 and a midpoint of 10.55.

10
corners in Nottingham Forest v Manchester CityCornerPro data, 7/11/2026; ESPN agrees
8 of 10
City away games with 10+ corners before this one
3-7
the corner split: City lost 3-1 and still had 7

The split is the part worth keeping. City lost the match 3-1 and still won 7 of the 10 corners, because corners follow where the ball is being played, not the scoreboard. A side that is behind, attacking down the wings, keeps forcing clearances.

Forest's own history pointed the other way and still fitted: only 3 of their last 10 home games had reached 10 or more corners. In other words, one side pulling the total up and the other pulling it down is exactly what produces a middling total like 10.

And one that missed, badly

The next day the same method pointed at the same zone and the game went nowhere near it. Sunderland v Chelsea, on 8 November 2026, had 4 corners in total.

Sunderland v Chelsea, 8 Nov 2026Number
Sunderland, last 10 home games in the Premier League9.20 corners per game
Chelsea, last 10 away games in the Premier League11.00
Range between the two averages~9 to 12
Actual corners4 (2-2)

Sunderland scored in the first half and were 1-0 up at the break, and from there the game emptied out. Each side managed 2 corners across the 90 minutes; even Chelsea, chasing the game, never got to a third.

That is the honest limit of any range. It summarises what the two teams have been doing; it cannot see the early goal, the red card, or the night a match simply dies.

A range is built from games that have already happened. The game in front of you has not happened yet.

The "today" part: why the prediction changes every day

A corner prediction is dated. The page lists the fixtures of the day, and the numbers behind each one shift as soon as a new round of games is played, because the last 10 home or away games are not the same 10 they were last week.

There is a second reason to care about the date. Once a game finishes, the prediction is marked right or wrong, and the site shows the hit rate of the last 30 days for each market. Those figures move daily, which is why you will not find a fixed accuracy number quoted in this guide.

Common mistakes

Most misreadings come from the same five habits:

  • Reading the range as a guarantee. Sunderland v Chelsea is the answer to that one: 9 to 12, and the game had 4.
  • Using three-game averages. Three games is one freak result away from a wrong number. Ten is the minimum worth trusting.
  • Ignoring home and away. City's away games averaged 12.40 corners; a single all-games average hides that, and the fixture was at Forest.
  • Forgetting how a game unfolds. An early goal, a red card or a two-goal lead changes the corner count within minutes.
  • Comparing ranges from different sources. One site's range may count both teams, another's one team or one half. Different things, same look. If you are deciding which corner prediction site to read, the comparison below says what each one is good for.

How we measured

The rule for both examples is the same: the home side's last 10 home games and the away side's last 10 away games in the same league, finished games only, with the total being both teams' corners added together over 90 minutes plus stoppage time. Forest, City, Sunderland and Chelsea each contribute a 10-game sample, which is why no small-sample warning is attached to them.

Game data comes from the CornerPro database (Premier League, finished games) and was checked against a second source, ESPN: corners 3-7 in Forest v City and 2-2 in Sunderland v Chelsea, identical in both. The league table above is CornerPro data to 16 November 2026, also checked against ESPN. Every figure here is league football; when a number does come from only three or four cup games, we always put the team's league figure next to it.

Frequently asked questions

What does a corner range like 9–12 mean?

It is the band the game's total corners tend to fall into, worked out from both teams' recent history. It is a probability, not a guarantee: some games land outside the range.

What does "over 9.5 corners" mean?

That the game has 10 or more corners in total, counting both teams. With 9 it is under. The .5 means the result can never tie with the line.

Are corner predictions always right?

No. An early goal or a red card changes the game. Sunderland v Chelsea on 8 November 2026 pointed at 9 to 12 corners and had 4. On CornerPro every prediction is marked right or wrong once the game ends, and the site shows the hit rate of the last 30 days.

Read next

All articles →