In my recent blog Does seasonality exist in football? I looked into goal scoring rates by calendar month in the top four European leagues.
My conclusions were that there was evidence of seasonality, despite the modern day improvement in pitch conditions throughout the winter months, and that goals are scored at varying but repeatable rates throughout the season.
Taking this further, I investigated whether seasonality was also prevalent in corner kick production.
In this piece, I will look into the possibility of seasonality in corners and also investigate the relationship between corner kicks and goals scored.
The following graphic shows monthly corner rates compared to the league average in the top four European leagues over a period of five years, ten years and the current season so far.
For example, if the average corners per match in the English Premier League was 10.5 and the average figure for the month of April over the same period was 10.5, then this is 100% (exactly average):
Although each league seems to follow its own pattern over both the five and ten year time periods, this seems to be league specific rather than similar across all four competitions. This leads me to conclude that generic catchall seasonality does not exist for corner kicks.
So if seasonality is not a factor in corner kick production, is there a relationship between corner kicks and goals scored?
If we think of goals scored and corner kicks as two possible outcomes of an attack, then I would think on a basic level there should be a relationship between the two.
There are obviously other eventualities of an attack that do not result in a goal or a corner, such as shots being blocked, defended or saved. Of course attacks can also break down without any shot output, with players being tackled or a poor final pass giving the ball back to the opposition.
If the outcome of an attack is a goal being scored then there cannot be a corner kick as well, so on one hand I would expect high scoring matches to produce lower than average corner kicks. Alternatively, if a match is producing a higher than normal amount of total attacks – perhaps in an open and free-flowing contest – then there is more scope for corners to be won, irrespective of how many goals are scored.
Tactics and style of play must also be a factor in corner kick production.
If a team sets out with natural wingers who are instructed to get to the by-line into crossing positions whenever they receive the ball, then they are naturally in better positions to win corners. With the introduction of StrataData, we are now collecting this type of information and beginning to understand how corners are won and which players are winning them.
The next graphic shows goals versus average and corners versus average numbers for the same four leagues over the same time period:
Looking at this graphic, the Bundesliga 1 plot shows no sign at all of a steady relationship between goals and corners, but there is some consensus in the other three leagues at least.
In the Spanish Primera and Italian Serie A there does appear to be a relationship, with the trends for both mirroring each other quite well. This would possibly suggest that when the league is going through a strong period of attacking play, then there are more goals and corners produced than average.
The English Premier League also shows some sign of this relationship too, with the corner rate rising above average during high goal spells from 2007/08 until 2011/12.
A possible explanation for corner kick production not mirroring the number of total goals being scored is the idea of Game State.
Game State reflects the position of each team in a match, with Game State 0 being when the game is level, -1 when a team is losing by a goal and +1 when a team is winning by a goal, etc.
The interesting thing here is that teams have unique corner and goal production rates within each state.
As one recent example, in Tottenham Hotspur’s 1-0 home defeat to Leicester City on January 13th in the English Premier League there were 24 corners, more than double the league average.
Tottenham produced 16 of these corners with the match at 0-0, with Leicester scoring their winning goal in the 83rd minute. This is one such occasion where corner production does not align with goal production, highlighting the varying factors at play.
In conclusion, and unlike with goals scored, I have not seen any evidence to suggest seasonality exists in corner kick production. It is my belief that there are too many factors involved in generating corner kicks and that they often conflict.
However, I still believe a relationship must exist between corners and goals and as a result this is an area that I will investigate further in my next blog.