It is important that we start the game well.

We must not let our focus drop towards the end of the first half.

It is imperative that we end the game strongly.

These are quotes that we often hear from different (sometimes even the same!) football managers about what they feel are the most important time periods in a match.

But, does anyone actually know what the most important segment of a football match is? The disparity in how managers talk about this topic clearly indicates that they have no definite insight.

We know that there is a strong correlation (statistical measure to find the relation between two or more variables) between scoring first and not losing. Analysis of the European Championships and the 2011-‘12 Premier League season have shown that if you score first, you end up winning 64% and does not lose 80% of the time when it comes to the European Championships. A team also wins 71% of the times and not lose 87.5% of the time in the Premier League when they score first. But what does this mean with regard to when that goal should be scored? Is it vital to start strong and break your opponent’s rhythm early on but leave them with enough time to recover? Or should a team begin/end a half strongly so as to give themselves a confidence boost?

Getting the data into place

I began to conduct my own research to try and find an answer to this question and I followed the approach below:

To further explain how these correlations are calculated, I would use a simple example from the English Premier League data set.

  • Analysed data from the top four leagues (English Premier League, Spanish La Liga, German Bundesliga, Italian Serie A) for the 2015-‘16 and 2014-‘15 seasons. This would ensure that results from one league would not skew the results in favour of one time period. Data for only these two seasons was available at this granularity.
  • Break the game into six 15 minute periods (0-15, 16-30, 31-45, 46-60, 61-75, 76-90). Breaking the game into halves/five minute periods would be too long/short a time period to derive any definite insights.
  • Calculate the goal difference (GD) for each team in each league for these six time periods. GD was taken as a parameter and not goals scored as it was not an accurate measure. If a team that had a lot of points scored a lot of goals towards the end of the match but also conceded many, it does not mean that the end of a match is a significant time period. Although, if that team scored many goals and also concedes few, that time period would be significant for that team.
  • Obtain the total points scored at the end of each season for each teams.
  • Find the correlation between the GD in each of these time periods and the overall points scored. The higher the correlation, the more significant the time period.

To further explain how these correlations are calculated, I would use a simple example from the English Premier League data set.

The table above provides the GDs for the four teams at either ends of the points table. Aston Villa had, by far, the lowest points total in the two seasons (the next closest team had almost double, 30) whereas Chelsea and Leicester City were the two champions in the two seasons under discussion. When looking at this table, one might conclude that the time period at the beginning of the second half (46-60) should be the most significant as this would explain the relationship between Aston Villa’s appalling GD of -14 and their low points total. However, we must look at the entire dataset.

The highest point scorers – Chelsea (87) in 2014-‘15, Leicester City (81) in 2015-‘16, Manchester City (79) in 2014-’15 – had a much higher GD in the last 15 minutes as opposed to time period 46-60. Chelsea and Manchester City had GDs of one and six in time period 46-60 respectively whereas this shot up to eight and 15 in time period 76-90. As a result, we would notice a higher significance for the final time period as opposed to time period 46-60.

‘Fergie time’

What we are interested in is finding out the most significant time period in each league. We are not concerned about the fact that the correlations are varying across leagues as inter league comparison is outside the scope of this article.

“Fergie Time” or scoring late to win a football match is a term synonymous with Manchester United’s Alex Ferguson, considered to be one of the greatest managers of all time. What is fascinating to see is the significance of this time period (76-90) across all leagues. Even when it is not the most significant time period in some of the leagues (La Liga, Seria A), it is always in the top two. Sir Alex had picked up on the significance of this time period a long time ago and his success may well be down to his emphasis on pouring everything into the final period of the game.

The strong correlation between ending the match well and overall points emphasises the importance of fitness and concentration levels for teams as these tend to fall during the last 15 minutes of games. A team that can maintain these throughout the match will tend to finish the match strongly and thus accrue more points.

Some managers stress on the importance of starting games strongly, some emphasise on not losing focus during the middle of the game and most use their half-time talks to try and elicit a response from their team and start the second half strongly. It is now clear that all managers should hammer home the message of finishing games strongly as this is vital to their success.

(Jayant is currently working as a Data Scientist and constantly finds himself trying to explain statistical quirks in sports)