Football analytics is becoming more and more a part of the modern game. With the huge sums of money floating around, especially in the Premier League, teams are looking for any way possible to gain marginal advantages over their competitors. Despite what it may seem from the outside looking in, according to Omar Chaudhuri, Head of Football Intelligence at 21st Club (a consultancy that advises football clubs and other sporting organisations), roughly 12 current Premier League teams are thought to be actively using analytics to inform decision-making.

Even with this seeming breakthrough at clubs, there is still little mention of it in mainstream media. Pundits and TV shows are becoming more welcoming of statistics, such as number of chances created, but they often included or referenced without any discussion of what they might mean. Even when football analytics does surface in the mainstream media, it is not always in a positive light; with negative reporting of Liverpool FC’s ‘stats-based’ recruitment and Brentford FC’s analytical approach facing criticism. This all leads to a lack of awareness or a general disinterest from the wider football public.

The main public availability of football analytics is on Twitter, and is where I originally stumbled upon it. There are a number of very clever people who produce interesting articles and analysis on a frequent basis. The site Statsbomb, for instance, published a bunch of great Premier League team previews in August which offer a completely different insight to the league.

With that being said, there is a clear gap in the market around Palace. As is the case with mainstream media, understandably the bigger clubs get more of the attention. I have been hoping for a while for more coverage of this kind on the Crystal Palace, but with none forthcoming, I thought I better crack on.

Measuring performance

A key concept with football analytics is that goals are not a very good way of predicting future performance. What this means is that just because you won your last two games 2 nil, doesn’t (necessarily) mean you will go on to win future games just as comfortably. Due to football being such a low scoring sport, and the effect of ‘luck’ or randomness in each game – the best team does not always win. This is one of the things that makes football such a great sport. The underdogs are always in with a shot.

Looking at the number of shots for and against is a much better indication of the actual performance of the team than just looking at the results. If you dominate more games and have more shots, although it doesn’t guarantee a win every game, over the course of the season this will result in wins. Going further, the amount of shots on target for and against is even better than just shots. This is because a shot on target is (generally) an indication of a better chance and so a team creating a large number of shots on target is probably a good team. For example, last season Tottenham and Arsenal took the most shots on target while West Brom and Villa took the least.

A good example of this is from the 14/15 Premier League season. Aston Villa, managed at the time by Paul Lambert, won 3 and drew one of their first 4 games, scoring 4 goals from just 6 shots on target but conceding just 1 from 10 shots on target. The Aston Villa board promptly rewarded Lambert with a contract extension and probably a pay rise. They were outplayed three of these games but the outcome of these games seemed to give Randy Lerner a false impression of the quality of the team/ management. Predictably later that season, after a dreadful run, Lambert was sacked. A closer look at the underlying numbers could have perhaps saved Lerner from this expensive mistake.

This flags another key concept from analytics (or perhaps common sense) – it is important to avoid drawing big conclusions from a small sample of games. A trend over 10 or more games is much more likely to be reflective of a team’s true ability rather than just looking at one or two games in isolation.

A lot of the analysis that will be in this blog over the season will be focussed around shot figures to give an insight into the performance of Pardew’s team. Things like the number of shots taken each game, the number of shots conceded, and also looking at total number of shots and shots on target across multiple games.

In the next blog I will explain further how shot numbers and other metrics can be used to dissect a team’s performance, and use it to have a closer look at Crystal Palace’s 2015/16 season.

Thanks for reading.