Stats – It’s how you use them

Heat MapStats are changing the way we view football and the way we digest footballing information. The emergence of Squawka and have contributed to what can only be described as a statistical revolution that has taken place over the last few seasons. This has been a god send to would be bloggers such as myself, changing the way we write about football completely.

No longer can we simply make bold sweeping claims about the ability of a player or a club. Now, it is essential to compound our beliefs with a wealth of stats, or risk being exposed as bias or flat our wrong. On Think Football I regularly use statistics to corroborate my views and formulate content for the week ahead.

Now when I watch a game I am forced to confirm my conclusions by checking Squawka or to see if my assumptions were correct. Was Michael Carrick really passing the ball around with ease? Was Martin Skrtel as aerially dominant at the back as I first thought? These questions probably resonate with a lot of would be bloggers and statistics have made it far easier at understanding the way games are being played.

The problem with statistics

That being said, statistics are not without their shortcomings. To rely on stats as the be all and end all is ineffective and problematic. One issue that I’ve had myself a number of times is that it is very difficult to analyse the performance of a defender based on statistics.

Whilst Squawka in particular have a plethora of defensive stats on offer such as ‘defensive actions’ per game and a comprehensive break down of ‘duels’ per game, this is still incomplete at truly analysing the performance of a defender. Of course it is useful to see the number of interceptions, clearances, blocks, tackles and headers a defender is making per game, but often it depends on the style of each given team.

Most importantly though, statistics fail to take into account the positional aspect of defensive play. If a defender steps out just slightly and loses his man it is not a glaring error so will not be recorded under ‘defensive errors’ but it will point to a lack of positional awareness, which is arguably the hall mark of a top defender.

With goalkeepers also it is not possible to assess their ability at commanding the area or organizing a defence. Lots of goalkeepers are good shot stoppers but this commanding and organizing ability often separates the best goalkeepers from the rest. Positional play is so important at assessing ‘keepers, defenders, midfield players and even forwards, yet there is nothing in stats than can quantify this, which means that the naked eye is still oh so important in assessing ability.

This isn’t to say stats are useless but it is probably the overwhelming limitation with regard to stats. In the book the ‘Numbers Game’, the author Chris Anderson argued that Chelsea should have tried to sign Darren Bent because his goals equated to more points than any other striker at the time. But, as Sean Ingle in the Guardian remarked “as most experienced football-watchers could tell you, Bent is weak at holding up the ball and releasing others – a role he would have been required to play at Chelsea.”

Manipulating statistics

Stats can also be wonderfully manipulated in order to create debates. On Think Football  I have done this myself on occasion. If say one week I am low on content or even in need of amusement I sometimes write articles, using statistics that are deliberately provocative.

One example related to the fanfare around Santi Cazorla last season- undoubtedly a very good footballer- and whether he is world class or not. At the stage of the season when I wrote the article Cazorla had scored eleven goals, but seven of those came against bottom half of the table opposition, and just two coming against teams in the top five.

With many fans deeming him at the time to be better than David Silva and Juan Mata I ran a piece ‘is Santi Cazorla good enough in the big games?’ The article had a table comparing Mata’s productivity to Cazorla’s and in short showed Mata had scored five times and assisted two versus top five opposition compared to Cazorla scoring just twice and assisting two versus the top five sides.

My conclusion was simplistic in stating that this showed that Mata was better. In the comments section it led to a huge deal of criticism and outrage. One reader rightly remarked ‘football is not top trumps.’ This has become a problem with stats, it is too easy to rely solely on them and use them as the be all and end all.

Even in a debate a bit more complex than this one cited above, statistics can be manipulated, or even open to interpretation. Someone may have a 95% pass completion rate but average 0.5 long passes per game compared to a player with an 85% pass completion rate with an average of 10 long passes per game. Who do we deem to be the better passer? It is very much open to interpretation.

Football would be incredibly boring if everything was as simple as being solely determined by stats. In fact, much of the beauty of the game (the debate and the divergence of opinions) would be lost and made redundant. If we can’t argue with each other over the merits of one player over another, over one style of play over another, what can we do?! Stats are brilliant, stats have added a completely new dimension to bloggers, to tweeters and even to many major news outlets who collaborate with stats websites.

But, we and even specifically myself, must always remember to use stats carefully and remember that they are not final. There is always a counter interpretation and nothing is as good as our eye at analysing a performance, stats just play a helping hand. As one manager said on the book ‘The Numbers Game’ “stats can’t tell me who to sign… they can’t measure the size of a player’s heart”.

Amit Singh (Think Football)

Author Details

Leave a Reply