The growth of football statistics have revolutionised the way many people think about football, but not everyone is convinced.
In the early parts of last spring, about a year ago, I came across a site named “Whoscored” via a forum discussion on just how good Gareth Bale was. On Gillette Soccer Saturday, Jeff Stelling had claimed he was as good as Cristiano Ronaldo, which naturally led to discussions on the merits of the statement. At this point one of the doubters brought up a diagram which showed a far less renowned player named Theo Walcott, playing in a similar wide offensive role to Bale, had a greater number of career goals and assists, despite playing a fewer amount of total games, showing that Bale perhaps wasn’t obviously much better than the top players in the league. This naturally drew contempt from others, who laughed off the meaning of “stats” and the meaning they have in evaluating a players overall quality.
One person did, however, point out that Walcott wasn’t even statistically superior to Bale. Obviously despite being similar positionally Bale and Walcott’s roles weren’t identical. Walcott is a far more direct player and doesn’t get that involved in Arsenal build up play, being more the icing on the cake to a move. Whereas Bale was the focal point at Spurs, where most moves came through him, and he also offered far more defensively. The second poster pointed this out by claiming that in nearly all other areas Bale was statistically superior to Walcott, sighting whoscored.com as the source.
For those who don’t know, Whoscored is a football site centred around statistics. They provide a great service for finding out results, fixtures and live scores all around the globe but their major quirk is the in-depth statistical analysis they provide for the major European leagues. For each match everything a player does is recorded and at the end they’re given a statistical rating based on their performance.
Upon first looking at the site I was intrigued. In-depth statistics had never really been covered in the mainstream media and I didn’t really know what to think of what I was looking at. Due to being a major follower of cricket, where raw statistics play huge roles in how players are rated and matters of selection, I was well equipped with the idea of carefully using statistics to analyse a sport. I also, however, new that such data could be misused by people in arguments and articles. What I found most remarkable, however, was that despite following football for nearly a decade, most of the stats were like a different language to me. I simply hadn’t been exposed to them before.
Arguably because of this, many football fans are incredibly skeptical of the role statistics can play in football. They’re a relatively new innovation that reflect the technological age we live in and are something people from certain eras have never experienced. Certainly the idea that a 17 year old kid in London could get a detailed statistical report of all games in leagues such as the Bundesliga or Ligue 1 moments after the full time whistle was blown would’ve seemed unthinkable not so long ago. One would think that it’s something we should enjoy and encourage. People can now look at football in greater depth and have another dimension to how they analyse and follow the game. Since finding Whoscored last year my understanding of football statistics have skyrocketed and it’s completely changed my understanding of football and how I experience the game. Not everyone sees it the same way, however.
For many people football isn’t something which can be read through a spreadsheet, graph or chart, it’s something that needs to be watched and experienced. A great attacking player’s contribution can’t be quantified into a few dribbles, key passes and a shot from a prime area. Pub debates about who the best midfielder in the country is aren’t about who can read the most impressive stats off their smartphone, it’s about forming your own opinions based on what you see with your own eyes.
They’re partly right of course. Statistics used without the right context, or presented badly, are close to meaningless. Using pass competition rates on their own as the way to decide who the best passer in the Premier League is foolish. Laurent Koscienly clearly isn’t the best passer in the premier league. But as a part of a whole greater range of stats, such as key passes, successful long balls, successful through balls and where the passes are made can give you a great idea of whether or not a player is a top notch passer or not. Furthermore, even the stat on its own isn’t useless. The top four for pass completion rate for this season are all Arsenal players, all of whom play centrally and primarily defensively, which speaks volumes about how Arsenal like to control the ball on the edge of the opposition half, liking to distribute the ball carefully out wide and are willing to be patient before seeking the ambitious ball.
They’re not perfect, no, but neither are human judgements. Humans can be swayed by personal preferences, stereotypes, conventional opinion and more which computers and databases are devoid of. Most humans would probably tell you the growling in your face Mathieu Flamini makes more tackles than the graceful Spanish passer Mikel Arteta. They’d be wrong (Arteta makes twice the tackles per game Flamini does).
Statistical ratings systems certainly create interesting results. Whoscored ranked Santiago Cazorla as the 3rd best player in the premier league last season, ahead of the likes of Robin Van Persie, Juan Mata and Eden Hazzard. Meanwhile, Michael Carrick a PFA player of the year nominee – Cazorla wasn’t – was ranked 62. Stop, I hear you cry. That’s a perfect example of statistics being flawed and not appreciating the role Carrick plays. Well, it’s worth noting Mikel Arteta played a very similar role for Arsenal last year, and he was ranked 14th. One could argue Carrick was the more impressive since he did it for a tittle winning side, but it’s also worth noting that players ratings for each match get a boost if the game is won; Arteta’s rating would’ve been even more superior to Carrick’s had the sides respective results been reversed. I’m not trying to diminish the role Carrick played in United’s title win last season, but perhaps he wasn’t the best holding midfielder in the league and his nomination was simply a case of people only really noticing a good holding midfielder in a title wining side (I’ve previously written about this before). That’s a basic human error and flaw which a statistical rating doesn’t have.
Another interesting result is just how far in front of everyone else Frank Ribery has been in the Bundesliga in the last season and a half. His Ballon D’or case was a peculiar one, mostly because people were unsure how to really mount his case. Most centred around the need for a Bayern player to be involved, although a large proportion felt that any of a large group of Bayern players had just as worthy a case. Ribery certainly didn’t standout like Messi or Ronaldo and was neither as incredible on the eye or in basic statistics as his two rivals. Yet for the period in question Whoscored had him as by far the standout for Bayern and not far behind Ronaldo and Messi. Statsbomb, a site created by someone who has worked in football betting since 2006, also revealed interesting statistical results for the Ballon D’or nominees. Messi and Ronaldo dominated goal scoring and shooting stats, but Ribery dominated nearly everything else; passing, dribbling, defensive contribution and was arguably more influential in the overall performance of Bayern Munich. In a side which won everything, that’s no mean feat. His Ballon D’or case was certainly a strong one, perhaps more so than most would’ve imagined.
There are still those who resist the influence of statistics like this. I myself have little problem with that. Football is not only a highly competitive game of which lots of statistical commodities can be derived from, it’s also an art and form of entertainment, arguably more importantly so. What of the negative narrative that many have towards the development of football statistics? That I have a problem with. Football statistics aren’t yet thrown at people. If you want to seek them you can, if it’s not for you, you can easily avoid them without having to go out of your way. I did it myself blissfully for years.
There’s also no reason why both can’t live in a happy harmony. I love the things we can find through good use of statistics. They not only provide a measure of quality, but also how players play and how certain people are unique. We can look into three great Arsenal midfielders but see they’re all incredibly different. We can look into the fascination of Frank Ribery’s role and how he and Cristiano Ronaldo, despite playing in the same position, have incredibly different roles and output. That doesn’t mean I can’t enjoy Barney Ronay describing the phenomenon that is Thomas Müller, in a truly unique and vivid style that no array of graphs, charts and tables could ever do. I rarely watch a football match without following it live on Whoscored, but conversely, I’d never follow a match online like that and not bother to watch it if I could.
Of course, like in most similar circumstances the best solution is an effective medium. Analysis not simply being a few tables and numbers, but also not being as airy fairy as a few creative adjectives, metaphors and cliches – with perhaps more extreme versions of each for the hardcore fans. Statistics have provided groundbreaking analysis of professional sport, and it’s a major shame that articles like this represent the views of so many. It’s a fascinating football world we now live in and it should be nurtured and encouraged, not diminished and putdown.