If data really is ‘the new oil’, then in the football business Opta are the Ewing family - furiously pumping statistical gold out of pitches around the world on matchdays. Do you see that straw? Opta are drinking your numbershake.
Thanks to a fortuitous accident of schooling, I was lucky enough to be invited to take a look around Opta’s London operation last year. There I was shown how the raw data is compiled during matches - in short, with a lot of very fast, very detailed clicking, which is then rigorously checked - and then got to muck about with the incredibly versatile query tool that analysts, both in the media and at clubs all over the world, use to search the Opta database.
A signed shirt from Arsenal's Invincibles season proudly displayed in the Opta office. Before long I was elbow-deep, on a personal quest to find a defensive midfielder, aged under 25, who wins more than 75 per cent of aerial duels, and has a high pass completion rate. And it turns out the ideal candidate is not Yann M’Vila, but Toulouse’s Etienne Capoue. Who knew, eh?
Before long I was elbow-deep, on a personal quest to find a defensive midfielder, aged under 25, who wins more than 75 per cent of aerial duels, and has a high pass completion rate
I also got to meet the editorial team behind the brilliant @OptaJoe account, which is led by @Orbinho - the man many of you will already know as the fount of Arsenal stat knowledge on Twitter. While not weeping over the number of times we hit the post last season, he was kind enough to answer some of my questions about what the numbers tell us about the current Arsenal side.
Note: answers are based on statistics from June 2012.
What’s your response to the statement (I’m sure you hear endlessly) that you can make the stats say anything?
I don’t think there are many absolutes with football data, but the data raises questions which people can assess using other measures, including what you see with your own eyes. For example, Arsenal had the worst crossing success rate [in 2011/12]. Some might say this means our players are bad at crossing (and would cite the occasional shank into the crowd by Theo) but actually perhaps it also has something to do with lack of bodies in the box or lack of a striker who is good in the air or the performance of the opposition defence. The truth probably lies in a combination of all of those things that together produce that stat. Context is everything.
How do you explain the high number of times we concede to the first shot on target?
The average for the Premier League last season was 10 games out of 38 conceding from the first shot on target faced. Arsenal conceded 13 times. It’s partly explained by the fact that Arsenal conceded a lot of goals not only in total but as a proportion of shots on target faced (37 per cent) and that as a team who enjoy a lot of possession, in most games the opposition’s first shot on target is from a counter-attack or a set-piece where the chance is perhaps easier to convert. Arsenal actually scored from their first shot on target more often - 16 times.
What do you think is the greatest myth about Arsenal that isn’t borne out in statistics?
The biggest myth is that Arsenal don’t convert their chances. They’re in the top four for chance conversion every season and always well above the league average.
This article first appeared on Arse2Mouse in June 2012
Is there a stat you feel is getting overlooked by the coaching team?
I’m sure that they’re aware of all of them. Clubs have access to a lot of data these days, but obviously they keep their cards close to their chests as to what they value and what they don’t.
The world seems to be obsessed with possession percentages now. Just looking at the data, what makes for good possession versus bad? (I’m thinking in the way we sometimes complain about the ball just being shuffled from left to right.)
Possession means different things to different teams at different times. It isn’t always how much you have but what you do with it. Barcelona dominate the ball which allows them to be patient and wait for a clear opportunity to create a scoring chance, but while they are passing it around it also means the opponents don’t have it and that reduces the chances of conceding. A team like Stoke are less worried about possession and more about position. Once they get the ball into a certain position of the pitch then they have ways of making the most of that possession. It’s the beauty of football that there is no one sure-fire way to score a goal or win a match.
Who would you say doesn’t get the credit he deserves in the current Arsenal squad, based on his stats?
It’s probably easier to highlight those who get too much credit. I think some players benefit from being fan favourites and get excused below-par performances or errors, whereas less popular types get unfounded criticism. For example Aaron Ramsey took a lot of stick, but only Van Persie created more scoring chances from open play last season for the club.
In 2010/11, I had the impression Jack Wilshere was fouled a disproportionately high amount of times - is that true?
He was fouled 48 times in 35 games - but fewer times than the likes of Charlie Adam, Cesc, Lucas, Arteta, Essien, Tiote, so not out of the ordinary for a combative central midfield player.
I think some players benefit from being fan favourites and get excused below-par performances or errors, whereas less popular types get unfounded criticism
Over a quarter of Arsenal’s league goals in 2011/12 were assisted by either Theo or Song - is that unusual to be so reliant on two creators?
Assists are an odd category in that of course you are reliant on the scorer putting the chance you created into the back of the net. Ramsey and Arteta actually created more chances than either Song or Walcott, and Rosicky about the same number, yet ended up with fewer than half as many assists. Van Persie got more assists than Walcott so there is a good spread among Arsenal players. Often, the player with most assists is also a set-piece taker, but then Arsenal don’t score that many from set-pieces.
Are there any prevailing trends in the Premier League you’ve noticed from the stats in terms of English football changing?
The last couple of seasons have seen record numbers of goals scored, this after a record low in 2008/09. There are now more goals than any season since the 1960s. Passing accuracy has been steadily on the rise over the last five or six seasons and took a big leap from 2010/11 76 per cent to 80 per cent in 2011/12. The number of fouls conceded has dropped by about 20 per cent in the last six seasons.
Interesting stuff I’m sure you’ll agree, and thanks again to @Orbinho and Opta marketing overlord @banouby for taking time to give me a tour of the data mines. Personally, I love this stuff.
No doubt some will retain Hodgson-esque suspicion of statistical analysis, but to them I’d say data in any walk of life is only as useful as the mind that’s interpreting it. Luckily, in our case that means Arsène’s big old brain. Let’s hope he’s furiously running searches right now, and the Park signing was the result of spilling his espresso over the keyboard.
The views expressed here do not necessarily reflect those of The Arsenal Football Club PLC or its associated companies (“Arsenal”). Arsenal is not responsible for, and accepts no liability in respect of, the content of external websites.Copyright 2014 The Arsenal Football Club plc. Permission to use quotations from this article is granted subject to appropriate credit being given to www.arsenal.com as the source 27 Feb 2013