Does Soccer Need More Stats?

By: Daryl | May 28th, 2009

I linked to a similar article in the Daily Dose yesterday, but wanted to get into this a little deeper. Chris Sprow recently wrote an article for ESPN, arguing that soccer needs more stats if it’s going to catch hold of the American imagination.

The basic idea is that American football, baseball and basketball all have numbers which fans and pundits can use to compare players. Take a look at the NFL numbers for quarterback as an example. Or the fact that Nate Silver has made a career out of analyzing baseball data.

But for the most part soccer stats are limited to goals and assist, which can only tell you so much.

I wanted to disagree with Sprow and argue that the beautiful game is just fine the way it is, and that said beauty lies in the intangibles, the moments of magic that don’t really correspond to a numerical value. But watching yesterday’s Champions League final changed my mind to some extent, as I became obsessed with Xavi’s midfield dominance.

Xavi ran the show last night. Which is typically how all football writers, fans, bloggers and TV commentators would describe his performance. But to a non-soccer-savvy and stats-dependent US audience, those words don’t really hold much meaning.

But if you told someone like Nate Silver (I have no idea whether he likes soccer or not by the way, just using him as an example) that Xavi completed 66 out of 71 passes (a 93% success rate) compared to Michael Carrick’s 43/57 (a 75% success rate) then those numbers are concrete evidence of Xavi’s maestro-like performance.

So a big part of me thinks that Sprow is right. Assuming ESPN keeps showing soccer (their Champions League coverage is finito, but they still have MLS and World Cup 2010 rights) then maybe they should start incorporating some numbers. The question is: Which ones?

I’d suggest things like passing (attempted/completed), interceptions, tackles (attempted/successful), headers (attempted/won), shots (attempted/on target) as well as things like dribbles (attempts, distance covered, number of players beaten). There are plenty of numbers out there, it’s just a case of relaying them to the audience.

Some Related Stories:

Category: World Football

« Daily Dose: May 27th, 2009 | Home | You Can Prove Anything With YouTube »

The Offside RSS Feeds

wengerball

Stats can have it uses in the beautiful game it's just a matter on context. The guardian chalkboards ( http://bit.ly/1alsly ) is a great tool to compare players.
ish

i think analyzing player stats can be useful such as passes completed, tackles won, headers won, distance dribbled and things like that. It wouldnt add to the game as analyzing which player is better(DM's would have close to 90% pass accuracy but they are passing 2 metres to either side or back) but it would be useful just to know like thats interesting.
Kevin M.

I'm of the belief that the stats we have in football right now are not good enough. For the purpose of broadcasting and watching a game, I partially agree with you that they can take away from the intangibles that are the reason this is called the beautiful game. But for the purpose of analyzing talent, some serious improvement is needed. I honestly think that teams depend too much on the gut feelings of a manager and the owner when spending big money on players. Simply put, they don't put the same kind of analysis into their signings as Theo Epstein of the Boston Red Sox or Daryl Morey of the Houston Rockets does, which is asinine considering the transfer fees paid for some of these players.
Weston (@wpazzurri)

the beautiful game doesnt need stats, us dumb americans need it.
Colin

I think a lot of football fans are skeptical of statistics, and for good reason - most of the stats most quoted in the media are worthless. Clean sheets are like wins and losses for a pitcher, they depend more on the rest of the team than anything. Assists are bad because they depend on someone else to actually score the goal (assist attempts, on the other hand, are quite useful). Goals per game could mean completely different things for starters and for bench players. Passing rates are ok, but only in a large sample. And so on. There is good stuff out there, we just need people to recognize it.

Over the past couple years, I've been developing an all-encompassing stat for evaluating players, similar to the aforementioned VORP. My formulas actually borrow a lot from Bill James' win shares, but have been adapted heavily to fit our game. My goals are to eliminate as much luck an reliance on teammates as possible, and to incorporate as many aspects of the game as possible. The result is a win share-type score that can be interpreted as "A team of 11 Player Xs will win Y games". Unlike the ACTIM index, my system is based on a tangible unit - wins - instead of an arbitrary number, and unlike the ACTIM folks, I'm very happy to share details on my methods.

Shameless plug for my website:

http://championshipatbest.com

I'm a Fulham fan, so most of the content is specific to Fulham, but I've been trying to branch out and cover more.
Ian

I think stats can be useful, as long as they're applied in the right way. Football will never be as stat-driven as baseball, because the constant motion makes quantifying it into simple numbers difficult and sometimes misleading, but statistics can still be a good tool, especially for managers. The statistics in a flowing game like football, though, will always be more complicated than in a stop-and-start game like baseball, and explaining some of these stats in a way that's actually useful to the people that matter in the game won't be an easy task.
Madridista

I'm reminded of the match report of a NCAA D3 soccer match I once read: "University X had more shots on goal, but University Y led 9-4 on corners won".

Do we really need this? We've been following this game for years, most of us, and it doesn't take much to realize that while there are general indicators that suggest how well a team or a player is performing, almost none of them are consistently predictive of the final score. Even then, how are you ever going to work with the statistics? It's difficult to isolate the value of a statistic from everything else that goes on in a game, and even harder when you're comparing different styles of play. Barcelona's style of play may inflate one statistic, Man Utd another, Aston Villa yet another, for instance.

In the end, no matter the statistics, the only thing that matters is the 90 minutes. And we've all seen games that turned all statistics leading up to that point on their heads. Stop thinking so much and just enjoy the game.
Kevin

What about total # of touches on the ball in a game? This would seem to include defenders, midfielders & strikers. It does not necessarily quantify the effectiveness of those touches, but shows involvment in the game. Refine the stat to include touches that go to teammates or win possession and it becomes more meaningful.

Detailed stats are only useful for true fans of the game. As Adam says above, looking at baseball stats is not really for the casual fan, it's for the already converted.

To use stats to entice casual fans they need to be simple, self-explanatary and comparable.

What about the stats that some of these fantasy sites use to judge points given to a player every week. Can any of those be used?
Daryl

I can definitely see how passing stats can be misleading, so not going to argue with that. But how about for defenders?

I know I've banged on about this before, but defenders don't always get the credit they deserve. Maybe some stats like number of tackles/blocks/headers would help highlights the top performers?
harken203

Statistics are like bikinis. What they reveal is suggestive, but what they conceal is vital.
Jem

Forgive me if this is a dumb question, but i always wondered how they calculate the amout of pitch covered (miles/Kilometers?
Adam

I'm not a fan of stats. The fact nobody talk about averages and percentages much in football is one of the things that made it possible for me to get into it. I look at a baseball stat table and want to take an aspirin and get some sleep.
SP

i've always been in favor of a statistic that would show the difference in average goal margin when someone is on the pitch as opposed to when he is not on the pitch for his team. never actually seen it though.
mlyons

Soccer needs better stats than that to be useful. Things like passing percentage are like batting average in baseball -- superficially interesting but ultimately misleading in terms of assessing a player's true value to his team. A sport as intricate and team-oriented as soccer would be better served by skipping over second-level statistics and going straight to more comprehensive overall value statistics, the equivalent to VORP or linear weights in baseball. Actim already attempts something like that, but it isn't terribly widespread yet, I don't think.
Brandon York

You are correct that there are a number of stats at play already in soccer, but the passing stats are not as key in my mind as shots on goal, time of possession, offside calls, etc. Those are the statistics that reveal an aggressive style of play.

To me, what will educate the casual fan who grew up playing the most rudimentary of football is telestration. At the risk of lionizing a buffoon like John Madden, the ability to diagram a set piece or reveal a brilliant run without the ball is key to showing a non soccer-savvy public why there are brilliant players who happen to not wear a number 10 in midfield.

I think--slowly but surely--soccer will infiltrate popular sport culture. Already ESPN shows highlights from free kicks and corners. If I can understand the concept of draft and yall in stock car racing, the layman can figure out the Brazilian Box.
kayla

Hi,

We have just added your latest post "Does Soccer Need More Stats?" to our Directory of Sports. You can check the inclusion of the post here . We are delighted to invite you to submit all your future posts to the directory and get a huge base of visitors to your website.

Warm Regards

Sportstrove Team
Nolan

No.

Statistics are always deceiving, but in football even more so than most American sports. Assists are misleading enough, yes they are something to look at but never tell the whole story about who is creating goals. As Ginge said, Xavi might be a positive extreme but overall the whole story isn't told. Take the example of Messi, who certainly had a good game, but passing numbers wise looks to only be slightly behind Iniesta, when Iniesta has an overall significantly more effective game. A lot of those successful passes coming from Messi and Henry lead to dead end plays.
Ginge

No, the main problem with statistics is that there is a tendency to rely on them to much, and I think it's always better to judge a players performance qualitatively rather than quantitatively. The Xavi example does seem to help the case for stats, but then if you look at the team passing stats overall, there wasn't that much difference between the teams (I think the completion rate may have been the same) and it doesn't really indicate Barca's complete dominance over United last night