It’s been a strange couple of weeks. Some of my work appeared in The Mirror and the Guardian on the same day, and The Mirror has asked me to continue to write for them, in my own style, about stats and football. The Guardian would like to keep StatsBomb as part of their sports network, but there are issues with republishing the data in our stuff that may or may not get sorted out.
I’m fairly impressed that six months after forming StatsBomb, two different national media sites want to feature our work, which is consistently riddled with stats and analytics.
Around the same time we leapt into national media, I also released the first player radar charts. These have, without a doubt, been some of the most cheered and controversial things I’ve done.
On the one hand, I have data journalists combining with owners of stats sites, and a raft of casual fans lauding them for being interesting and useful visualization of basic football numbers.
On the other hand, I have hardcore data visualization and stats people tearing them to bits for reasons various and sundry, many of which I explained in a piece I wrote a week before they started ranting, and most of which assume both my idiocy and their supreme knowledge of the discipline at the same time.
For someone who a) spent months learning about them before publishing and b) is fairly thoughtful about things in general, it’s odd.
I think the radars have a ton of value. The reaction people have to comparing shapes instead of stats is visceral. You can SEE the differences. You can use overlays to compare players. You can shop for similar shapes as replacement players, or even find better ones. You can (probably) use them in various ways to compare team statistical output, including finding strengths and weakness.
Maybe the coolest thing about them is: we don’t know all the uses yet.
The only thing I am certain about is that in many ways, they are useful. Imperfect? Definitely. But useful as well. That’s all I was going for. Actually, I was going for less. I just wanted to create them and see if they had any value at all. Full stop.
Create new thing. Test new thing. Learn more about new thing. Develop new applications, while discarding others. If it’s crap, then discard the whole thing and do something else.
It’s basic (data) science. How that would make people angry, I don’t know, but whatever.
That’s kind of the funny thing about writing about football and stats right now. People are sucked in to the learning experience, while at the same time castigating us for admitting we are still learning. They want to know more while shouting at us, “WHY DON’T YOU KNOW MORE!?!”
It’s odd. It’s even more odd when you think about the fact that basically none of us get paid to do this. We’re probably not even that useful in winning your pub quiz round.
I have worked in gambling for nearly a decade. I have an expertise in my subject matter (gambling and practical market economics) matched by maaaaybe a double digit number of people in the world. I still learn new stuff all the time.
Learning is awesome. Not learning would be boring and awful.
Listen… we have an infinite number of things yet to learn about football, stats, and analytics. It doesn’t mean that we (and therefore everyone else listening/reading) don’t know more useful info now than we did last month.
Not just pie in the sky shit like “possession is loosely correlated to more points”, either. I mean practical, help-you-win-more-football-matches type of stuff. Practical, help-you-buy-better-players type of stuff too.
It’s happening now. It will continue to happen into perpetuity. It’s progress.
How Dare You
Then you get pieces like the When Saturday Comes republication in The Guardian today. It’s not worth shredding the piece (although it’s filled with a number of blatantly wrong statements), when the sentiment is what actually matters.
Some people seem to get offended when you apply stats to the beautiful game.
I get it. That’s fine.
“I just want to watch football and drink a beer.”
“I want to trust what my eyes are telling me about what is actually happening on the pitch.”
“Thinking about numbers and math with regard to football is just wrong.”
I don’t mind those opinions at all. If you hold them, just tune out. Don’t click the clearly stats-based link. Roll your eyes when someone trumps your football argument in the pub with stats evidence. Please, sweet Christ, do NOT follow me on Twitter. You are welcome to your own opinion – no one is stealing that away from you.
I can tell you from personal experience, even if you do know the stats, you can still get into plenty of arguments with other people who have the exact same knowledge. Not enjoying statistics in football does not invalidate anyone as a fan. Even if we change the name of the game to Heatmap, as Barney Ronay so thoughtfully suggested.
I like stats. And football.
However, I still reserve my right to enjoy football and things like Santi Cazorla’s sublime first touch, even if I can’t quantify it. (Yet.)
I will say, the one thing that the WSC piece does well is complain about people using stats as a basis for awful opinions. The funny part is, this is no different than normal sports writing, or writing in the media in general.
There are plenty of terrible opinion pieces out there based on what the author saw or experienced. There are also plenty of people who are using stats who have no fucking clue what they are doing.
In short, there’s a lot of bad writing out there, and poorly thought out opinions. Some of it uses stats. Some of it does not.
The stats are not the root of this problem, the opinions are.
Or zonal marking is. One of the two.
I keep getting asked why I don’t argue with public skeptics toward stats and football more (especially on Twitter), and my answer usually takes the form of, “Why would I bother arguing, when I could spend that time working on new stuff instead?”
I’m not sure if I would have had this perspective before I got cancer, but I did and now I do.
I want to learn new things. I want to create new things. I want to teach new things.
And, critically, I want to foster an environment that helps other people do the same.
Negativity is fucking awful for doing all of that.
Skepticism is good. A healthy dose of skepticism is great for learning and getting to the bottom of things. Positive criticism is also useful, both as a tool to learn more, and a tool to improve what you’ve already done.
If you’re interested in this stuff, maybe just take a moment to think about the phrasing of your question before tearing in to someone’s new thing.
“Hey, I like this. Did you think about ? This place has data for that type of thing.”
Is a much different response than
“This is [bad/wrong/ignorant/random]. does this much better. I don’t know why you bothered.”
Or don’t consider your phrasing, and be like every other mouth-breather on the internet with a sense of self importance and a half-formed scathing criticism. Just don’t be surprised/shocked/appalled when you find yourself and your opinion lumped together with those other people.
Look… The Mirror* sought me out to write about football and stats for them. That’s either a sign of the apocalypse, or a fairly clear indication that this shit is actually happening.
It happened in baseball. In hockey. In basketball. It’s happening in the NFL and college football.
It will happen in football/soccer.
Teams that embrace analytics early will outperform those who are slow or stubborn at doing so, from now until the end of time. Even if the sole task is to help those teams to make fewer stupid mistakes, the value is there. (Assuming they are done correctly, which Damien Comolli taught us is not as easy an assumption as you might think. It is probably the most valuable thing Comolli has ever done.)
I choose to be relentlessly positive about the whole thing, because it makes me happier, and it means I can get more work done.
As a casual fan, or a mainstream writer, or a fellow stats geek, how you deal with it is your choice.
* For the record, I am happy to write for them. If they think it’s a good fit for what they want to do in the future, I’m cool with it too.