Advanced Stats in Hockey

One of the more interesting stories in hockey over the last several years has been the rise of “advanced stats”. The names of the new stats are familiar to many now (Corsi, Fenwick, PDO, etc), and they have given rise to a new war amongst hockey circles, that of the so called “advanced stats” view of hockey versus those of the old school “eye tests” or school old stats like plus/minus and purely point based stats.

When both sides get together, sometimes a calculator is used to slap the other person, and a duel is demanded.

What do these duels look like?

twitterfight1

twitterfight2

twitterfight3

Yes, it can get ugly sometimes.

This fight is interesting to me because it touches on many aspects of why both sides constantly get into fights with each other. So with that being said, here are my random thoughts on the rise of the advanced stats and the problems facing them. Keep in mind I fully admit my opinion means nothing, so don’t take this too seriously, nor should you get offended by it. Though if you want to twitter fight I’ll meet you at the back of the school at 3 pm.

Also I will be calling the two sides Advanced Stats vs Old School, please please please don’t give me a speech about how they really aren’t “advanced stats”, etc etc. It’s just an easy way of differentiation for the two sides.

Thoughts on Advanced Stats

- Both sides need to relax a little bit. The old school crowd needs to stop with the “nerd” this and “nerd” that talk, and the “have you even played the game, bro” talk. Sometimes the old school people can be incredibly dismissive of the new stats the advanced stats crowd is trying to showcase. There are some interesting theories and opinions to be found in advanced stats, so to just dismiss them as just “nerd talk” is foolhardy.

It is not a good look to be on TV or on twitter and do the old “the advanced stats guys would have you believe” and then start sneering at the camera like you’re Clint Eastwood about to shoot a stat in the face.

It is not a good look to have ONE example go against something an advanced stat theory has put forth and use that as a shining beacon that the advanced stats crowd are a bunch of lunatics. Do not get in a stats war with advanced stats with a small sample size. You will only look like an idiot. They will carve you up with stats.

- ┬áThat being said, the advanced stats crowd needs to take that chip off its shoulder and god damn relax sometimes. Look, I get that advanced stats are trying to make inroads in a culture that is incredibly hard to break into sometimes (more on that later), and that they face a lot of that dismissive behavior from many hockey fans/pundits, but man, sometimes you guys just go way too hard sometimes. It just reminds me of that Dave Chapelle sketch of “When Keeping It Real Goes Wrong”. I feel like sometimes stats guys “keep it real” at the drop of a hat and just go on an all out blitz in their efforts to prove that something a hockey pundit said was wrong.

I can’t count the amount of times I have seen stats guys nitpick and throw down with hockey pundits over things that seem extremely petty. Now to be fair, sometimes the advanced stats crowd can nail them out of the park and expose a line of thinking that is shallow and not very factual. Those are the good times, those are the times when advanced stats shine. But there are also times when I feel the stats crowd will write a 5000 word essay documenting how something someone in the media said was in fact off by .00003, and therefore, that person is a fucking idiot and should be kicked out of hockey and possibly beaten with a bat.

- One of the harder things to take from the Advanced Stats crowd is the fact you are basically telling people they are a bunch of idiots whose eyes cannot be trusted. So with that knowledge, advanced stats people would probably make a lot more headway if they approached things in a gentler manner. Instead of saying “fuck you, bend over and take my stats” why not take people out to a nice dinner of stats, explain to them what your thoughts are, then at the end of the night, make sweet sweet statistical love.

Wait….

Oh right, basically I think there are a lot of interesting viewpoints advanced stats can bring to the table. What I find annoying about the approach taken sometimes is the “well my viewpoint has god damn statistics behind it, yours just has your stupid puny eyes, so get the hell out of here” feel it has to it.

By all means explain your theory on why David Booth is better than I think. Please do not tell me why I am an idiot who has to realize how good Booth is. Take the nicer approach.

That is one of the biggest things I think a lot advanced stats guys ignore, the emotional aspect that comes with sports. People put a ton into their sports, so as a result, they can often lead with their emotions. So while I find it all good to boil away that emotion to try and find the underlying results that you think are more reflective of what actually happened in the game last night, I honestly feel the bedside manner of advanced stats can be improved by a fair margin.

One of the higher paying jobs in the IT industry is the guy who can talk to the techs and then talk to the project managers as the go between and let both sides understand each other. I honestly feel that kind of role is needed in hockey analysis, an advanced stats guys who can break down stats, make them approachable, and not alienate the general fan base. Daniel Wagner showcases this brilliantly in this piece:

http://blogs.thescore.com/nhl/2013/03/12/why-advanced-statistics-approaches-hockey-completely-wrong-but-still-gets-things-right/

He does a great job of talking about the new stats in an approachable manner, while still extolling the virtues of the new stats and putting forth a reasonable argument about their merit. Daniel also finds flaw in Corsi and Fenwick thoughts at times (http://blogs.thescore.com/nhl/2013/08/27/quality-quantity-and-why-hits-and-blocked-shots-still-matter/) which makes it seem like he has less of an “MY HAND PICKED ADVANCED STATS ARE EVERYTHING” agenda, and more of a “let’s see what we can learn with these new stats” agenda. I like this approach and feel it works best.

- To me, sports without emotions is boring as shit, so I enjoy both sides of it, the analysis as well as screaming at Torts for not putting Garrison on the ice on the power play. Just because I am screaming during the game at Torts doesn’t mean I can’t sit down and try and understand later if Torts was right or if I was right. ┬áLet people express themselves in the heat of the moment, that’s one of the best parts of sports, damn it!

- Sometimes stats can’t tell the whole story, in my humblest of opinions, and I think it’s ok that I think that. By all means, present your argument explaining how you think you are in the right, but I still feel stats don’t take everything in hockey into account, so sometimes your conclusions won’t match up with mine, and that’s ok. We can both live with that.

- I know this is probably feels like I am coming down more on the advanced stats, but I only do so because, like Torts, I know how much more they can bring to the table. To the old school crowd that dismisses advanced stats and doesn’t even want to see what they have to offer, I feel there is no hope for them. If you willingly shut your mind down to a potential new way of understanding hockey, and you’re in the media, then you’re potentially limiting yourself and I have no idea why you would do that. I find drilling down into stats really boring, I honestly do, but I still read up on advanced stats and learn about them because at the end of the day, there is a lot to learn from them, and at the end of the day, even if I hated them, they are a part of hockey culture and I would be stupid to ignore them. And while drilling into stats is boring to me, sometimes you can find interesting theories or trends out of them, which is always a good thing.

Advanced stats has a lot of interesting things to offer, so I hate to see it get dismissed due to the approach taken. One of the best things advanced stats can do is stop narratives from taking off and becoming lore. Sometimes a story line about a player will occur over the smallest of sample sizes and before you know it, that player is labelled with it for the rest of their career. Advanced stats does a very good job of keeping people more honest and making sure they solidify their own argument before presenting things as fact.

Just, you know. Don’t keep it real all the time.

wutang

PS: To get rid of the elephant in the room, I am not going after Charron or Drance or Canucks Army or the best guy of all Dimitri Filipovic, at all. I enjoy their work and consider them friends, so I hope they don’t feel like I am gunning for them. The trends I speak of I have seen all over the place, I just used the Price vs Canucks Army as a good view into what can happen sometimes in these spirited exchanges.

One thought on “Advanced Stats in Hockey

  1. *slow claps* Sometimes it seems both sides are more intent on being correct than painting an accurate picture of what is happening on the ice. Both sides need to calm down, or people will tune them out and find a more sensible person to listen to instead.

    Example: There’s a glazed donut, and a jelly donut in the lunch room. Lenny stands and argues with Carl about which one is better. Homer walks in and eats both delicious donuts. Be Homer.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s