Quants and Writers: Getting Along


Do you read The Diss NBA? If not, I’m sad. Assuming you have, you’ve seen Kevin’s great interview with Ethan Sherwood Strauss (@SherwoodStrauss) Ethan discusses the impact of Quant-bloggers on sports. I’ll let Ethan let me know if I missed his point, but I basically interpreted it as – sports bloggers bring a new set of information and view of the sport. But somehow, many lack the ability to make those stories compelling. Some of this could have to do with numbers themselves. Ethan posted another post on how adding a decimal changes how we view things. Saying “Shooting well from deep!” vs. “Averaging 1.7 shots per three-point attempt!” just aren’t the same.

Ethan sums it up best with this quote

There is that sense when I am reading what people write to get better informed, when I am forcing myself to gut it out and ingest information rather than enjoying it. I never felt that way when reading Hollinger’s stuff; that takes a rare talent.

And I’m gonna have to call us on it. Yesterday, Arturo released some of the best statistical analysis on the web. The first comment from longtime reader Statmandu about his work?

This in an amazing report, but it might take until the end of the season for my brain to process it all. :-)

Hate to say it, but Ethan nailed it. Us quants love our tables, our charts, our “X.X per Y”s to explain stuff! And the key is, many of us are different. You see, some will say we’re ruining the sport by taking out the mystery or the fun. I disagree! When you dig into stats, there is so much interesting stuff. The difference of a few missed shots or an extra bumbled ball night after night changes the tides for a player and as a result can change the history of their team. When we comb the data, we see that! It’s amazing! And Ethan sees that our issue is figuring out how to properly convey the beauty of that. It’s a rare skill, but writers like Michael Lewis show that it can be done. And yes, Arturo is right, we’re working on it!

Now, I’m gonna respond a bit though. You see, it’s hard to be an expert at everything. There are some excellent sports writers. You know what they have tons of experience doing? That’s right, writing! Finding people with lots of experience both writing and doing complicated statistical analysis is rough. And that’s where we need to work together.

Ethan lists Abbott, Hollinger and Tom Haberstroh as examples of good writers. I agree! Yet, Abbott has written about how numbers are just a tool. Right before last year’s Sloan he wrote:

The bigger point is I’m not into numbers. I just use them. This is the same relationship I have with lots of stuff, like my car, plumbing and DVR.

We’ve discussed how Hollinger (despite being a good writer and sometimes providing good analysis) still kept bad stats around just because they were popular.

On Hoopspeak,  Haberstroh once discussed a new defensive metric he was working on. When I asked what tests he’d applied to ensure it worked? He responded that he hadn’t! And understand he had formed theories and ideas and rankings already! Was his prose and explanation great? Absolutely! But that’s not the right way to deal with stats.

While not mentioned explicitly in the article, Zach Lowe and I once argued over a point that Lowe said was almost universally accepted. When I pointed at the data (done by none other than the amazing Kevin Draper), which contradicted popular belief, I was told I had a nice opinion.

All of these cases show the problem. Stats people (at least the good ones) can find beautiful and fascinating stories in the data. But we can very easily stumble over decimal points, weird scales, and hard to get numeric techniques to tell our stories. Good story tellers can often view the stats as little more than a tool to help tell their story. And neither of these approaches gives stats the treatment they deserve. The stats have fascinating insights to share. We’re all here because we love sports and the stats add another amazing layer to that! But for everyone to see that, we need the story tellers to give stats the respect they deserve, or the stat heads to find a way to show how beautiful the numbers really are.

The fact that we’re having this dialogue gives me a lot of hope. The truth is we’ll have to work together though. There are the rare gems that find a way to be good at both numbers and writing. But neither topic is trivial or easy. Arguably the most successful examples of this have been Freakonomics and Moneyball. Those worked when the two sides came together with mutual respect and found a way to tell compelling stories. Here’s hoping it happens more often in sports.


Comments are closed.