This Week's Show
We talk another poorly formed "advanced stat", some loopholes in the NBA CBA, why the NFL doesn't get marketing, and why shooting efficiency is hard to get.
You can watch us live at twitch.tv/nerdnumbers, we're deciding a new time, we'll keep you posted.
Please subscribe to Channel NerdNumbers on YouTube and like us on iTunes!
538 has a new metric that attempts to predict NBA player careers. We were unimpressed, to say the least. First, the data used was subpar.
- The use of percentage stats (e.g. assist percentage) have issues with how they're counted.
- Using height and weight are iffy, as the great book "The Sports Gene" goes over, it turns out wingspan is a better indicator, and this data is available.
- The draft has many issues that aren't related to player skill.
- You know how much I hate usage.
As we note, there are many better stat sources, which makes using these for the metric a somewhat odd choice.
We have some discussion on regression analysis and how a lot of the handwavy nature of CARMELO wouldn't even pass muster in a freshman level stats class.
We talk a lot more too. The good news is CARMELO is a fine lesson in how not to make statistical models. As if we needed another example from the "advanced stats" part of sports analysis.
Our takeaway is the 538 has an excellent graphics department with less than stellar analysis backing it. The good part is it reminded us of a funny "How I Met Your Mother" stats/graph joke.
As Brian notes, using good data was a vital part of Angus Deaton's Nobel prize.
A fun Salary Cap Hack from Larry Coon
Brian brought up a good QA with Larry Coon about a team could try and entice players by staying under the salary floor. One note I'll give this is any changes in the upcoming lockout/strike will likely not stop owners from spending money poorly.
Deadspin, the NFL, and how to Advertise
The NFL got Deadspin's Twitter account deactivated. Some may view this as poor marketing on the part of the NFL. Well, smart people at least.
Yep. The NFL is basically throwing away 50% of it's marketing. https://t.co/3JVqL7rq1A— The man with the #'s (@ArturoGalletti) October 13, 2015
Shooting Efficiency vs. Shot Volume
Despite the volume of pieces on the site, I do still do a decent amount of NBA analysis for fun. Sadly, some of it goes unfinished and "ends up in the desk drawer." I decided to show one to Brian this week. I made a heat map of shot volume vs. shooting efficiency as it relates to wins in the NBA. The gist is improving shooting efficiency by about a percent trumps getting an extra shot by the tune of a win a season. And, as we discuss, it's why shooting marginal three-point shots trumps being amazing from midrange.