Editor’s note: This post is part of our Sports Analytics Roundup series. You can read more about the series here. One of the goals of this series is to foster engagement within the analytics community. Thus, I encourage you to comment with your opinions and thoughts.
Quantifying chemistry: This one was quite a ride. ESPN The Magazine's Jeff Phillips ($) shared a proprietary model for evaluating team chemistry in the magazine's MLB preview issue. There were three key factors in the formula: demographics, isolation and ego. But there were precious few other details. That led to these two tweets from DJ Gallo (here and here) for actual snippets from the magazine. This formula was pretty ludicrous in that context.
SABR innovator Phil Birnbaum took on the topic and focused on the non-publication of the dubious regression model in question. He also asked: If this team chemistry is a value add or minus, then wasn’t it already embedded in previous year’s statistics? The Score's Jack Moore added on to Birnbaum's comments, accusing the piece of blatant SEO-type marketing to an analytics crowd. This formula dehumanized players and failed badly in its mission. There's a reason why it's difficult to really enumerate chemistry.
The value of steals: FiveThirtyEight has made quite the waves with their first two weeks of existence. This story was perhaps one of their more splashy hits so far, for better and for worse. FiveThirtyEight’s Benjamin Morris wrote about the hidden value of steals, finding that: “A marginal steal is weighted nine times more heavily when predicting a player’s impact than a marginal point.” The article is super interesting from the standpoint that it gets you thinking. It was pretty controversial for a number of reasons on Twitter.
Tangotiger had a quick post about how Morris’ math mostly addressed a problem in the need for better defensive metrics, but didn’t necessarily solve the issue well. HEAT.com writer Danny Martinez had a recent post on Chaos Percentage, a stat that also could spark further analytic dialogue. And I also was left thinking about this 2009 post from Jon Nichols, now the Cavs director of analytics, about diminishing returns of steals, among other box score stats. That’s a fun one to keep in mind when discussing steals.
Ex-NBA players v. analysts: And in a story written by “ESPN.com news services” but mostly reported by Chris Broussard, we learned that certain ex-NBA player front office executives “fear” that the reliance on analytics is keeping their brethren out of front office jobs. And that this market competition, that a “PhD in basketball” is no longer the lone requirement, is somehow a bad thing for the sport.
SB Nation’s Tom Ziller destroyed the hypothesis in his Wednesday edition of The Hook. His conclusion, after reviewing all of the new perhaps “suspect” front office hires, was the following: “Hybrids -- double-majors in Basketball and Math, not full-on quants -- are the real future of the NBA GM position.” Couldn’t agree more. If anyone still feels there’s a Moneyball-esque industry faction, then they’re missing out on the real value of analytics.
Sweet 16: March Madness continues tonight with four more games, including one with my beloved Dayton Flyers. FiveThirtyEight’s Roger Pielke Jr. wrote about forecasting accuracy as it relates to brackets. I know that I’m pretty terrible about bracket pools each year, despite (or perhaps because of) my math background.
The New York Times’ Billy Witz wrote about the analytics arm race going on now in college basketball. While more and more schools are using common tools like Ken Pomeroy’s website and Synergy, SportsVU presents a high-end next step. But only certain programs will be able to afford the very expensive military-inspired cameras. The rich could simply get richer.
NBA big men development: And sticking more with the basketball theme, GotBuckets’ Kevin Hetrick wrote about aging curves for big men. The charts were are fascinating. Hetrick concludes that big men peak early offensively (although with cumulative RAPM consistently below zero) and improve steadily defensively (while almost always having positive RAPM).
Seth Partnow updated his rim protection stats over at his Where Offense Happens blog. Using those numbers at ClipperBlog, he then offers up an update on DeAndre Jordan’s defensive value to the surging Clippers. The points saved per game metric is fascinating.
Jacob Rosen is a graduate of the University of Dayton, where he majored in applied mathematical economics and was the school newspaper's editor-in-chief and sports editor. Currently, Jacob (an Akron native) contributes to WaitingForNextYear, a website dedicated to Cleveland sports.