Skip to content

Two good news articles on trends in baseball analytics

Mark Brown pointed me to two recent popular articles on sports analytics:

1. The New Science of Building Baseball Superstars: review by Jack Hamilton of “The MVP Machine: How Baseball’s New Nonconformists Are Using Data to Build Better Players,” by Ben Lindbergh and Travis Sawchick.

This all seems very reasonable to me.

2. If baseball is any indication, the big data revolution is over: by Christopher Phillips.

The title and subtitle of this article (“The data revolution has largely disappointed—and we shouldn’t be surprised”) are ridiculous. The article itself makes sense, and the article does not say the big data revolution is over—quite the opposite, the article states that there will be more data revolutions in the future.

In short, I recommend both articles. The only thing I don’t recommend are the misleading title and subtitle of the second article, which I’ll attribute to busy headline writers under pressure to grab your attention with counterintuitive claims.


  1. Jonathan says:

    I found the first article weird: the concept that being really good in baseball, like the difference between ordinary and star, is something you can tinker with strikes me as bordering on the absurd. Especially when an example is Mookie Betts, one of the most gifted athletes I’ve seen play baseball: a 2nd baseman dumped into Fenway’s vast right field (or center), playing it perfectly, while swinging away. If Mookie benefitted from something, it’s that he came along when being huge was declining as a desirable attribute, while lean and flexible were rising. That is, he’s not a big guy, but he swings big, and he benefitted from the Sox having seen Pedroia, an even smaller guy with a giant follow-through. There has been a recognition that leaner, smaller guys can generate a ton of bat speed. And play the field. Yes, that’s related to more fly balls and swing angles, etc., but it’s also related to the size of pitchers, which has increased, and their increasing reliance on fastballs, and other factors, but there is no single causal chain, just a bunch of sloppy chains that fit together in various ways.

    Other examples make sense. Relievers tend to have more limited pitches, so an improvement to their main or secondary pitch can matter. It may be easier with tech, but it’s hard to tell, given that in the past relievers would suddenly get better or suddenly get worse (and they still do). I concede that yes it helps, but not to the extent implied, which is that old-fashioned coaching is invalidated like it never occurred. I can picture a number of ‘reclamation’ projects over the decades of baseball. Am I to believe that the success percentage of these has increased significantly? I dont see it. I still see bullpens as unreliable year to year, etc.

    In general, if data is able to transform ability from middling MLB to star MLB, then why isnt that true generally? I’m wondering aloud: you teach at Columbia and your tools have improved dramatically, so have you seen that data attached to traditional ‘coaching’ has this transformative effect? You see stars and non-stars, those who thrive and those who barely make the Mendoza line.

    • Ben says:

      > In general, if data is able to transform ability from middling MLB to star MLB, then why isnt that true generally?

      Yeah, this whole thing sounds like it’s in search of the Great Secret of Baseball.

      As a scientific prospect, I’m very nervous reading this. Given this is a sport and an entertainment product, I really look forward to everyone on 2nd base wearing red dresses.

      > Edgertronic, TrackMan, Proteus, Rapsodo, motusTHROW

      I vaguely remember reading about Moneyball and there was some joke about not caring about what pitchers looked like in a pair of jeans (but rather how they performed).

      The philosophy here seems to be more about measuring formally how pitchers look in their jeans. Perhaps with enough measurement we’ll finally discern the relation between jeans-wearing and baseball!

      > the founder of an expansive training empire called Driveline

      It sounds like there is vastly too much marketing in this field. I searched for “practice” in the article and didn’t find anything. That seems like an omission if we’re going to be talking about player improvement (there are other mechanisms of playing better baseball than steroids).

      > Lindbergh and Sawchik seem unsure how to reconcile Bauer’s visionary approach to professional improvement with his relentlessly appalling personal behavior.

      What is the reconciliation we’re seeking here? Why does it matter? If he’s just an asshole, maybe just talk about someone else?

      But skepticism aside, yeah, cool. Measure some things, make some plots, etc. etc.!

  2. Rahul says:

    I am waiting for the whole hot hands meme to die down….


    Money ball is based on old baseball.On base percentages.Great player get on base to score runs


Leave a Reply