Jump to content
  • Welcome to AngelsWin.com

    AngelsWin.com - THE Internet Home for Angels fans! Unraveling Angels Baseball ... One Thread at a Time.

    Register today to join the most interactive online Angels community on the net!

    Once you're a member you'll see less advertisements. Become a Premium member and you won't see any ads! 

     

IGNORED

OC Register: Hoornstra: Reckoning with the powers and limits of analytics in baseball


AngelsWin.com

Recommended Posts

It’s that time of year when baseball fans go to war over WAR.

In the National League, Mookie Betts and Ronald Acuña Jr. are essentially equals this season, according to Baseball Reference and FanGraphs, with 8 Wins Above Replacement apiece. The digit to the right of their respective decimal points could spell the difference between who wins the Most Valuable Player award.

I hold one of the 30 NL MVP votes this year, so I was eager to talk to a recently freed front-office quant about what the parameters of the war (and WAR) should – and should not – include. Scott Powers was the Dodgers’ director of quantitative analysis from 2017-21 and the Houston Astros’ assistant general manager in 2022. His teams have appeared in five of the last six World Series, winning twice (with the Dodgers in 2020 and the Astros last year).

What I learned was more powerful than the mere strengths and weaknesses of WAR.

What I learned was, when it comes to analytics in sports, it’s OK to ask: what is it good for?

Let’s start with the fact that Powers isn’t working in the baseball industry anymore. When we spoke by phone earlier this month, he was focused intently on women’s college volleyball. And to hear him tell it, that was a breath of professional fresh air.

“Something that I’ve really been struggling with, with volleyball, is trying to understand what place does my work have?” said Powers, who began teaching in Rice University’s Department of Sports Management in August. “I bring that up because in baseball there are a lot of consumers for player evaluation and projection. Major League Baseball teams have to make trades, they have to sign free agents, and so there’s a lot of money invested in terms of knowing how good these players are. Whereas in volleyball, the highest-revenue volleyball league in the United States is Division I women’s volleyball.

“Now you’re talking about colleges that, sure, there’s a transfer portal and NIL money but for the most part you’ve got the athletes that you’ve got. Player evaluation is suddenly less important than player development. In baseball, I do think the use of analytics has been very one-sided. Teams have benefitted from it more than players have.”

Unbound by the confines of pro sports, the first thing Powers chose to do with his talents was use analytics for the benefit of students who will spend little to none of their professional lives as athletes. (Not coincidentally, he considers volleyball his favorite sport.)

The oral presentation Powers prepared for the recent New England Symposium on Statistics in Sports – titled “Estimating individual contributions to team success in women’s college volleyball” – is a direct reflection of something front-office analysts do sparingly: help players improve their game.

“In volleyball, I want to be careful about the work I’m doing, not having it be one-sided for the team,” he said. “I want to have a positive impact on the sport. We’re talking about 18- to 22-year-old women. Trying to find ways to focus the research less on player evaluation than on in-game strategy.”

Powers’ moral reckoning is not unique. I’ve spoken to other front office expats whose post-MLB work has veered 180 degrees from anything that could be categorized as “salary suppression.”

Powers sees “Moneyball” as a playbook for how major league teams can (and did) win on the cheap. Michael Lewis’ seminal 2002 novel is also a dominant inspiration for the students who populate his classes, those who aspire to populate front offices in the future. Powers’ challenge to himself, and by extension to his students, is one and the same: to view analytics as a force for good, one that can empower athletes rather than suppress their future earnings.

If you can grasp that dynamic of tension within the sports analytics world, you can understand a lot about baseball today.

Think about the last time you watched the television broadcast of a game and heard a former player in the analyst’s seat bemoan the expansion of analytics departments. There’s a decent chance that player has first-hand experience with the salary suppression “Moneyball” wrought on the game. That might even be the extent to which he’s contemplated the growth of quantitative decision-making in baseball. What Powers is saying, in effect, is he understands where those former players are coming from.

There are also fans who will look at Betts’ lead in both versions of WAR – by tenths of a point through Tuesday – and experience cardiac arrest if Acuña’s historic home run/stolen base total carries him to the MVP award.

But Powers understands the limits of the publicly available versions of WAR. A tenth of a point is basically a coin flip. Why?

“The concept of a ‘freely available player’ is super messy,” he said. “There’s this pool of players, but trying to estimate who are the best ones who are freely available is a very messy process. The replacement-level definition by FanGraphs, they sort of sidestep that. Rather than getting at what replacement level is, they choose this round number which feels about right.”

Both FanGraphs and Baseball Reference calculate “replacement level” (based on the population of “freely available players”) as 1,000 wins per 2,430 games (the number of games played by 30 teams in a 162-game season). A team full of replacement-level players would, by definition, win 48 games over a full season.

“By no means do I want to impugn the work that’s been done by these folks,” Powers said. “The choice they made is very reasonable. I can’t argue against it or argue for something better. With the project I was working on recently, I was trying to do exactly that: what would be a more principled way to define replacement level if you don’t have access to the information that teams have access to about waiver-wire pickups and the population of freely available players? I didn’t really come up with a good answer.”

Player WAR evaluations, he added, are “very sensitive to the range of value that could be considered reasonable definitions for replacement level.”

The takeaway here is not that one of MLB’s most accomplished quantitative analysts has rejected analytics. After all, his current job title is “Assistant Professor, Sport Analytics.” The point is, it’s OK every now and then to ask “what are we doing here?”

Are fans really using two-tenths of a win to justify their pick in the MVP race? Are front offices using a similar differential in WAR to rationalize the decision not to incur a luxury tax when pursuing a trade or free-agent target? If that behavior feels intellectually icky, well, maybe it is. Neither should we shove the entire thrust of analytics research under the team bus.

As for my MVP ballot, it’s feeling like a coin flip.

View the full article

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...