clock menu more-arrow no yes mobile

Filed under:

Jerry Dipoto and his Love of the High Minors

What does a projection system tell us about Jerry Dipoto’s preferences?

MLB: New York Mets at Arizona Diamondbacks
As if we needed another reason to join the Mitch Haniger Fan Club...
Mark J. Rebilas-USA TODAY Sports

When he first took over the Mariners in September of 2015, Jerry Dipoto was left with quite a dilemma on his hands. It was his task to figure out how to improve the M’s despite a barren farm system.

Sure, there were a couple guys poised to make contributions, such as Edwin Diaz. But Diaz is actually the only player from the BP Top 10 M’s prospects released in December 2015 to have played in the majors. Given that, it’s really no surprise that Baseball Prospectus said “there’s as little talent in the upper minors as in any system.”

Amidst all of Jerry’s wheeling and dealing, one unquestionable hallmark has been the constant churn of players, all in the name of depth. But it’s also interesting to look at the kinds of players he’s targeted, especially this offseason, and that starts with guys in the high-minors.

First, I’m going to backtrack a bit. Have you ever heard of KATOH? Those of you who are especially sabermetrically-inclined probably have, but it’s still a bit of an unknown to many. KATOH was developed by Fangraphs’ Chris Mitchell with the aim of using minor league statistics to predict major league success. If you’re especially intrigued, go check out this article where Chris explains it in detail; for those of you who don’t have time to read 2,000+ words, I’ll do my best to be succinct. His methodology is roughly as follows:

  1. Use regressions on players through 2000 to determine which factors are most predictive of major league performance - think AAA strikeout rate, AA strikeout rate, homers in A+ ball, etc.
  2. Include data from defensive positioning and performances, as well as a second KATOH methodology that includes Baseball America Top 100 lists to add a human layer.
  3. Regress statistics based on luck and sample size - for example, a batting average of .500 means a lot less in two plate appearances than an average of .400 in 100 plate appearances.
  4. Generate probabilities that a given prospect will get to certain benchmarks, such as playing in the majors, earning 5+ WAR, etc.
  5. Then, weigh each probability and calculate an “expected WAR value” for each player.

It seems to make sense, right? Use statistics that have been predictive and calculate the probability of a bunch of outcomes, then average ‘em together.

Well, interestingly enough, Jerry Dipoto might think so too. Take a look at the Midseason KATOH Top 100 prospects, released last July.

KATOH Top-100 Mariners

Rank Player Position KATOH Acquired?
Rank Player Position KATOH Acquired?
23 Tyler O'Neill OF 7.0 Drafted
25 Mitch Haniger OF 6.9 Traded 11/16
57 Boog Powell OF 5.0 Traded 11/15
60 Dan Vogelbach 1B 5.0 Traded 7/16
64 Ben Gamel OF 4.9 Traded 8/16
98 Max Povse RHP 3.9 Traded 11/16

Huh. That’s interesting. Of the six Mariners in the top 100, five of them were acquired by Jerry Dipoto, and four of them were either midseason deals or from this offseason. Meanwhile, the M’s didn’t trade away a single guy on there.

I don’t pretend to know what this means, exactly. I doubt that Jerry was waiting for the newest KATOH list to be released, and once it was out, he adjusted his entire strategy on it. Rather, my guess is that his player acquisition strategy is similar to the things that KATOH values - stats like K% and players with proven track records at higher levels of the minor leagues.

That’s why he’s been willing to give up players like Luiz Gohara and Ryan Yarbrough, whom KATOH projects for 3.1 and 1.1 WAR, respectively. That’s why he could trade former top prospect Alex Jackson (projected for 0.4 WAR) for Povse (4.4) and Rob Whalen (3.3) - sure he might lose a high-ceiling prospect, but given the extremely low chance that he’d ever get there, better to gamble on a pair of decent arms who have a much higher chance of succeeding.

It’s also interesting to note that earlier this offseason, Dipoto said they’ve been “trading major league players for major league players or prospects for prospects.” This has certainly held up (with the exception of the Drew Smyly deal), but those prospects he’s been getting are much closer to the Whalen or Povse type, players available much sooner, than the Alex Jackson type.

This could mean a lot, or nothing. I think it’s just another way to conceptualize Dipoto’s focus on high-minors players (who will obviously do much better in KATOH than guys in A-ball). Part of that focus is out of necessity, of course - it’s hard to acquire highly touted prospects when there’s not much to deal from, so you probably have to target players whose top outcomes aren’t stratospheric. And when you’re trying to win now, acquiring Mitch Haniger and Daniel Vogelbach (and, yes, even Ben Gamel) makes obvious sense.

So let’s return to this article in a couple years, when Dipoto & Co. have had the chance to truly remake their system based on what they like. If they remain KATOH darlings, then perhaps we’re onto something. Otherwise, this is just proof that they’re willing to adapt their strategy to fit the organization - and after years of power-at-all-costs in the Jack Z regime, that’s something we can all get behind.

Thanks to John Trupin for the article idea and to Chris Mitchell for creating KATOH in the first place.