The Phillies and Sabermetrics

A friend of mine works at a hot dog slash burger joint in Chester, PA. I stopped by one day to taste-test the goods, and he recommended a special sauce they have. So of course I tried it and liked it. When I caught up with him later, I asked him what was in that sauce.

“I have no idea,” he said.

Interesting that the business does not even let its employees know what is in the sauce, like the recipes for KFC or Coca Cola.

Why do I bring that up? Apparently, the Phillies don’t utilize statistics in player evaluation. From Doug Miller on MLB.com:

“I think defensive statistics are the most unpredictable stats there are,” says Charley Kerfeld, a former big league reliever who now serves as a special assistant to Phillies GM Ruben Amaro Jr.

“And since I’ve been here, we don’t have an in-house stats guy and I kind of feel we never will. We’re not a statistics-driven organization by any means.

“I’m not against statistics. Everybody has their own way of doing things. But the Phillies believe in what our scouts see and what our eyes tell us and what our people tell us.”

I’m skeptical as to Kerfeld’s honesty here, though he is certainly correct about the uncertainty of defensive metrics. For the same reasons that Coke and KFC and my friend’s burger joint don’t want us to know their secret ingredients, neither do the Phillies want everyone to know their thought processes in player evaluation. Such publicity would adversely affect them particularly in trades and drafting.

For instance, imagine Franklin Gutierrez — the slick-fielding center fielder for the Seattle Mariners — is on the trade market and the Phillies and the Boston Red Sox are the finalists in the sweepstakes. If the Red Sox know that the Phillies are proponents of defensive metrics, then they know that the Phillies may be willing to get into a bidding war for Gut’s services. So, the Red Sox wouldn’t jump out with an offer; instead, they would wait for the Phillies to do so.

Similarly, imagine the Phillies are trying to trade Jimmy Rollins. He is known more for his glove than his bat (sans 2007). The Phillies, by publicizing their use of defensive statistics, may alienate potential suitors because they realize that the Phillies could be valuing Rollins more highly than other teams and thus would not be willing to submit an offer or engage in a bidding war.

Simply put, teams do not want anyone to have access to their thought processes. If anything, Kerfeld is likely running a misdirection. That’s because the Phillies have consistently been among baseball’s elite defensive squads going back to 2002.

Note that the UZR data on FanGraphs only goes back to 2002.

It’s highly unlikely that the Phillies lucked their way into teams as consistently elite as their defensive squads have been. That’s not to say that, secretly, Phillies officials are poring over UZR and plus-minus data. Instead, the Phillies likely use a little of everything to various degrees. They probably do consider the word of scouts highest, then video scouting, and various sets of defensive data whether it’s UZR, +/-, PMR, or their own brand.

Another Sabermetric principle is that walks have value. The Phillies, since 2002:

From 2002-07, the Phillies were either first or second in the NL in drawing walks. They can thank Bobby Abreu for that, of course. But again, it’s highly unlikely they just happened to find walk-prone hitters like Abreu, Pat Burrell, and Ryan Howard on the team.

I’ll conclude this with perhaps the most damning bit of evidence that the Phillies are Sabermetrically-inclined: base running.

According to EQBRR on Baseball Prospectus, the Phillies ranked first in all of baseball in 2006 and ’07 (13.3 each), second in ’08 (13.4), and sixth in ’09 (0.5). Last season aside — likely caused by Jimmy Rollins’ down year which cut his base running opportunities significantly — the Phillies have been not only elite but once again consistently elite. It’s one thing to have a fluke season here and there but the Phillies are incredibly consistent.

Sabermetric studies have shown us that the “break-even” point in stolen base success rate is close to 75%. The Phillies’ success rate on the base paths will astound you:

  • 2004: 79%
  • 2005: 81%
  • 2006: 79%
  • 2007: 88%
  • 2008: 84%
  • 2009: 81%

Once again, not just elite, but consistently elite.

Of course we don’t know for sure, but I’d be shocked if the Phillies weren’t heavy proponents of Sabermetrics. They show all of the symptoms. You’ll know for sure if you ever see Amaro sporting a pocket protector or using a slide rule.

Leave a Reply

*

11 comments

  1. Basil Ganglia

    January 12, 2010 02:23 PM

    High on-base percentage and great defense were also traits of the Pat Gillick Mariners.

    There’s little evidence that Gillick had any regard for any advanced numerical data and evaluation during his time with the Mariners. To the contrary, he expressed his disdain on several occasions.

    But he arrived at much the same end by intelligently applying his traditionalists instincts and approaches.

    If you give a good artisan good tools with which he is comfortable, you will get a good product. If you give a crummy artisan the best tools, you will still get a bad product.

  2. Tom Forsaith

    January 12, 2010 02:33 PM

    Very well stated. I think that there may be a little hubris involved here as well. Great article.

  3. Adam S

    January 12, 2010 05:02 PM

    Your conclusion is a result of many fallacies. Just because they rank highly in UZR doesn’t necessarily mean they use any defensive metrics. They scout defense, and they sign guys who are good defensive players. They can pay attention to defense without using defensive metrics, especially since I have yet to see a defensive metric that is more than a discussion point (I think UZR is an awful stat, personally).
    The Phillies’ propensity to draw walks can also be explained. Abreu was traded for in 1997, long before sabermetrics existed. Pat Burrell was a #1 pick who was drafted because he could hit dingers. Jim Thome and later Ryan Howard manned 1B because they could hit dingers. Nobody expected Jayson Werth to even be a full time starter. Our leadoff hitter was 7th on the team in walks, with the only other starter with less being 2007 FA signee Pedro Feliz.
    As for baserunning, you point to Jimmy Rollins as a big reason for their elite baserunning, yet he was drafted in 1996, also long before sabermetrics. Abreu was a good baserunner who was there before sabermetrics. Shane Victorino is just really fast.
    In the end, there’s really absolutely no reason to think that the Phillies use sabermetrics other than some high and mighty belief that you can’t be successful without sabermetrics, which is begging the question to begin with.

  4. Argive

    January 12, 2010 07:10 PM

    Adam S-

    Actually, sabermetrics existed in the 1990’s; they did not magically flash into existence when Moneyball came out. You are confusing the point at which sabermetrics became popular with the point at which they were created. For one thing, Billy Beane began running the A’s from a quantitative standpoint in the late 1990’s. Hell, OPS has been around since 1984.

    Of course we don’t know whether the Phillies use sabermetrics, but if they rank high in certain metrics every year, then either they use them or they’re looking for the same things that stat guys look for. This is not a bad thing.

    As for “some high and mighty belief that you can’t be successful without sabermetrics,” it’s not really high and mighty anymore. Teams that use advanced metrics have experienced a great deal of success.

  5. Travis Bickle

    January 12, 2010 11:35 PM

    So a stat geek has decided that teams must be paying attention to his warped statistics (UZR is worse than warped). Even if the team is very successful and claims to have little interest in the “new math”. Can anyone say Self absorbed.

  6. hk

    January 13, 2010 07:15 AM

    I wouldn’t be surprised if the current Phillies front office ignores UZR and other advanced defense metrics and relies solely on their scouts. If they relied heavily on UZR, I don’t think they would have given Raul Ibanez 3 years at $10.5M per following his -23.3 and -10.4 UZR/150’s in 2007 and 2008, respectively.

  7. Richard

    January 13, 2010 08:57 AM

    As much as I would love it if the Phillies were sabermetrically inclined, I think you’re engaging in some wishful thinking. You seem to have cherry-picked three areas that appear to support the conclusion, but I think there are reasons to be dubious about them.

    A truly good defensive team will probably score well on whatever metrics are currently in vogue. I think UZR is interesting, but I’m not yet convinced of its accuracy. Also, hk makes a good point re: the Ibanez signing (though compared to Burrell, perhaps this is still an upgrade).

    Stolen base percentage is likely a function of coach Davey Lopes, plus some existing player-skill, perhaps in Rollins’ case. Lopes was always a good baserunner, and he actually did play “before” sabermetrics (unlike the way offbase timeline implied by Adam S.).

    And the walks stats actually argue against your conclusion, I think. Placing 5th and 7th in the NL in walks is not good evidence in favor. If they had maintained their walks edge even after losing Abreu, if leadoff man Rollins’ OBP weren’t under .300, then perhaps… other than that, I don’t see it. Also, sometimes teams *do* just happen to assemble a bunch of players who walk a lot. Does anyone think the 1993 Phillies were sabermetrically inclined?

  8. Bill Baer

    January 13, 2010 12:51 PM

    You seem to have cherry-picked three areas that appear to support the conclusion

    Can you think of other Sabermetric principles that I’ve left out?

    I think UZR is interesting, but I’m not yet convinced of its accuracy.

    If you trust wOBA, you should trust UZR. That’s not to say that one season’s worth of UZR is reliable, but neither is it for wOBA or whatever statistic du jour you select.

    Stolen base percentage is likely a function of coach Davey Lopes

    I’m actually doing a study on this and Lopes did in fact have an impact on this, but it doesn’t explain all of it.

    And the walks stats actually argue against your conclusion, I think. Placing 5th and 7th in the NL in walks is not good evidence in favor.

    You say this after accusing me of cherry-picking? The Phillies were top-two in walks six years straight!

    if leadoff man Rollins’ OBP weren’t under .300

    This is a function of the manager, not the front office. Sure, the GM could’ve sent orders down to have Rollins bat seventh, but that’s a good way to unnecessarily ruffle some feathers of both the manager and Rollins.

    Also, sometimes teams *do* just happen to assemble a bunch of players who walk a lot. Does anyone think the 1993 Phillies were sabermetrically inclined?

    To quote myself:

    It’s one thing to have a fluke season here and there but the Phillies are incredibly consistent.

    1993 was a fluke season. The 2002-09 Phillies are not.

  9. Bill Baer

    January 13, 2010 12:56 PM

    There’s little evidence that Gillick had any regard for any advanced numerical data and evaluation during his time with the Mariners.

    The discussion isn’t solely about general managers.

    If they relied heavily on UZR, I don’t think they would have given Raul Ibanez 3 years at $10.5M per following his -23.3 and -10.4 UZR/150’s in 2007 and 2008, respectively.

    Again, to quote myself:

    That’s not to say that, secretly, Phillies officials are poring over UZR and plus-minus data. Instead, the Phillies likely use a little of everything to various degrees. They probably do consider the word of scouts highest, then video scouting, and various sets of defensive data whether it’s UZR, +/-, PMR, or their own brand.

  10. Richard

    January 13, 2010 04:22 PM

    I’m sorry, I didn’t mean to sound like I was “accusing” you of anything, though I’m aware that using the words “cherry-picking” wasn’t helpful.

    You are more attuned to the latest metrics than I am, and I defer to you on that front. I also thought your point that the Phillies would rather others not know what they’re up to is valid. However, to modify some of what I said, I can see how these data points support your thesis, but I think it was you saying you’d be “shocked” to learn they weren’t sabermetrically inclined that threw me, as if there were no other explanations.

    As for the 1993 Phillies being a fluke, were they not the same team in 1992 (offensively)? I haven’t checked the numbers recently, but I seem to recall them drawing a lot of walks for a few years running.

  11. Bill Baer

    January 13, 2010 04:33 PM

    Richard,

    No offense taken. I am certainly not perfect and apologize if that’s the impression I gave. I tried to be very thorough so I was legitimately curious as to other metrics you thought I should have included.

    I welcome any and all criticism, so please, have at me. Doesn’t mean I can’t respond to it though. :)

    I think it was you saying you’d be “shocked” to learn they weren’t sabermetrically inclined that threw me, as if there were no other explanations.

    I think the disparity between what I’m saying and what others seem to interpret what I’ve said is a matter of degree.

    What is “Sabermetrically-inclined”? Is it using Sabermetrics at all? Is it mainly using Sabermetrics? Or somewhere in-between?

    I certainly don’t think the Phillies are anywhere close to the level of Sabermetric thought that, say, the Boston Red Sox or Tampa Bay Rays are at. However, I also don’t think the Phillies have zero Sabermetric thought.

    As for the 1993 Phillies, like you, I haven’t gone over the numbers. It could certainly be true that they were “just as Sabermetric” as the Phillies. And I could certainly be wrong about them being a fluke. A quick perusal of their walk totals reveals that they were in the top-half of the NL (12 teams until 1993) for several years, including tops in 1993. However, they certainly didn’t have the consistency that the current batch of Phillies have displayed.

Next ArticleBDD: Brothers in Baseball