Marlon Byrd, Mike Moustakas De-Luck’d

A refreshed look at the data.

Specificity is both delightful and dangerous. The guys at The Book Blog have previously remarked about how UZR and WAR would be better for mass consumption without the decimal because neither stat can show a true talent level within a single season, but the extra decimal can make it appear more certain or accurate than it is. At the same time, though, the difference between 1.0 and 1.9 WAR can be the difference of a starting job or a bench role (or the difference between 1.6 and 2.4, if rounding is your thing).

Well, today we will err on the side of specificity. In the past, when a player’s BABIP was .498 or .93278, we would just say, “Well, he will regress to the mean,” and then resume our toiling lives. Now, with Fielding Independent wOBA, we can whip their numbers into shape, we can thrust them into the De-Lucker and find out where a regressed BABIP will take them — which is good news for Marlon Byrd, but bad news for Mike Moustakas.

Let’s examine it.

Combining Fielding Independent wOBA (FI wOBA for shortsies) and slash12’s xBABIP, we can get a specific wOBA calculation that strips away unusual luck, whether good or bad. It is important to remember these are both regression-based calculators, so they are backward-looking, not forward looking. Please do not pester me in 3 months when “[X Player] didn’t suck like you said he would! lulzlulzlulz, ur dum!”

As in finance and in life, past performance is not necessarily an indication of future success (or failure), so these numbers simply tell us what their wOBA would have been given the BABIP that probably should have had.

Is BABIP the only stat that has a lot of luck in it? No. Stuff like home run rates can be wild early too. And moreover, BABIP is many parts skill, several parts luck. But there is more luck (or random variation) in BABIP than probably any other hitter stat out there. That is why it is worth focusing on it here.

No more ado; here’s this:

The De-Lucker!


Some assorted reactions to the De-Lucker results:

    • Mike Moustakas and Marlon Byrd both appear to be at clearly opposite ends of the regression spectrum. When Byrd’s BABIP normalizes, he will below average, but acceptable assuming his home run rate normalizes too. Meanwhile, Moustakas is riding high on a .337 BABIP, and if that comes down to match his xBABIP (a surprising .244), then he will be miserable to watch. Of course, I suspect that when his BABIP comes down, his approach will change, and his xBABIP will settle closer to a more normal .300, but for now, his success appears the least sustainable.
    • Kansas City fans concerned by the potential drop-off from Moustakas: Take heart! Eric Hosmer‘s cold start appears to be entirely BABIP-based. His peripherals are in line for a solid season.
    • White Sox fans hoping to see Brent Morel‘s production go up with his BABIP might want to look away. Even if he does hit his whopping .345 xBABIP, he will still be a massive disappointment with the bat (.257 FI wOBA).
    • Brewer-man Jonathan Lucroy is another player poised for a hard crash. His xBABIP (.311) isn’t terribly low, but his production is too singles-heavy, and therefore BABIP contingent, so when his BABIP normalizes, his production will likely tank.
    Bryan LaHair, Paul Konerko, David Wright, and David Ortiz are all in the “Regression, so what?” Club. Each of them should still have wOBAs above or at .400 if their BABIP normalizes, and (big if) if they can maintain or improve their walk, strikeout, homer, and stolen base rates.
    Josh Hamilton has a BABIP of .407 and an xBABIP of .370. His FI wOBA is .539, but even if you put his BABIP at .300, he still has a .495 wOBA. The man is using the turbo button for sure.

NOTE: I went ahead and made it for a minimum of 100 plate appearances. But let’s be honest, 100 is just as arbitrary as 50 or 72 or 0.1. If you have a hankering to see this applied to a different grouping of players — or an updated grouping of players (these numbers are current through Tuesday), then simply do the following:

Thing the first. Go to the leaderboards section and do a custom leaderboard like so:

This is just an image. Don’t freak out and click it.

Thing the second. Scroll back up the page and download Excel file:

Encircled and enarrow’d.

Thing the third. And then combine those two spreadsheets by your copy-paste method of choice. I recommend browsing over to the Excel tab entitled “FanGraphs Leaderboard (15)” before pasting, but as long as you unhide everything on the other tab, er’thing will be fine.


Print This Post

Bradley writes for FanGraphs and The Hardball Times. Follow him on Twitter @BradleyWoodrum.

18 Responses to “Marlon Byrd, Mike Moustakas De-Luck’d”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Lg4ny says:

    This is pretty amazing. Great work corralling some important numbers in a clean format to produce a better look at regression. Love this place.

    Vote -1 Vote +1

  2. sprot says:

    I think he just had an aneurysm, someone call a doctor.

    Vote -1 Vote +1

  3. guesswork says:

    LaHair, Konerko, Wright, and Ortiz have a collective stolen base rate of 36% (4 out of 11). Even for those guys, I think they can maintain or improve that rate.

    Vote -1 Vote +1

  4. Relddem says:

    I wouldn’t be so confident Moustakas’s xBABIP will normalize. It may not be .244 low, but it might also eventually settle in closer to .270 than .300. His career minor league BABIP was .295, which doesn’t seem that odd out of context, but minor league BABIPs (especially for bit-bat prospects) tend to be much higher than in the majors. Poor infield conditions and weaker defenses yield more hits on weaker contact. Hosmer’s career MiLB BABIP was .351. So is Anthony Rizzo’s. Moustakas may just be one of those moderately high-strikeout, fly ball hitters who yields a lower than league average BABIP. Maybe Edwin Encarnacionish.

    Most of this is at least semi-anecdotal based on experience examining minor league players, but in general it would be interesting to see a study on minor league BABIP and how it translates to the majors.

    Vote -1 Vote +1

    • Good insights. I think you’re on the right track.

      There’s always a chance, though, that Moustakas will start walking more as his BABIP goes down, but there’s certainly no guarantee of that.

      Vote -1 Vote +1

      • TFINY says:

        Also something to consider, as per Tango, league average BABIP is .283 this year, and not .300. Something to consider, especially when looking at how many players looked to have had “bad luck” in the chart above.

        Vote -1 Vote +1

  5. Jim Lahey says:

    Looks pretty much like the buy low / sell high list of fantasy baseball. Nice.

    Vote -1 Vote +1

  6. Mel says:

    There has been some discussion about the shifts being used on Hosmer and how his BABIP is beng suppressed as a result. Would the use of a shift in theory limit his upside even if his luck begins to turn?

    Vote -1 Vote +1

    • Tom Rigid says:

      Eventually there will be a UZR extraction to crossref with BABIP; until then, Hoss just gotta chow some Charlie Lau-Lau.

      Vote -1 Vote +1

    • Paul says:

      The shift works especially well when you swing at absolutely everything and completely ignore situations. Call it luck if you want. For the most part from what I’ve seen Hosmer gets himself out most of the time. Interestingly, Moustakas has the lower BB rate and has a reputation as a hacker. But his ABs starting the last month of last year, and now especially this year, have been outstanding.

      While it would be a smart play to sell high on Moustakas, the smarter play would be to feign being the knife catcher, because there is zero chance his BABIP gets close to what this parody of analysis predicts.

      Vote -1 Vote +1

    • AJ says:

      Contributor Jeff Zimmerman has looked into Hosmer’s numbers versus the shift. It shows an effect on groundballs because he is a dead pull hitter:

      I’m not taking much heart entirely because more teams are going to use the shift and he has shown little ability to go the other way.

      Vote -1 Vote +1

  7. Tom Rigid says:

    Did you ever think that you might get too close to the essential volatility of this universe and simply go mad? I suppose it’s sample-size discipline which keeps us from darting to every particle of data until our eyes are spinning like marbles in a rock-polisher. One complaint though: You didn’t nerdulate the geekometer.

    Vote -1 Vote +1

  8. jcxy says:

    I’ve read that one of the primary issues with using the xBABIP calculator is it’s reliance on outcome data besides GBs…which seems easy to determine vis a vis LD/FB data (which is more subjective).

    1. Is this a reasonable concern?
    2. How are LD/FB data determined currently on FG?

    Vote -1 Vote +1

  9. Nick says:

    Could we also create a stat such as this that takes into account HR/FB%, like FIP vs xFIP for pitchers? Not all players are created equal in terms of HR/FB so using a player’s career average might be more appropriate than using the league average as in the case of xFIP. For example, we wouldn’t and shouldn’t expect a 41% HR/FB rate from Josh Hamilton (but we also shouldn’t expect league average) so his xFI wOBA would be significantly lower than his FI wOBA. In the case of Melky Cabrera on the other hand, a 6.3% is probably about right for a him, so his FI and xFI wOBA’s would be a lot closer together.

    Vote -1 Vote +1

  10. Chomp says:

    Wait…where are Chase Utely’s numbers??

    Vote -1 Vote +1

  11. Kevin says:

    “the Excel tab entitled “FanGraphs Leaderboard (15)”

    hahaha thumbs up. I’m glad I’m not the only one

    Vote -1 Vote +1

  12. thalooch says:

    um, so should I be trying to sell high on Moose from what you’re saying?

    Vote -1 Vote +1

    • Skob says:

      Mous went into a 2 week slump after this post, so… well done Mr. Woodrum, but more to the point, Mous is a bit more leveled off than he was before. You wanted to sell high when his average was .300, not now because he is a .275 hitter anyway. The power is real and the kid is good, so if you like having him today, keep him!

      Vote -1 Vote +1