Friday, July 5, 2013

De-Lucker X: Buy Navarro, Sell Machado

The De-Lucker X needed a new home, and I've decided to give it a couch here at Cubs Stats. The following article was written for a broader audience, so pardon the limited Cubs-slant.

Players who began the season as starters are losing playing time; bench players are earning playing time; rookies are flopping and popping. It's the 2013 MLB season midpoint.

For a little over a year now, I have been crunching a sort of FIP for hitters -- the De-Lucker X -- that adjusts a player's wOBA based on their periphery stats. Now, behold!, the results from the first half are in! And if xBABIP can be fully, blindly trusted (and it can't), then it's time to bench Manny Machado before things get worse. Just kidding. I don't things will ever get that bad for Machado.

Why DLX?

DLX uses the core concepts behind the Should Hit formula — essentially:

0.09 + 1.74(HR%) + 0.39(BB%) – 0.26(K%) + 0.68(BABIP)

The De-Lucker part comes in when we plop an xBABIP in the place of yonder true BABIP. Jeff Zimmerman and Robert Boden (slash12) have been working on and promoting what I believe is the best xBABIP formula out there. So we take that formula, and we insert the xBABIP into the DLX formula, compare to the original wOBA, and then presto magico, we have a more discrete way to estimate BABIP changes on specific players' performances.

Here is the data:
De-Lucker X!
NOTE: You can download this spreadsheet -- and I encourage you do to so (with yonder green Excel icon) -- and you can meld that spreadsheet with this leaderboard to change the minimum PA level (currently at 100 PA) or update the numbers, should something horrible happen to me in the near future. (The present data is accurate through 7/4/13.)

DLX Analysis

It is important to note the DLX estimations tend to be a touch optimistic. As a whole, this above 100 PA group averages a 12-point improvement in production (a median 11-point improvement). I have studied this phenomenon for some time, and I think it sources largely from the descending league production:

I would use statistics from the future -- when presumably this offensive depression has settled -- but those are yet available. So I have built this regressive, backward-looking tool using historical numbers, and it is unable, therefore, to accurately anticipate the steadily-changing offensive environment.

It's a rough, frisky thing to do, but I threw in a little FIP-constant sort of adjustment -- a -0.012 in the DLX formula. Keep that in mind.

  • Sort the "Diff" column by Manny Machado has a .327 xBABIP despite his .373 BABIP. That would drop his .359 wOBA to a .300 wOBA/DLX. When I run these calculations, I always have a few players in mind with regards to who will improve, who will suffer, and who will stay steady in the heavy, burning eyes of xBABIP. Manny Machado was not one of those players. But upon further investigation, it makes sense.

    Machado never maintained a .300 BABIP or higher in the minors. Normally I don't give minor league BABIP much credit when looking for hints as to a player's true-talent-level BABIP. That's because most prospects and most MLB-quality hitters lathe out an easy .320+ BABIPs in the minors, and that MiLB BABIP has almost no correlation with their MLB BABIP. But in the reverse direction, I do give BABIP history in the minors a little credit. Rarely have I seen an MLB player maintain a BABIP higher than their MiLB norms.

    Machado's doubles-heavy, BABIP-heavy 125 wRC+ could sink -- if these xBABIP prognostications prove true -- to around a 104 wRC+. But Machado is a defensive prodigy at third, and he's only 20; his bat will improve, and so will his periphery stats. He is not at risk of losing playing time, but I think the DLX is right to suggest his bat will regress if his approach does not change or his home run power does not increase.
     
  • Let us inspect backup catcher Dioner Navarro. Navarro is of interest to me not only because of his affiliation to the Cubs and Rays, but also his affiliation to Joey Votto which has seemingly reinvigorated his once-kinda-promising bat. His present success with the Cubs (138 wRC+ entering play on Friday), is decidedly not BABIP-heavy. It is home-run-heavy. He is one trot away from matching his career high, despite 316 fewer plate appearances.

    Navarro was definitely one of the players about which I had expectations entering this study. I expected to see his wOBA decline. And indeed it may and probably will -- unless he can sustain this new, absurd ISO (a .252 ISO compared to his pre-2013 .111 ISO). This season, his steady powah and altogether uninspired defense have left him behind Welington Castillo in the catcher depth chart, but atop the Cubs pinch hitter / designated hitter list. And the slash12 xBABIP forecasts his BABIP improving a whopping 53 points.

    That would bring his wOBA from an already un-catcher-like, un-Navarro-like .377 to a Vottoian .430 wOBA (thereabouts of a 157 wRC+). This, again, is a tough assumption to make; it is an assumption that his LD rate will sustain and that he will set not only new career high in ISO, but also BABIP. Color me unconvinced, yet thoroughly entranced.
     
  • Yasiel Puig is single-handedly keeping my Scoresheet team -- the Chicago Orphans -- alive. But like Machado, Puig's numbers portend a drop in production. His wOBA should drop a heavy 73 points, leaving him with a mere .397 BABIP and .424 wOBA/DLX. If that feels a little crazy still, it probably is.

    Comparisons to Jeff Francoeur have drizzled across the baseball blogosphere lately. Both Puig and Frenchy started their careers with strikezone abandon and home run nirvana, so I threw Frenchy's 2005 season into the DLX (unadjusted). I got:

    .337 BABIP
    .372 wOBA
    .310 xBABIP
    .350 DLX

    So let this work as a warning of the limitations of both DLX and the limitations of assuming BABIP regression does the heavy lifting in a player's ups and down. Here, in the DLX world, Frenchy was still a great hitter. But pitchers learned to adapt to his aggressive zone; managers learned to never send lefties against him; the league adapted, the BABIP normalized, the power disappeared, and then -- years later -- Francoeur got DFA'd.
     
  • The Marlins have the worst offense in the league -- and it's not close. But, good news!, 6 out of 12 position players above 100 PA are poised for positive regression, some more than others, and another two hitters are in that funky 12-point error / league run-environment range. After/if the BABIP clouds clear, Giancarlo Stanton -- if'n he stays healthy -- will still be the team's top slugger, but surprisingly-productive infielder Derek Dietrich may not be far behind. Dietich, the bounty from the Yunel Escobar trade, made a sudden leap from Double-A to starting second baseman earlier this season. The Rays have to be happy with Escobar's production as well as their glut of second baseman, but this is a trade that may look unfavorable for the Rays in the very near future.
     
  • How about that Miguel Cabrera and Chris Davis? Those guys are nuts.

    There's a bagillion other players and storylines to study here in this data -- Norichika Aoki padding his potential trade value; Raul Ibanez earning an impossible xBABIP (seriously, it's impossible for him to reach that xBABIP at his age -- right?); Jean Segura straight blowing my mind; Derrick Robinson building an absurd, near-.400 xBABIP; Jose Iglesias's production being possibly somewhat legitimate; Josh Willingham, Aramis Ramirez, and Jacoby Ellsbury all poised for trips over the productive cliff (beyond just injuries). There is not enough Internet to cover all of the intrigue here.
  •  

No comments:

Post a Comment