The 2009 Yogi Berra Award

If you followed the blog I run for a couple of months before making it to The Show (i.e., The Hardball Times), you might remember a post about the bad ball swingers of 2008. Since Vladimir Guerrero is the poster boy of the third millennium greatest chasers, I gave the nickname “Vlad Guerrero Award” to the standings I produced.

Here I’m going to revisit that post, but I have decided to change the name of the award. Since awards are named after retired players, I went to Yogi Berra, because many sources report him as a batter who used to swing at anything within his reach, from his eyes—and higher—to his toes—and lower.

Before proceeding to the 2009 standings, let me remind you what I did a few months ago. First, I calculated the probability that a pitch is called a strike given its location. I calculated two sets of probabilities (one for right-handed batters and one for lefties), after standardizing the vertical component of the location taking the batter’s height into account.

Then I classified as a “bad ball” every pitch that had a probability lower than 10 percent of being called a strike. The cut point was arbitrarily chosen, so if anybody has a better option, the suggestion is welcome.

Note: Repeating all the work I did on 2008 data I discovered a few errors in my code, so I urge you not to look at last year’s result anymore.

Here is a graphical representation of the bad-ball zones.

The first table I will present shows the top 10 bad-ball chasers; that is, the 10 batters with the highest percentage of swings on bad balls.
Note: from now on I’m using 300 bad balls seen as the qualifier.

Table 1 - Swinging percentage on bad balls. Top 10.
          last    first pct bad balls
         Olivo   Miguel  39 574
      Guerrero Vladimir  36 559
       Soriano  Alfonso  36 877
     Rodriguez     Ivan  36 529
      Sandoval    Pablo  36 972
      Gonzalez     Alex  35 542
        Molina   Bengie  35 612
         Aybar    Erick  34 628
    Pierzynski     A.J.  34 635
        Cedeno    Ronny  33 417

Some usual suspects come on top.

A top 10 chart is often followed by a bottom 10 one, and I won’t make an exception here, especially since the guys at the bottom of this list have the great virtue of letting pitches unlikely to be called strikes go by.

Table 2 - Swinging percentage on bad balls. Bottom 10.
          last   first pct bad-balls
      Castillo    Luis  10 725
         Abreu   Bobby  10 901
       Ramirez   Manny  10 610
    Willingham    Josh   9 625
      Bautista    Jose   9 488
       Scutaro   Marco   9 827
      Iannetta   Chris   9 448
          Drew    J.D.   9 719
         Jones Chipper   9 820
          Cust    Jack   9 855

After what I wrote this summer on Marco Scutaro we should have expected him to finish down on this list.

Do players mantain their chasing tendencies? In the following chart I have plotted the players’ percentage of swings on bad balls in 2008 versus the same quantity in 2009.

It’s quite clear that chasing balls is a permanent trait: To confirm what’s already apparent in the above figure, the correlation between the plotted variables is 0.84 (with a 95 percent confidence interval of 0.79-0.87)

The first time I wrote about the subject, I introduced the net run value on bad balls to quantify the cost of swinging at those pitches.
For each bad ball I calculate the net run value as following:
{exp:list_maker}if the batter didn’t swing, assign the run value of the pitch (likely the run value of a ball; but if the ump called it a strike, then the run value of a strike);
if the batter swung, assign the run value of the outcome minus the expected run value of the pitch had the batter not swung (that is something like 90-percent-plus-something times the run value of a ball, plus 10-percent-minus-something times the run value of a strike). {/exp:list_maker}

So here’s what players have cost their teams by not refraining from swinging the bat.

Top 10, that is low cost for their team.

Table 3  Net run value on bad balls. Top 10.
         last   first net run value
       Suzuki  Ichiro -1.99
    Keppinger    Jeff -3.67
      Iwamura Akinori -3.97
     Castillo    Luis -4.17
         Ruiz  Carlos -4.28
        Gerut    Jody -4.37
        Lewis    Fred -4.52
       Millar   Kevin -4.59
      Paulino   Ronny -4.61
       Hinske    Eric -5.00

Bottom 10, i.e. high cost.

Table 4 - Net run value on bad balls. Bottom 10.
         last   first net run value
     Longoria    Evan -23.53
       Teahen    Mark -23.61
        Pence  Hunter -23.65
    Francoeur    Jeff -23.78
      Cabrera  Miguel -23.98
      Morneau  Justin -24.14
      Cuddyer Michael -24.88
      Peralta  Jhonny -25.29
      Soriano Alfonso -28.24
       Howard    Ryan -30.43

Some of the players in the second list more than make up for their discipline shortcomings thanks to their powerful lumber; sometimes they might be deliberately giving up the chance of a walk because they feel their power is more needed. Soriano (I didn’t want to point the finger at Francoeur, everybody’s whipping boy when it comes to plate discipline), on the other hand, is listed at Fangraphs at 8.6 batting runs below average: his season would have been less disatrous had his bad balls swing percentage been closer to the MLB mark of 20.4 percent.

What about pitchers? Do they take advantage of opponent hacking tendencies? You would expect that the less-disciplined hitters get the highest percentage of bad-balls.

The correlation between bad-ball swing percentage and bad-balls seen percentage (out of all pitches seen), while significantly different from zero, is not very high at 0.255. Anyway there’s at least one factor confounding the relation: the hitter’s proficiency. Manny is in the bottom list of bad-ball chasers, but that’s not a good reason for throwing him a lot hittable pitches.

Here are the top 10 pitchers at making opponent fish.

Table 5 - Swinging percentage induced on bad-balls. Top ten.
         last   first pct  bad balls
       Wuertz Michael  33  507
         Pena    Tony  30  352
       Madson    Ryan  30  369
    Gregerson    Luke  30  406
       Nathan     Joe  29  441
     DiFelice    Mark  29  302
       League Brandon  29  391
     Halladay     Roy  28 1067
       Rivera Mariano  28  342
      Buckner   Billy  28  432

And now the bottom 10.

Table 6 - Swinging percentage induced on bad balls. Bottom 10.
          last   first pct bad balls
     MacDougal    Mike  14 309
        Wilson   Brian  14 420
       Aardsma   David  14 370
         Cecil   Brett  14 474
        Miller  Andrew  13 438
          Hill    Rich  13 342
         Lewis  Jensen  13 334
       Cabrera  Daniel  13 371
    O'Sullivan    Sean  12 321
       Swarzak Anthony   8 306

The first list contains three notable cutterballers: Mo, The Doc, and… well.. DiFelice; the other seven have their top weapon in either a slider or a mid-90s fastball (or a combination of the two). A couple of flamethrowers (and a former one in Daniel Cabrera) appear on the second list as well, but for the most part you can not point out a dominant pitch in the repertoire of the pitchers among the bottom 10.

Hey, I was forgetting to actually give away the Yogi Berra Award. We could give it to Miguel Olivo, because it’s the batter with the highest percentage of swings on pitches way out of the zone. However, while Yogi used to throw his bat to any spherical object that happened to transit in the area where the game was played, his swings were successful more often than not, leaving opposing batteries wondering what they should have thrown him to avoid a double.
Thus Ichiro could be crowned for having totaled the highest net run value with his hacks.

But it doesn’t sound good to me to attach an award to a negative contribution, so I will simply end this article with the list of the players with the most hits on bad-balls.

Table 7 - Base hits on bad balls. Top 10.
         last   first bad-ball hits
    Sandoval    Pablo 68
    Phillips  Brandon 64
     Polanco  Placido 61
      Molina   Bengie 60
     Ramirez   Alexei 60
    Gonzalez     Alex 55
      Suzuki   Ichiro 54
        Cano Robinson 52
       Loney    James 51
       Aybar    Erick 51

The Award is yours, Panda!

References & Resources
PITCHf/x data from Sportvision & MLBAM.

I used batter handedness and the coordinates of the pitch only when crossing the plate to estimate the likelihood of a pitch to be decreed a strike. Pitcher handedness, balls and strikes count and pitch type also have some weight on the umpire’s decision. And the knowledge of the umpire tendencies may change the batter’s threshold on pulling the trigger.