clock menu more-arrow no yes

Filed under:

Bases Per Plate Appearance

New, 6 comments

It's been a statistical kind of week for me.

First I spend Friday learning about the deficiencies of Prospectus' Fielding Runs Above Replacement, and the better path shown by the Probabilistic Model of Range, now, I spend (some of) my Sunday looking at BPA.

Although Ray Kerby's Astros Statistical Software remains my statistical tool of first resort, the fact that its data files only cover the Houston team leaves a niche for Lee Sinins' Complete Baseball Encyclopedia.

It's very Astro-centric around here, and most times I don't care which Phillies OPSed over 1.000 in their rookie year, or whatever.  But from time to time I wanna check out the statistical doings of major league players outside the great city of Houston, and although I'm not happy about the price, and think that maybe its interface is a little too complicated for its own good, Mr. Sinins' Encyclopedia otherwise fits the bill.

The preceding has been my own long-winded way of introducing to you the fact that I received my 2006 copy of the Encyclopedia last week, and that I went ahead and installed today.  

And when I did so, I noticed a stat included in the interface that I never noticed all last year:  BPA, or Bases per Plate Appearance.  The formula looks complex,

but the concept is simple, like slugging percentage if slugging counted walks and hit by pitches and stolen bases, or like OBP if OBP cared when you walked and stole second (not that anyone on the Astros does that).

BPA seems a simple concept that also seems to convey a good deal of information, and by now I'm saying to myself, why didn't I think of that?

But having spent half of my previous post griping about accountability in quoting stats, my next step was to see whether the numbers churned out by Mr. Sinins' Encyclopedia in this BPA stat made any kind of sense. So I took a look. Here are 2006's major league leaders in BPA, minimum 450 plate appearances:

       Player                        BPA    
 1    Travis Hafner                 .703   
 2    Ryan Howard                   .700   
 3    Albert Pujols                 .694   
 4    David Ortiz                   .681   
 5    Carlos Beltran                .667   
 6    Jim Thome                     .659   
 7    Lance Berkman                 .658   
 8    Manny Ramirez                 .654   
 9    Barry Bonds                   .647   
10    Jermaine Dye                  .637  
Well, if I do say so myself, there does appear to be a good degree of correlation between a high BPA and your having had a kick ass season hitting the baseball. Hafner would have been American League MVP if the writers hadn't been derelict of their duty, and (although I definitely feel Pujols was the right choice) Howard/Pujols were 1-2 on anyone's list for the National League. All further runners-up (including our own Lance) had themselves very creditable seasons, as well.

Do you see that one of the things that's nice about BPA vs. OPS is that it's not double counting batting average?

So anyway, at that point having concluded that this BPA stuff might be worth looking at, my thoughts turned, as they so often do, to the Astros.

Here's your 2006 Astros, minimum 200 plate appearances, so we could see Luke:

                                          BPA        PA     
  1    Luke Scott                    .667      249   
  2    Lance Berkman                 .658      646   
  3    Morgan Ensberg                .562      495   
  4    Aubrey Huff                   .510      261   
  5    Mike Lamb                     .485      421   
  6    Chris Burke                   .484      413   
  7    Jason Lane                    .458      345   
  8    Craig Biggio                  .442      607   
  9    Willy Taveras                 .420      587   
 10   Preston Wilson                 .403      417   
 11   Adam Everett                   .390      566   
  12   Brad Ausmus                    .318      502   
And here (although I bet you couldn't see it coming) is where this piece becomes another Better Let Aubrey Walk And Keep Ensberg post.

This BPA business is showing me that it wasn't just Morgan's defense that we've been underrating. What the table above shows pretty clearly is that even in a year where he kind of sucked, Morgan still brings more to the table offensively than the ex-Devil Ray who was imported to try and supplant him. Aubrey hit .250 and slugged .478, while Morgan hit .235 and slugged .463, yes, and the difference is the walks.

Everyone (including myself) was saying forget the walks, Ensberg needs to show power, but when all of it was said and done, Ensberg had provided almost 19 homers worth of bases over what Aubrey did. 75 extra bases--101 to 26--through the walk. Last year's team needed an RBI guy after Lance. And Morgan wasn't it after April. But we shouldn't have taken that to mean he didn't make contributions.

Now, I also don't want to let Morgan skate for having lied to the team. And I don't want to sound as if I don't understand the deep offensive doldrum he went into May and June where even the walks stopped coming so quickly. I don't want to gloss over that. But the point is, even in categories where it appears Huff has the advantage, he doesn't.

Tim Purpura, if you're reading this, you need to install Morgan Ensberg as your 2007 third baseman.

(Some more BPA numbers after the fold)

Ensberg's "crappy" 2006 season is actually 34th in BPA in franchise history among the 254 qualifiers with a minimum 450 plate appearances. Here's your Franchise Top Ten:

       Year Player                    BPA   PA 
   1. 1994 Jeff Bagwell            .768  479 
   2. 1997 Jeff Bagwell            .682  717 
   3. 1999 Jeff Bagwell            .676  729 
   4. 2000 Richard Hidalgo         .661  644 
   5. 2001 Lance Berkman           .658  688 
   6. 2006 Lance Berkman           .658  646 
   7. 2000 Jeff Bagwell            .652  719 
   8. 1996 Jeff Bagwell            .651  719 
   9. 1999 Carl Everett            .640  535 
 10. 2004 Lance Berkman           .636  687
Sinins' software doesn't normalize the standard statistics, but since Ray Kerby's Engine does, and since I had the formula, I was able to go back into ASS and plug it in. Morgan drops to 70th of 254 normalized, and the top ten looked like this:
       Year Player                    BPA   PA 
   1. 1994 Jeff Bagwell            .724  468 
   2. 1972 Cesar Cedeno            .679  617 
   3. 1969 Jim Wynn                .678  652 
   4. 1973 Cesar Cedeno            .657  579 
   5. 1968 Jim Wynn                .640  664 
   6. 1997 Jeff Bagwell            .639  699 
   7. 1965 Jim Wynn                .635  668 
   8. 1967 Jim Wynn                .626  682 
   9. 1996 Jeff Bagwell            .617  710 
 10. 2001 Lance Berkman           .608  674

You may recall that Wynn's 1969 season was also the one where he reached base in 52 consecutive games, and one thing I implore you not to do, is do not mess with Jimmy Wynn in 1969. He was the absolute shit.

As an afternote, I feel as if I need to apologize for my long-windedness. I aways get the feeling that a better writer could convey the same info in half the space. Please understand, I do my best. . . .