Tuesday, April 6, 2010

The Luck Police


April 6, 2010

By Bret Sayre

There are few things that spark more of an argument in baseball these days than the proper (or improper) use of advanced statistics.  You’ve read the articles and you’ve seen the arguments.  This is no place for such arguments – if you don’t at least incorporate the most basic of advanced statistics into fantasy baseball, you’re operating at a disadvantage.  Personally, I’m into the advanced stats, but I’m a math guy.  Always have been.  The question is – which of these stats can really help you win your league?  Fortunately, I have your answer.

Batting Average on Balls In Play (BABIP) and Batting Average on Balls In Play against (BABIPa) are two of my favorite stats to help predict future performance in the most difficult of fantasy categories.  Let’s face it, ERA and Batting Average can be a real crapshoot from year-to-year even with the most predictable of players.  So when a player is hitting .300 in mid-May and is a career .260 hitter, what can we look at in order to guess whether he’ll come back down to Earth or if he’s made improvements to his approach?  Has he cut down his strikeout rate?  Has he improved his walk rate?  Do we think this is just a luck-driven small sample size?

The difference between Fantasy Beards and other sites in this regard is that I don’t believe that everyone’s BABIP or BABIPa regress to a universal mean.  There are inherently players who make “better contact” than others.  There are pitchers who leave too many balls up in the zone.  That’s the beautiful thing about baseball – it takes all types.  So what we do here is compare a player’s current BABIP/BABIPa to their own personal history, using 5-year and 3-year rolling averages.  If I told you that Prince Fielder and Jason Bay both had .315 BABIP’s in 2009 – would you think that they would both slightly regress to the MLB mean of approximately .300?  Maybe, but if you look at their 5-yr rolling BABIP’s (Fielder - .300 and Bay - .322), you’d think a little differently.  

Same deal with pitchers.  Cole Hamels had a .317 BABIPa in 2009 and Matt Cain’s was .263 – pretty straight forward case of Hamels being unlucky and Cain being lucky, right?  Well, you’d be right about Hamels who has a 5-year BABIPa of .284 (and this is a big reason why a lot of people expect him to bounce back from his comparatively poor 2009 season).  But Cain, not so much.  As the immortal Denny Green might say, he was who we thought he was (his 5-year BABIPa is .269).

Now just like a lot of other stats, BABIP and BABIPa are not completely cut and dried.  For pitchers, defensive statistics can play a role in determining whether a lower than historical BABIPa can be replicated.  My calculations show that Felix Hernandez and Ryan Rowland-Smith are both candidates for regression because of their lower than historical BABIPa.  However, I believe that in their case, the majority of that can be attributed to the great improvements Seattle made as a defensive ball club – so I’m not concerned there.  For hitters, I tend to not be concerned about younger players who appear to have “figured it out” – although this is a much more subjective area.  Miguel Montero outperformed in 2009 as compared to his historical BABIP, but he also got more at-bats in 2009 than he had in his entire career up to that point.  Plus, I personally think he’s improved as a hitter (take that, objective analysis!)

On Tuesdays, I will be comparing players’ current year BABIP and BABIPa with their historical data to try and give some extra insight as to whether that .300 hitter will help your title run or if you should probably try to trade him before he implodes.  Of course, we’ll be starting this in May since anything before that would just be pointless due to small sample size, but until then, I’ll leave you with my thoughts on a few guys who deviated from their performance notably last season.

Pitchers to target

Kevin Slowey – 2009 BABIPa of .345, 5-yr BABIPa of .307
Matt Capps – 2009 BABIPa of .360, 5-yr BABIPa of .291
Derek Lowe – 2009 BABIPa of .326, 5-yr BABIPa of .292
Cole Hamels – 2009 BABIPa of .317, 5-yr BABIPa of .284
Ricky Nolasco – 2009 BABIPa of .316, 5-yr BABIPa of .297

Pitchers to Avoid

Randy Wolf – 2009 BABIPa of .251, 5yr BABIPa of .290
Leo Nunez – 2009 BABIPa of .242, 5yr BABIPa of .288
David Aardsma – 2009 BABIPa of .254, 5-yr BABIPa of .289
Edwin Jackson – 2009 BABIPa of .275, 5-yr BABIPa of .305
Huston Street – 2009 BABIPa of .240, 5-yr BABIPa of .265

Hitters to Target

Kelly Johnson – 2009 BABIP of .212, 5-yr BABIP of .290
Geovany Soto – 2009 BABIP of .246, 5-yr BABIP of .305
Ian Kinsler – 2009 BABIP of .239, 5-yr BABIP of .285
Chipper Jones – 2009 BABIP of .287, 5-yr BABIP of .332
Grady Sizemore – 2009 BABIP of .274, 5-yr BABIP of .316

Players to Avoid

Felipe Lopez – 2009 BABIP of .357, 5-yr BABIP of .321
Jason Bartlett – 2009 BABIP of .360, 5-yr BABIP of .325
Michael Bourn – 2009 BABIP of .362, 5-yr BABIP of .327
Rajai Davis – 2009 BABIP of .359, 5-yr BABIP of .325
Casey Blake – 2009 BABIP of .326, 5-yr BABIP of .305

No comments:

Post a Comment