It Could Be Maybe
August 4, 2006

William Goldman once famously said, "Nobody knows nothing."  While pithy, it sort of glosses over the truth.  Obviously we know some things; otherwise, life itself would be impossible.  Perhaps what he was trying to say is that no one knows anything so well that he/she can predict outcomes 100% of the time.  While that interpretation is also pithy, I'm not sure it's true either.  I know if I started my car a billion times that there are certain outcomes which would never happen.  It would never, for example, turn into a squid,   This is something I know and can predict with 100% accuracy.  What I can't predict with 100% accuracy is what will actually happen.  I can eliminate at least a billion alternative results but even with only a few true possibilities, I'll concede that absolute certainty is elusive.  Nearly 100% of the time, the car will start but there is that .0001 % where it might just sit there.  Still, even with that scintilla of doubt, I'm ok with saying we do know some things

How much, however, is subject to considerable debate.  For example, we don't really know how much we actually know because we keep forgetting stuff we do know.  For example, both the Greeks and Romans of two millennia ago had surgical instruments very similar to those we use today.  They used these instruments to do fairly complex surgeries including basic brain and eye surgery.  They used pins to help heal broken limbs, traction and dozens of other "modern" techniques to help people recover from injuries.  We learned this from recent archeological finds: healed broken bones with pins still in them, instructions on papyrus found in jars, etc.  We probably could have known this from books had there not been so many book burnings during the Dark Ages.  It had been thought that modern medicine invented these techniques and tools in the late 1800s.  Now it appears they merely rediscovered them.  So more than a thousand years of were lost because a particular segment of the human population felt so threatened by someone else's knowledge that they actively destroyed it.  Unfortunately, not much has changed in 2000 years.

So what does all this have to do with baseball?  Well, stats are becoming an increasingly popular tool for evaluating players.  A growing number of people who use them denigrate the value of the contributions of scouts and anecdotal evidence.  Of course, some of this is probably just revenge for the scouts denigrating the value of the numbers.   But nevertheless, both are necessary for talent evaluation.  Using one without the other would be a lot like burning those medical books back in the Middle Ages.  So when I say stuff about the stats not telling the whole story, it's not that I'm against stats.  I use them all the time and they can provide great insight.  The problem I have is when people try to use them to sum up the value of a player.  There are simply too many things that happen in each game that aren't covered by the numbers to rely exclusively on them.  I don't believe one can sum up a player with numbers any more than one can sum up a person by their tax return.  And even when the numbers paint a fairly clear picture, there's still the very real possibility that they will fail miserably to predict future outcomes.

One of the most popular modern statistics in use is batting average on balls in play (BABIP).  The conventional wisdom is that all pitchers gravitate to a league mean.  More likely is that they gravitate toward their own career mean that resides in the vicinity of the league mean.  There are a number of pitchers who consistently perform well under the league average and plenty who are over.  It seems to me that the proper way to use BABIP is using both an individual's context - how what he's doing this season compares to what he's done previously - and a team context - how does he compare to other pitchers on his team and are his successes (or failures) part of a team pattern.

For example, Oliver Perez is having a pretty tough year.  It's easy to look at his control issues and say he's a lost cause.  He can't pinpoint his pitches so when he doesn't walk guys he's wild in the strikezone and gets hammered.  His .336 BABIP would seem to back this up.  But here's the thing: his career BABIP is .288 and that includes several years when he struggled to find the strikezone, just like this year.  It would be reasonable to expect that his performance forward would be better as he regresses toward his own mean.  But then that would ignore the fact that of the Pirate pitchers who've thrown over 25 innings this season, only Matt Capps has a BABIP below .310.  If league mean varies between .285-.300 in any given year, this indicates that team defense might be the possible cause of at least some of Perez' struggles.  One possible culprit is Jack Wilson, who is posting career lows in both fielding percentage and zone rating and nearly as bad in range factor.  All of those defensive stats have serious flaws, but what can be observed with certainty is that Wilson bulked up this winter and hasn't displayed the same quickness as before.  So even with their flaws, the numbers may be telling a true story.  Not to throw the blame squarely on Wilson, the Pirates have also had a problem with injuries at third base and a revolving door in center field.  All that to say that while Perez might not bounce back this year, it was reasonable to think that with a more stable defense that better times were ahead even before he was traded to the Mets.

So with that as a lead-in, what I'd like to do is point out a few pitchers who're on the bad side of both the league average BABIP and their own career rates but whose circumstances point toward probable improvement.  And although there's no guarantee that their luck will change with the flick of a switch, they probably do make a good place to start when you're looking for guys who are going to surge down the stretch.  I'll also toss in a few guys who've had everything go right so far who might be good guys to trade at their peak value. 

One guy who was pretty universally dismissed as over-hyped this winter when he signed as a free agent has posted some pretty nice peripheral fantasy numbers yet not so much in wins or ERA.  His K/BB ratio is 4.82 Ks for every walk.  His G/F rate over the last two years is over 2.00, 1.73 this year.  But he's been plagued by what looks like some bad luck - his BABIP is .339 well above his career average of .283.  League average ratio of homers/flyballs is around 1/10, but 9 of his 55 flyballs have left the yard.  The team he plays for ranks in the top 5 in the AL in offense and he has electric stuff.  So you'd think that AJ Burnett would be due for a rebound over the last two months.  Also working in his favor is that he pitches very well on turf - 2.77 on turf (0.967 WHIP), 3.86 on grass (1.284 WHIP) over the last three years, and his next four starts are on turf against the White Sox, Twins, Devil Rays and As.  He might even get an end of the month start against the Royals.  As much as people like to bust on him for his contract and health, he's a guy I would be looking to trade for if I didn't already have him.

Andy Pettitte has an unbelievably high .358 BABIP.  I've never been a big Pettitte fan, but even I will acknowledge that can't possibly continue.  For most of his career he's been comfortably over .300 in that regard but this is well beyond the pale.  Maybe it has something to do with Preston Wilson playing regularly in left.  There was a time when he was an above average flychaser, but watching him last year in RFK was almost painful.  His knee surgeries have robbed him of more than a little speed and perhaps some of his confidence.  Willie Tavares has been pretty average in center and Jason Lane/Lance Berkman don't have much range, so maybe Pettitte is being victimized by balls to the gaps.   His G/F rate is right in line with his career average so even if his struggles can be blamed on his outfield defense, it still looks like the ball hasn't been bouncing his way.

CC Sabathia hasn't been quite as unlucky, but his .315 BABIP is well above his career average of .290.  It's interesting to note that of Cleveland's starters, only Jake Westbrook is under .300.  He's the team's most prolific groundball pitcher, getting more than twice as many as #2 on the team, Sabathia.  So maybe it's the Cleveland outfield range that's costing the team so many extra runs.  Maybe Coco Crisp wasn't so expendable.  And maybe Cleveland can carry a second rate glove (like Peralta) or two on the infield with so many flyball starters.  Still, it stands to reason that a natural outfielder like Shin-Soo Choo would do a better job of chasing down flyballs than a guy who spent much of his career at third base (Casey Blake).   but So if anyone shows significant signs of improvement in Cleveland, it will probably be Sabathia.  Other Tribe starters are more flyball prone, but none are as far off their career norms as big CC.

Odalis Perez BABIP in LA was .372.  For a guy whose career rate is .284, this would be what statisticians would call "an outlier".  KCs defense hasn't been great.  In fact, of their active starters, Luke Hudson's .309 BABIP leads the team.  But Luke Hudson is no Odalis Perez so the guy known as "the other O. Perez" should be able to bounce back and at least challenge for the team lead.  This would produce acceptable numbers - innings that won't kill your team ratios - down the stretch, even with a terrible team behind him.

I don't have anyone specific in mind, but the starters for the Devil Rays might show significant improvements over the final two months.  Most of them have pretty bad BABIPs, but with the exception of Travis Lee the infield defenders that have backed them up have been among the worst collection of non-gloves in quite some time.  Now with Zobrist at short and Upton at third, they have a chance to be pretty decent everywhere but second base.  Fewer groundball singles could mean significant improvements to the ERA and WHIP of a healthy Scott Kazmir or a Casey Fossum, but mediocre talents like Jamie Shields and Tim Corcoran might benefit even more.  Or maybe they are simply too mediocre even for defensive improvements to help.  Shields didn't look too bad tonight against the Red Sox, though.  Only time will tell.

One thing I found especially surprising was how poorly Twins' pitchers have done this year.  But maybe having Juan Castro and Tony Batista on the field everyday for two months wasn't such a good idea.  I have to believe that the promotion of Jason Bartlett and the use of Nick Punto has played more than a small role in the Twins resurgence into the Wild Card race.  Even with the changes, only Johan Santana and Francisco Liriano have BABIPs under .300 in Minnesota.  Even Joe Nathan is at .306.  Even without making any trades at the deadline, Twins pitchers will be better in the second half.

Bobby Abreu "won" a gold glove last year but is it merely a coincidence that Philadelphia has won four straight and seven of their last eight since he was traded?  Brett Myers is the closest thing they have to a groundball pitcher at 1.33, so are Shane Victorino and Dave Dellucci that much better defensively than Abreu?  No Phillie starter had a BABIP under .300.  Cory Lidle was the closest and he is a groundball pitcher (1.59 this year, 1.75 for his career).  Sure, it would have helped had they had Aaron Rowand in center every game, but he only missed about a dozen games from his collision with the centerfield wall.  Jon Lieber is my bet to show the most improvement over the balance of the season. 

OK, so who is likely to fall off during the stretch run? 

Well, #1 on my list is Jered Weaver for a lot of reasons.  The first is that he's had a pretty easy schedule to this point and in August and September it's going to get tough.  He'll be facing much better teams in a pennant race so the pressure will be greater to make good pitches or else be punished.  The second is that he really isn't that great.  His fastball is pretty average both in terms of velocity and movement and he struggles to throw his breaking ball for quality strikes.  He's got good deception, but his delivery is so long and so involved that mechanical issues will always be a concern.  The third is that his BABIP to this point is a ridiculously low .232.  Even pitchers who consistently beat the league average rarely go below .250 for a full season.  Look for a lot more outings like his last one against Texas the rest of the way.

The Mets think they've found the solution to their rotation problems in John Maine, but he's not.  His stuff is pretty average, so for him to be posting a BABIP of .224 is pretty much a dream sequence.  The Mets do have a good defensive team so he will be better than expected.  But by season's end it wouldn't be too surprising to see his totals look a lot more like Steve Trachsel's than Pedro Martinez' as they do now.  One caveat: the last time he was up (for 8 starts in 2005) he posted a BABIP of .248, but was undone by the 24 walks and 8 homers he gave up in 40 innings.  Shea will take care of some of the home runs, so if he just throws a lot of strikes maybe the rest will take care of itself.  He's an interesting case but I wouldn't bet on him to continue.

Mike Mussina is probably due for a bit of a correction as well.   New change-up grip or not, his BABIP of .266 is nearly 40 points better than his average of the last 5 years.  Maybe it's all those ARod errors that caused his surge.  Those balls in play would have been hits but because ARod got to them and fumbled them, they were counted as errors and thus not applied to his BABIP.  Or maybe he's just been lucky.  Or maybe Miguel Cairo is really that much better defensively than Robinson Cano.  Or maybe he's just been lucky.  Or maybe... he's just been lucky.

I wouldn't look for much improvement from guys in Milwaukee, which doesn't bode well for their playoff chances.  David Bell is an improvement at third but they still have problems defensively just about everywhere else.  Only Tomo Ohka has a BABIP under .300 and that's unlikely to remain so low given his track record.  

Only two Tiger pitchers who've thrown more than 25 innings have a BABIP above .300.  So either the Tiger's defense suddenly became otherworldly, or they've had a lot of balls bounce their way.  My guess is the latter.  They might continue to enjoy this magical season; after all, Sean Casey hit a home run in his first game with Detroit after hitting just 3 in the previous 61 games.  But the safe bet is that the Tigers - both their hitters and their pitchers - will come back to reality no later than next spring.  And the two Tigers who are posting .300+ BABIP?  Todd Jones (.320) and Roman Colon (.330).  But I think most people could have told you that.