Oswalt’s Performance Can Teach Important Evaluative Lessons

Roy Oswalt, the right-handed ace of the Houston Astros, has been, without question, one of the best and most consistently solid pitchers in baseball since his debut at the beginning of the decade.  He has never been a strikeout machine, but punches batters out at a very respectable rate, and exhibits enough control to limit his free passes.  With an above average strand rate in his career, it is no wonder how he has been so successful.  He works quickly, complements a 94 mph fastball with a 70 mph curveball, does not surrender many hits, limits his walks, strikes out enough batters, and prevents a good portion of those that do reach base from scoring.  This year, however, things have been different.  Or have they?
Overall, in 29 starts, Oswalt’s 3.54 ERA and 3.81 FIP are off of his proven track record, but evaluating his season on these two metrics alone will result in an inaccurate and incomplete analysis.  In his first 21 starts, Oswalt pitched in 126.1 innings, about 6 IP/GS.  He surrendered 146 hits, 20 of which left the ballpark.  He fanned 106, walked 32, and allowed 67 earned runs.  All told, this produced a 4.77 ERA, an approximate 4.34 FIP, and an other-worldly 1.43 HR/9.
None of those numbers looked particularly great, but the HR rate was, or should have been, a big red flag.  Too many ignored this well below-average number, though.  Additionally, his HR/FB% was well below the average 10-11%.  His K/9 in this span was 7.55, which is actually higher than any year in his career.  On top of that, his 2.28 BB/9 was very respectable and not disproportionately high.  So, while Oswalt was clearly struggling, his K and BB rates were great; he was merely getting “unlucky” with balls falling in for hits and was doing something to allow plenty of balls to leave the yard.
As Dave Cameron discussed last week at Fangraphs,  if you change Oswalt’s HR rate in the early part of the season to that of a more recent stretch, it is the difference of 1-2 runs in ERA and FIP.  The reason, as he astutely pointed out, is that home runs are fickly by nature and that future success is much more stable when based on K and BB rates.  If you see a pitcher with poor K and BB rates, but great home run suppression numbers, be wary.  On the flipside, a pitcher with great K and BB rates, but very, very poor home run suppression numbers, expect some form of regression.  Essentially, Oswalt had to get better.  He just had to.
Over his last 8 starts, this regression occurred, big time.  He has pitched 64.1 innings, about 8 IP/GS.  Just 37 hits have been surrendered, of which only one has been a home run!  Only 8 earned runs have scored, and his K/BB is a tick under 4.0.  Interestingly enough, while his HR/9 has drastically dropped to 0.14 in this span, and his BB/9 has dropped from 2.28 to 1.54, his K/9 has gone from 7.55 to 5.87.  He has been so good lately that his 4.77 ERA/4.34 FIP is now a 3.54 ERA/3.81 FIP.  These numbers are right in line with his pre-season Marcel projection of 3.61 ERA/3.63 FIP.  Something to keep in mind, though, is that the HR rate has a significant effect on FIP.  The Hardball Times publishes a stat called xFIP, which merely normalizes the HR aspect of FIP to offer a better indicator of controllable skills success.  While Oswalt appears to be declining somewhat, would you believe that his controllable skills are right in line with dominant years past?
In 2004, his xFIP was 3.73.  The next year, 3.56.  2006 brought with it a 3.71 xFIP.  Last year, it rose to 4.08, which brings us to this year, when it is… 3.71!  When home runs are normalized, Oswalt’s controllable skills have been equal to, or better than, those over the previous four seasons.  An earlier study of mine showed that pitchers give up the majority of their home runs on pitched 2-3 mph below the normal velocity for that pitch, with very poor location, IE, the general vicinity of right down the middle.  It isn’t exactly rocket science.  Slow pitches + right down the middle = home runs.  Comparing these “mistake pitches” of Oswalt in the 21 starts to those in the more recent 8 starts shows that he has actually thrown more mistakes recently.
The biggest difference, however, is that the mistakes early on were more often than not, hanging off-speed pitches that followed other off-speed pitches.  More recently, they have been 92-94 mph fastballs following an off-speed delivery.  A fastball at that velocity becomes very hard to hit when its predecessor is nearly 15-25 mph slower.  Oswalt has also increased his fastball usage from 62% to 73% over these last 8 starts, cutting back his usage of sliders from 14% to 7%.  With no real differences in velocity, movement, or location, it seems that Oswalt’s sequencing and selection loom large in his recent “turn-around.”
I use quotes only because his true talent level told us he couldn’t be as bad as he “looked” early on.  Oswalt is a very good pitcher, who was experiencing an unsustainable home run rate that deflated his numbers.  He isn’t as good as he looks right now, but is much better than he looked early on.  Since his K and BB rates stayed on par with the past, but his HR rate crazily increased, he should continue to have success.  Now, if we find that, three years from now, he is still posting higher HR rates, then it might be a case of him doing something differently on the mound that is resulting in more longballs.  For now, though, it just seems like an extreme outlier screaming regression.  Hopefully, Oswalt’s season will convince some that true talent levels and unsustainable counts in certain metrics need to be taken into account when evaluating a player, but more likely than not, it will just end up another very solid season in what will end up a very solid career.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: