In my last column, I introduced Ultimate Value Index (UVI) and used it to evaluate several of the hitters in the Oakland A’s system.
One of UVI’s greatest assets is its flexibility. That is, there is a metric for pitchers as well as hitters, so one can get a per-plate-appearance idea of how a pitcher performs relative to a hitter.
However, there are a few differences between pitcher UVI and hitter UVI. Primary among these differences is that while hitting UVI simply states what a hitter has done, pitcher UVI normalizes balls in play to show what a pitcher should have done. In effect, it doesn’t matter if a pitcher has been lucky or unlucky on balls in play, because UVI removes the luck and defense elements.
In addition to UVI itself, my system also generates several other statistics. It produces expectations for innings pitched, hits allowed, WHIP, RA (a metric I call Adjusted Run Average (ARA); note that this includes projected unearned runs as well as earned runs, so it runs a bit higher than ERA), BABIP, BA against, OBP against, SLG against, and OPS against.
Also note that UVI does not include baserunning against a pitcher.
Let’s begin looking at some pitches and how they’ve done. All stats are through games on Tuesday, May 6.
Joe Blanton has a .417 UVI so far. He has gotten slightly unlucky on balls in play, but Blanton continues to do what he has always done, which is eat innings while delivering slightly above-average performance. His low strike-out rate is worrisome, because it is even lower than his normal rate, but he’s kept the ball in the yard and is showing his usual great control. It all appears to add up to a No. 2 starter package.
Dallas Braden is not thought of highly by A’s fans in general because of his abysmal ERA in his rookie year. However, Braden got extraordinarily unlucky in 2007, in large part due to a .347 BABIP. That was 31 points higher than he should have had, and thus hurt him badly. According to my translations, Braden’s Adjusted Run Average (ARA) for 2007 was 4.62, and that includes unearned runs, so it runs slightly higher than the more traditional ERA metric. His 2007 UVI was .464, which is pretty much average.
This year, Braden has continued to excel, albeit in a relief role. He has only thrown seven innings, and he has allowed four earned runs on nine hits, but if you remove the luck, two of those hits should have gone for outs, and his ARA is an excellent 3.21. Braden’s .344 UVI to date ranks third on the 2008 A's, and he deserves to remain with the club.
Andrew Brown is the opposite of Braden: A’s fans are comfortable with him late in the game due to his low ERA, but he has been very lucky this year. His BB:K ratio is 7:12, which isn’t particularly good, and his K/9 is fairly low as well, especially for someone with Brown’s power stuff. With all those mitigating factors, it doesn’t take a genius to realize that Brown has been extremely lucky to allow just seven hits in 17 innings.
He’s not a bad pitcher by any means, but with average luck he allows 14 hits in 14.7 innings, which is far more pedestrian. Even with correcting for Brown’s good luck, his UVI is still .455 and his ARA is 4.59. As a side note, the next time you hear people talking about how the A’s defense is killing the pitching with all the errors, remember this.
Santiago Casilla, like Brown, has been dominant this year. Unlike Brown, however, Casilla hasn’t been very lucky. His translation involves a whopping .308 UVI and a ridiculous 2.76 ARA. How great is that? Well, there weren’t any pitchers in the majors last year that I know of (I’ve run the numbers on all the good ones, Joba Chamberlain is lowest at .332) who pitched 20 innings and have a UVI below .330, so if Casilla keeps this up, he’ll easily be the best pitcher in baseball this year.
Joey Devine is second to Casilla and just ahead of Braden with a .333 UVI thus far. An interesting statistical tidbit about him that I noticed is that for his career, he has always been an extreme flyball pitcher (posting GB%s of 40% or lower) every year except for 2007, when his GB%s were in the 50s% and 60s% (at different levels). Devine easily had the best year of his career in 2007, so it’s very tempting to attribute his success to getting more grounders. However, this year he has reverted back to his fly-balling ways with a GB% of just 43. Has he gone back to his old approach or is this just a small-sample fluke? It’s too early to tell, but whatever he’s doing, it’s working, because look back at the Casilla comment to realize exactly how well Devine has pitched so far.
Justin Duchscherer has posted a good .407 UVI so far with an ARA of 3.84. If he can stay healthy and pitch deep into games, there’s no statistical reason he can’t be every bit as valuable as a starter as he was in the bullpen the last four years.
Dana Eveland has a solid .412 mark so far, and his projected slugging against is a measly .338. However, Eveland’s 18 walks make his projected OBP against fall at .333, which is a bit below-average. Watching Eveland pitch, it’s not hard to see that his delivery is very difficult to repeat 100 times a game from either the windup or the stretch, so it’s difficult for him to maintain his release point. Eveland tends to be either very good or very bad in an inning, and I think losing his release point is the problem. If Curt Young calms Eveland down a bit on the mound, I would expect the walks to drop, making his line look even better. Even with messy mechanics, he’s pitching pretty well.
Chad Gaudin has a rather perplexing statline. On one hand, he’s gotten a bit lucky this year, as four of his outs should have gone for hits. However, what is most intriguing about Gaudin is that he normally exhibits fairly pronounced groundball tendencies, but he has not done so this year. He has kept barely 40% of balls on the ground. Since Gaudin had been starting this year, his sample size is a bit bigger than Devine’s, but it is still a bit early to tell if this is a 36-inning fluke or a real change in approach. One good sign from Gaudin is that he has only walked 10 batters in 36 innings, so he is on pace to come nowhere near triple digits in walks like he did last year. It all adds up to a .458 UVI, which is a bit below where he needs to be, but there are some real positives here.
Greg Smith is perhaps the pitcher most affected by UVI not including baserunning, so he doesn’t really deserve to have a .462 UVI like he does. Basically, Smith the pitcher (not Smith the pickoff artist) has pitched as well in 2008 as Dallas Braden did in 2007. It doesn’t matter how good your pickoff move is; you’re never going to be able to sustain a .226 BABIP. My translation calls for Smith’s BABIP to rise to .327 though, and that seems a bit extreme even to me. It’s been suggested that my translations are biased towards groundballers (something which I’m looking into), but Smith has the highest projected BABIP in the organization, and he doesn’t have anywhere near the highest flyball rate. Anyway, even if his BABIP doesn’t go to .327, it’s not going to stay where it is. Expect some regression, although exactly how much is unclear.
Huston Street has been the worst pitcher on the A’s, with the exception of Keith Foulke (who has pitched so little that it’s hard to take his .500 UVI for anything). That says more about the A’s pitching staff than it does about Street, whose .477 UVI is just slightly below-average. His ARA stands at 4.56, which is also fine. However, it’s looking more and more like it’s Santiago Casilla who belongs in the shutdown role.
About The Author: Nathianel Stoltz is a statistics major at James Madison University in Virginia. He is the creator of the "Ultimate Value Index" or "UVI" baseball statistic. He hopes to some day work in the front office of a major league team. You can e-mail him with questions or comments by clicking here.