David Price is having a fantastic season. As of September 17th, his ERA is 2.75, his FIP is 3.46, his WPA is 3.14 and his WAR is 3.9, good for 3rd, 8th, 6th, and 11th in the AL. 2010 has been a year of improvement for him in what are generally considered the "Big 3" of pitching ability indicators; GB%, K/9, and BB/9.
|David Price 2009||David Price 2010|
Looking at the changes, Price hasn't really drastically improved in any one area. His improvements have been more incremental in all three categories. This leads to an xFIP of 4.02 for Price, which, while better than last year's 4.49 mark, isn't quite as much of a breakout as the above measures (ERA, FIP) might indicate. What is xFIP, and why do we care? Simply put, xFIP tries to measure how good a pitcher will be going forward by regressing things that are generally considered "fluky" for pitchers (BABIP, HR/FB, LOB%), and we should care because it's been shown to be the single best predictor of pitcher performance. Indeed, looking at the generally unsustainable measures, Price's numbers suddenly look a little less wonderful.
|League Average||David Price|
*This 9.4% is the lowest HR/FB has been in several years. It's been higher in the last few years in the 10-11% range, so if you see someone suggesting regression to that point, that's a slightly outdated measure. HRs are way down league-round this year.
Recently I was talking with a friend, however, and he made the claim that Price might be able to maintain these three things at better levels than the rest of the league because of his blazing fastball. It's long been accepted that good pitchers can maintain a LOB% outside of league average, and this makes a load of sense. Certainly, better pitchers have an ability to leave men on slightly base more often. That being said, a LOB% of 77.9 is abnormally high. Over the last 3 years, only Johan Santana has maintained a higher LOB rate. We'll look at Price's high LOB% in Part 2. For now, we'll look at my friend's claim that Price's fastball would be harder to square up, and lead to weaker contact both making fewer hits and fewer long balls. While my immediate instinct was to dismiss this claim, I decided to take a further look.
Below is a graph of BABIP against fastball velocity for all pitchers with enough innings this year.
We see immediately that there's almost no relationship. The R-Squared is less than 1%, and the line of best fit has a positive slope. Price isn't about to be preventing balls in play from being hits because of his blazing heater.
That being said, Price's low BABIP might be sustainable, even if not because it's a skill of his. The Rays' team BABIP at this point is third in the majors at .286. As a result, when predicting Price's BABIP, we should project towards this .286 mark rather than league average of .302. How about his ultra low HR/FB?
This relationship is immediately more promising. While the correlation is still VERY weak (-.23) there does exist an inverse correlation between HR/FB% and fastball velocity, i.e. harder throwers have lower HR/FB. Plugging Price's fastball velocity into the linear regression, we find that Price's projected HR/FB% is a much lower than league average 7.6%.
It would be immensely foolish, however, to accept this as his true talent HR/FB%. While some fireballers like Ubaldo Jimenez manage to keep amazingly low HR/FB%s in high run environments, there's also Josh Beckett who, despite throwing flames, has an 11.7% HR/FB. There's a reason the correlation is so weak, and although it might be due to the inherent flukiness of something like HR/FB, the correlation didn't improve over larger sample sizes. That being said, however, the existence of this correlation may not be entirely meaningless. Tim Lincecum's increased HR/FB might also be the result of a drop in velocity as much as it is regression, for example.
Based on what we know, it's nearly certain that Price's true talent HR/FB% is above 7.6%. Since 2002, just 2 pitchers have maintained HR/FB%s at or below this mark over a large sample size. That being said, however, perhaps when we're projecting Price, we should not regress Price's HR/FB all the way. HR/FB seems to take roughly a long time to stabilize, and based on some napkin calculations that use estimations of when HR/FB would become stable, I would project it to be roughly 9%. Additionally, his BABIP is sustainably low (for now) because of the Rays' stellar defense.
So while it would be foolish to look at Price's sub-3 ERA and project that going forward (barring an improvement in true talent level), it would be equally foolish to simply take his xFIP and project that going forward. Most likely, his true talent lies in between, somewhere in the 3.3 to 3.6 range.