Changes in Starting Pitcher FIP since May 10th

(Thanks to commenter schmenkman for pointing out an error in the % change calculations. The article has been edited to reflect the new values)

We all know that players' day to day statistics are subject to some degree of volatility and random fluctuation throughout the season. All of those fluctuations even out by the end of the year and we are left with an aggregate picture of a player's performance.

I wanted to get a sense of which starting pitchers have been throwing better over the course of the second half of this still young season. To do that, I used Dave Pinto's wonderful Day by Day Database to calculate Fielding Independent Pitching (FIP) numbers for starting pitchers before and after May 9th--roughly the mid-point of the season to date.

To be fair, I calculated the percent difference between the two time periods to get a better sense of who has really improved and whose declined the most. There's lots of interesting pitchers on the list, but the one that stands out the most to me is Cole Hamels.

Over his first seven starts this year, Hamels posted a 2.99 FIP. His K/9 (9.25) and BB/9 (2.08) were very impressive. Over his past eight starts, Hamels strikeouts dropped a tad (8.63 K/9), but his walks improved by nearly one per nine innings (1.28 BB/9). The biggest difference for Hamels in terms of his FIP has been home runs. Through his first seven starts, Hamels gave up four home runs (.76 HR/9), whereas the last eight starts has seen Hamels give up only one home run (.16 HR/9). 

All of this adds up to a 33% improvement in Hamel's FIP--2.99 vs. 1.99. With that 1.99, he's also been the best starting pitcher in terms of FIP since May 10th. Hamels is also the only sub-3.00 FIP starter that improved from the first to the second time period.

Some other observations:

• Erik Bedard has also been pretty impressive the past month or so. Bedard started this year with a 5.19 FIP, but has more recently posted a 2.54 FIP. That's a 51% improvement. Pair that with Felix Hernandez and Michael Pineda and it's no wonder the Mariners are pushing the Rangers in the AL West.
• A number of starters that got off to good starts have seen their fortunes reversal significantly. Matt Garza saw the most drastic reversal, going from a 1.83 FIP to a 4.94 FIP, a difference of 169%. A pair of Athletics also saw a negative change. Brett Anderson (3.24 to 5.39) and Trevor Cahill (3.14 to 5.38)  both saw negative changes of over 66%.
• The most interesting starter that's posted a big decline since May 10th is Roy Halladay. Halladay's post-May 10th FIP has increased 71% versus his pre-May 10th FIP (5th largest increase). The funny thing? He still posted a 3.00 FIP since May 10th. Gives you a sense of how ridiculous his first month or so was.
• Josh Beckett has been pretty good--and pretty consistent. FIP prior to May 9th: 3.22. FIP after May 10th: 3.22.
• Of the 12 starters that have posted sub-3.00 FIPs since May 10th, only two of them saw their FIP increase from the previous period (Jaime Garcia and Roy Halladay). 
• There was some nice symmetry to the data, which just reinforces the idea that players will regress towards their mean. Of the 113 starters in the data set, 57 saw their FIP increase from time 1 to time 2 and 56 saw their FIP decline.

Data set can be found here.