I went through and calculated the final DIPS for Oswalt, Moehler, Wolf, Backe, and Rodriguez (Wandy). These were the five guys who pitched a significant amount of the season as starters, project to be starters next year who we have decent enough data on, and are all possibilities or sure things for members of the 2009 starting rotation. While they are all possible candidates for the rotation next year, I hope that my analysis of their DIPS presents a compelling argument for Wolf not to be resigned and adds emphasis to our need for quality starting pitching for 2009.
While some of you may remember an article I wrote in my first week at the CFB, I'll refresh your memories and provide a link to the far more detailed explication of why DIPS is the measure I'm choosing to utilize for next year's forecasts. DIPS is calculated by applying league average rates for pitchers balls in play, negating the effects of defense and luck, by adjusting a pitchers' line drive rate and then adjusting the other rates accordingly. The adjusted number of batted ball results are then assigned league average results to determine the number of 1B, 2B, 3B, HR, and outs a pitcher would have allowed in a league average park with a league average defense behind them. It, in my opinion, provides the best inference to the the "skill" a pitcher demonstrated on the mound. It is further useful because DIPS have a large correlation with the pitcher's next year ERA/RA -- more so than FIP or ERA could wish to possess. In essence, it's one of the best tools to analyze a pitcher's season and then forecast the next season's performance.
While I'm uploading a spread sheet with all the numbers and calculations included, I'll go through a quick break down of the aforementioned pitchers' DIPS, and their DIPS-RA allowed difference -- something we can essentially call the luck/defense factor in their performance.
(Source: Astros2008DIPS (me) and fangraphs.com)
Interestingly, our pitchers don't exhibit a universal tendency towards having been worse than their ERA or RA indicate, given that our defense was one of the best in the league. I'm not sure how valid Roy's numbers are due to his obvious mechinical issues earlier in the year, but perhaps we can't expect his total dominance during the second half to be the Roy we see all of next year, but certainly he'll still be our ace. For those of you who want to argue about Randy Wolf being signed, I urge you to download the spread sheet I've attached. I calculated Wolf's DIPS and differential for the full 2008 season and his partial season with the Astros. His split for just the Astros gave him a DIPS above 5.00 and he had a over a full run differential from his RA and DIPS. What I'm saying is: dude got lucky. I do not want the Astros to bank a few million on him getting that lucky again. Let's let him walk and take the compensation pick (something which could off-set the compensation pick we could lose if we sign a Type A FA pitcher...the upside on that transaction would be ridiculous).
The person who surprised me was Brandon Backe. He's been awful this year, but a 5.30 RA is both acceptable for a number five starter and above the replacement level ERA of 5.61 for NL Pitchers (Source: Baseball Between the Numbers, "Why Is Mario Mendoza So Important?" pg. 169). If Backe is really pitching at a 5.31 RA level, then we should expect his ERA to be slightly lower, say a 5.20 ERA. For less than a million dollars, I'd take Backe as an insurance policy against the likes of Jack Cassel and the rest of the Round Rock staff.
So what can we glean from our projected ERA's next year? We'll we can make value judgements as to whether or not we should sign these guys or let them go. Given the information I came up with, my rotation, if I were in charge, next year would be:
- Roy Oswalt
- Wandy Rodriguez
- Brian Moehler (though he's only in here because we inked him to an extension)
- Brandon Backe