Before we get to this week’s edition of Sabermetric Tuesday, I’d be remiss if I didn’t correct a mistake in my player blurbs earlier this week. Amidst a long-winded rant about my beloved Bears not turning in a draft card with the name Brian Brohm on it, I mistakenly referred to the last great Bears QB as Sammy Baugh. A number of readers wrote in to note the mistake, correctly guessing that I meant Sid Luckman. While admittedly I probably don’t know my 1940’s football as well as I should this was an inexcusable mistake for a Bears fan as Sammy Baugh and Sid Luckman battled each other in a number of title games and for some reason when I was grasping for Sid Luckman my mind reached for his counterpart Sammy Baugh. This reminds me of the time last year when I referenced the great job Tommy Lasorda would do managing the Cardinals (instead of Tony La Russa). Last year I blamed those damn slim-fast commercials and their subliminal messaging. So this year with Sammy Baugh’s marketing team not doing their jobs well enough I’m going to blame Anthony Perri for letting me write un-edited. All I know is someone is to blame here and it can’t possibly be me! Am I in denial? I’ll let you be the judge while I sit here in my gorgeous brand new Navy Blue #12 Brian Brohm jersey.
Now that I’ve wasted 200+ words trying to rationalize my idiocy, let’s get down to the real reason you’re here today, Sabermetric Tuesday!!! As my good buddy once said while enjoying a beverage at an establishment by the name of Fat Tuesdays, “they don’t lie… there are fat girls here, and it IS Tuesday!” Well neither do we, it is Tuesday and as such we’re going to bring some Sabermetrics to the table. Today we’re going to focus on a statistic called: Disaster Starts. We touched on this topic last year after a subscriber wrote in looking for some more information on the statistic defining it as any start in which a starter allows more ER’s than Innings Pitched. At the time I hadn’t seen the term or the stat published anywhere but my recent research into the stat suggests Jim Baker of Baseball Prospectus established the stat a few years back. So here’s a hat tip to Mr. Baker for the idea, which in last year’s article I built upon a bit further. I thought limiting the statistic to just outings in which a starter allowed more ER’s than IP wasn’t capturing all of the data we’re looking for. Since the goal is to identify which pitchers are more likely to “blow up” or have a “disastrous start”, I felt it was important to try to collect all of the starts which could really hurt your team’s ERA. In my humble opinion, I’d like to avoid at all costs any start in which the pitcher has an ERA above 5.50 because in my mind they’re “disastrous”. So I simply added in any start that a pitcher allows 5 or more ER’s, since almost every outing in which a pitcher allows 5 ER’s or more results in an ERA of 5.50 or higher (add in the fact that we don’t see many 5 ER CG’s and I feel safe adding this minor modification to Jim Baker’s original stat). So with that minor addition, we have our definition of a “Disaster Start”, which is simply: Any start in which a pitcher allows more ER’s than IP or any start in which a pitcher allows 5 ER’s or more.
With the definition out of the way let’s talk about the practicality of the stat and how we should use it going forward. When I think of the times I most need to avoid a “Disaster Start” it’s one of two situations either 1) I’m looking for a spot starting option or 2) I’m looking for a serviceable back-end of the rotation type that will preserve the good work my Ace(s) has done. So when we’re looking at the Disaster Start numbers, the players we really want to focus our attention on are the middle of the road starting options. If we can identify the “safest” option (those with the smallest % of Disaster Starts), we can play the percentages a bit more with our back-end starters by employing those less likely to act like kamikaze bombers on our WHIP and ERA.
With the 2008 season underway and starters getting through their 5th time or so in the rotation, it’s a bit early to start analyzing this year’s data to see which starters can be trusted for those last spots in our rotation or as spot start options going forward, so let’s instead focus on 2007’s data and see who stands out as players that can be trusted.
In order to make sure we didn’t have any sample size issues with last year’s data any pitcher who started 18 or more games was included in the sample, which gave us a 125 pitcher sample from last season to work with. Of that group of 125 pitchers, approximately 30 had a Disaster Start % below 10%. The 10% threshold isn’t statistically significant just a barrier I’ve decided to use to separate, plus it fits in nicely with the Top 30 overall starters at avoiding Disaster Starts. Let’s take a look at this list (note: Bolded/Highlighted Names are those of the back-end of the rotation types we discussed above).
class=shape>
style='language:en-US'>
border-collapse:collapse'> | ||||
Highlighted Names Include: Jeff Suppan, Tom Gorzelanny, Paul Maholm, Noah Lowry, Brian Bannister, Aaron Cook, Jake Westbrook, and Greg Maddux.
A couple things stand out to me: 1) The list is loaded with Ground Ball Pitchers and 2) The majority of the back-end pitchers that qualify are in the NL. I haven’t had the chance to run the data, but it wouldn’t surprise me if there’s an inverse correlation between Ground Ball Rates and Disaster Start %. Ground ball pitchers are able to get out of jams a bit easier by inducing double plays and tend to avoid big innings thanks to limiting HR’s. Similarly pitching in the NL against offenses with a pitcher’s slot in the batting order and often a weak-hitting 8th hitter as well, it’s much safer to use pitchers from the NL in the back-end of your rotation or for spot starts. We’ll do a bit more analysis on which indicators lead to lower Disaster Start % later in the season along with an analysis of the 2008 pitching logs, but for now when looking for potential spot-starting options either look for names on this list or look for pitchers with high GB Rates that pitch in the NL.
One last name to keep in mind: Joe Saunders
The one name that stood out to me on this list but just missed the 18 starts required to qualify was Joe Saunders. In 15 starts last season Saunders didn’t post a single Disaster Start. Saunders also qualified on my Ground Ball screen earlier in the season as a potential sleeper. While his hot start to the season is unsustainable (2.55 ERA, 1.02 WHIP heading into today’s start), Saunders isn’t going to go away and makes for a nice back-end of the rotation option or even spot-starting option in shallow leagues.
