Andy Pettitte is going to announce his (second) retirement this afternoon. Much will be written (again) about Pettitte’s career and, of course, his Hall of Fame prospects. Others are better at the history and biography stuff, and, well, at pretty much all of the other stuff, too. Personally, I am not interested in predicting whether a player will get into the Hall of Fame. Analyzing players is one thing. Sociological and psychological evaluations of the Hall of Fame’s voters is another (that is not a commentary on the voters, just on my interests). When it comes to stuff like this, I prefer to focus on a player’s worthiness, that is, whether he should get into the Hall of Fame.
Much will be written regarding Pettitte’s Hall of Fame contributions now and in the off-season, just as much was written about his post-2010 retirement. I am not going to cover every angle or offer a final verdict. Rather, I want to to discuss two or three tough angles for the sabermetric evaluation of Pettitte’s case that make it an intriguing topic.
For most (but not all) of this post I am going to use Wins Above Replacement (WAR) as a basis for discussion. It is not the only way to get into these matters, and there are ongoing debates about WAR and its variants. For the sake of getting to the point, I am going to leave those aside for now, since I think WAR, for all its issues, is a good rough way of understanding a player’s value.
There is no simple algorithm for determining a player’s sabermetric Hall-worthiness, of course. Not everyone is going to agree on general principles. As a general guide and starting point, we can start with overall value. Something around 60 career wins is a good guide, although, again, it is not an absolute baseline. Pettite passes the 60 WAR test, having about 68 career WAR at the moment.
But even with more advanced metrics like WAR, most of us do not want mere accumulators in the Hall of Fame. We want players who were more than just average or above-average for a significant chunk of their career. Now things get a bit fuzzier. Sometimes we use graphs to illustrate this when we compare players. That is one good way, especially when comparing the player in question with another whose worthiness is generally acknowledged. Another quick way I prefer it to look at a player’s best three to five seasons. Without getting into all the different viewpoints, I think that a player needs at least three really awesome seasons on his Hall of Fame resume. And by “awesome,” I do not mean “above average” (three or four wins) or great for a season (five or six). I mean really, really great. Something like seven wins or more in a season.
As an example, while Duke Snider‘s total career WAR is seemingly just barely in at at 63.5, he is still a no-doubt Hall of Famer largely due to his awesome peak from 1953 to 1956, four seasons in which he averaged about eight wins a year. He was very good before that, but that sustained peak (in a short time) really puts him over the top, statistically speaking. Sure, WAR is not the most subtle of tools, so seven wins a season in a peak is not a hard and fast rule, but conveys the general idea.
The peak aspect of Pettitte’s case is not very strong in this light. He only has one season with around seven wins (7.2 in 1997), and no other season really comes close. He only had three other seasons over eighteen seasons with more than five wins. Pettitte never really had an out-and-out bad season, and was pretty much always above average, but we are looking for more than just “above average for a really long time.” Longevity is a good thing, but is it enough to outweigh Pettitte’s lack of an truly impressive group of peak seasons?
FIP, RA, and Peak Value Revisited
While one might buy FIP-based WAR (as implemented here at FanGraphs) on a season-to-season basis, one might question whether it is really appropriate as an evaluation of a player’s career. After all, Pettite has pitched 3300 innings, we probably have about as good an idea of his non-DIPS skill as any pitcher. At first glance, this might seem to hurt his case for the Hall a bit, as his RA9-WAR is around 62 as opposed to the 68 with FIP-based WAR. His career 86 ERA- is a bit worse than his career 84 FIP-. These are not precision instruments, so the difference is not that big in practical terms, but it is not really helping his case.
Unless, that is, we return to the peak value issue. If we look at Pettitte’s seasonal RA9-WAR, his peak much more impressive. His 1997 RA9-WAR (7.5) is about the same as his FIP-based WAR that year. However, the switch in perspective adds 2005 as an awesome, Hall-worthy peak season of 7.8 WAR. putting him up with the best pitchers in baseball that year.
Is it enough? Well, it is just one more great peak season, and no third season really stands out. And, as those of you have already done the math in your head have already figured out, while RA9-WAR makes his two best seasons look a better, it also makes most of the rest of his seasons a bit less impressive. So I am not sure how much it really helps or hurts him, other than bringing people’s attention to the amazing 2005 he and Rogers Clemens had for the Astros.
The above discussion all dealt with Pettitte’s regular season performances, but as anyone who has paid attention knows, Pettitte has had plenty of postseason experience as well. He has pitched in eight different World Series, including five winners. He was the 2001 ALCS MVP. All together, Pettite started 44 games in the playoffs, and totaled 276 innings. For most of today’s starting pitchers, that is almost another season-and-half of work. He ERA- (84) was actually better in the postseason than it was in the regular season, even if his FIP was worse.
We do not have WAR (or other general value stat) for the postseason play at this time, and whether and how it should be done is another issue, but those innings clearly have some value. The competition was tougher than in the regular season. And while Pettitte’s amount of postseason experience was boosted by him having very good teammates, one cannot discount the fact that Pettitte made big contributions to his teams making the playoffs (and advancing so far) so often. It was not a mere coincidence.
How much should we value Pettitte’s (mostly good) postseason play? How should that be weighted against the peak and longevity issues? I do not know, but it matters, and is one more reason his Hall of Fame case is fun to discuss.
Print This Post