Earlier this week on Twitter, I was part of a discussion comparing Troy Tulowitzki and Evan Longoria, two of the best players in the game. I personally give Longoria a slight edge, but obviously Tulowitzki is great, too. If someone prefers him to Longoria, that is fine, and I could probably be talked in to it. What really spurs this particular post is the discussion we had about comparing their offense. Keeping in mind that this was a casual discussion rather than a deep evaluation of “true talent” involving all of the necessary regression and adjustments, someone noted that over the last three seasons (2009-2011) the two players have had virtually identical offensive value per plate appearance: Tulowitzki has a 137 wRC+, and Longoria has a 136 wRC+. I argued that Longoria’s performance was more impressive given that the American League has superior pitching relative to the National League.
However, Dave Cameron made an interesting point: the Rockies play in the National League West, where hitters seemingly face s larger proportion of stud pitchers — Dave mentioned Tim Lincecum, Matt Cain, Madison Bumgarner, Clayton Kershaw, and Mat Latos in this connection. He also pointed out that Longoria did not have to face the Rays’ own excellent pitching staff. So I decided to look at it more closely. The point is not to settle the Longoria versus Tulowitzki dispute. Rather, I am interested in whether individual hitters face (or do not face) particular pitchers enough that they require a “divisional” adjustment of some sort.
Let’s get a couple of things out of the way. First of all, I am not going to get into the overall American League versus National League dispute here. I realize that even seeing a birth certificate is not going to convince some people, so I will simply reference (again) MGL’s recent study and his more extensive 2006 study. If you want to complain about the assertion that the AL has better talent overall, please direct your complaints there or elsewhere. While the Senior Circuit has caught up and perhaps taken a slight lead in hitting talent (at least through 2011, who knows how this off-season might have changed things), the AL still had more talent overall because of a much greater lead in pitching.
Second, the best way to do this would be to take all the pitchers each hitter faced and get projections (or retro-jections) of the true talent in 2011, including all of the relevant adjustments for environment, and then compare. Well, I’m not doing a whole set of pitcher retro-jections and then matching them up per plate appearance with two hitters for one post. I am claiming that this is some super-duper rigorous study. I am simply going to look at the small group of “really good” pitchers as given by my Twitter interlocutors. There were other qualifications I could put in here, e.g., the problem of park factors and interleague comparison, but I’ve already gone on too long with prefatory remarks. Caveat lector.
Did Troy Tulowitzki face such a great proportion of NL West studs (as listed by Dave and others) such that he should get a special “adjustment” for difficulty level? For the sake of keeping things relatively simple, we will just look at 2011 matchups. (I just used MLB.com’s pages, which do not have PA listed separately, in a convenient way, so “plate appearance” should be understood as AB+BB for the purposes of this post. Yes, this probably led to an insignificant arithmetical error or two.) Tulowitzki had 596 plate appearances (AB+BB) in 2011: six versus Linceum, six versus Bumgarner, nine versus Cain, nine versus Kershaw, and 10 versus Latos. That is a total of 40, just under seven percent of his total plate appearances.
While that might be a slightly greater proportion of good pitching than hitters for other NL teams and divisions may have faced, given that 93 percent of his plate appearances came against the “rest” of the league (I know it gets complicated because of interleague, but let’s just stick with the basic premise for the sake of simplicity), I do not think that requires us to make NL West hitters like Tulo a special case. We do not take 40 plate appearances to be a significant sample for almost anything of which I am aware of off the top of my head.
How about if we expand this off-the-cuff selection of great pitchers to include the Diamondbacks’ Daniel Hudson and Ian Kennedy, as Paul Swydan suggested? (With all due respect to Paul and the great years and Hudson and Kennedy had, I am not sure who would put them on the same “true talent” level as Kershaw, Lincecum, or Cain, but let’s humor Paul. Dude’s gotta be exhausted from running all those “After Dark” chats.) That adds another 16 plate appearances to Tulo’s list of “stud” opponents. Again, without doing the math, I am pretty sure that those 16 plate appearances do not push things over into “cancel that league adjustment” territory, as the total is still under ten percent of Tulowitzki’s plate appearances.
What about Longoria not having to face the Rays’ staff? That’s a bit trickier, since we do not have the numbers for how many times Longoria would have faced individual Rays pitchers in an alternate universe. With a bit of digging, I found an AL East hitter who faced the Rays quite a bit: Dustin Pedroia. Now, I was not given a list of “official studs” on the Rays’ staff, but for 2011, I will count David Price and James Shields. Pedroia had 721 plate appearances (AB+BB) in 2011; 16 versus Price, 17 versus Shields. That comes to a total off 33, or less than 5 percent of Pedroia’s plate appearances.
Oh, I’ve forgotten one Tampa Stud: Pedroia also faced Matt Moore twice, pushing things all the way up to… 4.8 percent. Obviously, if I am no inclined to accept that Tulowitzki’s proportion of stud pitchers faced should change the way we thing of his “difficulty adjustment,” I am not going to be doing the same for Longoria based on an even smaller proportion of hypothetical plate appearances.
[Side note: Longoria faced CC Sabathia 16 times, more than Tulowitzki faced any one of the NL West pitchers listed above.]
The 2011 National League West (and the San Francisco Giants in particular) did feature a number of tremendous pitchers, and it is understandable that one would be tempted to judge the hitters in their division on a different grading curve. However, even taking both those pitchers that Tulowitzki faced (including Hudson and Kennedy) and that Longoria did not face together, we are talking about less than 80 plate appearances between two players versus more than 1000 PA versus of all the other pitchers they collectively faced during the season. If someone wants to do a more detailed and mathematically rigorous account of that proves otherwise, that would be both great. Until then, I do not think Tulowitzki and other NL West hitters get bonus points, or at least not a significant number of them.
Our current measures of opponent strength are imperfect and somewhat crude, and can probably be improved upon. I can understand the why Dave and others want to note that hitters in the 2011 NL West faced a great number of excellent starters. But as we have seen, a hitter like Tulowitzki faces a group of seven starters less than 60 or 70 times a season, and given all the other evidence about the league-wide skill level of the far greater proportion of hitters he faces, those league-wide evaluations likely come closer to the truth about the difficulty level faced by a individual hitters. In any case, Tulo is a great player, he doesn’t need the extra credit.