It’s time for the annual lineup assessment based on Runs Created (RC). Why RC? Well, I like it because it doesn’t discriminate, it isn’t based on what other players do, and it isn’t position dependent. It is also relatively easy to understand where the number came from, unlike some other statistics like VORP. But the best part about it is that we can use one single number per player to compare their hitting abilities. RC comes via a (relatively) simple formula using some of the easier to understand counting stats, and it helps to predict how many runs each specific player would be responsible for making. The formula is as follows:
Making this a realistic predictor of total runs that the team will score is a bit problematic. Besides attempting to predict how many plate appearances each player would have, you also have to do this for every pitcher that comes to the plate. The more variables we throw in here, the more inaccurate we’re going to get. Rather than approach it that way, I’d rather see, based on predictions, how many runs the different players would create if given the opportunity to bat every day. Again, since this is a philosophical exercise, I’m going to assume everyone gets 650 plate appearances. Originally I had this set at ABs, but the more PAs one has, the more runs they create, and someone like Nick Johnson, who has alot of walks, would have an artificially high number of PAs if he has the same number of ABs as everyone. Since this is an exercise, I’m just going to make everyone have 650, even the catchers. That will never happen, but we’re using our imagination here. We’ll look at how everyone fares based on 3 sets of predictors and see what the ideal lineup.
The first “predictor” isn’t a predictor at all, it’s their last season’s performance, again extrapolated out (or in) to 650 PAs. For Nick Johnson’s case, we’ll look at ’06, but Guzman’s limited ’07 data is enough to work with. The second predictor is Baseball Prospectus’ PECOTA stat predictions for 2008. As they say “PECOTA is BP’s proprietary system that projects player performance based on comparison with thousands of historical player-seasons.” So there. Finally, I will use the 2008 player predictions from Bill James’ Baseball Handbook.
That is a list of some of the guys who should be mainstays in the lineup, in descending order of RC according to Baseball Prospectus. They see Kearns as the biggest run producer on this team and that may not be too much of a stretch. Especially if he returns to form. Nick Johnson is expected to do alot for this team, and past performance indicates that if he is healthy, he can. Zimmerman’s numbers should go up as he has more people to knock in. As for some big ups and downs, not surprisingly they think Milledge, Pena, and Dukes will all do much better. Lopez is expected to have a bounce back season, as all of us are hoping. Also not a shocker, they don’t think Christian Guzman’s numbers from early last year will be repeated. In terms of the optimal lineup with this, here is what I found:
Estrada and Lo Duca can be argued about here, which one is better is dependent on which source you go with. And their average score of the 3 sources is the same, but since Estrada take 2 of the 3, I’m going with him. They are relatively interchangeable though, so try not to sweat it too much. I don’t think Wily Mo’s numbers translate so well to this exercise because of his dramatic righty/lefty splits. He’ll sit against some guys, which will bring down his total RC, and whoever bats in his place probably won’t be as good as he is against lefties. Anyway, just something to keep in mind. It’s telling as to how much improvement BP and Bill James see when we see that they predict the lineup to produce 50-80 more runs this year than last. Now if we look another potential lineup, with Guzman at SS, Dmitri at 1B, and Lo Duca at C, it looks like this.
Remember, even if these predictor stats were on the money, this wouldn’t be the teams totals. . We gotta bat pitchers, there’s the Wily Mo’s issues, catchers don’t play as much, and lots of other things. But the message is still important here. What you’re seeing is a loss of 20 runs, which over the course of the season, would translate to a few less wins. Maybe it’s not so applicable to this team this season, but a few less wins is a big deal if you’re in a playoff race. Last year 2 wins separated the Yankees and Red Sox for the AL East, and Milwaukee, New York, and San Diego all missed the NL playoffs by 2 wins or less.
So you can see what the best lineup is, and any time they are using anything else, they may be hurting their chances to win games. Coincidentally (I think) the Bill James predictions are the same as what the totals would have been last season with proper at bats. This lineup based on ’07 stats is clearly “better,” but that is thanks to Christian Guzman, and if he can hit like he did last April, he should starting. Just don’t bet on it.