"Luis Castillo may bat leadoff for the Mets. Potentially, Carlos Beltran could bat second and Jose Reyes third. It may make more sense to bat Carlos Delgado fourth and David Wright fifth in that scenario, since otherwise the Mets would be stacking a ton of lefties at the bottom of the order."

Seeing as this quote has generated a lot of discussion on AA, I've decide to use the PECOTA forecasts as well as this handy tool to arrive at the Mets Optimum Lineup. So, armed with the A-Rod of beers (I consider it awesome, others see it as overpaid and overrated) and one of my favorite albums, let's do some posting.

Plugging the PECOTA projection, with a lineup that's inspired by the above quote, into the Lineup Analysis:

We get 4.873 runs per game, which is 789 runs scored over a full season. PECOTA has the Mets scoring 828. So, the lineup generator did a pretty good job considering baserunning (and the Mets have some real standouts) is not considered as well as pinch hitters, etc. However, more interesting is what the lineup generator spits out for the Mets optimal lineup:

Wow, this is a lineup even Tony LaRussa would have a hard time loving. This "optimal" lineup showcases one of the problems with this tool (besides the aforementioned baserunning consideration). The Lineup Analysis considers the statistical numbers of each batter as static. Therefore, it doesn't consider the impact on Delgado's numbers by having a pitcher batting behind him. I'm sure that somehow this is technically the Mets' best lineup, but why would any pitcher give Delgado a pitch to hit when a double play can be all but assumed. The tool only generates the top 30 batting orders with all of them having the pitcher batting sixth, so unfortunately more cannot be gleaned. Just for kicks, here's the worst lineup predicted:

Lucky for you guys, I decided to take a look at what happens when I toss in CHONE projections instead of PECOTA. The results are very intriguing.

This lineup will generate 4.819 runs per game and 781 runs over the season. But the real intriguing part comes from the optimal lineup with these numbers:

That's right: not only is the pitcher in his proper (debatable) spot, but Castillo is slated to bat first (in all of the top 30 lineups) and Reyes should bat third. Check it out for yourself if you don't believe me. Could it be that Omar, Jerry, or someone in the Mets front office did a similar study using CHONE numbers? Is it possible that Jerry says all of this "statistical numbers, need to learn to be clutch" stuff because he knows that's what the majority of fans as well as writers want to hear? Maybe he's covering his bases, knowing us Vorpies won't call for his firing if he says he hates numbers but then manages the way the numbers say to, while if he said Stats 0wn, there would be unneeded controversy... probably not, but I guess anything is possible.

Anyway, the main difference between the two projections is Castillo's and Reyes's OBP, which causes a huge impact on the lineup generation. Now, obviously we can debate about splitting up R/L, but to me the more important point is getting our best players the most ABs as possible and ignoring minor problems with the lineup distribution, especially if you consider that Tatis replaces Murphy vs LHP.

Another debatable point are the projections themselves, which I feel are a bit harsh on Church based on the second half of his season.

The last point I'll bring up has to do with Reyes. There has been a lot of talk about whether it would be a waste of Reyes's speed batting him third. Personally, I don't see how batting third would hurt Reyes's speed value. He can still steal when it makes sense to and I don't see Reyes catching up to Castillo/Wright when he hits a triple and they were on first. To me it's more important to recognize that while Reyes has speed, his OBP is somewhat lacking (for a perfect leadoff hitter) and he has more power than he is given credit for. Really, though, there is no right answer and the more important question is, is it Opening Day yet?