Mets new lineup order of Castillo, Beltran, Reyes?

http://www.nytimes.com/2009/02/19/sp...r=1&ref=sports

Manuel is planning to set Castillo as 1st, Beltran 2nd, and Reyes 3rd.

Now, I liked what he tried to do with the bullpen last year(though it didn't help), but is anyone else completely perplexed by this? Hanley is batting third but he has more power and less speed than Reyes, as Reyes himself admits, and it'll take away his SB totals.

Castillo leadoff is fine to start the season and hopefully give him a boost, but I don't see how you can bat Reyes behind Beltran :confused:

Hanley, Reyes, Rollins, and even Alfonso Soriano all don't have a perfect fit in the lineup, but their combination of power and speed makes me always bat them 2nd.

Of course, this all stems from the fact that the Mets have no true cleanup hitter, as David Wright is an ideal 3rd bat, and Delgado is a solid 5th, provided he can put up the numbers again.

Reyes, Beltran, Wright, MANNY, Delgado, Church, Castillo, Schneider/Castro would dispel any and all doubts of a playoff run for at least 2009, but i digress...

Sounds to me like a bad idea simply because it places too much pressure on Castillo to rebound. I would say bat the man sixth and let him get his confidence back...then move him to first if that's the best place for him.

Manny the Met? Oy.

http://www.fangraphs.com/blogs/index.php/oh-manuel/

Quote:

---

There is not a person alive who thinks Castillo is a better hitter than Reyes. Over the last three years, Castillo has gotten on base at a .358 clip, that’s slightly better than Reyes 0.355. Compare their slugging percentages, where Reyes’ lowest is .421 and Castillo’s highest is .362, and there’s just no contest on who should be getting more plate appearances.

If the Mets want to go to a more OBP conscious leadoff hitter, that’s fine. Hitting Reyes second or third in favor of Carlos Beltran makes sense. Throw David Wright behind those two and you have quite the top of the lineup. Batting Luis Castillo first is taking away plate appearances from your best hitters. Not enough to make a difference, but if Manuel is serious about winning, shouldn’t he be utilizing every asset of the game to its max in order to help his cause?

---

pretty much.

Yep. If this happens, poor move. Probably won't matter much in the grand scope of things, but Castillo should not be leading off for a team like the Mets. Reyes-Beltran-Wright-Delgado is probably ideal. Castillo should be hitting 7th or 8th...maybe 9th if you want to get creative.

See, but that lineup makes Delgado hit cleanup, which I do NOT want to see. No sane person can believe he's going to hit like he did in 2008 again, and his avg, obp, and hr totals are no longer good enough for a cleanup hitter for team with high aspirations.

The lack of a solid cleanup hitter is forcing everyone to play outside their "set of skills" and try and fit the mold, but this lineup is not the answer..

Actually, I wouldn't mind as much if Reyes just batted 2nd and Beltran hit 3rd

Perhaps no sane person would believe he's going to hit again like he did in 2008, but every projection system thinks that while he may not repeat, he'll still be a good hitter. ZiPS and PECOTA, not listed at FanGraphs, concur as well, so that makes 6 projection systems all projecting an OPS in the .810-.870 range. He probably won't be as good as 2008, but he won't be as bad as 2007 either. He'll be a solid 115/120 OPS+ hitter...not IDEAL for a clean-up hitter, but that's more than offset by maximizing the plate appearances of three elite hitters in Reyes, Beltran, and Wright.

So, I tried the Lineup Analysis tool, with Tatis/Church as the corner outfielders and a league average hitting pitcher (.177 OBP, .175 SLG). I used ZiPS projections.

Three different lineups came out with 4.996 runs per game, which was the highest this assortment of players comes up with. None of them are close to be conventional...

1) Luis Castillo, David Wright, Ryan Church, Carlos Beltran, Jose Reyes, Fernando Tatis, Brian Schneider, Carlos Delgado, Pitcher
2) Jose Reyes, David Wright, Carlos Delgado, Carlos Beltran, Ryan Church, Pitcher, Brian Schneider, Fernando Tatis, Luis Castillo
3) Luis Castillo, David Wright, Jose Reyes, Carlos Beltran, Ryan Church, Fernando Tatis, Brian Schneider, Carlos Delgado, Pitcher

This tool is blind to speed, obviously, so it's only a rough estimation...so make of it what you will.

It's interesting to see how the pitcher's spot can be moved around, but it's like in football where the coach wants to go on 4th and 1, but he's afraid of fan criticism.

However, it seems flawed on many aspects, including not being able to distinguish lefty/righty splits, speed as mentioned, etc.

And I feel like Brian Schneider's a 2nd pitcher's slot =/

The lineup analysis tool IS very basic. It's based just off the 9 hitters you input and their OBP and SLG. It's just something to use as quick estimate.

Quote:

---

Originally Posted by HoustonGM

The lineup analysis tool IS very basic. It's based just off the 9 hitters you input and their OBP and SLG. It's just something to use as quick estimate.

---

Uh, yeah...considering it spit out a lineup featuring a 6th-place hitting pitcher as one of the three "ideal" lineups, it's safe to say it needs work. :)

This is as a clear case of statistics losing to common sense as one could ever make...

Where is Howard Johnson going to bat?

Quote:

---

Originally Posted by oriole^

Uh, yeah...considering it spit out a lineup featuring a 6th-place hitting pitcher as one of the three "ideal" lineups, it's safe to say it needs work. :)

This is as a clear case of statistics losing to common sense as one could ever make...

---

The difference between those 3 lineups and some of the other ones is in the order of .001-.002 runs or so, well within any margin of error. :p

Quote:

---

Originally Posted by HoustonGM

The difference between those 3 lineups and some of the other ones is in the order of .001-.002 runs or so, well within any margin of error. :p

---

This was going to be my question. For any team, does it ever matter what order you put the players in the lineup. If you had the pitcher hitting leadoff, and then moving to your next worst hitter and so on, what would that get you?

A lot of plate appearances for your worst hitter, and in real life, allowing the pitcher to set in comfortably from the start.

Quote:

---

Originally Posted by kenny1234

This was going to be my question. For any team, does it ever matter what order you put the players in the lineup. If you had the pitcher hitting leadoff, and then moving to your next worst hitter and so on, what would that get you?

---

Does it matter? Absolutely. The top of the order sees more plate appearances that the bottom of the order, so if you're maximizing the plate appearances of a pitcher, you're going to negatively impact your run scoring ability. Generally, the difference between the best and worst lineups that would realistically be used is minimal, in the order of 5-10 or so runs a year (off the top of my head from a study I read in Baseball Between the Numbers)...but if you "deoptimize" the lineup as much as possible, such as batting them in reverse order of their ability, it's going to cost a lot.

Using the same players I used above, the worst lineup comes out to 4.570 runs per game - Tatis, Pitcher, Church, Schneider, Castillo, Beltran, Delgado, Reyes, Wright. The difference comes out to 69 runs in a season, which, using the established conversion of 10 runs = 1 win, is a difference of 7 wins between the best and worst possible lineup.

So, lineup order could have an immense impact, but most commonly accepted lineup structures are within 1 or so win of each other. Teams should, though, always optimize as best as they can, because any gain of a couple runs here or there is good.

Quote:

---

Originally Posted by HoustonGM

Does it matter? Absolutely.

---

Sorry, I realized that it probably matters in real life - I was interested in the sensitivity of the Analysis tool. My feeling is that such a tool is far too imprecise to be useful - the difference of 10 runs a year is worth it, but just using OBP and SLG isn't going to provide results that are accurate enough to differentiate lineups that are within 10 runs of each other.

What this tells me is that everytime a manager says that they are going to make a particular lineup choice based on player preferences, such as Soriano hitting leadoff, that the difference in productivity may easily be less than the unmeasurable boost caused by player happiness.

Quote:

---

Originally Posted by kenny1234

Sorry, I realized that it probably matters in real life - I was interested in the sensitivity of the Analysis tool. My feeling is that such a tool is far too imprecise to be useful - the difference of 10 runs a year is worth it, but just using OBP and SLG isn't going to provide results that are accurate enough to differentiate lineups that are within 10 runs of each other.

---

Of course, it's just a tool that produces a VERY rough estimate.

Quote:

---

What this tells me is that everytime a manager says that they are going to make a particular lineup choice based on player preferences, such as Soriano hitting leadoff, that the difference in productivity may easily be less than the unmeasurable boost caused by player happiness.

---

Perhaps.

Quote:

---

Originally Posted by HoustonGM

The difference between those 3 lineups and some of the other ones is in the order of .001-.002 runs or so, well within any margin of error. :p

---

Which misses the point completely. When numbers are produced which have no basis in objective reality, and you aren't in a math or an astrophysics course, then it doesn't matter what the margin of error is, nor any optimization, nor any conclusion made from it. You simply discard the numbers, and disregard the tool.

This isn't something that produces a "very rough estimate"...it produces obvious nonsense.

It provides an estimation of a lineup order's run scoring ability based on its players' on-base percentages and slugging percentages. It's not "nonsense."

Quote:

---

Originally Posted by HoustonGM

It provides an estimation of a lineup order's run scoring ability based on its players' on-base percentages and slugging percentages. It's not "nonsense."

---

No, I'm sorry, it is, and you are wrong. If something tells you to bat a pitcher sixth, it's clearly, obviously, and unequivocally incorrect, and therefore unreliable as a means of predicting or estimating anything.

Quote:

---

Originally Posted by oriole^

No, I'm sorry, it is, and you are wrong. If something tells you to bat a pitcher sixth, it's clearly, obviously, and unequivocally incorrect, and therefore unreliable as a means of predicting or estimating anything.

---

It's not "telling" you to bat a pitcher sixth. It does not know what positions the players are. The only things it knows are 9 pairings of OBP and SLG. It just gives you an estimation of how many runs a specific ordering of the 9 OPS's would score per game. You're free to think that the entire tool is nonsense because of one result related to one group of players, but the tool has its uses for providing a rough estimation.