Tuesday, March 22, 2011

Taking a Stab at Lineup Optimization - Twins Edition

There has been a lot of blog-angst in recent years about how Ron Gardenhire has figured his daily lineups, most often about the first three spots in the lineup. It's almost impossible to predict what he'll do on a day-to-day basis and it's a safe bet to say that he doesn't subscribe to every rule in the unwritten lineup optimization handbook (actually, there is a "book", but it wasn't written until 2007). Almost everything you read about lineup optimization will tell you that over the course of a season it will probably only account for maybe 1 win, or a few runs, but hey, we were shown in 2008 and reminded in 2009 that 1-win can decide whether our beloved team goes to the playoffs or not. I'm not going to go into any calculations, but rather just use the general guidelines that were laid out very nicely by Sky Kalkman a couple of years ago; I'll be quoting his post liberally from here on.

1. Lead-off (candidates include Denard Span and Tsuyoshi Nishioka)
According to 'the book', the lead off spot is about two things: Speed and On-Base %. You want this hitter to be able to get on base frequently and advance, via steal, hit-and-run or bunt, especially late in games when you're trying to manufacture the tying and/or winning runs. You also don't want this hitter jamming up the bases with power hitters coming up behind him. This is why you wouldn't put someone like Jim Thome in the lead-off spot, despite his high OBP. For the Twins, they have a couple of options for lead-off hitter, including Denard Span and new-comer Tsuyoshi Nishioka. Personally I think D-Span should get a shot at it before Nishi. In 2008 and 2009, Span was an OBP monster (.387 and .392 respectively). He was also a threat on the base paths, making him an ideal lead-off man. Last year he struggled mightily, thanks in large part to an uncharacteristically low BABIP, which has made many doubt Span's lead-off capabilities. Even if he rebounded only somewhat this season, to around a .350 OBP, that would make him a good lead-off candidate because he still possesses a good eye for stealing bases (26 out of 30 attempts last season). I'm not saying you couldn't put Nishioka here, but I think he has to prove himself a little before you replace the guy who's had the job for past two seasons. If we were really going "by the book" here, I'd have to nominate Mauer for the lead-off, but that's a little too progressive even for me.

2. Two Hole (candidates include Joe Mauer and Tsuyoshi Nishioka)
Kalkman's write up says the following about the ideal 2-hold hitter:

"The Books says the #2 hitter comes to bat in situations about as important as the #3 hitter, but more often.  That means the #2 hitter should be better than the #3 guy, and one of the best three hitters overall.  And since he bats with the bases empty more often than the hitters behind him, he should be a high-OBP player."

The system that Gardenhire seems to operate under is what would be considered the "old-school" of thought which basically says that the 2-hole guy's job is to "set the table" (get the leadoff runner over) for the 3rd and 4th hitters. Knowing that, I fully expect him to put Nishioka in the 2-hole, when really it's Joe Mauer he should have there. With Mauer's high OBP (.407 career mark), he makes an ideal 2-hole hitter because of all the people on the team, he's the most likely person to get a hit or take a walk. Just think, if you had Span getting on-base at a .350 clip and Mauer getting on-base at a .400 clip, that equals a lot of runners for the 3-5 guys in the lineup to knock in. The reason it doesn't make a whole lot of sense to put Nishioka in this role is because when you do that, you negate his speed. If Span gets on via a single, and Nishi comes up and walks or singles, it's unlikely you're going to execute a double-steal so you've negated some of Nishi's value, he's better off in the 9-hole. 

3. The Third Spot (candidates include Joe Mauer and Danny Valencia...go with here for a minute)
Ron Gardenhire will almost certainly put Joe Mauer in this spot as he's done consistently over the past years. I thought it would be fun to pretend if Gardy didn't do that. I was looking at the players the Twins have that could fill a 3-spot role and after much consideration I came up with Danny Valencia. It's important to note two things about the #3 hitter.
a) this hitter comes to the plate with two-outs and nobody on more often than any other hitter and
b) this hitter, not surprisingly given the last point, comes to the plate with, on average, fewer men on base than the #4 and #5 hitters
In other words, the #3 hitter should not necessarily be the high-average, solid-power guy that old-school lineup composition rules would suggest. Valencia would fill the #3 role admirably, I think, especially if he can duplicate his performance from last year. He's a decent on-base guy (.351 mark last year and .353 career MiL mark) and he showed he has some decent pop as well. Another thing I like about him is his consistently high BABIP marks over his career. That tells me that he not hitting a lot of weak grounders that turn into double-plays, he makes good contact. He also doesn't strike out a lot... I'm just saying, it's something to think about. 

4. The Clean-Up Hitter (there is only one candidate for this job, Justin Morneau)
This is where old-school and new-school rules come together, for the most part. This is the most important spot in the lineup (probably tied with the 2-hole in terms of importance) as the clean-up hitter comes to the plate with, on average, the most chances to do damage. I cannot think of a better option for the Twins than a healthy Justin Morneau, and he's excelled in this role for a number of years. He's a decent average guy, he's a solid on-base guy and he's a great power guy. There really isn't much debate here. 

5. The Number Five Guy (candidates include Delmon Young, Jason Kubel and Jim Thome)
The book says this guy is the 4th most important  hitter in the lineup, after the 1, 2 and 4 positions. Old school would have you believe that this lineup position is for, as Kalkman put it, the "wannabe cleanup hitter." This position, for the Twins, gives one a lot to think about. The new-and-improved Delmon Young makes a good candidate and some combination of Jason Kubel or Jim Thome makes an intriguing option as well.
I think if Thome's in the lineup, he's a lock for the #5 slot. His combination of power and patience makes him a great on-base threat which is exactly what you want to keep a rally alive. I also think his all-or-nothingness is an attribute at this spot in the lineup because with him, you have a lower risk of the double-play...especially with the shift other teams put on him.
If Thome is not playing, I think Young or Kubel both make sense as the #5 guy. Both are higher average hitters (usually) and both have some pop in their bats. Kubel has a slight edge in the on-base department, but Young has a little more speed. Either is a better option than Cuddyer who's streakiness leaves something to be desired.

6. The Six Hole (candidates include Young, Kubel, Michael Cuddyer, Nishioka)
From here on out, who you put in what spot matters less and less. Any of the 4 candidates for this position would fit here in the 6th hole, though I suspect if Nishioka is not high in the lineup, he will be in the 8th or 9th spot. Young or Cuddyer make the most sense here. The general thinking is that from 6 to 9, you stack the hitters based on talent with the least talented hitter batting 9th. If I had to rank the remaining hitters, I'd probably put Delmon Young here in the 6th hole, followed by Cuddyer, Nishioka and Casilla.

7 - 9. Back end of the lineup (candidates include Nishioka, Cuddyer, and Casilla/Hughes)
If I were the Twins, with Nishioka and Casilla still left to slot, I would put Casilla at 9 and Nishioka at 7, to stagger them a bit since they are likely very similar hitters and both have a little speed. I would sandwich Cuddyer in the 8th slot because he has a little pop and a decent career on-base percentage.
So, after all that, my ideal lineup would look something like this:
1. Denard Span
2. Joe Mauer
3. Danny Valencia
4. Justin Morneau
5. Jim Thome/Jason Kubel
6. Delmon Young
7. Tsuyoshi Nishioka
8. Michael Cuddyer
9. Alexi Casilla

There's obviously a fair amount of room for debate here. I could easily see someone else besides Valencia in the 3 hole and there's a lot of play towards the back end of the lineup as well. I think having Mauer batting 2nd is very enticing simply because of his ability to get on base in front of the power the Twins have behind him. Just for fun, I ran the numbers at this site to see what it would come up with in terms of an optimized Twins lineup:

1. Joe Mauer
2. Jim Thome
3. Danny Valencia
4. Justin Morneau
5. Michael Cuddyer
6. Alexi Casilla
7. Tsuyoshi Nishioka
8. Delmon Young
9. Denard Span

A couple of caveats. 1) I chose Thome over Kubes. 2) It only has you input two data points, OBA and SLG. For Nishioka and Valencia I had to project a little, for the rest of the players, I just used their career averages.
I'm glad to see that the simulator agreed with my Valencia slotting, but having Mauer and Thome at the top of the lineup would be problematic in terms of log-jamming on the basepaths. Both of them are slow and thus not ideal in front of faster players. I probably should have adjusted Thome's numbers a little, his career averages look to be skewing things, and he's a part time player anyway, perhaps an average of his and Kubel's projected production. My problem with this "system" is that it doesn't account for speed. I know that the new "book" downplays speed, but you can't ignore it when you've got guys who are as slow as Thome.

It's a fun look anyway, I'd love to hear what other people think.