Fun With the Lineup Tool

A funny thing has been happening lately. Not really funny ha ha, you understand, but more funny strange.

The funny thing that has been happening is a sudden, oddly specific fascination a certain percentage of the Cardinal-observing world seems to have developed in the topic of who exactly will be leading off for the Redbirds in the upcoming season. What's so funny, so odd, about this development is how specific it is; it seems nearly every major outlet for the following of the Cardinals has put out some piece of speculation regarding El Birdos' leadoff situation.

I suppose, in a way, it's understandable, of course; we've reached the point in the offseason where there are very likely no further moves coming, aside from perhaps some last-minute minor league deal with a spring training invite to some middle relief type. In other words, the roster we see the Cards with right now is, in all likelihood, the roster with which they will attempt to defend their NL Central crown, at least for the early stages of the season.

Given that fact, it's probably not all that surprising that the predominant speculation about the roster has moved on from the particulars of how it will be composed to how it will be arranged. After all, if you can no longer argue about the things you have, you can at least argue over how to place the things you have.

It seems odd, however, to me that the leadoff spot in the lineup specifically has become such a focal point for debate. I suppose perhaps it has something to do with Kolten Wong announcing he would like to bat in the leadoff spot this coming season, but really should get the job in spring training to get used to it, to which I would reply to Kolten, how about getting on base at a reasonable fucking clip before you start demanding you be allowed to sink the offense with your career .303 OBP, but it probably also has a lot to do with the fact that the best hitter the Cardinals have at this moment also happens to be the guy hitting leadoff. Oh, and also that whole thing where Matt Carpenter, leadoff hitter and bell cow for the Redbird attack in general, might have turned himself into a power hitter right in front of everyone's eyes, and in an offense as starved for runs as the Cardinals of 2015 having a 25+ home run guy hitting first seems like an almost shockingly wasteful luxury.

And so the Cardinal leadoff spot has become a source of consternation, discussion, and speculation. Leaving Matt Carpenter there has been suggested. Putting Kolten Wong there has been floated. Stephen Piscotty has been suggested, as a somewhat outside the box solution, while Matt Holliday was put forth in these very electronic pagesas a further outside the box solution.

Well, luckily for those of us who enjoy the discussion -- or perhaps just those of us looking for an easy, breezy (beautiful) idea for a Sunday-morning blog post -- about lineups, there's a thing invented on the interwebs to help us out with our endless speculation. We have the Baseball Musings Lineup Analysis Tool. Colloquially known as just 'the lineup tool', there are hours of fun to be had playing around with said tool.

For those of you who don't know how the lineup tool works, you just type in the names of the players likely to be in your lineup, their OBP and slugging percentages, and then voila! the power of math takes over and tells you the numerically best way to order your deck chairs. Perhaps even more entertainingly, the lineup tool will also run some calculations and tell you approximately how many runs a given lineup should score on a per-game basis. Of course, the changing offensive eras makes the runs scored portion of things a little less reliable than something you maybe want to build your hypothetical baseball franchise entirely on, but hey, it's still pretty fun to play around with.

So what I did was pull up the lineup tool, and start plugging in numbers for the Cardinal offense in 2016. I went with my best guess as to the starters at each position, and used the ZiPS projections for 2016. Just for fun, I ran the simulation with both Tommy Pham and Randal Grichuk in the role of starting center fielder, since that is the position I personally feel most interested in, in terms of a couple players possibly competing for time. (Though I understand the organisation's preference right now is Grichuk, for a variety of reasons I tend to only somewhat agree with.)

I also used Yadier Molina as the catcher, rather than messing around with part-time players. I know Yadi is going to miss a fair amount of time this season, but still, he should be the guy taking the majority of the plate appearances once he gets healthy. And finally, I just pulled Jaime Garcia's numbers to represent the pitchers.

The lineup I put in, using Randal Grichuk, looked like this:

1. Matt Carpenter
2. Stephen Piscotty
3. Matt Holliday
4. Brandon Moss
5. Jhonny Peralta
6. Randal Grichuk
7. Yadier Molina
8. Kolten Wong
9. Jaime Garcia

That particular lineup I entered is, well, not great, but not exactly terrible, either. The Cardinals, as currently constructed and with ZiPS being infallible, would score just 4.145 runs per game utilising that lineup. That translates to 671 and a half runs in 2016, which would actually be a decent little bump of 24 runs over the 2015 offense. Not a huge improvement, but every little bit counts, right?

On the other hand, the idealised version of that lineup would score 4.176 runs per game (so, you know, not a huge difference), and would look like this:

1. Holliday
2. Carpenter
3. Peralta
4. Piscotty
5. Moss
6. Wong
7. Molina
8. Grichuk
9. Jaime

Interesting to note Holliday slots into the first slot, suggesting perhaps Joe knows what he's talking about. In fact, of the top ten best lineup arrangements, Holliday hits first in nine of them, with Carpenter right behind him. In the only variant that doesn't have Holliday in the leadoff spot, Carpenter hits first and Holliday hits second. I guess it's just as simple as wanting your two best hitters to hit the most, huh?

Anyway, I then moved on to my Tommy Pham version of the lineup. Pham doesn't pheel like the same type of hitter to me that Grichuk does, so I actually slotted him into that number two slot.

1. Carpenter
2. Tommy Pham
3. Holliday
4. Moss
5. Peralta
6. Piscotty
7. Molina
8. Wong
9. Jaime

By the way, isn't it weird that the Cardinal lineup has suddenly become so insanely right-handed, after much of last season having five of eight starters in the lineup batting from the left side?

Anyway, the lineup above scores just 4.102 runs per game, a little over a third of a run less per contest than having Randal Grichuk starting in center, so at least by this measure it would seem Randal wins out. However, I feel the need to say I don't think Pham's projection of a .314 on-base percentage and a .411 slugging is all that likely to occur, and Tommy's circuitous path to the majors probably makes him extraordinarily difficult to project in general. If pressed, I would say he beats that OBP by 40 points, at least, but hey, the numbers are what the numbers are.

The optimal Pham lineup scores 4.139 runs per game, and goes like this:

1. Holliday
2. Carpenter
3. Peralta
4. Piscotty
5. Moss
6. Wong
7. Molina
8. Pham
9. Garcia

Again, interesting to note how large the difference is in how I see Tommy Pham (as a two-hole hitter), and where the projections place him. It should be obvious by now, also, but I feel I should point out the lineup tool has no way of knowing batter handedness, and so take the results with a grain of salt, in that even optimising the lineup in this way you would still probably try to alternate lefties and righties a bit more.

But, damned if Matt Holliday isn't still the leadoff hitter! Fascinating.

Now, just for fun, I felt like doing something a little wacky, so I decided to see what the club would look like offensively if we were already in the NL DH era. Rather than retyping a bunch of names and numbers, I just took the Pham lineup and plugged Grichuk in for the pitcher's spot.

And, well, it made a difference. A big one. The lineup arranged exactly as my Pham version above, but with Grichuk in place of Jaime Garcia, scored 4.639 runs per game. That's over half a run more, and translates to 751.5 runs over the course of a 162 game season. The optimised version is even better, scoring 4.730 runs per game, and goes a little something like this:

1. Holliday
2. Carpenter
3. Pham
4. Piscotty
5. Moss
6. Wong
7. Peralta
8. Grichuk
9. Molina

So...there's something weird about this lineup. Do you see it? It's the fact that Tommy Pham is now hitting third, for some reason. I honestly don't know enough about lineup theory minutiae to understand why that happened, but it did.

Oh, and Holliday is still leading off.

Finally, just for real shits and giggles, I thought I would do something really fun. So, I ran the DH version of the Cardinal lineup, but instead of ZiPS projections I just plugged in the numbers each player produced last season. So it went:

1. Matt Carpenter .365/.505
2. Tommy Pham .347/.477
3. Matt Holliday .394/.410
4. Brandon Moss .304/.407 (his projection was actually better)
5. Jhonny Peralta .334/.411
6. Stephen Piscotty .359/.494
7. Yadier Molina .310/.350
8. Kolten Wong .321/.386
9. Randal Grichuk .329/.548

Yes, I was still too lazy to move Grichuk, even with having to change a bunch of the fields already.

Before I tell you how many runs the above lineup would score, I'll tell you it optimised thusly:

1. Holliday
2. Carpenter
3. Pham
4. Piscotty
5. Grichuk
6. Wong
7. Peralta
8. Moss
9. Molina

Which, I have to tell you, makes me sad to see Yadi all the way down as a number nine hitter.

The horribly unoptimised version of the line with Grichuk hitting ninth would score 5.074 runs per game, good for roughly 822 runs in a season.

The best version of the lineup? It would score 5.251 per game, or ~861 runs over a 162 game season. Which, interestingly, would have made the Cardinals the second-best offense in all of baseball last year behind Toronto. Of course, that's hypothetical runs, and based on a run-scoring model that seems a little shaky in our current low-offense environment according to the guys who actually developed the lineup tool.

Still, though. I would take that. Wouldn't you?

The real takeaway here: Matt Holliday is apparently the only choice to hit leadoff. Hopefully as a DH.