In and of itself, the way batters are ordered in the lineup doesn’t really matter. It might be the difference of a win over the course of a season. It’s more likely a few runs added. But ignore enough things that don’t matter in isolation, and there could be problems. At the same time, paying heed to all of the supposedly insignificant things can add up to a bigger positive—even if it’s just a slightly modified way of perceiving baseball or a grudging encounter with inherited wisdom.
Following Ryan Freemyer’s article from last weekend about getting the most out of Jorge De La Rosa, I thought it would be a good time to revisit lineup optimization. I first take a look at the Rockies’ lineup splits over the past two years under Walt Weiss. My go to statistic here is weighted runs created plus (wRC+), which is park and league adjusted and measures how much better or worse a batter was compared to league average (100 is average). We’ll use these figures to detail what is and is not optimal. I’ll then offer my ideal lineup.
The Rockies did not have a good offense in 2013. The team’s wRC+ was 88, which is hitting at a clip twelve percent below league average. That mark ranked 26th in baseball. In a weak year at the plate for the Rockies, Walt Weiss didn’t do the team any favors in constructing the lineup.
A key component of lineup optimization is to get your best hitters as many plate appearances as possible. The collection of batters who garnered the most plate appearances throughout the season hit eight percent below league average. The default choice for the number one spot was Dexter Fowler. He did quite well. Fowler had a wRC+ of 108 from the leadoff spot in 2013. What sunk the slot below league average was all of the plate appearances given to Eric Young Jr., D.J. LeMahieu, and pre-breakout Corey Dickerson. Fowler did miss time due to injury, so this one’s forgivable.
The same cannot be said of the two slot. In 2009, Sky Kalkman at Beyond the Box Score summarized lineup optimization according to a sacred sabermetric text, The Book. He indicated that the team’s best hitter should bat second. In 2013, the Rockies’ production from the two-hole was the worst on the team. The wRC+ outcome was tied with the seven spot, but the second spot had 91 more plate appearances. In 2013, the second spot of the Rockies’ lineup didn’t hit all that much better than the ninth. LeMahieu and Josh Rutledge got the most plate appearances from this spot, and both also had on base percentages under .300 there.
Stacking the top of the lineup with below average production can’t but prevent the well-above average hitting from slots three through five from creating runs. That goes not just for the lack of traffic on the bases for the best hitters, but also because the hitters in those two slots received more plate appearances. The six through eight holes are less significant when optimizing a lineup, so the only crucial observation is that LeMahieu and Rutledge should have hit anywhere there instead of in the two-hole.
In terms of lineup optimization, Walt Weiss displayed improvement in his second season as Rockies manager. Part of the improvement was in the general improvement at the plate. The Rockies finished the season with a 98 wRC+, which was good for 13th in baseball. There is not a lot to take issue with here.
Again, though, the quibble is with the leadoff position. In 100 percent of the games Charlie Blackmon started, he hit leadoff. Charlie Blackmon is an average hitter. In fact, Blackmon’s wRC+ in 2014 was 100—average, defined. The smattering of other leadoff hitters, such as Corey Dickerson, Drew Stubbs, Brandon Barnes and Josh Rutledge, kept the collective wRC+ of the leadoff spot below average.
I’m happiest with the personnel Weiss chose to hit in the two-hole. Among them were Michael Cuddyer, Corey Dickerson, Nolan Arenado, Carlos Gonzalez (twice) and Stubbs. But they also included guys who shouldn’t sniff the top of a lineup card, such as Barnes, LeMahieu Charlie Culberson (once), and Rafael Ynoa.
It’s better, but there is still room to improve. Almost 400 plate appearances, for example, were given to Justin Morneau, Wilin Rosario, Corey Dickerson, and Nolan Arenado in the sixth spot. In isolation, it would have been better to take whoever was in the leadoff spot, bat them sixth, and move everyone else up one. But it’s not like that would have been the difference between playoffs and 96 losses.
So we have seen that from 2013 to 2014, the Rockies improved with regard to the second spot in the lineup, but there’s still room to be better at the leadoff spot.
This is how I would do it.
|7||Nick Hundley/Michael McKenry|
|7||Nick Hundley/Michael McKenry|
We’ll start with the lineup against righties. Corey Dickerson is a natural fit to hit leadoff because he gets on base a lot. It’s really that simple. He was tied for second on the team with Justin Morneau in on base percentage, and his wRC+ trailed only Troy Tulowitzki. Tulowitzki, the team’s best hitter, should hit second. From there, it’s a matter of stacking the best hitters at the top of the lineup. That means getting CarGo, Arenado, and Morneau as many plate appearances as possible.
Like Dickerson, Charlie Blackmon helps bring the best hitters to the plate with the most traffic on the base path. He can just do it from the ninth spot. In the Sky Kalkman article I linked above, Kalkman notes that batting the pitcher eighth is a "smart, albeit insignificant, strategy" that might produce a few extra runs a year. Tonly LaRussa employed that strategy to put traffic in front of Mark McGwire. In the lineup against righties, Blackmon essentially serves as a second leadoff hitter. Once the pitcher makes an out by looking foolish, the opposition will then face Blackmon, Dickerson, and the murderer’s row of "O" inflected names. While it does hurt to give more plate appearances to pitchers, that spot becomes a de facto designated hitter after the fifth or sixth inning anyway.
Much of the same logic applies for the lineup against lefties. Corey Dickerson does not need to be platooned, so he’s worth a lineup spot against southpaws. But his 2014 on base percentage against lefties was just .306. Stubbs, on the other hand, had an on base percentage of .395 against left-handers, which contributed to his 148 wRC+. He fits atop the lineup here. Rosario fits in the three spot for the same reason. Dickerson is a decent option to bat ninth against lefties, but I went with LeMahieu because I think he is a candidate to improve next year. I’m hanging my hat on Rob Arthur’s idea that improvement can be predicted by mapping the tendency for pitchers to avoid the middle of the strike zone when throwing to a particular hitter, and LeMahieu saw one of the ten largest increases in zone distance over the course of 2014. If LeMahieu can manage to get on base more and hit more doubles, as Arthur's method predicts, then he fits perfectly well as the second leadoff hitter.
I’m hopeful that the Rockies will, at least, take advantage of their best hitters and bat them second. ROOT has only televised two spring training games, but in both of them Drew Goodman and Ryan Spilborghs made a point to say that we might see the louder bats hit second this season. They might know more than we do, because to me it sounded like they were trying to gently modify the audience’s preconception of what the two-hole hitter should look like. They’ll know what the deal is when they see, for example, Troy Tulowitzki there.
Hoping to see the Rockies bat the pitcher eighth for the season is hoping for too much. It’s too bad, because lineups like the ones I suggest will soon be relics of the past. Given that the current CBA expires in 2016, MLB’s desire for more runs, the necessity of daily interleague play, and the agreeability of creating more jobs on major league rosters, I have a gut feeling that this year will likely be either the final or penultimate season without the designated hitter in the National League. For that reason, I want to see these lineups even more.
So: Go get those few extra runs, Rockies. Who cares whether or not it really matters?