It's that time of year, when we crack out the spread sheets and guess what the ideal batting order should be. It's time for rosterbation. Joe Maddon, of course, will probably use a new line up in 150 of the 162 games, based on hitter profiles and pitcher types, but there's a few locks to set us up, and it's nice to have ideals.
Maddon confirmed to reporters his intention to use Desmond Jennings and David DeJesus in a righty-left platoon at the top of the order this season, as we saw him use down the stretch. Furthermore, Maddon has become a huge fan of batting Yunel Escobar ninth:
"There are better people before the nine-hole from the right side that we have not had before," Maddon said. "I'd probably be inclined to keep him more in the nine-hole regardless, just because I like him there and I think he does good work and he feeds into the top of the batting order, but that's not finalized yet. Against a right-hander, probably, but I'm not there yet." [tbo.com]
From there, Maddon is again reconsidering where to bat everyone else. Gotta love it.
Every discussion about building a batting order should start with Tom Tango, Mitchel Lichtman, and Andrew Dolphin's study on the importance of each batter, which can be found in The Book. The basic strategy, as savvy readers will already know, is to bat the most important hitters in the lineup second and fourth. The research shows that the best balance between more plate appearances and the highest leverage plate appearances (re: more runners on base) occur in these two slots.
For the same reasons, the clean up and lead off hitters are considered more important in a standard line up than whoever is hitting third. I was not always convinced that third is a less advantageous opportunity than fourth, but some research by R.J. Anderson persuaded me. Maddon may still mix up his line up and move Longoria to the third slot, though, based on an appeal to player mentalities, or how he treats his lead-off situation.
Put these rankings together, and your batting order by talent looks like this:
4, 2, 5, 1, 3, 6, 7, 9, 8
The prototypical lead-off hitter has contact and speed, and the guy batting ninth should probably be fast and have some ability to get on base in front of the best hitters when the line up comes around.
C: Ryan Hanigan, Jose Molina
INF: James Loney, Ben Zobrist, Evan Longoria, Yunel Escobar
OF: David DeJesus, Desmond Jennings, Wil Myers, Matt Joyce
UTIL: Logan Forsythe, Brandon Guyer, Sean Rodriguez
Yesterday afternoon we learned Guyer was the 25th man for the Rays roster. The only bench players with options are Forsythe and Rodriguez, and the only real challenger was Jayson Nix. Waiver claim Jerry Sands has already been optioned to the minor league side, and Wilson Betemit is not currently challenging anyone with his defense, but he'll fulfill the Shelly Duncan role from last year well enough.
The Front Five
With lead off and ninth hitter essentially finalized headed into the season, the Rays need to determine batters 2-5. To do so, let's look at some overall performance expectations for the best on the team. As a reminder, the final column will be a linear regression. PECOTA uses True Average, which scales to batting average, and Fangraphs uses wOBA, which scales to on base percentage.
Here are the PECOTA projections for the Rays' top four hitters:
Likewise, here are the Fangraphs hybrid projections, which overlays Steamer and ZiPS:
As mentioned above, The Book would argue the best hitter should bat second, but there are legitimate reasons to bat your best a litter further back in the order: player mentality, and batting men around the bases.
Longoria underperformed with men in scoring position last year, but is easily the best hitter on the team. I will continue to want him batting with as many guys in front of him as possible, so I will vote Longoria takes the No. 4 slot. From there, I will continue to love the switch hitting Zobrist batting No. 2 overall. Where you'd slot Myers and Joyce is a more interesting conundrum.
Splits and Projections
Fangraphs has a depth chart that presents a hybrid of Steamer and ZiPS, and PECOTA has it's percentile projections, but neither considers regressed splits for hitters. This aspect is rather important for projecting the Rays roster, as up to all three of the bench fielders are expected to rotate into the Rays line up against southpaws this season.
The three positions expected to be platooned are David DeJesus in left field, Matt Joyce playing out of the designated hitter spot, and if he struggles with lefties like he has in his pre-2013 seasons, James Loney at first base.
Brandon Guyer is the natural platoon partner for David DeJesus, as he would bring quality defense in the corners and speed on the base paths. Although, there is an argument for Logan Forsythe to get his turn in the outfield when he makes catches like this. Forsythe had a .439 wOBA and 187 wRC+ vs lefties over 99 PA's in 2012, and then was injured in 2013 and gave an underwhelming offensive performance. It's hard to know where to put your stock. Projections favor Guyer, but maybe the Rays and some more playing time can revive Forsythe's 2012 performance.
Joyce and DeJesus are locks to be platooned, but the question here is whether Loney will need it. The Rays clearly value his defense, so chances are he's starting 95% of games, but there is a possibility he needs more help this year than last, particularly if you believe his 2013 was an outlier.
With all that in mind, here are our regressed split projections for 2014:
| Proj. wOBA
| Proj. wOBA
Platoon splits numbers were generated by a tool developed by Ian Malinowski and Jason Hanselman that incorporates Bojan Kopravica's splits regression research, and ZiPS projections.
League average wOBA was .314 vs. RHP in 2013, and .312 vs. LHP.
The Batting Order
All of this leads us to the batting order. Using the quotes from Maddon, I've locked the lead off and ninth positions to what was used at the end of last year, chose fourth as the ideal place for Longo to bat players around, and refused to bat Zobrist lower than third.
| Proj. wOBA
| Proj. wOBA
The projection tool scored Loney low enough that we've removed him from facing left handed pitching, which again, could either be too conservative a projection (due to adjustments Loney made to his swing last year not measured by ZiPS), an inaccurate assumption (given the Rays apparent preference for fielding the best defense possible), or possibly insulting for Loney's psyche.
These lineups also fail to consider opposing pitcher nuances and player tendencies. If a southpaw is on the mound with a killer curve, are you batting Myers second? Do I want to move Zobrist out of the two-hole in the line up? There's much to consider.
Let us know your ideal line up below.