For much of the past 15 years or so, the Twins have had very easy decisions to make with the first spot in the batting order. Jacque Jones settled there, then handed the position over to Shannon Stewart, who was a fixture atop the order until 2006 when Luis Castillo joined the team. He held the role until 2008 when it was briefly handed over to Carlos Gomez, who promptly hit .246/.281/.345 and lost the job to Denard Span. Span locked the role down, holding it until he was traded prior to the 2013 season. (Strangely in 2011, despite Span’s prowess leading off, the team felt that Ben Revere deserved nearly a full-season look in the lead-off spot; he wasn’t quite as bad as Gomez, hitting .291/.322/.349, but still not good enough to keep the team from reverting back to Span.)

Since Span’s departure, Brian Dozier has been the Twins’ lead-off hitter of choice, though at various times he has shared the gig with Danny Santana or Aaron Hicks. Dozier has been a fine choice at the top of the order, posting a .243/.313/.459 line when hitting first, but he’s a little more in the mold of Jones (.289/.328/.471) than Span (.285/.352/.389). Does one type of hitter make more sense than the other given how the rest of the Twins' lineup shakes out?

The primary job of a lead-off hitter is to get on base and set the table for the more slugging-heavy hitters to come. This is not news. However, in addition to getting on base, a lead-off hitter is often asked to work the starting pitcher a little bit -- not to the extent of laying off of a hittable pitch, but a nine-pitch plate appearance that ends in a strikeout isn’t an entirely negative outcome since it provides the rest of the team with a sense of what that day’s starter has on offer. As I mentioned earlier this season, taking pitches has been one of the few things that Twins hitters have been doing well, so they have some options:

http://i.imgur.com/z6mWJNu.png

League Average: 3.89 P/PA

So the Twins have some options for patience at the top of the order, but how many of these guys can fulfill the primary task of getting on base consistently?

http://i.imgur.com/pnFx3cz.png

The obvious name on neither of these two lists is Santana’s. He owns a team-worst (among qualified hitters) 3.56 P/PA and his Gomez-esque .256 OBP is the spared from being the low-water mark by the recently demoted Eddie Rosario’s almost unfathomably bad .227 OBP. Santana’s speed makes him interesting when he’s on base, last night’s TOOTBLAN notwithstanding, but he’s not giving the team enough additional value to justify giving him more plate appearances than literally anyone else. Paul Molitor has had Santana’s usage right for the last week or so: Hit him ninth to limit the damage he can do, and if he does get on base, so much the better.

Given the way the season is progressing, it would be interesting to see how Byung-Ho Park or Miguel Sano would do leading off, but that’s not a serious suggestion for the long-term, if for no other reason than it would virtually guarantee than any homers hit would be solo home runs. Oswaldo Arcia and Trevor Plouffe offer only fractional upgrades at best over Dozier, and while Dozier’s no Byron Buxton, he is notably faster on the bases than either of the others.

This leaves three serious candidates who can both get on base and take an extra pitch or two on their way: Mauer, Dozier, and Nunez, who have combined to lead off in 29 of the Twins’ 40 games.

Nunez has been tremendously fun to watch this year, and he’s as good a bench bat as the Twins have had in recent years, but the other shoe is going to drop with him. His .395 BABIP is built largely on flyballs falling in and isn’t supported by an increase in his hard hit percentage. Even if his BABIP dropped to .320, which would still be good for a career high, it would vastly diminish his suitability to lead off. This all assumes that Nunez even gets consistent playing time once Eduardo Escobar returns, which is far from a given.

I have been on the bandwagon to bat Joe Mauer second for about as long as it had wheels and a seat, but I understood the theory behind hitting him third. On those early-2000s teams where “get ‘em on, get ‘em over, get ‘em in” was less a team mantra and more a foundational principle, Mauer was clearly in group three and tasked with driving in runs, then being driven in himself by the actual power hitters. Not his ideal usage, perhaps, but his lot in the lineup anyway.

This year’s team was never designed to “get ‘em on, get ‘em over, get ‘em in” and they certainly aren’t tricking anyone into thinking that’s their path to glory. They’ve already hit nearly as many home runs in May (17) as they did in all of April (21) and baseballs fly further in warm weather, so it’s reasonable to assume they’re going to actually start scoring a reasonable amount of runs that way; however, if they continue to hit solo home run after solo home run, they’ll need to hit an astronomical number of home runs to actually challenge their opponents. Far better to pull a page out of crazy Earl Weaver’s playbook and hit home runs with people on base.

Maximizing the average value of a home run hit by the 3-4-5 hitters means ignoring the relative value of slugging for a lead-off hitter -- the increase in expected runs of having a runner on second or third base with no outs -- in favor of fully optimizing for OBP. It also means pairing the hitter with the highest probability of being on base with the hitters most likely to drive him in. So, not only should Mauer be leading off for the foreseeable future, but Park and Sano should be moved up in the order to maximize the likelihood of both of them getting a chance to hit with Mauer available to drive in.

Ideally, the Twins would have another high-OBP hitter with slightly more speed to hit ahead of Mauer, as a lineup of XXXX, Mauer, Sano, and Park would be a nice setup, but that hitter doesn't appear to be on the Twins’ roster right now. As long as the hitter in the second slot, between Mauer and Sano, doesn’t have a proclivity for ground balls, having someone hit there shouldn’t change the expected value of setting the lineup this way much at all. This makes Dozier an ideal candidate to bat second. Yes, he’s not a “get ‘em over” guy, but that doesn’t matter. He has the team’s second lowest groundball rate and there’s virtue in giving Dozier the second most plate appearances on the team since he has typically been an offensive asset.

Molitor has sent the Twins out with 36 lineups in 40 games, trying to find something that works, but not once has Mauer lead off with Dozier batting right behind him. A top four of Mauer, Dozier, and either Park and Sano or Sano and Park stands to give the Twins the best shot of scoring an early run and of taking advantage when the lineup turns over thanks to Santana. Given how often the lineup is shifting, it seems fairly likely we’ll see this top four sometime this season, hopefully sooner rather than later.