What to do with pitchers who can hit

... but only kind of pitch.

The Reds can boast a respectable history of two-way players in the modern era. Since 1950, three Reds pitchers rank in the top 20 in OPS+ among all pitchers with at least 30 PAs in a season. A fourth pitcher on this list, Micah Owings, was under the team's employ the season after he recorded a 1.033 OPS with the Diamondbacks. In fact, the only appearances he made for the Reds the year he was shipped to Cincinnati in the Adam Dunn trade were at the plate (2-4, with a double).

The Reds also have Mike Leake in their starting rotation, who swings a good stick. Or at least, he did. No one seemed to notice that he wasn't very good at the plate in 2011. Which might say a lot about pitchers who hit well.

In the case of Dontrelle Willis and Micah Owings, though they hit well enough to be credible bench bats, neither performed particularly well as pitchers. In fact, their stat lines are nearly identical in Red: both posted an underwhelming 78 ERA+ and 1.5-something WHIPs. Owings is the better hitter over his career (.286/.313/.507 vs. .244/.287/.378 for Willis), though Willis has hit much better than Owings recently.

More to the point, Willis actually has a chance of being with this Reds in 2012 in some capacity.  So how does his bat - or the bat of any pitcher capable of hitting well enough to be a pinch hitter - factor in determining whether he comes back and what role he plays? I think the most important thing to remember is that, as funky as Willis' 2011 was at the plate, his natural level is somewhere closer to his career line. 

His bat was valued at 0.7 bWAR last season and 0.9 fWAR. Over a full season, with pinch hit appearances, he might be knocking on the door of 2.0 WAR with another season like that. The replacement level for pitchers is low. Homer Bailey had a decent year at the plate last season and was close to a half-win above replacement with a .282/.300/.333 in 45 PAs. Almost anything can happen in 50 PAs or less, so I'm not sure we can say Willis' bat would be much more than a tiebreaker when it comes to putting him in the rotation. 1.0 WAR is the most Willis has done with his bat in his career, so if he hews closer to his career line, that's probably the most advantage he can expect over his competitors for a #5 starter job.

That's also assuming the opportunity cost of not using Willis in the starting rotation is losing his bat entirely. That doesn't need to be the case. Willis would almost certainly have less PAs as a reliever, but he could still hit in games that he was unavailable in the bullpen or if his hitting didn't interfere with his warmup (this would probably mean coming in to pitch, then staying in at least long enough to make a plate appearance if the pitcher's spot came up). So you'd really be losing 10-20 PAs. Having Willis available as a pinch hitter also provides flexibility on the bench, but that advantage wouldn't be entirely lost with him out in the bullpen. 

Unless Willis is judged to be very closely matched, pitching talent-wise, with his competition, his bat probably shouldn't play a role in whether or not he makes the starting rotation.  But it could play more of a role in whether or not he's on the team at all, since it's probably a bigger jump from having him as reliever/pinch hitter to "not on the team" than it is from starter to reliever. Just as Willis would avoid burning a position player by PHing on his off days, he would avoid a double switch on days he made relief appearances. 

Willis is also likely to be more evenly matched with his competitors for the left-hand reliever or long reliever role, so his bat skills could become more meaningful there. His career as a starter may not be over, but if he re-signs with the Reds, the best question might be how they can make best use of D-Train as a multifaceted sub: a guy who can hit righties, pitch lefties, do long relief and generated 1,000-word rambling Friday blog posts.