What should the Pirates do at shortstop?

Clint Barmes has struggled with hitting recently like Sisyphus struggled with boulders. Since signing with the Pirates, Barmes has hit .221/.260/.306 in 666 plate appearances. His on base percentage has been above .300 only once in the 198 games he has played for the Pirates: In his first game with Pittsburgh he got on base one time (via being hit by a pitch) in three plate appearances.

Barmes has remained the Pirates shortstop largely because of his defense, the hope that his offense would improve, and the lack of better internal options.

The defense has been very good. Since the 2012, Barmes' UZR/150 score (an all-encompassing metric of defensive proficiency used by Fangraphs) ranks fourth among all qualified shortstops; his RngR score (range) ranks first.

The offense has not been serviceable. Since 2012 Barmes ranks 186th out of 187 players with at least 600 plate appearances, ahead of only Brendan Ryan. And, until now, there has not been any indication that the Pirates have anyone else that can provide more value at the position.

I say "until now" because recently Jordy Mercer has shown some signs that he may be ready to get regular playing time. In his stint with the Pirates last year, he provided little offense (.636 OPS in 68 plate appearances) and was a poor defender. This season, however, Mercer has shown signs of being a productive bat (.761 OPS in 85 plate appearances) and, at least as Clint Hurdle sees it, his defense has improved.

Mercer's play has led to calls from Pirates fans for him to become the regular shortstop. While that is unlikely, it does appear that he will receive increased playing time. Here is what Hurdle had to say about Mercer earlier this week:

If we're going to keep the young man here, the young man's got to play. I got to find opportunities for him to play. To stay engaged. He's got to develop here. This is a completely different situation than it was last year. He's brought more to the table. His confidence level has never been higher. His game is fresh, it's as solid as it's ever been. So we have a lot of confidence.

The most likely situation in the short term is that the Pirates will give Mercer more starts, but continue to spot-start Bames in games that one of their ground ball pitchers is starting (say, Morton) and against difficult right-handers. Mercer has hit left-handers very well, but whether he can hit righties is an open question.

For the sake of argument, though, let's assume that the Pirates did hand over the full time starting position to Mercer. How much of an increase in production could they expect? We can make an educated guess using Fangraphs' positional depth chart projections. (Here's more on how the projections are calculated.)

Pirates' shortstops are projected to receive a cumulative 410 more plate appearances this season. Of those 410 PAs, Barmes is expected to get the bulk of the playing time, with 308 plate appearances to Mercer's 62.

From those 308 plate appearances, Barmes is projected to cost -11.4 runs relative to league average offensively and -.2 runs from baserunning, and to save 3.3 runs defensively. So, in total, Barmes' is projected to cost the team -8.3 runs relative to league average.

If we take Mercer's projections as they are calculated for 62 plate appearances and scale them to 308, thus giving Mercer, Barmes' plate appearances, we get the following: offensive -6.46 runs, baserunning -.49 runs and defensive .99 runs saved, for a total of -5.9 runs.

In other words, according to Fangraphs' forecasting system the difference between giving Mercer 306 plate appearances as opposed to Barmes is 2.4 runs, or 25 percent of one additional win. Not much.

The small difference does not mean, however, that there aren't good reasons for going full steam ahead with giving Mercer much more playing time. First, the projections for Barmes are far more likely to be closer to accurate than Mercer's because of his much more extensive Major League body of work. We know what Barmes brings to the table; there is a bit more of a fog of war surrounding Mercer. He could provide significantly more than 2.4 runs in added production. Second, if they head down the road of giving Mercer consistent starts and it doesn't work out it is unlikely they will lose much production in the process, and they always have the option of changing directions again. Finally, Barmes is a free agent after this season and the Pirates are unlikely to re-sign him. There seems to be little reason not to see what they have in Mercer this season, especially when there is so little to lose in terms production.

That said, it is unlikely the Pirates will turn the position over to Mercer to the tune of 306 plate appearances. More likely, it will be something like a 60-40 or 70-30 split in the short-term, with Mercer also getting playing time at second to spell Walker. That seems to be a reasonable approach.

One of the most powerful forecasting systems we have suggests that ultimately Mercer provides little increased value. It is unlikely that Mercer will provide as much of an upgrade as Pirates fans want to believe. But he could, and the way to figure that out is by slowly moving in that direction.

The fact is the Pirates are twelve games over .500, while getting decent defense and little offense from their shortstops. There seems little reason to radically change direction under such circumstances. Rather, by moving slowly, they hedge their bets against lost production if Mercer is noticeably worse (particularly defensively). And they still give themselves, and Mercer, the opportunity to see if he is legitimate option at shortstop in 2014.

Follow on Twitter David Manel @DavidManel and Bucs Dugout @Bucsdugout