I could very easily write a few thousand words on how foolish I think it is for the Mets to carry Jenrry Mejia as a reliever this year.  Many others have already done so.  You can read about it here, here, here, here, here, and here, and I'm sure in many other places as well.  Instead of rehashing many of those ideas, I'd like to work under the assumption that Jenrry will break camp in the majors (though that's hardly written on high) and discuss a way this might actually work .

Disclaimer:  I cannot emphasize enough how strongly I oppose putting Mejia in the major league bullpen, this is simply a hypothetical designed to explore how this situation might play out in the interest of all parties involved.

First things first, lets establish the objections and problems that would need solving in throwing Mejia into the Major League fire:

1. The long term goal with an arm like Mejia should still be starting pitcher.  It would be wasteful to relegate an arm like his to relief at such a young age.  
2. If he spends a full year in the major league bullpen, his innings total would likely be reduced from last year's, despite the fact that he was a starter and is yet to throw 100 innings in a season.  If the long term development goal is still starting pitcher, his innings total should be increasing in a controlled manner, not decreasing in an erratic manner, as it would if he was held to the standard of typical Major League reliever.
3. He needs to work on his secondary pitches.  His fastball is Major League ready, but his curveball and changeup are not.  The best way to get them there is to put him in a starter's role where he's forced to sequence his pitches against the same hitters multiple times per game instead of just using his fastball to blow through three outs at a time. 
   
4. His command was a serious issue in Double-A and to expect an immediate improvement against Major League batters is wishcasting at best.  His performance this spring--specifically the fact that he's yet to walk a batter--is encouraging, but realistically does no more than offset his poor performance in the Arizona Fall League.

Of course, the oft cited advantage of putting Mejia on the roster is that he may legitimately be the best option, at least talent-wise.  Command aside, his "stuff" has certainly been as impressive as anyone to come through Mets camp in recent memory.  There's certainly no one competing for a relief role who has Mejia's upside, and no pitcher in camp period this side of Johan Santana. 

So now the question is whether there's a way to both solve these issues and let Mejia start the year in the major league bullpen.  The near universal answer (Kevin Kernan aside) amongst writers, pundits, bloggers, and people generally more knowledgeable than I is a simple no--there is no way to resolve these problems with Jerry Manuel's desire to have Mejia in Flushing.  Being the contrarian that I am, I'm going to do my very best here to disagree and suggest a plan that might be too crazy not to work.  Here it is:

Jerry Manuel should be told he has no more than 20-25 innings and until June 15th to use Mejia as a reliever, at which point he will be sent back to the minors and very cautiously stretched back out into a starting role.  Through June 15th, the onus will be on Jerry to both carefully use Mejia as an ace-in-the-hole reliever and figure out exactly who else he can rely on in high leverage, non-closer situations.  After June 15th passes, Mejia will be given five minor league appearances and 20 innings before the All-Star break.  He could make one appearance of two innings, one of three, one of four, and two of five, which would total 19.  Then we hit the All-Star break and its decision time.  If Mejia was effective in his limited Major League role and the big club is need of a starting pitcher, they can seriously consider re-promoting him and putting him in the rotation.  If the starting five are healthy and pitching well or Mejia was ineffective as a reliever, he should be left in Buffalo for the rest of the regular season to make regular starts.  In either case, he will have a total of roughly 40-45 IP through the break, which is a bit more than halfway through the schedule.  He threw 94.2 innings in 2010, not counting the AFL.  That means his 2010 innings total should not under any circumstances significantly exceed 130, and that might even be pushing it.  This leaves the organization about 90 more innings to play with.  In 15 starts, that would equate to six innings per start, a mark Mejia is unlikely to average as a major league starter this year anyway, and a number which is much more easily controlled if he's in the minors. So 15 starts it is.  If the team happens to make the playoffs, he could be re-inserted into the bullpen--especially if he's still well under the 130 total innings mark for the year--and used sparingly.

This plan actually has a number of advantages, and at the very least addresses three of the four concerns listed above in some fashion (#4, command, isn't something that can necessarily be addressed by anyone other than Mejia anyway).  It would give Jerry an early season cushion to determine how he's going to use some of the unknown quantities in his bullpen--such as Ryoto Igarashi, Bobby Parnell, and Kiko Calero.  It's also a means to limit Mejia's innings total for the year without either shutting him down after 20-25 starts (also preventing a September callup) or limiting the number of innings he could throw per start.  It could even buy some time for a wild card like Kelvim Escobar to get healthy.  It could also give the Mets a few weeks before the trade deadline to evaluate whether they need to make a trade for an arm or consider riding it out with Mejia.  And it could set a precedent for next year.  If a plan like this were to work, a similar approach could be taken in 2011, except with Mejia getting sent down and stretched out even earlier, allowing him to come back and make, say, 20 starts instead of 15.  Then he could become a full time member of the rotation in 2012, by which time his arm should be built up well enough to handle a full-season workload as a starter.  As an added bonus, this would delay his service time by at least two months, thus extending the period under which he would be cost controlled and potentially even holding him back from Super-Two status, similar to what we've seen with players like Evan Longoria and David Price in recent years.

Now if you'll excuse me, I'm going to go look for some fire retardant clothing while you all ponder this suggestion.