A Note Of Caution On Prospect Projection

We’ve been talking a lot lately – given the paucity of good answers in the free agent market and the high price of veteran talent in trades – of the possibility of acquiring this prospect, that prospect or some other prospect from outside the organization to fill one of the Tigers needs this offseason.  We’ll no doubt be continuing this discussion until Spring Training at the least.  Before we all get carried away, I’d like to take this opportunity to inject a note of realism into the discussion:  minor league numbers are a far from perfect guide to predict what a player could bring to a big-league club.

Hah.  Made you think I was going to be talking about pitchers, didn’t I?  No, that’s our own Don Kelly in his one appearance out of the ‘pen last year (courtesy of the Pittsburgh Post-Gazette).  Why Don Kelly?  He’s the perfect case-in-point to show that a players minor league numbers (and the specific skills those numbers seem to indicate) don’t just need to be deflated when translating into big league projections (like BP does in their Prospectus) sometimes they simply become irrelevant.

Don Kelly was never anything like a premium prospect, he wasn’t young when he was drafted, he didn’t rocket through anybody’s system and he never made any Baseball America lists.  But his numbers in the minors looked pretty good.  Through age 25 Kelly was a prospect in the Tigers system and it was during that last year that he made his first appearance in Toledo.  He had played mostly shortstop, with a few games at other positions sprinkled in, and stolen a few bases.  Most relevant to the Tigers predicament at the moment, his one true talent seemed to be getting on base.  Kelly’s minor league numbers through 25 were .297/.376/.400 with more walks than strikeouts.  Couple that with speed and the ability to play middle infield positions?  Looks like the perfect solution to the Tigers present-day needs, right?

I published a post in mid-season about the projectability and deflation of minor league stats – for the average player – that you can find here.  The gist of it was that BABIP, power, walks and strikeouts all get deflated:  the average guy loses 32 points of BABIP, 29 points of isolated power, increases his strikeout rate by 5.6 percentage points and decreases his walk rate by 1.2 percentage points.  For Kelly that would have meant a BABIP dropping from .323 to .291, ISO dropping from .103 to a measly .074, K% rising from 8.7% to 14.3% and BB% dropping from 10.8% to 9.6%.  We would have expected a slash line from Kelly (in his first year as a big-leaguer) in the neighborhood of .253/.328/.327.  Of course, this is only intended to predict what we’d likely see from a prospect in his first year in the bigs, it doesn’t account for potential growth and development as a big-leaguer.  We also know that not all prospects are created equal, we might figure that guys who were younger for their league would have a better shot at beating the odds (and deflate less, like Miguel Cabrera).

Most guys aren’t going to be successful big leaguers if they can’t ‘beat the odds’ and perform better than such a deflator would suggest, and Kelly has not (at least to this point).  Whenever you’re giving a roster spot to a rookie, you’re hoping that he can elevate his game – like Brennan Boesch – and prove doubters wrong, but you’re aware that he could tank – like Delmon Young – and make those deflated projections look relatively rosy.  Kelly’s projected OPS isn’t all that far from what he’s actually done as a big leaguer a .655 projected and a .648 actual.  If you gave him the roster spot and installed him in the infield what you would have been hoping for is that his BABIP would decline by less than expected and/or his strikeout rate would rise by less than expected and he’d be able to give you the OBP of .345 or so that makes a solid leadoff hitter.  If he couldn’t you’d have to be ready with a backup plan.  The kicker is that while Kelly projected as a guy with ‘leadoff skills’ (though potentially not enough of them) that isn’t what we’ve seen from big-league Don Kelly.  As a major leaguer, Kelly has been a .240/.285/.363 guy.  Overall that’s not much worse than the projection, but it makes for an absolutely awful leadoff candidate.  His walk rate plummeted from 10.8% as a minor leaguer to 4.5% in the show.   His BABIP dropped from .323 to an Inge-like .259.  His strikeout rate rose by about what we would have expected but his ISO actually increased from .103 as a prospect to .123 as a major leaguer.

For reference, you can see in the table below what other Tigers regulars did while in their prospect phase (prior to age 25) and then compare that to what you’ve seen from them in the bigs.  A lot of them looked like they could have been a leadoff guy, no?  Really, everybody but Brandon Inge and Brennan Boesch showed a decent on-base percentage as a minor leaguer and while Kelly may be an extreme case he’s far from the one and only.  So… do I hate prospects?  No, of course not.  But to fill a specific need on a championship caliber team?  Give me a guy with a track record.

Oh… and lest I forget?  Once upon a time Brian Hunter put up a .432 OBP in AAA while stealing a pile of bases.  He was in BAs top-10.  Anyone remember how that all turned out?