On Mark Minicozzi, or Complementing Projections with Reason

Last offseason, the Yankees signed infielder Yangervis Solarte — who’d left the Rangers by way of minor-league free agency — to a decidedly more robust minor-league deal than is the standard in the industry. The result for New York was ultimately a positive one: not only did Solarte record the highest major-league WAR figure in 2014 of any player who’d departed his club the previous offseason by way of minor-league free agency, but the Yankees were able to parlay him (along with right-handed prospect Rafael De Paula) into a trade for Chase Headley.

As noted by Kiley McDaniel earlier this month, the Solarte signing wasn’t an anomalous one for the Yankees: they’d completed similar deals with reliever Jim Miller and catcher Bobby Wilson, as well. And while neither of those players did much of anything at the major-league level, the strategy was ultimately a very profitable one for New York based on Solarte’s production alone — profitable enough, as McDaniel notes, to fund 10 seasons of such an experiment.

“Who might be the next Solarte?” one wonders. On Tuesday afternoon, the St. Louis Cardinals signed one possible candidate, infielder Dean Anna, to a major-league contract despite the fact that Anna enters his age-28 season with just 25 major-league plate appearances ever. The projections offer some logic to St. Louis’s decision: despite that limited major-league track record, Anna’s projected to produce an 89 wRC+ and slightly above-average second-base defense — a combination of skills which, when taken together, produce a nearly league-average major leauger. That’s a considerable value for a league-minimum contract.

This post was originally going to be called something like The Top-Five Minor-League Free Agents by the Projections — written with a view, that is, towards identifying those minor-league free agents most likely to receive that Solarte-type money and produce the Solarte-type production. In fact, it’s quite possible I’ll publish a post along those lines next week. But what becomes clear as one inspects the issue more closely is that, while the computer math of a projection system like Steamer is capable of translating without too much difficultly a player’s minor-league batting numbers to a major-league environment, producing a defensive projection for that same player is more difficult. And those defensive projections, if taken without any sort of healthy skepticism, can alter one’s understanding of a player’s value.

In particular, this is true of those players who’ve lingered long enough in the minors to have reached free agency. For, while an 18-year-old shortstop prospect is likely to remain a shortstop into his age-19 season, those players who began their careers a decade ago have very possibly become a different sort of player.

Consider the case of Mark Minicozzi, for example. Selected by San Francisco in the 17th round of the 2005 draft out of East Carolina University, Minicozzi made every one of his minor-league starts in the Giants system between 2005 and -07 at either second, third, or shortstop. Following a elbow injury, however, he was released by the organization and spent the entirety of the next three seasons in the independent Atlantic, Can-AM, and Northern Leagues — after which he was re-signed by the Giants in 2012.

Now, he enters his age-32 season with a decidedly different physique and set of skills than that 23-year-old version of himself originally drafted by San Francisco. And while he’s made starts at both second and third base as recently as 2013, he made the majority of his defensive appearances in 2014 at first base and left field.

With that thought in mind, let’s consider his Steamer projection for 2015, prorated to 600 plate appearances:

Minicozzi’s projected batting line is nearly major-league average — and this oughtn’t be too shocking. Over a span of three seasons since his return to the Giants organization, Minicozzi has produced a batting line about 30% better than average relative to the various leagues in which he’s played. This year with Fresno — his first at the Triple-A level, incidentally — the 31-year-old Minicozzi produced reasonable walk and strikeout rates (13.2% and 22.4%, respectively) while also hitting 12 home runs and exhibiting above-average batted-ball skill (.371 BABIP) in 370 plate appearances. That he would profile as a slightly below-average major-league batter is entirely reasonable.

The question of Minicozzi’s defensive value is a more difficult one to answer, however. For while, as noted, Minicozzi played mostly first base and left field in 2014, he did make those starts at second and third base in 2013, and has, at points in his career, played exclusively those more difficult infield positions. It’s possible that his starts at less demanding positions are reflective of a loss of agility and athleticism — not an unreasonable proposition given Minicozzi’s age. (Defensive skills peaks quite early in player’s career.) That said, it’s also possible that the organization has moved Minicozzi to those less demanding positions to allow younger, more promising players their due repetitions at the position (second or third, for example) they’re likely to play in the majors.

The way Steamer handles Minicozzi, specifically, is to offer no projection of fielding runs saved (which is typical for Steamer of minor-league players) and to assess the positional adjustment of a second or third baseman (+2.5 runs per season). That explains the +2.2 figure in the defensive column (Def) of Minicozzi’s projection above.

From what we know of Minicozzi, projecting him to be a major-league-average second or third baseman — or a decidedly above-average left fielder, for example — seems ambitious. In the specific case of Minicozzi, it almost certainly is. Producing 4000-plus projections, however — such as are available by means of Steamer for position players alone — necessarily requires the employment of some assumptions — chief among them, what position the player in question is likely to play adequately. And while those assumptions might work well for the majority of players, it’s also true that in specific instances — such as the case of players like Minicozzi — that one will be required to apply some reasonable alterations.

For Minicozzi, as we’ve noted, a probably reasonable assumption is that he’s more well-suited to left field now. Applying the generic positional adjustment for a left fielder (-7.5 runs per season) to Minicozzi’s projected 2015 line (again, prorated to 600 plate appearances) yields this result:

Is that ultimately more accurate? I don’t know — although, it’s probably true that, in the case of an older minor leaguer, assuming that he’s closer to a replacement-level player than league-average one is probably smart. Ultimately, though, this post isn’t about Mark Minicozzi and his 2015 season. It’s about projections — and specifically it’s about how one needn’t entirely abandon a projection merely because one aspect of it (like the defensive part) appears not to fully depict the reality of the player’s situation. The projections are wildly useful as a starting point to a discussion about a player’s ability. In those cases where more major-league data is available, those starting points are decidedly accurate. In those instances where no major-league data is available, however, the projections still have their uses — so long as they’re complemented by reason on the part of the one considering them.