Of the six or seven major potential prizes in MLB free agency this winter, B.J. Upton probably presents the most uncertainty. That statement is more loaded than it seems, too. Upton plays the outfield. He’s a position player. Zack Greinke, Edwin Jackson and Anibal Sanchez all are in the top tier of available guys, but those guys–pitchers, mind you–are easier to reliably forecast than Upton. Josh Hamilton, a 30-something with a long injury history and the cloud of a once-crippling drug addiction following him, presents less uncertainty, although perhaps more risk. Most teams will have serious reservations about signing any of Greinke, Jackson, Sanchez or Hamilton. All teams could conceivably use B.J. Upton, as evidenced by the eagerness of the entire NL East to outbid one another for him. (Okay, three fifths of the NL East. But come on.) Pricing Upton, though, is as fascinating and difficult a challenge as any front office will face this winter. Whereas valuing the others listed will be a question of degrees of expected production (will he be worth one win in 2015, or two?), Upton presents a more fundamental question: What kind of player are we going to have if we sign him? Who is B.J. Upton, and which shape will his performance take as his contract plays out?

I published a post Monday morning examining the viability of Darwin Barney as an everyday player. The principal strike against him: Barney reached base at a paltry .299 clip in 2012. Here’s the thing: B.J. Upton came in at .298. Now, Upton has more power (not a little, but a ton more) and posted that figure both in a tougher park and in a tougher schedule than did Barney, but it demands note. Once a star with terrific OBP skills, Upton now has a more one-dimensional offensive game. In order to determine how much offensive value he will contribute, we need to dig deep into his numbers and find out what they’re telling us.

Upton has traditionally walked about 10 percent of the time in his career. That figure had crept to a steady 11 percent in each of 2010 and 2011. In 2012, however, it suddenly dropped to just over 7 percent. His approach had always been more ‘selectively aggressive’ than truly patient, but Upton morphed into a singularly free swinger this season. He waved at over 35 percent of all pitches outside the strike zone in 2012, according to FanGraphs, and over half of all offerings.

There was also a shift in Upton’s zone preferences for swinging at fringe pitches in 2012. He had always attacked pitches near the top of the zone and away from him. He didn’t really stop in 2012. However, he got much more aggressive on pitches below the zone, and low in the zone on the inner third. Upton’s angularity at the plate and his quick swing make him a natural to slug fastballs on the inner half, especially down, and he always has done so, and he continued to do so.

However, he has always had a hole in his swing when pitchers can locate hard stuff below the belt and from the outer half of the plate away. That continued as well. Counting 2012, Upton is now hitting .189 with a .253 slugging average in 200 career plate appearances resolved on fastballs low and away. For some perspective, Justin Upton has batted .268/.451 on similar pitches in his career, and Michael Bourn (a fellow free-agent center fielder) has batted .283/.335.

I don’t mention all of this as fodder for criticism. It’s simply true. Upton is a good hitter, but of late, a very aggressive one, and he has a hole in his swing it seems beyond him to correct. The data (specifically, in this case, the decrease in swings at high pitches, the increased overall swing rate and the falling contact rate) imply that Upton might be losing some bat speed (to age, injury or both) and trying to cheat for power, but his sustained success against even good fastballs in most areas of the hitting zone refutes that thesis. All I want to point out for certain is that the hole exists, so Upton may have a chance to feast on a new primary opponent set until the advance scouts find that weakness, and that his aggressiveness has risen markedly in the past two seasons, dampening his on-base skills.

Let’s zoom out and look at some larger samples of less granular data. Upton fanned nearly 30 percent of the time in 2012, a viable but far from ideal rate. Nonetheless, it’s probably where Upton will hover going forward. His swing rates might come back to earth, but the miss in his swing spiked in 2012, and that is not likely to turn itself around. His walk rate could easily rebound to its former heights, but I’m not convinced that Patient B.J. is the best possible version of the player anymore.

For one thing, Upton batted .413 with a .696 slugging average on the first pitch in 2012, 46 percent better than the adjusted league average hitter did on first pitches. With two strikes, though, he managed a meager .423 OPS, just 63 percent of a league-average performance in those spots. Letting pitchers get deep in the count doesn’t seem to behoove Upton.

To frame the same argument another way, look at Upton’s season in chunks, based on his changing approach. He got a late start to the season (which, by the way, though I find it to be a watery narrative, could help explain his overall unrefined approach, since he lost not only the first fortnight of the season itself but virtually all of Spring Training to back soreness) and didn’t debut until April 20. From then through May 5, in 13 games, he walked exactly once. On May 6, though, he walked three times, and he’d rack up 10 total walks in his second 13-game stretch. Beginning May 20, though, and moving into mid-June, he walked twice in his next 90 trips to the dish. Hearing it from his agent (it’s a walk year, B.J., and walks pay now, so walk!) or from someone in the organization, Upton turned his approach back to ‘patient’ from June 14-July 26, walking 19 times in 168 plate appearances. Then, near the trade deadline, someone turned off the shock collar, and Upton walked 13 times from July 27 to season’s end, in over 250 times up.

Call the extraordinarily impatient Upton in three of those chunks Type A B.J., and the passive walkster Type B. Here are their stats:

Type A: 408 PA, .267/.294/.519, 24 doubles, 3 triples, 22 home runs, 16 walks, 104 strikeouts, 3.9-percent walk rate, 25.5-percent strikeout rate

Type B: 225 PA, .205/.305/.324, 5 doubles, 6 home runs, 29 walks, 65 strikeouts, 12.9-percent walk rate, 28.9-percent strikeout rate

Honesty compels me to report that I’m not sure that means anything. I chose the endpoints for that study. It sure seems there is a correlation, but these results are very, very unscientific, and possibly illusory altogether. Still, this point shines through: B.J. Upton is not at all likely to rediscover the balance that made him an elite center fielder for a year or two. He’s a good but flawed player, and it might not be wise to sign Upton if the object of doing so would be to change him back to the patient, top-of-the-order hitter he once was.

The results when Upton does make contact might be as indicative of his changing approach as his newly dubious command of the strike zone. Once a 52-54-percent ground-ball guy, Upton has that number down to around 40 percent, the culmination of a years-long trend toward more flies. Since 2009, his percentage of outfield flies leaving the park has also risen, from 6.8 percent that year to 16.7 percent in 2012. His batting average on balls in play was sky-high in 2008-09, but for the last three years, it has leveled off: .304, .298, .294. The noise in any BABIP sample, especially one as small as Upton’s (home runs, walks and strikeouts have accounted for two fifths of his plate appearances the last three years), prohibit too much analysis of that small downward trend. He might be losing half a step, but those numbers are essentially the same. Still, paired with the clear effort to trade contact for power and the fact that his stolen base attempts have tailed off (60, 56, 51, 48, 37) since 2008, the evidence of Upton losing trust in his quickness and young player’s skills is piling up.

It might be as much about not wanting to pound his body further as about simple aging. Upton has had three ankle sprains; four quad strains; surgery on one shoulder and trouble with the other; and lower back soreness over the past few seasons. His trouble keeping his legs healthy may have him considering a game based much more on power than speed, simply as a survival mechanism.

His defense also reflects the possibility of smart tradeoffs helping him paint over injury accumulation or the physical toll of aging. Three of the four important models of defensive evaluation have generally liked Upton as a center fielder over his career, but have trended downward in their grades for him the past few years. The one that consistently panned Upton before, Baseball Info Solutions’ Defensive Runs Saved, still has him below average, but has trended positively for him the past three years.

The difference has been Upton’s improved handling of deep fly balls. Baseball Info Solutions’ system tracks outfielders not only on their overall coverage of their space, but on the specific types of balls they handle well or poorly. Although a generation of center fielders (Jim Edmonds, Andruw Jones, Steve Finley, etc.) counted it a point of pride to play shallow enough to steal bloop singles, they often overestimated their ability to go back and get balls hit far enough to be doubles and triples. Thus, their hubris only hurt their teams. So it once was with Upton, who from 2008-10 was -18, -12 and -17 plays against average on deep balls, per the system. However, in what I can only read as a concession to his body’s depreciation, Upton has adjusted, and he rated -7 in 2011, then +3 in 2012, on deep balls. He may be losing a step, as the other systems show, but he seems to be addressing the problem properly.

All that considered, Upton is not likely to be a plus defender in center going forward, but he can stick there at least halfway through a five- or six-year deal. An efficient base runner who has always stolen with at least the requisite success rate to be worth trying it, Upton also adds value by taking the extra base a dozen or so times per year more often than is average. Base-running is as much about instinct as it is about sheer speed, so Upton should hold that (admittedly small) value well.

It comes down, then, to what each team’s plans and hopes for B.J. Upton are. The Braves need an OBP guy if at all possible, and should key in on Angel Pagan instead of Upton. The Phillies could use a right-handed bat to break up Ryan Howard and Chase Utley, but Upton is not cleanup material on a first-division team. The Nationals make the most sense, as they need a center fielder and can stand to bat him fifth or sixth.

Of course, one other important variable in where Upton lands and how he produces will be where he plays his home games. Tropicana Field is a great pitchers’ park, and not for any one reason, but for many. It depresses hits on both ground balls and fly balls by right-handed batters by about five percent. It depressese singles, doubles and triples by right-handed batters by about five percent. It depresses home runs by right-handed hitters by nearly 10 percent. It increases strikeout rate and decreases walk rate for right-handed hitters. Almost every skill Upton has will probably be augmented by his team change this winter.

Given all the things Upton might be able to do, and given how wide the margin of error in projecting a player like him is, it would be hard for him to sign a deal that surprises me. A diligent team will do their homework on him; determine what adjustments to make (or not make) in order to maximize his considerable potential; make sure he buys into that plan; and make him an offer. Just don’t ask me what it will be worth, or whether Upton will earn it.

Matt Trueblood writes Arm Side Run mostly at night, and mostly on his phone. Baseball Prospectus, Baseball-Reference and FanGraphs provide the wonderfully detailed statistics used here free of charge, but for the BIS Runs Saved detail, which is available to subscribers at Bill James Online. StatCorner.com provides the component park factors, and truckloads of other data, for free, too.