More photos » Julie Jacobson - AP

"Don't celebrate my departure too soon (at a baseball game), boys. I hear Goldman wants Ca$h to bring in Cody Ross. Yeah, you heard, me, Cody Freaking Ross. He has a fan club and everything."

Browse more photos »

In a recent Pinstriped Bible blog posting,  Steven Goldman argued that while Melky Cabrera might be adequate for the moment as the starting center fielder for the Yankees, that the club should look elsewhere for next season. Goldman's main point:

It would be unrealistic to expect even a team with the financial capabilities of the Yankees to field an All-Star or MVP candidate at every position, but that's not a reason to keep striving to get better wherever you can, however you can. It is axiomatic that a team that accepts that it is "good enough" at a position will not win consistently, or at all. Complacency will kill a nascent dynasty faster than hamstring pulls and cigarettes combined.

There is much to comment on in this blog posting (see the Think Factory discussion). For example, calling Nick Swisher, he of the 3.6 WAR (that's about the same as Jason Bay or J.D. Drew at this point in the season), "inconsistent" is Joe Morgan-esque.  Ignoring the Yankees other significant holes for 2010 is also curious. Indeed, the whole post, with its discussions of league-average offense by position, vague references to defense without numbers, etc., strikes me as a very "pre-FanGraphs" sort of approach to internet player analysis.

However, I'm not going to be issuing a bunch of snark regarding a good writer's problematic blog posting. Rather, taking Goldman's post as a jumping-off point, I want to look at three options for the Yankee 2010 CF that he mentions: Cabrera, Brett Gardner, and the Marlins' Cody Ross (for whom Goldman suggests the Yankees trade.). I will compare them utilizing less vague and more helpful concept of WAR (Wins Above Replacement).

I've done a number of these "true talent" postings lately (on Jason Bartlett, Jamey Carroll, and Matt Murton, among others), so you can look around the blog to see how it works. I won't review all of the "methodology" here -- it's pretty simple weighted averages and regression to the mean, etc. using publicly available data. There are no big secrets. I just don't want to make this more boring than it has to be. Feel free to ask in the comments section. I'll go component by component for all three players. These are "true talent" estimates based on information as of last night. I'll go component-by-component for all players, just to mix things up.

Hitting

Melky Cabrera is the youngest player of the group, being only 24 at the beginning of the current season. His major-league wOBAs from 2006-2009 are .333, .317, .285, and .328. Regressing to the mean and applying a slight age adjustment, that projects to about a .327 wOBA for next season. Given this season's run environment, that means he projects as  a minus 1 run (one run below average) hitter for 2010.

Cody Ross is 28 this season and the best hitter of the three. He will be 29 next season. From 2006-2009 his wOBAs are .310, .446, .345, and .342. He projects as a .344 wOBA (+9) hitter next season.

Brett Gardner (who is only briefly considered in Goldman's post) is 25 this season. Gardner hit terribly in the majors in 2008, posting a .282 wOBA. He's been very good relative to expectations this season, sporting a .349. But given the sample size and his performance in the minors, I decided that I needed more data. So I tried something new. I got his MLE from the minors from 2007 and 2008, converted them to wOBA, and incorporated them into my data. I'm not sure how great the method it, but the result I got  -- a .318 projected wOBA (-6/700 PA) -- is close enough to his preseason projections from CHONE, etc. that I don't think it's distorted.

Fielding

For the fielding portion, I used UZR as implemented at FanGraphs as my primary stat not only because I suspect it's the best defensive stat out there, but because it's publicly and freely available. I heavily regress outfield projectoins. I regress UZR based on Speed Score, as has often been suggested for OFs. Defensive stats being what they are, I also incorporated the Tom Tango's Fans Scouting Report Converted to runs. I usually do a weighted average 75% regressed UZR + 25% FSR, but for Gardner, given the extremely low sample, I did 50% regressed UZR + 50% FSR (his defense would project even better without the FSR!).

The Milkman has been up-and-down, defensively.  While he's been decent according to UZR his season (+3.1/150), in 2007, when he had the most playing time, he was dreadful (-10.7). The Fans Scouting Report  is even less enthusiatic, having him at -4.8, -2.4, -2.7, and -8.0 (2006-2009) according to my "translations." So it's a mixed bad. Altogether, he projects as a -2/162 defender in CF for next season.

Cody Ross has been similarly up-and-down. He's -7.1 UZR/150 in CF this season comes after an excellent +18.3 in 2008. He was terrible in limited time in 2007, but good in 2006. The FSR hasn't been too impressed, with a -9 in 2006, a -9.5 in 2007, and a -3.5 2008. But Cody's fan club must have been working overtime this season, as he's all the way up to 0.2. PUtting it all together, and adjusting for age as Ross enters his late 20s, he projects as a -1 /162defender in CF in 2010.

Like Nyjer Morgan, Brett Gardner is the kind of player who would have been dismissed a Juan Pierre wanna-be prior to the wide public availability of defensive stats. And his numbers are so outstanding in brief playing time that they have to be taken with a grain of salt. Or, as grains of salt are expressed here -- with plenty of regression. UZR has Gardner was an insane +56.8/150 in the OF in 2008 (in only 39 games), and +14/150 in 2009. The Fans Scouting Report shows its value here by not forcing us into either taking these on face value nor dismissing them completely. The FSR has Gardner at +4.2 in 2008, and +5.8 in 2009. Putting it altogether (note his impressive speed scores), Gardner projects as a +9 defender in CF for 2010.

Baserunning

Don't worry, almost there. I used Baseball Prospectus' excellent EqBRR for this portion. Since FanGraphs' version of wOBA includes steals, I substracted that portion out. Briefly -- both Melky and Cody project as pretty much average ("0") baserunners, while Gardner, even after heavy regression due to low playing time, projects as very good at +2 (keep in mind that the "spread" is very small).

Putting It All Together

The difficult in the "old days" would have been how to weigh offense (and then how to weigh OBP and SLG relative to each other, etc.) vs. vague assertions about defense, positional value, etc. Baserunning would probably be totally ignored. But the beauty of WAR is that, while one can debate over whether the projections were done correctly, how replacement level is set, positional adjustments, etc., one can at least get it all out on the table. So here it is... in a table (ahem). Since all three players are being projected at CF, the positional adjustment is +2.5 runs per 162 games. To account for the difference in league difficutly, following Tom Tango I use 20 runs as replacement per 162 games for the NL (Ross) and 25 runs for the AL (Cabrera and Gardner). It's easy from there: I sum all those toger to get Runs Above Replacement (RAR). Then I convert to WAR by multplying RAR times 85% playing time, then simply dividing by 10 to convert runs to wins.

Goldman writes that Gardner is only "fractionally" better than Melky, and given the nature of WAR, 0.5 isn't a huge deal. But he does look better. What is interesting is that Gardner also projects to be better than Cody Ross. If you don't but the league adjustment, well, I'll just ask you to compare the NL East's pitching staffs with those in the AL East. Despite recent comedic assertions that Brett Gardner should only be at the head of a list of terrible baseball players, he's seems to be above average here (2 is about MLB average), and the best candidate of the three to play center field for the Yankees. He's another potential Endy Chavez (another defensive standout player underappreciated in the pre-UZR era), and might actually have a shot at having a league-average bat. Gardner might not be 1998 Bernie Williams in 1998, but he might be 2009 Nyjer Morgan.

But this isn't all about Gardner. Melky is still almost as good as Ross. Don't get me wrong -- Ross is a decent player that could help a lot of teams. But why should the Yankees trade talent for a guy who, if he's better, is himself "fractionally" better than the alternatives (at best), older, and, yes, getting more expensive? Even the Yankees operate on a budget, and as we've seen, even they occasionally need to rely on some homegrown talent (e.g., Robinson Cano, Joba Chamberlain, Phil Hughes, etc.) rather than free  agency. Even if the Yankees should panic (they shouldn't) over Austin Jackon only hitting four home runs this season, they have cost-controlled talent that should be more than adequate in place.

Finally, given that they have a massive holes opening up in left field and at DH with the free agency of the declining Johnny Damon and Hideki Matsui (not to mention their relatively shaky pitching), it seems to me that the Yankees considerable organizational resources would be better spent elsewhere, say, on Matt Holliday (easily the best free agent outfielder this season -- far superior to Jason Bay, but that's another post) for LF. If they must trade, make it for Adam Dunn as DH (although I'd suggest trying to sign Jim Thome or especially Russell Branyan for that position -- Branyan in particular can actually  play the field a bit and will be cheaper).

These are much bigger issues than being "stuck " with a player who is "only" average or above-average at center field, a problem that wouldn't be solved by trading for an older player who isn't any better. On the other hand,  if the Yankees really share Goldman's low appraisal of both Cabrera and Gardner, I know a team short a couple of outfielders...