Orioles Player Values for the 2008 Season

It's a topic that lots of people like to speculate about—how overpriced was Ramon Hernandez? Is Brian Roberts worth $15 million a year? $10 million a year? Should we just trade him and get his salary off the payroll? One of my favorite sites, FanGraphs, has made these discussions really easy to resolve—for outfield players at least.

Here are last season's data for all Orioles batters with at least 10 plate appearances. The trifecta of Markakis/Roberts/Huff stands head and shoulders above the rest—each of those players was worth more than $17 million in salary last season. Meanwhile, our shortstop platoon combined for a total of -$14.9 million, relative to replacement level (in the 2008 season, Seattle's Yuniesky Betancourt was exactly a replacement level player, for comparison).

Another interesting note is that Ramon Hernandez was actually underpaid during his time at Baltimore. You can click on his name on the chart and scroll to the bottom of the page to see a summary of his predicted value vs. actual salary—and very quickly you'll notice that in 2006, his value was a staggering $13 million! That more than makes up for the modest losses we incurred over the past two seasons.

I should mention, though, that catcher fielding is completely ignored in this system—all catchers are considered exactly average fielders. This obviously creates a problem when you have guys like Brad Ausmus, whose value lies almost entirely in his skills behind the plate.

Two players who were decidedly overpaid at Baltimore are Kevin Millar and Jay Payton. Millar, for all of his "clubhouse leadership," was a below average fielder and a below average hitter at a position laden with talent—first base. Meanwhile, Payton was a mediocre fourth outfielder being paid a starter's salary. Based on this analysis, Payton and Millar have rightfully been sent out into the free agent market.

That said, you have to remember this system isn't perfect. The most obvious problem is that catchers are generally misvalued, because the system doesn't account for catcher defense past the most basic level (pop-ups, foul balls, and bunts). There is a more subtle problem, though, which has to do with fielding.

As we all know, it's hard to create a reliable, unbiased measure for fielding ability. FanGraphs uses UZR/150 (ultimate zone rating per 150 games), which compares a player to the average at his position. So when FanGraphs sticks a "3.6" under the fielding column on the value chart, it's saying Markakis was around 3.6 runs above average as a right fielder.

That's all well and good, but look at Juan Castro. Was he really 5.6 runs below average in the field? I doubt it. The reason for that anomalous number probably has to do with random luck. Castro played only about 400 innings in the field last season, and was supposedly a well below average fielder in that time.

But take a look at some previous data and it certainly looks like an anomaly—in close to 2,000 combined innings between 2002 and 2007, he was well above average at shortstop and second base (the only positions with significant data). FanGraphs says Castro was worth -$4.7 million compared to replacement level; that might be true, but it's also likely that he just had a statistically fluky year and is actually worth substantially more than that.

What about the future? What kind of deals do Markakis and Roberts in particular deserve? Here are Markakis's career values:

2006: $6.9 million
2007: $13.3 million
2008: $22.5 million

A clear improvement over the three years, and he hasn't even reached his peak yet. Markakis is a great all-around player who can hit, field, and doesn't seem to have injury problems. You have to remember also that UZR/150 doesn't account for a player's arm, and Markakis has an amazing arm in right field—that probably adds a few bucks right there.

It's hard to say that Markakis deserves $22.5 million a year. Can we assume FanGraphs' value projections actually correlate with reality? It's a lot easier to compare him to players who are similarly valued, and see what kind of contracts they got. Among the outfielders who were valued around $22.5 million last season, there are a few big contracts:

Alex Rios ($24.9 million), signed six-year / $64 million deal in 2008
Randy Winn ($21.9 million), signed three-year / $23 million deal in 2006

Other guys like Ludwick and Quentin are still under club control, and haven't signed extensions as far as I know. Anyway, this would seem to imply Markakis actually deserves a deal closer to $10 million a year, $15 million at best.

I'm sure I've used up all of the readers' patience at this point—you can probably do similar calculations for Roberts and anyone else you might care to think about. I personally think this tool, and all of FanGraphs for that matter (check out the lefty/righty split graphs!) is really nifty. Does that mean GMs should use it when they're making deals? No. But can armchair GMs make judgments on their teams' real GMs using these data? Most definitely.