Table of Contents

Thursday, January 15, 2009

What's Adam Dunn worth?

We're snow delayed this morning, so I thought I'd take a moment to re-visit Adam Dunn's value (here's my piece on this from last summer).

Word 'round the campfire is that Dunn is still looking for a Carlos Lee-sized contract, despite some of his comparables signing for much less than expected. Pat Burrell, for example, signed a 2 year, $16 million contract with the Rays. Let's look at Pat Burrell quickly:

CHONE has Burrell projected at +13 RAA/150g and 85% playing time, which projects to ~28 RAR. Defensively, Burrell projects to be terrible: -14 RAA in a corner outfield slot, which also incurs a -7.5 runs/season penalty due to being a weak defensive position. Overall, that puts him at -21.5/season on defense, or -18 runs/season given the playing time estimate. Alternatively, if you rate him as a DH, he gets a -17.5 runs/season penalty * 85% playing time = -15 runs/season. Given that he signed as a putative DH, let's go with the latter estimate.

So, total value on Burrell is +28 RAR on offense and -15 runs on defense = +13 runs above replacement in 2009 (and +8 runs above replacement in 2010, given 5 runs/season aging). If free agent contracts are continuing to increase at 10% per season, that puts a two-year contract for Burrell worth 1.3*4.85 + 0.8*5.3 = $10.5 million. So, given this, even in a normal economy (and this probably isn't a normal economy), Burrell's contract seems like it might be a slight overpay. Tango came up with a slightly higher estimate of his value, but even then, in a normal economy, the contract comes in at even value.

Now, let's look at Dunn. CHONE has Dunn projects as a +16 runs/150g hitter and an excellent 90% playing time, which puts him at +32 RAR on offense. That's quite a dropoff from where I had him last season, which was closer to 40-45 RAR after aging. Why? I think it's because CHONE downweights home runs substantially when they happen in GABP, more so than my rough runs-based park adjustments do. If you think it's too dramatic, we can add 5 runs to the CHONE estimate...

Defensively, Dunn's -13 runs/season in a corner slot, + 7.5 run penalty for playing an easy position = -20.5 runs/season * 90% = -18 runs in 2009. Or, as a DH, he gets -17.5 runs/season * 90% = -16 runs/season.

So, overall, we get +32-37 RAR on offense minus 16-18 runs for defense depending on which figure you use. That equals somewhere between +14 and +21 runs above replacement in 2009. I'm going to split the difference again at put him at +17.5 runs above replacement.

Assuming 10% increases in free agent value, that puts his value at:

2009: 1.75 * $4.85 = $8.5 m
2010: 1.25 * $5.3 = $6.6 m
2011: 0.75 * $5.9 = $4.4 m
2012: 0.25 * $6.5 = $1.6 m

So, that puts his value at 1/$8.5, 2/$15.1, 3/$19.5, or 4/$21.1.

My guess is that the offers he's seeing are right around these values, or maybe slightly better than this.


I've been thinking about this more today, and I think I undervalued both Burrell and Dunn above. The reason is that I was using CHONE's R150 numbers as their full season contributions, and then calculating a percentage of that based on an estimate of playing time. The problem with doing this is that R150 is already 92% playing time (otherwise it'd be R162!), so I can't use that as a full season estimate. It's designed to just be a good, optimistic estimate of full season production for most players (few if any players are projectable for over 92% playing time that aren't named Cal Ripken).

I also probably should be using a runs/9.4 conversion instead of a runs/10 conversion to get runs into wins. Offense isn't at 5 runs per game per team anymore.

So, with that in mind, here are revised estimates for Burrell and Dunn's values:

Pat Burrell

+13 RAA/150g / 0.92 = 14 RAA/162 g + 20 runs = 34 RAR * 85% = 29 RAR. Fielding is still -15 runs/season for DH. So, 29 RAR - 15 RAA = 14 runs above replacement total value.

2009: 14 RAR / 9.4 = 1.5 WAR * $4.85 M/WAR = $7.2 M
2010: 9 RAR / 9.4 = 1.0 WAR * 5.3 M/WAR = $5.3 M

2 year value = $12.5 M

Not much different, but this is closer to being accurate, I think. And a little closer to what he got and what Tango came up with.

Adam Dunn

15 RAA/150g / 0.92 = 16 RAA/162g + 20 runs = 36 RAR * 90% = 33 RAR hitter. If you want to be biased and think Dunn's better than that (I think he probably is, but I'm biased), you can add 5 runs and make him a 37 RAR hitter. Fielding remains -16 to -18 runs/season. So we're now at 15 to 21 RAR. Splitting the difference puts us at 18 RAR.

2009: 18 RAR / 9.4 = 1.9 WAR * $4.85 M/WAR = $9.2 M
2010: 13 RAR / 9.4 = 1.4 WAR * $5.3 M/WAR = $7.4 M
2011: 8 RAR / 9.4 = 0.85 WAR* $5.9 M/WAR = $5.0 M
2011: 3 RAR / 9.4 = 0.3 WAR * $6.5 M/WAR = $2.0 M

That puts a 2 year deal worth ~$17M, a 3 year deal worth $22M, and a 4 year deal worth $24M. As I said in the comments below, you'd probably spread the money evenly across the years, but it should sum up to these totals.

Again, not a huge difference, but the small adjustments net him an extra $1 million or so a year.

Adam Dunn at these rates would be a nice improvement for the Reds. He's better than Willy Taveras. If he wants more than that, given that he won't singlehandedly get the team into the playoffs, it's probably the right move not to sign him.