Manny on the Market
For the last month or so, we have primarily been focusing on dollar valuations and factors outside of offense that need to be added into any type of evaluation. This has enabled us to determine the fair market values for several players as well as figure the fee certain teams are paying for each added win. One player left out of such posts is Manny Ramirez. Per reader request, we will figure out exactly what Manny’s fair market value is, taking into account his offense, defense, and baserunning.
Offense
One of the best hitters of all time, Manny is still one of the best hitters today, even at 36 years old. In 2008, he hit .332/.430/.601, good for a .432 wOBA and +56 runs above average, his best offensive season since 2002. In fact, via WPA/LI, 2008 was the best offensive season of Manny’s entire career! Moving forward, Marcel projects Manny to produce a .389 wOBA in 2009, which would translate to +30 runs above average. Not as potent as last year by any stretch, but still more effective offensively than the likes of Adam Dunn, Pat Burrell, Bobby Abreu, and the recently-signed Raul Ibanez.
Defense
Defensive excellence and Manny Ramirez, quite simply, do not belong in the same sentence. Unless, of course, the sentence is sarcastic and designed to point out how much of a liability he can be with the glove. UZR has him all over the place since 2002, ranging from +4 runs back then to -23 runs in 2005. From 2005-2008, his UZR numbers are -23, -20, -18, -5. His defense looked substantially better in 2008, but that does not make him solid in the field. A convervative UZR projection would peg Manny as a -10 runs fielder in 2009.
Baserunning/Adjustments
Via BP’s equivalent baserunning runs, Manny was worth +0.25 runs on the bases last year, and -2.05 runs the year before. Then, for whatever reason, the data does not seem to be there. Let’s call him -1 run on the bases, not terrible, not impressive.
As a corner outfielder, Manny should be docked -7.5 runs per 162 games. Projected to play 150 games next season, Manny’s prorated adjustment would be -6.9 runs. To determine value above replacement, and not average, the +20 runs per 700 PA also needs to be added. Prorating that, based on 584 projected plate appearances, adjusts this to +16.5 runs.
Overall
Adding the rounded results together: +30 + 17 – 10 – 7 – 1 = +29. Manny is a +2.9 win player in 2009. Assuming a going rate of $5 mil/win, Manny’s fair market value right now is $14.5 mil. A 3-yr deal, in a vacuum, would be worth closer to $42 mil. Suffice it to say, he is going to make much more than this. It is too difficult to quantify the merchandise and ticket sales that he will add to a team, but even with that, he does not seem to be worth the $22.5 mil/yr the Dodgers offered him earlier. Then again, accounting for the value of the actual marginal wins he is adding, he could very well be worth that fee.
Then again, perhaps the team feels that his -5 defense in 2008 is more a sign of things to come than the putrid numbers posted from 2005-07. If that is the case, then they are valuing Manny as a +34 run player, or +3.4 wins. That fair market value would be $17 mil/yr, and with the ticket sales and merchandise that would follow, it does seem more realistic that he could be worth about $22 mil/yr. For him to truly command a deal that impressive, he would have to be worth closer to four wins offensively than three, show that his defensive shortcomings are closer to a half-win docked rather than two full wins, and really keep a fanbase entertained.
Assuming -10 runs defensively, for Manny to truly command $22.5 mil/yr, he would need to be +45 runs on offense/baserunning combined. 45 runs offense + 17 over replacement – 10 defense – 7 position = +45. +45 runs = +4.5 wins. +4.5 wins * $5 mil/win = $22.5 mil.
Eric is an accountant and statistical analyst from Philadelphia. He also covers the Phillies at Phillies Nation and can be found here on Twitter.
Fenway has a known park affect on defensive numbers in left field. The wall prevents Fenway left fielders from making plays on fly balls that left fielders in other parks catch for outs.
He could… DH?
This is primarily from the standpoint of the Dodgers. If he DHs, he would be docked something like -15.5 runs as an adjustment. That would put him +30 offense, +17 over replacement, -15.5 adjustment, -1 for baserunning for +30.5, which we can round to +31 to make things easier.
If he DHs, he is worth +3.1 wins, if he is -10 in the LF he is worth +2.9 wins, and if he plays -5 defense like this year, +3.4 wins. No matter how you slice it, his FMV comes between $14.5-17 mil.
The only way for this to really work, in giving him a $23-26 mil/yr deal, is if he performs offensively closer to 2008 than his true talent level. This is possible, but for Manny to truly command $23 mil before taking into account tickets and merchandise, assuming -10 runs defense in LF, he would need to be +45 runs on offense/baserunning.
In that scenario, +45+17-10-7 = +45 runs, +4.5 wins. 4.5 wins * $5 mil = exactly $22.5 mil.
Defensive excellence and Manny Ramirez, quite simply, do not belong in the same sentence.
And yet, there they are… 🙂
But wait! I had the qualifier – “…designed to point out how much of a liability he can be with the glove.” I think it just might meet that.
If he signs a long term deal, don’t we have to consider the deal as a whole, not just the first year? The deal could be okay the first year, but then become an albatross in the future as Manny ages.
For whatever reason, teams don’t tend to sign players like that. They see what he is worth the next year and give him a contract accordingly. And how much of a star the player is generally combines with age to determine the years.
But wait! I had the qualifier – “…designed to point out how much of a liability he can be with the glove.” I think it just might meet that.
Well played, sir!
The Dodgers have traditionally been a pitching team (at least since they won the 1965 pennant and World Series with a .245 team batting average). Hitting is the scarce commodity in Dodgerland.
So the LA fans go ga-ga when they see anyone that can hit at all, never mind as well as Ramirez. Look at the way they reacted in 2004 to the “dumping” of Paul Lo Duca, a league average hitter (outside of LA), who was a liability on defense and off the field.
One reason that some teams systematically undervalue defense is the greater “fan value” of offensive over defensive “production.” A win is a win, no matter how produced, but not to some fans.
I don’t think it’s too far-fetched to think that Manny might be worth +45 runs with his offense & baserunning. Other than 2007, Manny’s lowest wRAA in the last decade was 45.6. Yes, he’s 35 years old, and the projection systems assume a normal aging curve, but he wouldn’t be the first superstar to defy the normal effects of aging.
He may be the first to do it without chemical assistance.
Well that’s how you get dreadlocks, do it without chemical assistance.
Eric,
I have a question. In this Manny piece, you are comparing him vs. a league average player (correct?). So your measuring his Wins Above Average.
I noticed in the Burnett Article written by Dave Cameron, he is calculating WAR in order to figure out what he is worth.
Why for hitters are you comparing to league average, but for pitchers its compared to replacement player?
Also… since most fielding systems compare to league average, since replacement means a whole different thing for defense, how do you use UZR/150 for example when calculating WAR?
Thanks for all the greaet work.
Josh,
I’m also doing WAR. Dave and I do the same type of analyses in terms of dollar valuations. The thing is, though, that when we use wOBA and UZR to calculate offensive runs and defensive runs, they are BOTH relative to the league average.
So, if we have a player with +25 wRAA and +20 UZR, he is 25 runs above average offensively and 20 runs above average defensively. In order to calculate value over Replacement, and not Average, we prorate +20 runs. So, if a guy had 700 PA, he would get +20 runs added because an average player is about 20 runs above replacement. In Manny’s case, because of his projected playing time, he gets +16.5 runs above replacement.
The big difference is that with pitchers, we use a distinct FIP for a replacement starter and reliever. For hitters, we calculate above average and then adjust above replacement.
Geez, I don’t know how I missed that, thanks Eric.
So the best strategy is to use RAA for offense, defense, base running (does wRAA include base running on this site?) and then convert it to RAR using the position adjustments outlined by Tango or Sean Smith?
Thanks
Eric,
one more comment, it is my understanding that the fangraphs wRAA includes SB and CS in the calculation.
However, you are using BPs eqBRR which already calculates SB and CS, etc.
Josh, for the first comment, yeah what you said is exactly right. And then Tango/Sean have the same adjustments: -12.5 runs for 1B, +12.5 for C, +2.5 for 2B/3B/CF, -7.5 LF/RF, +7.5 SS.
As far as the second, yeah, using the baserunning metrics isn’t concrete yet. However, in Manny’s case, it makes very little impact. Ideally, the Bill James baserunning metric could be turned into runs since that factors in the advancements from 1st to 3rd, etc.
what i decided to do was use BPs EQBRR-EQSBR which should eliminate the double counting. I hope.
Yeah, that would do it. Add in the EQBRR sans steals. For the future I’ll likely do that. Though it is a tad incomplete given we only have it for 2007 and 2008. That could be enough, but I like 3 yrs.
I agree, preferring 3 years… the other thing, i noticed, is i am plugging in marcels projected wOBA and using the formula for wRAA which i do not believe includes sb/cs…. its basically a zscore for wOBA… its (wOBA – .328)/121 * PA
I am using that, so that I can adjust the plate appearances (which is marcels real weak spot if you ask me). So in this case, I can use eqBRR without removing the steals (i think).
Josh, wOBA here includes SB and CS, so no matter what you use, if you get wOBA from this site, those are included. I think the formula you meant, if it’s the one we used here before introducing wRAA, is as follows:
((wOBA – avgwOBA)/1.15)*PA)
The league average is usally like .332, at least that’s what Dave and I tend to use.
I used the one linked to on tango’s site that had the specific 2008 data because its supposed to be more accurate to use current data.
I got that idea from the article and following comments from the wRC and wRAA article
http://www.fangraphs.com/blogs/index.php/wrc-and-wraa/
Although i guess when forecasting it helps to use a standard
what are your thoughts?
this is where i got my league wOBA and scale which seems like its generally higher then 1.15
http://tangotiger.net/bdb/lwts_woba_for_bdb.txt
Josh, the wOBA and all wOBA stats in the player pages and leaderboards are calculated using those values you linked to (with possibly extremely minor differences) and using that particular year’s “wOBA Scale”.
I will note that the wOBA used in the Marcel calculations is whatever was provided in the original 2009 Marcel projections. To calculate wRAA and wRC for Marcels I use the league average .332 that was provided and the 1.15 wOBA Scale. Obviously you’re free to calculate wOBA however you want off the projections, but we provide the projections with the same numbers as they’re submitted and try not to alter anything that’s already calculated in the projections unless otherwise advised by the authors.
Thanks David,
Question, are you saying that the projected 2009 marcels to use is a league average of .332 (wOBA) and a scale of 1.15. I simply used the 2008 (whcih i realize is somewhat inaccurate as you are not going to get an identical season in 2009).
For some reason I naturally used the latest one.
Josh, wOBA in the Marcel projections appears to be calculated using the “standard” wOBA formula which is:
(.72* NIBB + .75 * HBP + .9 * 1B + 1.24 * 2B + 1.56 * 3B + 1.95 * HR) / (AB + NIBB + HBP + SF + SH)
For this formula, the wOBA scale is 1.15 and the league average happens to be .332, so those are the numbers you’ll want to use if you choose to use the wOBA projection provided with the Marcels and found here on FanGraphs.
If you are using the raw Marcel projections to calculate wOBA differently (say using the 2008 values found in Tango’s spreadsheet), you’ll want to calculate wRAA etc… using whatever the Scale is for that year (1.21 in ’08).
Where do you get $5 million/win? According to the book Diamond Dollars (ISBN-13: 9780977743636), each win that does not significantly impact a team’s playoff potential is worth aprox. $1.1 million. The difference between 89 wins and 90 wins is obviously far more significant, and is likely to be somewhere around the $6 million range.
The point being that Manny Ramirez is worth different amounts to different teams. That said, I completely agree with your assessment that no matter who signs him, he is nowhere near worth the 20 something million he will probably get (actually, no single player is according to the affore-mentioned book, great read by the way). I suspect that with the current economic climate that this will be more true than before.
Testing, Testing