Down Goes the Price on Wins
It has been well documented that free agent salaries have fallen since last offseason. Right now, the average dollar amount per marginal win that MLB teams are paying is $3.56M for hitters and $3.42M for pitchers, according to the Free Agent Tracker over at Sean Smith’s www.baseballprojection.com.
Has this dollar value persisted throughout the entire signing period, or have we seen a decline since free agents began signing in November? This graph will show us our answer (does not include Orlando Hudson’s contract).
The blue and red lines signify the means for batters and pitchers respectively.
There isn’t any discernible pattern or flow to this market. As we proceed into March, it’s possible that remaining talented players such as Johnny Damon and Felipe Lopez may have to accept below market deals in order to find roles on teams. However, this won’t have a large impact on the market as a whole – based on CHONE projections and the signings that we’ve seen this winter, the market for wins has indeed fallen below $4.0M, and seems to be settling around $3.5M, regardless of when the contract was signed.
As we push towards the end of the recession, we should see an upswing in this amount. If free agent salaries don’t rise towards or above $4.0M next season, it means that baseball (and likely America as a whole) is still feeling the effects of recession. Some indicators do suggest that the recession is headed towards an end (or over already), and so a surge in the winter of 2011 is likely.
39
I like how fangraphs always pushed the boundaries. Who needs headlines? What is the marginal value of a head vs its marginal cost?
I kid, I kid. I like the work.
^What is the marginal value of editing vs rushing to be the first poster? Probably about 10 thumbs down for failed snarkiness.
Headlines probably have more marginal value than content……….
David Appelman moonlights at The Onion.
Doubt it, Jack. No way is the recession over nor is it close to being over. 2011 will not be an accurate indicator.
With economists suggesting unemployment will hold above 10% into 2011, and with MLB attendance down 5mill from ’08 to ’09 (no idea about merchandising/ad revenue, but I’d assume those are down too), isn’t it likely that attendance and advertising will continue to be suppressed far enough into 2010 to drive down the value of a marginal win even further next winter?
As an economist, let me assure you that economists have clue what direction the economy’s headed this year.
Unemployment fell to 9.7% in January.
Of course UE only fell to 9.7% b/c more people stopped looking for jobs, i.e. discouraged workers. Total employment fell.
Not a bullish sign for the economy at all.
I’m concerned about the 5-year ARMs that are just now starting to adjust to higher interest rates. We’ve had a lot of time to fix those and still far too many people are saddled with unreasonably high looming interest rates.
Why doesn’t it ever seem like the SABRmetric community takes credit for the overall decline in salaries?
Remember just a few years ago, ridiculous offers were being given to the likes of Zito, Soriano, Lee, and others?
Now even the Yankees say that certain players’ asking price is too high. The Yankees. They have $200MM+ a year to throw around and told Johnny Damon / Scott Boras to screw off.
From aging curve research suggesting a peak at ages 26-27 (well before most players even sniff free agency) to a generalized better understanding of player worth, it just seems like many a GM would rather just play the value game. Part of the reason marginal win values were so high in the 00′s, IMO, was because teams would throw insane dollar values at mediocre players. As evidenced in this Old but good Jeff Passan piece
SABRmetric research on value will affect relative salaries (player type x vs player type y), but I don’t see how it could produce an overall decline in salaries. Overall salary is determined by the amount of money owners budget for payroll.
I’m just implying SABRmetrics are causing teams to be smarter with their money.
Hejuk-Yes, that is true.
But I believe payrolls will go down as GMs are choosing equally talented younger players who are still in their arbitration years and getting paid a reduced rate over their older counterparts that have no artificial constraints on their salaries.
SABRmetrics has played a significant part in changing GMs philosophies in terms of the value they put on veteran/younger players.
If teams find that development is more rewarding than signing free agents and put more money into draft signing bonuses, international signings, and other developmental things (coaches, equipment, technology, analysts), then the money spent on free agents can go down.
Very interesting graph. Thanks.
One question: If you remove all signings by Jack Zduriencik, does it change the dollars per win averages by anything significant?
I don’t think Jack Z has signed enough talent to have any significant effect on things.
For a site that holds stats in such high regard, it would be nice to see some confidence intervals or error bars once in a while. Especially with such small sample sizes.
I’ll do it for you.
According to CHONE, the Standard Deviation of the $/MW for hitters in 2010 is $1.5MM.
For pitchers, it’s $1.3MM
More importantly, at least using this data, no proof that teams pay more per marginal win for good players than average-bad ones.
Even though that factors in the CHONE projections that likely underrate players like Mike Cameron / Adrian Beltre / etc.
CHONE underrates defense? I had paid close enough attention to this to realize. Or does it just heavily regress players’ UZR, even with established high levels of defensive performance? Or maybe it just underrates Red Sox. ;-)
I keed, I keed!
Not so much underrates defense as I think it’s giving a harsh aging curve to them.
It’s projecting Cameron to go .235/.318/.393, and post average defense in CF. Adjusting that Ron Gant played in Colorado for most of his age 36 year, Cameron’s best age 36 comp is probably Darrell Evans (since Cameron is a TTO guy, a lot of his comps fell off quicker than he has). Evans went .241/.354/.459 in his 37-38 years, and recorded 1,077 PA. I see no reason given Cameron’s superior conditioning that he can’t record similar numbers and be around the +4 UZR/150 CF I predict.
Essentially, I’m thinking +7 bat, +4 glove, +2 position ~ 3.3 marginal wins high end-average scenario, and CHONE as almost a worst case scenario.
Okay, gotcha. That does seem particularly harsh on Cameron.
“More importantly, at least using this data, no proof that teams pay more per marginal win for good players than average-bad ones.”
This data doesn’t have the necessary information to determine that either way, so that’s kind of a pointless statement.
Yes it does.
It has projected marginal win values, and how much teams paid per marginal win. From there, you can run a regression using Marginal Wins as the X and $ per Marginal win as the Y.
0 falls well within the range of the slope.
So, if my 15 year old stats knowledge is correct, the 95% confidence interval is +/- about $2.5M. The error bars overlap. The difference from last year could be due to chance. It may be real, but the small sample size means you don’t know.
“since Cameron is a TTO guy, a lot of his comps fell off quicker than he has”
Because he’s a better defensive player than most? Or because he keeps himself in better shape than most?
I’m not sure why you would expect other TTO guys to fall off quicker than him.
P.S. (SABR)metrics. I like how you spell it that way.
Down goes the price on (Randy) Winn, also.
There seems to be an inter-season negative trend as well. That is to say, the earlier a player is signed, the more $/win he’s gotten.
- Is this unique to this season?
- Is is just noise?
- Is it a function of the market moving down from a higher starting point where last season left off? (greater fluidity than year-to-year)
- Is it a function of higher win players going earlier in the season combined with each marginal win being of increased value? I think this is interesting — do teams really treat each marginal win the same or does it scale within a given player’s contributions (e.g. win1=$2.5, win2=$3.0, win3=$4.0, etc.)
We kind of saw that with the corner out-fielder market last year, with Raul Ibanez first and everything declining after that. Though that may also have been the first signs of the shift towards valuing defense more highly.
I lean towards the statiscal anomaly side as things as there wasn’t a real bumper crop of big ticket free agents. There was a few upper tier free agents that were snapped up quickly and later Holliday signed. Was that indicative of the price per win or of the availability of higher value players?
My opinion, it was an odd sample set.
Could we just be seeing the effect of the Yankees making so many big-ticket signings last off-season and mostly standing pat this one?
It’s just that younger potential Free Agents don’t hit the market much anymore. We’re seeing them locked up well past their arbitration eligible years. Most of the Free Agents hitting the market these days are past their prime.
IMO, when calculating value per win, it should include the arbitration class of players that are setting the salaries, such as Josh Johnson, King Felix, and Justin Verlander.
Good point. It certainly seems as if we’re seeing a lot more extensions that buy out FA years than ever before. Also, I think teams are being more cognizant of the value of calling up players later to delay their service clock.
I (the author of a book, “A Modern Approach to Graham and Dodd Investing” that called the 2008 crash) am calling for a “double dip” recession. If I’m right, the value of a marginal win will continue to fall, below $3 million.
Sorry to say all prices may go down but never on games. thaks