Revisiting Last Year’s Free-Agent Signings

Before all our attention is focused on the post-season, I thought I’s take a quick look back at free agent signings from the past year and how those deals worked out in 2012. The focus here is just on what teams got for their money. In other words: Did the players meet or exceed the expected value of the contracts they signed?

I focused on major league signings only, so the analysis does not include myriad minor league deals — many of which resulted in players accumulating playing time in the majors this year.

To get a sense of the how the deals turned out, I compared players’ expected values — which are based on their positions and the annual average value (AAV) of their contracts — to their actual values. I uses Matt Swartz’s research on the differences in dollars per Wins Above Replacement (WAR) by position, rather than assume an average dollar-per-WAR, as is typically done.

For reference, here are the positional averages from 2007 through 2011:

$/WAR 2007-11 POSITION
$10,700,000.00 RP
$6,300,000.00 CF
$6,100,000.00 1B
$4,900,000.00 LF
$4,600,000.00 DH
$4,400,000.00 RF
$4,000,000.00 SP
$3,900,000.00 3B
$3,100,000.00 C
$3,000,000.00 SS
$2,500,000.00 2B

Using these multipliers, I calculated expected WAR for 2012 by dividing the AAV of each player’s contract by the estimated $/WAR based on their position.

On balance, teams did a pretty good job of valuing players and signing them to appropriate deals. The 111 signings under examination here were expected to produce 102.8 WAR and so far have produced 115. Fifty-one of the signings produced more value than expected, while 60 produced less value. Looking at the frequency of differences between expected and actual WAR, we see that there were more players outperforming their expected WAR by greater than 1 WAR than there were players who underperformed by more than 1 WAR (18 versus five).

Let’s take a look at how the signings turned out when considering positions:

POSITION N Exp WAR Actual WAR Difference
3B 8 0.8 1.4 0.7
LF 10 0.5 1.1 0.6
2B 6 1.2 1.6 0.4
DH 3 0.6 0.9 0.3
CF 2 1.0 1.3 0.3
RF 7 1.4 1.6 0.2
SP 18 1.6 1.7 0.2
RP 30 0.3 0.3 0.0
C 11 0.5 0.3 -0.2
1B 7 1.5 1.2 -0.3
SS 9 1.9 1.5 -0.4

Shortstop was expected to produce the most average WAR — driven mostly by the Marlins signing Jose Reyes. Reyes was initially valued at 6.2 WAR this year, but he’s only accounted for 4.5 WAR. And while Reyes was the biggest disappointment at his position, he wasn’t the only shortstop to underperform his contract. Alex Gonzalez and Jack Wilson all underperformed by 1.1 WAR.

The greatest additional value this year has been provided by third basemen (+.7), followed by left fielders (+.6). Eight third basemen were signed and were expected to produce .8 WAR, on average. Outside of Aramis Ramirez, no other third basemen was expected to produce more than .8 WAR. Ramirez has certainly outperformed his expected WAR this year (+3) — but right on his heels, in terms of additional value, has been Tampa Bay’s Jeff Keppinger (2.7 WAR versus .4 expected WAR). Greg Dobbs has been the biggest disappointment, giving -1.2 WAR while Miami assumed .4 this year.

The Twins grabbed one of the greatest bargins out of left field with Josh Willingham and his 3.8 WAR — 2.5 more than expected. Ryan Ludwick has been huge for the Reds this season, considering the team’s offense has been average (+2.5 WAR). The New York Mets resigned Scott Hairston for $1.1M, but have been treated to a 2.1-WAR season (+1.9) thanks to his production against left-handed pitching (137 wRC+). On the negative side of the ledger, Endy Chavez was supposed to produce roughly .3 WAR for Orioles but posted -.9 WAR for the season.

We can also look at how teams did, overall, looking at their signings as a portfolio:

Team # of Free Agents Average WAR above Expected
Texas 2 1.8
Oakland 4 1.4
Boston 3 1.0
Cincinnati 2 0.8
Tampa Bay 5 0.7
Milwaukee 3 0.6
Arizona 8 0.5
Chicago Cubs 4 0.4
NY Mets 4 0.4
Minnesota 6 0.4
Detroit 4 0.4
NY Yankees 3 0.3
Seattle 4 0.2
LA Dodgers 9 0.1
Kansas City 4 0.1
Philadelphia 7 0.1
Baltimore 5 0.0
San Diego 2 -0.2
Pittsburgh 4 -0.3
Toronto 2 -0.3
Chicago Sox 1 -0.3
San Francisco 4 -0.3
Houston 1 -0.4
St. Louis 3 -0.5
Washington 4 -0.5
LA Angels 3 -0.6
Cleveland 2 -1.1
Atlanta 1 -1.1
Colorado 3 -1.2
Miami 4 -1.3

The Rangers, Athletics and Red Sox were the only teams that averaged at least 1 additional WAR per signing. Suprisingly, Oakland’s numbers were not fueled by Yoenis Cespedes — who has been about even in terms of expected and actual WAR — but by Jonny Gomes (+1.3), Coco Crisp (+1.5) and Bartolo Colon (+2.1). Boston benefitted mostly from batt-flipping Cody Ross (+2.0), who posted almost 3 WAR manning right field for the Sox. Texas saw the most additional value from its two free-agent signings: Joe Nathan (+1.1 WAR) and Yu Darvish* (+2.4 WAR) have both been outstanding — posting a 60 and 74 FIP-, respectively.

Miami’s spending spree didn’t work out as planned. There will be no postseason games for the Marlins, partly due to the fact that every one of their free-agent signings underperformed, relative to their contracts: Mark Buehrle -2.0, Jose Reyes -1.4, Greg Dobbs -1.2 and Heath Bell -0.4.

Splitting up the signings by those where players were re-signed and those who switched teams, we see something interesting. Contra Matt’s hypothesis (backed by analysis by MGL) that teams that re-signed players tended to see better value — relative to signings where players switch teams — 2012 saw players who switched teams produce an additional 12.2 WAR above expectations (96.9 versus 84.7). Players who re-signed produced almost exactly what they were expected to (18.1 WAR). This isn’t to say Matt is wrong (hint: he’s not), just that in this particular year the league did a pretty good job of getting value from players that switched the teams. The Diamondbacks re-signed six players — and thanks mostly to Aaron Hill‘s 5.5 WAR — benefitted from an additional 4 WAR above what they expected.

All things considered, 2012 was a pretty good one for free-agent signings. Considering that there were some players who didn’t get a chance to step on the field this year — like reliever Ryan Madson — it’s reasonable to say that the numbers could have been even better.

——–

Data for this article is current as of Sept. 27, 2012. Free agent signings compiled mostly from ESPN.com.

*Darvish’s AAV does not include the posting fee. If it did, Darvish would still be outperforming his expected WAR (+.2), and the Rangers would drop to fourth overall in terms of average WAR above expected.




Print This Post



Bill works as a consultant by day. In his free time, he writes for The Hardball Times, speaks about baseball research and analytics, consults for a Major League Baseball team and appears on MLB Network's Clubhouse Confidential. Along with Jeff Zimmerman, he won the 2013 SABR Analytics Research Award for Contemporary Analysis. Follow him on Tumblr or Twitter @BillPetti.


28 Responses to “Revisiting Last Year’s Free-Agent Signings”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Vinny says:

    Love this post. Great work Bill!

    Vote -1 Vote +1

  2. Matt says:

    I wonder how the analysis looks – and whether positions like RP do any better or worse than expected – using WPA instead of WAR. Is there even an “expected” WPA?

    Vote -1 Vote +1

  3. Julian says:

    Was there any difference for players signed to multiyear deals? Theoretically players signed to longer-term deals should have managed to outperform their AAV by more than those signed to 1 or 2 year deals. Maybe this is why the free agent class appears to have done so well.

    +9 Vote -1 Vote +1

    • Bill Petti says:

      Good question:

      Deals >1 year: +12.2 WAR above expected
      Deals = 1 year: 0 WAR above expected

      Essentially, on average 1-year deal players performed right to expectations.

      Vote -1 Vote +1

      • vivalajeter says:

        Aren’t >1 year deals expected to be above their salary in the first year though? When you sign Pujols or Fielder to a monster deal, you’re expecting to get most of the value early in the contract in order to offset the bad years at the end.

        +5 Vote -1 Vote +1

      • Preston says:

        Yes, and one year players should get exactly as expected. So these numbers show that on average FA’s produced as expected.

        Vote -1 Vote +1

  4. Table says:

    I wanted the Dodgers to go after Aramis Ramriez for third (with Uribe becoming a super sub). Re sign Jamey Carroll or if he leaves replace him with any of the other decent free agent 2nd basemen (Ellis, Barmes). Sign Josh Willingham and Cody Ross for 1st/LF. If enough cash left sign Ramon Hernandez, if not AJ alone would be fine. Resign Kuroda. Fill in the back of the rotation with cheap signing like Francis.

    I’d say for the most part I had the right ideas.

    Just don’t ask me about who I was rooting for them to sign in the 2006 offseason………………Dice K cough cough

    Vote -1 Vote +1

    • Jason B says:

      Hrmmm…I never expected to see the word “super” and the word “Uribe” in the same sentence. At least not a baseball-related sentence. Maybe something akin to “Uribe did a super job demolishing that buffet at Ryan’s Steakhouse last night.”

      +7 Vote -1 Vote +1

      • Table says:

        Super sub as in he would not be only backing up one or two positions in the infield, but all of them. Phrase has nothing to do with his offense.

        Vote -1 Vote +1

      • steex says:

        I think it’s one of those terms without a real clear definition. I look at a “super sub” as someone like Allen Craig entering the year – someone with a bat that needs to be put in the lineup, but doesn’t have a clear starting position and has some ability to back up at several. Someone who backs up all the infield positions is just what I’d call a utility infielder. However, there’s no way to suggest my interpretations of the terms are more correct than yours.

        Vote -1 Vote +1

  5. Robbie G. says:

    There is no discussion of the Jonathan Papelbon contract here. It appears to me that teams can, if shrewd, obtain the biggest bang-for-their-buck by signing undervalued relief pitchers (and there are generally many such players, each offseason) to bargain contracts and then utilizing them properly. In which case, even if Papelbon is technically delivering anticipated production, the Phillies nonetheless grossly overpaid him (relative to his peers), since they could have simply signed, say, Fernando Romney for significantly less money and received comparable (if not superior) production.

    It just never seems very bright to give a bunch of guaranteed money to a relief pitcher whose name isn’t Mariano Rivera, and Tampa Bay is surely not the only team that routinely puts together a dirt-cheap relief pitching core that outperforms expectations. So it can be and often is done, which makes you wonder why any team is willing to hand out huge sums to relief pitchers, ever.

    Vote -1 Vote +1

    • bpdelia says:

      ha. we have this discussion every year. the only defense for reliever spending is the premium value based on a, known quantity. for a team with a good sized market and read realistic aspirations of a title it would make sense to overpay for a sure thing. because trade deadline reliever moves are almost guaranteed to, be overpays. for instance the soriano signing for ny or the paps signing for Philly. with the ability to have payroll flexibility it makes sense to overpay a SURE THING RP to make sure you aren’t desperately giving away prospects in July because you tried to save 4 percent of your payroll in January. so big market contending teams should overpay reliable relievers imo. obviously MLB teams value sure thing RPs differently than we do. its still going on well into the mainstream saberetric era so I thinkits safe to assume clubs are using different metrics to set elite RP value

      Vote -1 Vote +1

    • vivalajeter says:

      Robbie, the problem is that you typically don’t know which relievers were undervalued until after the season is over. Nobody knew Rodney would have this type of season, and it easily could’ve been money down the drain with him. Sometimes you get a random reliever and he has a phenomenal year, and sometimes you get one that gets lit up and released in April. If people knew Rodney would be this good in 2012, he would’ve had much better offers out there.

      Vote -1 Vote +1

      • Robbie G. says:

        I realize that Fernando Romney has far exceeded expectations but you still have teams like Tampa Bay that routinely (as in, every single year) put together very good bullpens on the cheap. I believe Papelbon is making around the same amount of money this season as Tampa Bay’s entire bullpen, and Tampa Bay surely has a better bullpen than Philly. And, again, teams like Tampa Bay are doing this on a regular basis. So it is just never going to make sense, in my mind, for a team to throw $30+ mil at a single relief pitcher. I realize that teams are making this mistake on an annual basis but I still say that it qualifies as a mistake.

        Vote -1 Vote +1

      • SKob says:

        If Tampa Bay does it then it must be easy! No way they are just far superior at coaching pitchers. Go on the cheap and be awesome… duh!

        Vote -1 Vote +1

  6. Matthias says:

    Fernando Romney—Mitt’s nephew that he wants sent back to Mexico (despite him actually being Dominican).

    Vote -1 Vote +1

  7. Calvin says:

    This would be more interesting limited to people intended to start (or be 25-man rostered RPs). You usually don’t want your FA backups to exceed their WAR/$, especially on a 1-year deal.

    Vote -1 Vote +1

  8. Doug B says:

    Aramis Ramirez illustrates the difficulty in doing what you are doing. In theory you are right saying he outperformed the contract by 3.0 WAR… a huge number. But a very large amount of his money is back loaded. So the Brewers EXPECT him to outperform in year 1, match in year 2, and underperform in year 3. I would be shocked if Ramirez outperforms the contract in year 3, because in addition to being 2 years older his pay is more than double what it is in 2012.

    Vote -1 Vote +1

    • Bill Petti says:

      That’s true–it’s unlikely that multi-year deals will outperform the farther out you get from year 1. However, if they are already underperforming in year 1 it’s a pretty clear signal that something was way off with valuations.

      Vote -1 Vote +1

      • Doug B says:

        I suggest just taking an average salary for the contract. Some are more front loaded and some are heavily back loaded. This would at least remove such a big factor wrecking the data.

        Vote -1 Vote +1

    • Bill Petti says:

      That’s what I used–AAV for the life of the contracts. It mitigates the effect you are talking about, but not completely.

      Vote -1 Vote +1

      • Doug B says:

        Aramis Ramirez was listed in the fangraphs article of the worst free agent signings this year (for being heavily overpaid).

        While I realize that was not the same author… something seems amiss. Are you saying Ramirez outplayed the average of his 3-year contract by 3 wins? I think he’s getting close to $12,000,000 per season if it’s averaged out.

        Vote -1 Vote +1

      • Bill Petti says:

        (Sorry, can’t reply to your message for some reason).

        Yes, the AAV is $12M. Free agent 3B have averaged $3.9M per WAR since 2007. Ramirez was therefore expected to produce 3.1 WAR this year based on AAV and he’s produced 6 WAR.

        Vote -1 Vote +1

  9. Doug B says:

    wow. thanks for correcting me. I guess he was NOT one of the 10 worst free agent pickups then.

    Vote -1 Vote +1

  10. mullahjim says:

    great work. but i do question your opinion of josh willingham. he certainly produced on offense. but he cost us several games due to his poor defense.

    Vote -1 Vote +1

  11. matt w says:

    I’m a little confused by the math on Clint Barmes. According to Fangraphs he has 1.7 WAR. At $3M/WAR for shortstops that means he was worth $5.1M, which is just about in line with his AAV of $5.25M, isn’t it?

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>