Designated HitterMarch 16, 2006
Long-Term Free Agent Contracts: A Historical Perspective
By David Regan

As the old English proverb goes, hindsight may be 20/20, but there's a lot of evidence that signing another team's free agent to a long-term, big money contracts is poor business. My interest in this topic was set in motion last year when the Dodgers shelled out Darren Dreifort money to perennially-injured outfielder J.D. Drew, 5 years $55 million. Right away you heard the cries: "DePodesta waaaaay overpaid," "He'll never live up to that contract," and then the comments regarding his walk year being a "free agent push."

When Toronto GM J.P. Ricciardi doled out a combined $102 million to free agent hurlers A.J. Burnett and B.J. Ryan this off-season, similar comments were heard uttered. Perhaps J.P. just likes guys with initials for first names (look for F.P. Santangelo to get a coaching job). It's hard to fault Ricciardi though, as the Blue Jays needed a closer and a front-line starting pitcher and got arguably the best of each that were available (sorry Billy Wagner fans).

All this got me thinking: exactly how true is the statement that signing another team's free agent to a long-term, big money free agent contract is often a losing proposition? It's easy to look back on some of the more noteworthy and controversial contracts of the recent past and wonder, "What was that GM thinking?" As a Dodger fan, I was thrilled that the team added Kevin Brown to the rotation after the 1998 season. So what that they spent $105 million over seven years on a 34 year-old pitcher? It wasn't my money. Brown, of course, was coming off a great 1998 in which be put up a WARP of 9.9 while pitching the Padres to the World Series. He then went on to have good seasons for the Dodgers in 1999 and 2000 (WARPs of 7.7 and 7.9) before being hurt for a good portion of the following five years. I won't even get into the Dreifort contract.

I thought it would be interesting to dig into the numbers and determine whether the "walk year" phenomenon held up to the underlying data. I hypothesized that the good deals would balance the Kevin Brown-type deals and that there might be a slight regression in terms of the free agent's year-by-year performance over the life of the contract as the player aged.

In order for this analysis to encompass an adequate amount of data, I selected all free agent contracts signed during the 2000-2004 off-seasons that met the following criteria: 3 years or more in length at a minimum of $5 million per year OR 4+ years and a minimum of $15 million total. Totals much less than that wouldn't impact a team's budget too significantly, save the bottom-feeders such as the Devil Rays and Royals.

The Results

                             Hitters     Pitchers
Total players in sample         29          17
Average age when signed      30.5 yrs    31.6 yrs
Average contract length       5.7 yrs     4.1 yrs
Average total contact         $55.5M      $43.7M
Average annual salary          $9.7M      $10.7M

The Hitters

Sample average:

          Contract Yr   Year 1   Year 2   Year 3
OPS          0.914       0.831    0.830    0.802
EqA          0.307       0.290    0.283    0.273
VORP          48.1        40.4     36.9     27.6

The good signings: Alex Rodriguez (although at $25 million, you could make the argument that he's still overpaid), Manny Ramirez (see A-Rod comment), Hideki Matsui, Cliff Floyd, Miguel Tejada, Vladimir Guerrero, Richie Sexson, Johnny Damon, Ellis Burks (at least until the final year of his 3 year $20 million deal), Ray Durham, Carlos Delgado (so far).

The bad: David Segui, Todd Hundley, Edgardo Alfonzo, Charles Johnson, Edgar Renteria, Roger Cedeno, David Bell, Kaz Matsui.

The in-between: Moises Alou, Tino Martinez, Mike Cameron, Jim Thome.

Jury is still out: J.D. Drew, Adrian Beltre, Carlos Beltran, Jason Giambi, Ivan Rodriguez, Magglio Ordonez (although there is an argument to be made that this already is a bad signing).

As one might expect, the data gives credence to the oft-uttered "contract drive" theory. Clearly, free agent hitters had excellent years in their walk year and followed that up with ever-increasingly worse years. The biggest offenders (a.k.a. the guys who took a dive) were, by far, catchers Charles Johnson and Todd Hundley. After posting a terrific .954 OPS in 2000, Hundley signed a lucrative four-year $23.5 million with the Cubs and proceeded to put up OPS numbers of .642, .722, and .735 in several injury-riddled years with the Cubs and Dodgers. Johnson, meanwhile, put up a .961 OPS, also in 2000, before signing a five-year $35 million deal with the Marlins (back when they had money). He never came close to 2000 again, posting OPS totals of .771, .670, and .775 between 2001-2003.

More recently, the Dodgers' Adrian Beltre and Carlos Beltran had career years in 2004, posting OPS numbers of 1.017 and .927, respectively. Those years netted Beltre $64 million over five years from the Mariners while Beltran, the prize acquisition last season of the Mets, took $119 million on a seven-year deal. 2005 OPS totals?: Beltre: .713 and Beltran .744. Not exactly what the Mariners and Mets were hoping for. David Bell was given a four year $17 million deal by the Phillies after he hit 20 home runs in 2002, but he's been nothing but a headache for the Phillies since the signing and now they can't give him away.

Guys that continued to build upon their free agent year with their new teams (note that "N/A" refers to future years - i.e. Guererro has been an Angel for just two years):

                      Contract            OPS
                        Year    Year 1   Year 2   Year 3
Carlos Delgado          0.907    0.981    N/A      N/A
Richie Sexson           0.915    0.910    N/A      N/A
Vladimir Guerrero       1.012    0.989   0.959     N/A
Miguel Tejada           0.808    0.894   0.864     N/A
Alex Rodriguez          1.026    1.021   1.015    0.996
Manny Ramirez           1.154    1.014   1.097    1.010

Unlike some of the guys whose performance took the proverbial dive after signing a big-money free agent contract, these superstars of the game proved to be worth the money. I think it's worth noting that none of the players who suffered declines after signing their deals (particularly Beltre, Beltran, Bell, Hundley, Johnson, and Roger Cedeno) have ever been classified as true superstars. Beltran is the closest, but even he had never had a .900+ OPS season prior to 2003. Seems safe to say that the superstars have been, in most cases, worth the big investment.

The Pitchers

Pitchers have been an even riskier investment, and perhaps that isn't surprising given their greater injury risk.

Sample average:

          Contract Year   Year 1    Year 2    Year 3
Stuff         14.3         12.4       5.0      (0.4)        
DERA          3.71         4.02      4.43       4.57
VORP          41.6         22.4      21.6       16.4
For those unfamiliar with the metric "Stuff," it is a measure of a pitcher's overall dominance and factors in strikeouts, home runs allowed, runs allowed, walks, and innings per game. A 10 equates to a league average pitcher. For reference, in 2005, Johan Santana had a 44 Stuff rating while a seemingly mediocre pitcher such as Brian Lawrence had a 6.

The good signings: Mike Mussina, Jason Isringhausen, Tom Glavine (barely), Greg Maddux, Andy Pettitte, Bartolo Colon.

The bad: Aaron Sele, Keith Foulke, Kevin Appier, Andy Ashby, Mike Hampton, Denny Neagle, Steve Karsay, Chan Ho Park.

The in-between: Kelvim Escobar.

Jury is still out: Carl Pavano, Pedro Martinez.

While it's not surprising that our pitchers' years following the signings of new contracts show a decline, the speed at which that decline occurred was a surprise. By the second year of their new contracts, the 17 pitchers in our sample had seen their average VORP sliced nearly in half and by the third year they were teetering on replacement level status. Hitters meanwhile had seen their VORP decline by approximately 23% by the second year. Worth noting here is the impact that a specific pitcher had on these results. His name: Chan Ho Park. After winning 33 games between 2000 and 2001 while making the All-Star team in his last season as a Dodger (2001), Park signed a five-year $65 million contract to go from pitching in pitcher-friendly Dodger Stadium to its polar opposite in Arlington. His new environs coupled with the pressure of a $65 million contract, combined to result in an increase of nearly two full runs in the Korean's ERA (3.79 as a Dodger to 5.76 with Texas and San Diego). Could A.J. Burnett be the next Park? History tells us it's a possibility. J.P. Ricciardi is gambling $55 million that Burnett will help Toronto close the gap in the AL East. In Ricciardi's defense, the team's locale basically forces the team to overpay for free agents.

So what conclusions can we draw from all this?

  • Hitters have historically been a better investment than their counterparts on the mound for teams looking to spend big money in free agency. There's not much risk in signing an under-30 superstar hitter to a long-term deal.

  • Home-grown is the way to go. Instead of overspending on guys who stand a great chance at underperforming once they sign, develop young, cheap pitching talent.

  • Contracts longer than three years for pitchers aren't a good idea. We've seen the rapid drop-off in years two and three of a deal, and it likely won't get any better in year four unless, of course, year four is another contract year.

  • Lengthy and lucrative free agent contracts are not going to go away.

    It is likely that teams will continue to view the later years of a long-term deal as essentially "sunk costs." For instance, the Mets know that it's highly improbable that Billy Wagner will be worth eight figures in 2009, but it took adding the extra year to get the benefit of what are likely to be very good 2006 and 2007 seasons. Unfortunately for most teams, it's the type of deal that would bust budgets in and of itself, so the Wagners of the world will be reserved for the big-market teams only.

    David Regan is a freelance writer whose previous work has been published at a variety of sources, including Baseball Prospectus, InsiderBaseball, and RotoAmerica.com. He welcomes comments, suggestions, and questions via e-mail here.

    • Comments

    If "Contract Drive" is a year's worth of clutch hitting, does it exist?

    Posted by: Rev Halofan at March 16, 2006 2:08 AM

    Interesting article. One thing to note: since you are selecting players with very lucrative contracts, you are bound to see a Contract Year effect. The big contracts come because of big Contract Year performance, then regression to the mean will tend to lead to worse performance in the years following the contract.

    Posted by: John Walsh at March 16, 2006 3:35 AM

    The sample does not demonstrate that the contract year was special in terms of the player's career in that it does not show that year in the context of the preceding year. Thus we do not know that he was playing better than usual to that point.

    Is it possible that the reason for the decline in evidence is a function of age? Note the average age of the free agents studied, both groups over 30. Generally, players will not reach free agency before their late 20s, and often later, so it is natural to assume some drop in production. Imagine if any of the players had been given 5 year contracts at age 25! It would likely be a different story. It seems to me to make sense that teams ought to negotiate with their young stars before free agency on long-term contracts, paying more than they ordinarily would for those years, but locking up the best years of play, rather than paying for declining years. Even risking money on lesser players seems more sensible than gambling on the big ticket older ones. For example, wouldn't the Yankees have done better to give Jake Westbrook a 6 year deal at his market value or more than the 2 years they paid for Kevin Brown?
    A team like the Devil Rays may be in that exact position and might be wise to continue their recent effort to sign their young players to extensions now.

    Posted by: Bob R. at March 16, 2006 3:46 AM

    One thing I would add to this study - it might be useful to also compare the years before the contract year - i.e. Contract -2, Contract -1, Contract Year, New Contract 1, New Contract 2, New Contract 3.

    I feel that this would give us a better perspective - are the players overperforming in their contract year? if you threw out that year, do the previous years make their subsequent performance more understandable? etc. etc.

    Good stuff though. Nice article.

    Posted by: Will at March 16, 2006 5:48 AM

    Another reasonable adjustment to the study would be to use park-adjusted data, since most of the players move teams over the period. In that light, Johnson's and Park's drop-offs are less extreme (though still significant).

    As several of the posters have pointed out, adding the seasons prior to the contract year for context makes a lot of sense. In Johnson's case, a player who only had one season of OPS+ greater than 100 before his 141 OPS+ contract year was not a big surprise to regress back to his career average, posting a 100 OPS+ his first year in Florida.

    Is this a walk-year (causal) effect or outlier (random) effect? Causal or not, it seems to be a pretty bad idea to pay top dollar for an extreme season. (Not that that's news to most of us.)

    Posted by: Doug at March 16, 2006 6:41 AM

    Foulke had a terrible 2005; but the Sox would not have won the World Series the year before without him. How is that a bad signing?

    Posted by: Drew at March 16, 2006 8:13 AM

    Good study. But I would also like to see more years before the walk year.

    Don't teams realize that they are signing guys over 30 who will decline? Could these players still be a good deal even with their declines? That is, are you still getting a number of marginal wins (or Win Shares) that is worth the cost based on market averages?

    I like the comment about Foulke. Maybe it is worth it to pay a little more for a guy if it brings you a Championship or just gets you into the playoffs. That extra revenue might be worth it. For example, if you don't "over pay" some guy, you have zero chance to make the playoffs. But if you do "over pay" him, you have a chance. I guess you have to make sure that the increased chance of making the playoffs or of winning a championship is worth the amount you "over pay."

    Posted by: Cyril Morong at March 16, 2006 8:26 AM

    The problem with "over paying" is that a team doesn't know if a player like Foulke will bring them a championship at the time of signing. We only know in hindsight if: (1) the player performed up to his contract and (2) the team ended up winning a championship due in part to that player.

    Posted by: Rich Lederer at March 16, 2006 8:40 AM

    I agree, looking at contract year -1, -2, etc. would be an interesting exercise. Perhaps you'll see that in a follow-up.

    Posted by: David Regan at March 16, 2006 9:02 AM

    Interesting stuff....I've been crying "DON'T DO IT" in the direction of GMs since I was in high school and saw Jose Lima get paid after his good season with the Astros. It always bothered my parents when I yelled in the house.

    I'm curious why Bagwell didn't make the list? I believe he signed his monstrosity of a deal after the 2001 season.

    Posted by: AgRyan04 at March 16, 2006 9:57 AM

    Bags wasn't signed by another team. He's been with Houston his whole career and I didn't look at those contracts. Just on observation alone, it seems that teams re-signing their own free agents or giving them extensions that bought out free agents years has been far more successful (Pujols, Jeter, Santanta, etc).

    Posted by: David Regan at March 16, 2006 11:40 AM

    Buying out free agent years often times also involves buying out arb years, too. As a result, teams invariably lock up their star players for less money than what they would make in the open market. The combo of getting great players when they are young and getting better coupled with paying them less than, on average, what they could command as free agents is an obvious benefit and isn't really comparable to the study at hand.

    As far as re-signing players who would otherwise be free agents, I'm not so sure these contracts work out any better. You mentioned Darren Dreifort. He would be Exhibit A against this idea. In addition, the Angels, among many other teams, have been guilty of showing perhaps too much gratitude with their star players (like Tim Salmon and Garret Anderson).

    Posted by: Rich Lederer at March 16, 2006 12:03 PM

    Besides the pre-contract year info, find comps of similar age/value and see what is the normal progression for non-FA's. Maybe this is not far from typical regression and GM's may be aware they are not going to get 3 years like the last one and are valuing what 95% of that performance on their team (and not on a competitor) might do to their W-L outlook.

    Posted by: Gilbert at March 16, 2006 12:32 PM

    Ah Darren Dreifort. I see his contract as the epitome of an outlier. I can't think of any instance of a team re-signing its own free agent and having it work out nearly as bad for the team as it did for the Dodgers and Dreifort.

    Posted by: David Regan at March 16, 2006 2:31 PM

    To me, Foulke is still a jury-out signing (if you look at stats only and not his role in '04). If he performs this year like he did in 2004, it remains a good signing because the Red Sox get their money's worth two of three years. If not, well sure, it's a bad signing.

    In either case, it's certainly better than the Yankees' Pavano signing, listed as a jury-out move, especially as he has only put up one good season (a push year), looks to be DL-bound yet again to open 2006 and has produzed zero good years for his new team.

    Posted by: CatNamedManny at March 16, 2006 6:02 PM