Very few things in baseball receive quite as much derision as large contracts given to relievers. Relievers have come to be seen as fungible, volatile assets who are poor investments. Many view the contracts given to established closers as being entirely based on saves, a stat that is rightfully maligned and makes a poor basis for a multi-year multi-million dollar contract. However, the logic underlying these complaints has holes large enough to push Phil Hughes through, and a closer look suggests that the truly large reliever contracts may actually make a modicum of sense.

My theory is that general managers who hand relievers big money have not been looking for saves per se. Rather, they have been looking for pitchers who have provided consistent performance on a regular basis. To test this hypothesis, I decided to take a look at the largest contracts given to relievers since 2000, as well as the most consistent performers over the same time period. For the contracts, I limited my search to 3+ year contracts worth at least $7 million per season. 3+ year deals tend to reflect a level of trust by the club in the player, and $7 million struck me as a reasonable cutoff between the deals handed to top players and to those a level down on the talent chain. For measuring performance, I used a simple ERA+ and IP combination to try and isolate the most consistent performers (a search for relievers who have racked up 35+ saves on a yearly basis unearthed a similar list. Players who provide that many saves regularly tend to have strong underlying numbers, so saves can serve as a proxy for performance when addressing a multi-year sample).

Here’s the list of pitchers who had at least 3 seasons with an ERA+ of 150 or better and at least 65 innings pitched:

1 Joe Nathan
2 Billy Wagner
3 Mariano Rivera
4 Francisco Rodriguez
5 Keith Foulke
6 Mike Adams
7 Joakim Soria
8 Carlos Marmol
9 Jonathan Papelbon
10 Jonathan Broxton
11 B.J. Ryan
12 Juan Rincon
13 Brad Lidge
14 Francisco Cordero
15 LaTroy Hawkins
16 Luis Ayala
17 Eric Gagne
18 Jason Isringhausen
19 Octavio Dotel
20 Armando Benitez

It is important to note that when the search was expanded to players with at least 2 seasons of this sort of performance, an obvious drop in quality could be perceived. To my eye, 3 seasons turned out to be a very good parameter by which to evaluate consistent success. Looking at the list, Adams, Soria, and Marmol have not yet reached free agency, while Broxton, Dotel, Rincon, Gagne and Ayala all suffered injuries that hurt their performance and value before they could cash in. That leaves us with 12 pitchers relevant to our purposes.

Here is the list of relievers who have received large contracts, meaning deals for 3 or more seasons at an AAV of at least 7 million dollars (this is the list I was able to construct. It may not be complete. Please correct me if possible):

Jonathan Papelbon
Mariano Rivera
Rafael Soriano
Francisco Rodriguez
Francsico Cordero
Joe Nathan
Heath Bell
Brad Lidge
Billy Wagner
BJ Ryan
Armando Benitez
Jason Isringhausen

Soriano and Bell are the only players on the “got paid” list not on the “consistently performed” list, and Bell has two seasons of requisite performance and a third that falls just short (146 ERA+). Soriano is the only real outlier here, as he has never had a season meeting the performance criteria yet was paid like the more consistent elite performers. Conversely, Foulke and Hawkins are the only two of the 12 relevant players from the “consistently performed” list who failed to make the “got paid” list, and Foulke missed it by .25 million (3 years, 20.75 million).

Basically, when looking at the two lists, we find that the pitchers who have performed at a high level on a regular basis are the ones who are getting the big money. Now, correlation is not causation, but it does seem reasonable to say that large contracts for relievers have been largely reserved for pitchers with established levels of consistency and performance. Now, the next question to ask is whether it makes sense to be giving those pitchers large contracts. The obvious retort to this is that:

1) relievers are a volatile commodity, and
2) past performance does not guarantee future results, and
3) relievers are fungible and good relief can be acquired cheaply.

As for #1, Stephen Rhoads addressed this very issue in this space a few weeks ago:

In any walk of life, one quick way to open yourself up to embarrassment is to assume that those around you are either unable or unwilling to comprehend the complexities of your worldview, to borrow a turn of phrase from Confederacy of Dunces. I’d wager that most General Managers have a pretty good idea that relievers are volatile creatures, and that they are also aware of the failure of these relievers to live up to the contracts given to them. So, avoiding the arrogance that would suggest that they’re just irrational actors, what would drive a GM to pay a premium for a reliever? It boils down to predictability.

Paradoxically, the volatile nature of relief pitchers drives GMs to pay big money for relievers whom they don’t believe will be volatile. Thus, relievers with a long track record of health and consistently superb performance are the most likely candidates to get big money.

Essentially, reliever volatility actually makes handing big contracts to those relievers who have proven to be more of a sure thing a logical decision. As for #2 and #3, they can both be answered by the same point. While it is easy to look back at the end of a season and find relievers who provided great results for few dollars, it is much more difficult to identify those pitchers ahead of time. For every Joaquin Benoit there are 10 Buddy Carlyles and Lance Pendeltons, pitchers who are blanks in the game of reliever roulette. Additionally, while some of these large contracts have flopped, that is a risk that comes with any free agent contract. In the right context, it makes sense for clubs to take that risk rather than cross their fingers and hope to stumble upon the right reliever. Although past performance does not guarantee future results, it does make good results significantly more likely and predictable.

Relievers being fungible and volatile does not mean that their talent changes yearly. It means that in a small sample, you can often get statistical anomalies in both directions. Since relieving is by nature a small sample, there is more volatility and more risk. But if you have identified relievers who you think are more talented and more consistent, you lower that risk of volatility. There is value in that certainty, such that it makes sense to pay those relievers more than a pure talent to dollars evaluation might suggest. This added level of predictability is why general managers have been paying a premium for top relievers on the free agent market.