Offseason Spending on Relievers Isn’t Working Out

While this past winter moved slowly for a number of free agents, the offseason’s available relievers actually found work pretty quickly. By the time the calendar turned, 13 relief pitchers had received multi-year contracts worth more than $10 million, totaling more than $250 million overall. Addison Reed and Greg Holland would later ink deals for more than $10 million, as well. Much has been made of the fiscal restraint exercised by teams this past winter, but teams didn’t really apply that same sort of caution to reliever deals. Perhaps they should have.

In total, there were 30 deals in excess of $1 million dollars signed by relief pitchers this past offseason. With half the season having passed, it seems like an opportune moment to review how those deals are working out for the players and their clubs. Because of how relievers are typically utilized, we are necessarily dealing with small sample sizes, but that’s also just how things operate with relievers: the difference between a good and bad season might be a few rough innings.

The graph below shows WAR and the amount of guaranteed money the player signed for in the offseason.

Teams would hope that the trendline here slopes up and to the right. That would suggest a general correlation between the money received by a player and his on-field production. The graph above, however, doesn’t look anything like that. Indeed, if a slope exists at all, it goes down and to the right. And even if we omit the Rockies from it — they were responsible for the winter’s three biggest relief contracts — this graph would still pretty much look like a jumbled mess. Seven of the 11 players with more than 0.5 WAR this season signed contracts for less than $10 million total. Of the 13 relievers at replacement level or below, eight received eight-figure guarantees. There appears to be little rhyme or reason at all when it comes to money and performance.

Perhaps the total money skews things somehow. To see if that’s the case, here’s a similar graph, except with average annual value instead of total money.

Still nothing, right? It appears that way.

One note: while our main version of WAR accounts for leverage, it also uses FIP instead of ERA. It is possible for relievers, meanwhile, to provide totally satisfactory results even while recording less than ideal FIP numbers. To examine such results, we can turn to Win Probability Added (WPA), which will show how much a pitcher’s results added or detracted to a team’s chance of winning.

Even by this measure, however, no pattern emerges. That’s probably partially due to the fact that WPA and WAR generally line up pretty well. The chart below shows the players who signed deals of at least $10 million and their performance thus far.

For roughly $135 million in salary this season, the biggest free-agent relievers of the offseason are currently on pace for three wins. Even if every single reliever on that list above were worth another win going forward — that is, about 10 times what they’ve produced so far — that production would still cost about $8.2 million per win. Of the players where WPA departs from WAR, the Rockies might be a little happier with Davis’s performance to date than WAR shows. Elsewhere, the Phillies with Hunter and the Mariners with Nicasio aren’t actually receiving the benefit of those good pitching performances due to events that occur in small sample sizes. Of the pitchers who have pitched well and have been good by WPA, we are probably limited to Cishek and Morrow, with the latter just coming off the disabled list.

It would be fair to say that we are dealing with small samples. As I note above, however, we are almost always dealing with small samples when it comes too relievers. Last year, Holland’s WPA on the season was 1.43. It took him a whole season to accumulate that, and with the help of his manager, he’s already been worth negative that amount this season. In order for Holland to help his team this year as much as he did last year, he’s going to have to pitch the second half like Edwin Diaz pitched the first half, which is not very likely. On the list above, only four players have at least 0.2 WAR and 0.2 WPA, something 77 relievers have done this year and 40% of relievers who have pitched 20 innings. There’s a pretty low bar here, and this year’s free-agent class hasn’t delivered.

How does that compare to their cheaper friends? In the graphs above, we looked at 30 relievers, 15 with guarantees of at least $10 million, and 15 with guarantees between one million dollars and $10 million. Here’s how the lesser-paid group has performed.

For a quarter of the cost, the cheaper relievers are providing top-quarter performance. Not all of these relievers have done well, with Logan already dropped from the Brewers and Koehler and Benoit yet to pitch this season, but the group does just as well as the highly paid set of players. In summary:

It’s possible that this group is just a random assortment that won’t repeat itself, but how much confidence is there behind that assertion? The higher-priced free agents from last year are way behind the lower-priced ones. Even if they were even — and it isn’t close — that still wouldn’t provide much of an argument for signing relievers to big contracts. More study over previous years is probably necessary for confirmation, and we’ll have to see how things play out with the players above, but so far, teams that spent a decent amount of money on relievers probably have some regrets.