Aroldis Chapman signed a record deal for a reliever on December 7, with the Yankees inking him to a five-year, $86 million contract. Kenley Jansen decided to return to the Dodgers for five years and $80 million on December 12. Even Mark Melancon landed a four-year $62M contract with the Giants back on December 5. Are top-end reliever contracts getting out of hand, or is the market just now starting to catch up to how valuable a truly dominant back-of-the-bullpen pitcher can be? This isn’t the first time we have seen contracts in this range, at least as far as AAV (average annual value) goes. A quick glance at this SB Nation piece from 2014 shows that Rafael Soriano got $14 million a year, Jonathan Papelbon found someone willing to pay him $12.5 million over four years, and, of course, Mariano Rivera was making $15 million over his last five seasons with the Yankees.
The willingness to pay elite relievers ten-figure contracts is not a new development in baseball. It’s the length of the contracts that is unusual. Rivera may have made $15 million in each of his last five seasons, but it was over two separate contracts. Papelbon’s deal with the Phillies was for four years, and the reaction at the time was similar to that of the deals being signed this winter. People, for the most part, are convinced these are bad deals because of the length. They are justified as the cost of doing business, and the concern over handing any reliever eight figures over four or five years is certainly understandable given the volatility of relief pitching from season to season. But is it worth going that extra mile, to a fourth or fifth year, to land your guy for the next three?
It’s a tough question, and the answer may not line up too well with how these contracts play out in the end. That’s because we still don’t really know how to peg the value of relievers relative to other positions on the roster. According to holistic metrics like WAR, relievers come up very short in comparisons with position players or starting pitchers. As an accumulating stat, it’s not one that is ever going to make relievers look terribly good since they just don’t spend that much time on the field. It’s completely valid to penalize players who spend so little time producing for their lack of innings, but as a specialized position, there needs to be an adjustment in our thinking since the very best will be performing in almost exclusively high-leverage situations.
On the other hand, stats like ERA+ will overvalue relief pitchers since they care not for trivialities like innings pitched. A great reliever will almost certainly have a higher ERA+ than all but the most elite starting pitchers. For example, in 2015 Clayton Kershaw was unquestionably the best starter in the majors. His ERA+ was 230 (that’s 130% better than league average) in what was one of the best seasons we’ve seen from a starter since Pedro Martinez was dazzling fans around the turn of the century. Last year Zach Britton was unquestionably the best reliever in baseball and had one of the best seasons we’ve seen from a reliever since Mariano Rivera was at his peak in New York. Britton’s ERA+ was 827. He broke the metric and it was still only the 9th highest ERA+ of all time.
Statistics on both ends of the spectrum (accruing and rate stats) seem to have quite a bit of trouble figuring out how these specialists compare to other players on the field, so it’s no wonder we have trouble wrapping our heads around how to value them. Most stat-savvy fans tend to be cautious around relievers. It’s not uncommon to come across the sentiment that a solid #4 starter who gives you 200 innings of league-average pitching is more valuable than even the very best relievers because of the number of innings they spend on the mound. The problem with that line of thinking is that it doesn’t adequately take into consideration just how important being able to trust your bullpen in high-leverage late inning situations can be. To illustrate this, we will take a look at another highly paid back-of-the-pen reliever: Craig Kimbrel.
Kimbrel was traded to the Boston Red Sox going into the 2016 season. The value of that trade, specifically the cost in prospects, has been debated quite a bit, but is related to the issue we are looking at here. Kimbrel’s first season in Boston was met with mixed reactions. His results in save situations were spectacular, but his ability to pitch in non-save situations (an issue that existed before his arrival in Boston) was suspect. He had a 135 ERA+, the lowest of his career, but he converted 31 out of 33 save opportunities (good for a 94% conversion rate) and sported a 5.44 strikeout-to-walk ratio and a .499 OPS against. He was utterly dominant in save situations. In non-save situations, his K/BB dropped to 2.07 and his OPS against jumped to .676.
For whatever reason, when deployed in save situations, he maintained his historical dominance and when used as a late-inning reliever in non-save situations, he struggled. Save situations accounted for about 64% of his production, so the non-save situations were not an insignificant portion of this season. That said, the numbers don’t tell the entire story here.
When Boston traded Manuel Margot, Javier Guerra, Carlos Asuaje, and Logan Allen for him, they were terrifyingly thin in the bullpen. Previous closer Koji Uehara had just completed a season in which he barely pitched 40 innings and setup man Junichi Tazawa had posted a 4.14 ERA and a 1.330 WHIP. Tazawa’s FIP was 3.05 suggesting he might bounce back in 2016, but he had been ridden hard and the thought of coming back with those two as the back end of the pen was not appealing.
Enter Craig Kimbrel and Carson Smith. Smith was coming off of an excellent 2015 season but, sadly, missed almost the entirety of 2016 after tearing his UCL. The Red Sox were left with just Kimbrel and the questionable duo from 2015 as set-in-stone pieces to lock down games. Predictably, Uehara and Tazawa both spent time on the DL in 2016. Tazawa logged his fewest innings since 2012 and Koji just 47, his second lowest in Boston. Kimbrel himself needed a knee scope mid-season. In short, the bullpen thinned out considerably and forced Dave Dombrowski to trade for Brad Ziegler.
Was Kimbrel worth the trade, or his $15 million salary? According to WAR, no. Baseball Reference’s version of the stat had him worth 0.9 wins. That means if you take him off of the Red Sox and replace him with a replacement level player the team would have lost only one more game all year. And this is where WAR breaks down as being useful for relievers.
Even if we conclude that Kimbrel’s season was less than great, removing him from that bullpen entirely would have had a far greater impact than WAR implies. Think about the impact of having Uehara as the closer and Tazawa as the primary setup guy before Ziegler arrived. How much harder would each reliever have been pushed? How many nights would manager John Farrell have left them in to face one more batter, or put them in one earlier? And how many of those nights would that have gone poorly? How many games would he have tried to squeeze one more batter out of a laboring starter for fear of what his thin pen would do in a high-leverage situation? It’s not hard to imagine more than one of those situations leading to losses. The effect of removing Kimbrel from the pen cascades over the entire pitching staff. It has a systemic impact that you simply can’t measure with something like WAR.
If Craig Kimbrel was replaced by a replacement level player in the Red Sox bullpen, it is highly unlikely that they would have won the division and is entirely possible they would have missed out on the playoffs. You won’t see this in his WAR, and that doesn’t mean that WAR is a useless stat (although, it is often misused). If the Red Sox had not traded for Kimbrel, they likely would have made some other move for a reliever, but that isn’t the point here.
The point is that the way we value relievers is inadequate.
And perhaps it is only us, the fans and the media, who are doing so. If this offseason isn’t a blip on the radar, then it is possible that teams are already valuing ace relievers as highly as ace-level starting pitchers; at least, on an inning-by-inning basis. Consider that David Price hit the free-agent market last year and landed a contract with an annual value of $31 million. He pitched 230 innings in the first year of that contract. That innings total seems like a pretty good estimate for what you can hope for from your ace. Rick Porcello threw 223 in his Cy Young year, Clayton Kershaw threw 232 in 2015 when he won his.
The best relievers in the game throw up to 70 innings each year. Aroldis Chapman was the only elite reliever on the market without a qualifying offer depressing his value, so we’ll look at his contract as a point of comparison. He received $17.2 million per year from the Yankees.
Chapman threw 58 innings in 2016, 66 ⅓ in 2015 and 54 in 2014. This is a three-year average of 59.44 which we’ll round to 59 ⅓ since a pitcher can’t throw 44% of an inning. So Chapman threw 25.7% of the innings that David Price threw in 2016. His $17.2 million per year is 55% of the dollars Price received on an annual basis. Both players have an opt out, so it’s not likely that such a clause had a radical impact on the total dollars for one and not the other. That’s 55% of the dollars (per year) for 25% of the innings, so Chapman is making significantly more for the innings he is expected to pitch than David Price.
Of course, contract length is a major factor when trying to peg the value of a player, but no one would argue that Aroldis Chapman is more valuable overall than David Price was last winter. We are just trying to understand what relievers are worth, and looking beyond total dollars and years may provide us with a bit more perspective.
Even if we look at Jansen, whose total dollars were almost certainly depressed by the fact that the only team that could sign him without losing a pick was the Dodgers, he got paid 1.65 times as much as David Price on an inning-per-inning basis using Price’s 220 ⅓ innings pitched in the season before he signed his contract.
So is the market for high-end relief pitchers spiraling out of control? It might be, but it is probably equally possible that it is finally catching up to the value a truly dominant back-of-the-bullpen pitcher offers a contending team. The stability afforded by inserting one of those arms into a pen, extending it by at least an inning most nights, can be felt throughout the pitching staff and that, perhaps more than anything else, is why teams are suddenly willing to go four or five years at premium rates to land one of the elite bullpen pitchers on the market.