The story of this offseason is the outrageous price of elite relievers. Smart teams are giving big value, both in money and in trades, to acquire the game's top bullpen arms, and there's a good chance that the two best bullpens in baseball will be in the American League next season.
So what should the Rays do about it?
Should the Rays be bullpen buyers too? Should they sell off their more valuable bullpen pieces while prices are high? Should they abstain entirely? To answer those questions, they need to think both about the value of relievers, and about the shape of that value.
Bullpen Value
The biggest question regarding top-flight relievers should be whether they are currently overvalued, or if they're correctly valued in the new bullpen marketplace. I don't have an answer.
For many years, one of the more universally-recognized ideas within the sabermetric community was that closers were overvalued because of the emphasis placed on the save statistic, and that smart teams could and should deal their closers if they found a less savvy buyer willing to pay. Saves were not a good measure of reliever quality, and most decent relievers, given the right opportunity, could finish games and rack up saves.
The save is not what's driving the current market, however.
Guys like Craig Kimbrel and Aroldis Chapman are valued because they are literally among the best relief pitchers in the game, and have been dependably so for years. Ken Giles and Carson Smith, two other relievers moved this offseason, are similarly excellent pitchers based on their peripherals, rather than on the save numbers.
Moreover, teams are stockpiling multiple quality relievers, rather than focusing on the need for a "closer," perhaps attempting to copy the success of the champion Kansas City Royals.
The other trend raising the price for relievers was what we Rays fans saw pushed to the extreme due to necessity -- teams are finally experimenting with limiting the exposure of their starting pitchers within the game. Even if a pitcher hasn't reached a preset pitch count, and even if he isn't visibly tiring, batters gain an advantage as they see the same pitcher multiple times, and this advantage becomes pronounced the third and fourth times through the order.
Kevin Cash and his early hook helped the beaten-up Rays starters to an AL-leading 3.36 ERA, but the overall results left something to be desired because the Rays bullpen was unable to take over at a high enough level. Obviously, the better way to implement this strategy is to have more, better pitchers.
Still, all of that is not to say that the current high price of top relievers is correct. All of these factors are real, but they might be overemphasized.
Probability Distributions of Value
The other reason that smart teams have previously avoided huge investments in "closers" is volatility. Because even the most-used relievers generally pitch under 80 innings, their statistics are more prone to fluctuate from season to season than other full-time players. That means that, while a high-leverage reliever can surely give good value to his team, it's difficult to know which high-leverage reliever will be the valuable one.
This is another aspect of the new market. The teams paying top dollar don't expect to have bought the top-performing reliever for next season, but they do expect to have bought a good one. The "best relievers" are the ones that are almost certain to be "among the best relievers."
Rays fans understand the concept based on their experience with Jake McGee and Brad Boxberger.
In 2014, Boxberger posted a 2.37 ERA, a 2.84 FIP, and a 1.95 xFIP. That's a better xFIP than anything Jake McGee has ever put up, and on the whole, it's a dominant season. The next year, though, Boxy regressed, giving the Rays a 3.71 ERA, a 4.26 FIP, and a 4.00 xFIP, while McGee continued his four year trend of excellent performance.
In 2016, it's entirely possible that Boxberger will put up better numbers than McGee, but McGee is far more likely to be a top-10 reliever, while Boxberger could either be that or be something much less. It's the McGee's of the world that are now in such high demand.
And this is what I think makes the market so interesting. Back in August, I wrote about David and Goliath strategies. The short of it is that the Goliaths want to pursue strategies with low variance in outcomes, while Davids want to pursue strategies with high variance, as it will take some fortunate outcomes for David to beat Goliath. Over the long term, the Rays are a David.
Conclusion
I don't know if the Rays should be taking part in the new reliever market or not. If relievers are currently overpriced, as they may be, the Rays should sell. But if the pricing is accurate, then their path needs to be more nuanced.
Trading the dependable Jake McGee for his true, large value and replacing him with lesser but high-volatility pieces is a net win for the Rays, as they're unable to count on having a higher average performance expectation than their richer rivals. Rather, the Rays need to go for a combination of high expected performance and high variance, so that with a bit of very possible luck, they will be able to outperform the teams that enter the season knowing they'll be playoff contenders.
Quality is still important, especially because the Rays appear to value increased bullpen use as an in-game strategy, but they should embrace the inherent variance of relievers and seek upside while concentrating on acquiring sure things in other areas of the team.