Position-by-position tiers: C | 1B | 2B | SS | 3B | OF | SP
My favorite metric when it comes to slotting relief pitchers is win probability added (WPA). According to FanGraphs.com, the top five relievers in that category last season were: Oakland's Blake Treinen, Milwaukee's Jeremy Jeffress, Seattle's Edwin Diaz, Boston's Craig Kimbrel and Milwaukee's Josh Hader. If you followed last season closely, you don't even have to go to the numbers to realize that's a solid listing of the best relievers in baseball.
In most good statistical categories, there is a certain amount of year-to-year consistency. There are always changes, but the leaderboard from one season to the next doesn't look as if you've accidentally subbed in names from another sport. But you know where Treinen -- the reigning WPA leader -- ranked in that metric in 2017? That would be 595th. Jeffress? He was 555th. Even Diaz, who was already established as a premier reliever, ranked just 62nd.
With relievers, you just don't know how the landscape will look from one season to the next.
Building a bullpen is hard, for just this reason. No position group varies as wildly in its performance from one season to the next as relievers. It's not even close. That's why, as much as anything, the best reliever of the current decade -- Craig Kimbrel -- has yet to find a home via free agency.
With that rather large caveat, we conclude our position tiers series with relievers. Squirrelly as they might be, their prominence in baseball has never been greater, in innings, in sheer number of pitchers, in the impact on deciding individual games.
We've reached this place with relief pitching in large part because of analytically driven strategic evolutions over the past 30 years or so. But we've also gotten here because the supply of big-armed relief pitchers has never been greater. That's certainly true in the major leagues, and by all accounts, it's just as true in the minors.
There used to be an adage in baseball that if what you have in the bullpen isn't any better than what you've got on the mound, then you stick with what you've got on the mound. Nearly all of the best arms in baseball were starters, so the goal for every manager was to get as many innings out of his rotation as possible.
Now that's not really a thing. Starters aren't judged so much for their stamina as they are for their pitch efficiency. Can they get through the order twice without burning through too many pitches? Good enough, because there is a crew of flamethrowers ready to come on and finish it off. Managers don't dread going to the bullpen -- they can hardly wait to do it. It's the most fundamental change to the experience of taking in big league baseball over the past decade.
As with the starters, we can take a snapshot of teams' bullpen situations by seeing how many relievers they have in each tier. This is a role-independent evaluation except in this regard: By awarding more points for the better tiers, we can in effect replicate the effect of leverage on how bullpens are set up. The better relievers, if used in an optimal fashion, should have a disproportionate impact on their team's chances to win in a given game. That's what leverage is.
It's not a perfect way to look at it, but it's worth checking out, even if that year-to-year volatility that I mentioned is a giant red flag that these rankings probably will look very different by the end of the 2019 season.