Last week we pointed out a bug in conventional valuation theory, and we discussed how you can leverage it to your advantage when assembling and managing a winning fantasy baseball team. This week let's look at a related wrinkle in the valuation process: The latent utility of the dominant, non-closing reliever.
As a reminder, the valuation quirk is the assigning projected earnings as if the player is going to be in the active lineup all season long. However, in a practical sense, in leagues with the freedom to move players between active and reserve, a good portion of a fantasy roster will be deployed strictly for favorable matchups. This is not captured via conventional valuation, thus the estimated earnings and associated ranking derived from the season-long projection is not an accurate reflection of how many players affect a fantasy team.
Dominant, non-closing relievers are an excellent example of this valuation foible. In many cases, conventional valuation places non-closing relievers ahead of back-end starters when converting season-long projections to projected earnings. The result is giving the impression it's better to draft an outstanding set-up man over a lesser starting pitcher.
The key to this line of thinking is understanding you're not drafting a player, but rather a roster spot. Sometimes, the expectation is for the drafted player to occupy the spot all season. Others, the plan is dedicating the spot to the player with the most favorable matchup that week. However, you are "paying" for the spot with the player initially drafted to that spot.
This brings us to the optimal deployment of a dominant middle reliever. Unfortunately, there is not a universal approach. Non-closing relievers are least useful in shallow leagues while they are most advantageous in deep, single league (AL or NL only) formats. In terms of valuation, it's a balance between how much a reliever helps ratios versus the strikeouts and win potential of a starting pitcher. The inventory of available starters in a 10-team Mixed League is clearly superior to those in a 12-team AL or NL only draft.
Ideally, you want to be deploying starters and closers, but sometimes you aren't comfortable with all your starters' matchups. In these instances, the optimal active lineup includes a dominant setup man, or maybe even multiple. The thing is, you don't want to draft a player of this ilk, at least not as part of the initial active lineup. Ideally, they are picked up as reserves or on a needs basis via waivers or free agency. The exception is AL or NL only when you may require a setup man in your lineup every week. In this scenario, it's best to use a non-closer with a pathway to a few saves (more on this in a bit).
Valuation landing non-closing relievers in the draft-worthy pool and it being inefficient to draft them is a Catch-22. In a perfect world, the draft-worthy pool would be driven by the stats most likely to be on an active roster for the season, but this is impractical. As such, we have tweaked our pitching prices to devalue middle relievers with the starters most likely to be streamed in their stead. The simplest way to put it: It is much better to roster the starting pitchers you intend to stream before adding the middle relievers whose usage will be less frequent, and only when necessary.
There is another reason why focusing on starters is the sager approach. Since relievers compile fewer innings than their starting pitcher counterparts, reliever projections and performance are more volatile. Our projections are skills-drive, which helps flesh out the luck bias, but fewer innings still result in wider error bars. In terms of in-season production, a starter can bury a poor outing in their season-long numbers much faster than a reliever. This is why in-season analysis should also focus on skills and not surface stats like ERA and WHIP. Bringing it back to draft day, it's better to pay more for predictability, which is another reason to devalue middle relievers. Again though, the exception is deep, single league formats. In these instances, target the more established, hence trustworthy setup men as opposed to those with one season of strong, possible misleading numbers.
Let's discuss the most trustworthy stat to use when evaluating middle relievers, though this also applies to starters. In an era where the number of meatballs a pitcher surrenders is tracked (yes, that's what it's called, and it is tracked), the simple metric K-BB% is most useful. The more K% contributes is better for fantasy, since strikeouts are a category in rotisserie and points leagues scoring. However, a strong K-BB% usually correlates well with ERA and WHIP, regardless of whether it is driven by high strikeouts or low walks.
By means of perspective, last season, the league average K-BB% was 14.1% for starters and 14.5% for relievers. Only 29 starting pitchers exceeded the 20% mark while 48 relievers tossing at least 50 frames surpassed the 20% level. Of those 48, only seven collected at least 30 saves, leaving a bevy of non-closing relievers from which to choose when desiring to bench a starter with a risky matchup.
Of the 41 non-closers in the group, 35 logged at least one save while 11 notched at least eight. Of course, you need to be fortunate to have the reliever active when they nabbed a save, but over the course of the season you’re apt to pick up a handful. That said, you can increase your chances to vulture some saves by focusing on relievers from bullpens without a primary closer, of which there are several. Things could change, but the best teams to target are Tampa Bay, Seattle, Minnesota, the Dodgers, Atlanta, St. Louis, and the Mets. They are all replete with a strong bullpen without a definitive closer, at least for now.
Many say Moneyball is about on base percentage. It isn't; it's about how there was a market inefficiency for players whose primary asset was getting on base. Dominant relievers are the "Moneyball" of today's fantasy baseball.