Some Thoughts Regarding Holds

Allow me to start by saying that I really dislike holds as a category. Saves plus holds makes plenty of sense – you’re basically telling owners to draft the best relievers regardless of role. Separating the two baffles me – at least in a standard league. I’m sure there are some unusual league configurations where it can work. I don’t like the category, but I still have to live with it in some leagues. What follows is how I cope with with holds.

Holds are relatively flatly distributed, meaning that it’s hard to select a reliever who will outperform the crowd. Last season, over 30 relievers recorded between 13 and 20 holds. Only 23 relievers had over 20 holds with only four reaching the 30 hold mark. Pitchers don’t often repeat near the top of the leaderboard – only Tyler Clippard and Joel Peralta have in recent years. Let’s compare to saves.

Saves vs. Holds (2014)

Number Saves Holds
40 or more 7 1
30-39 10 3
20-29 7 24
10-19 11 51

Last season was typical. A few things should jump out. Saves are comparatively scarce, but it’s easier to find a pitcher who can deliver a lot of them. It comes down to the nature of the statistic. Holds can occur in any inning, but it’s easiest to accrue them with an setup ace like Clippard or Wade Davis. These same relievers are closers in waiting, so if anything happens to the primary guy, they lose their holds (and gain saves).

This dynamic can reward the ol’ running back handcuff approach. If you roster Dellin Betances and Andrew Miller next season, you should get plenty of saves and holds. If either hits the disabled list, just call upon David Carpenter to fill the void. In this scenario, neither Miller nor Betances would reach the 30 holds threshold, but you might still get to 30 holds from Yankee relievers.

In a daily roto league that allows streaming, my preferred approach is to keep a spot or two open and chase matchups with fresh relievers (i.e. guys who didn’t throw the previous day). I want the reliever to matchup well with the opposing lineup, and I want his team to have a decent chance at winning. I’m not going to chase a Pirates setup man if they’re throwing Jeff Locke against Clayton Kershaw. In this type of league, it should be relatively easy to post a high holds total by simply remaining active on the wire. You don’t need to get lucky with a 40 hold guy to take the category.

In H2H leagues, the dynamic changes considerably. Many H2H leagues limit streaming. You’ll want to target more important positions and statistics than holds. Because holds are hard to predict, I usually hope my opponents will waste relief resources on them so I can consistently win saves. Each opponent has different strengths and weaknesses, so it can be relatively easy to pick and choose when to go after holds. If you have a weekly moves limit, plan ahead by picking up setup men on the Sunday before a matchup.

Weekly roto leagues offer the biggest challenge. The opportunity cost of chasing holds is readily apparent – you either skimp on closers or starters. My preference is to approach holds as a Plan B. I don’t want to punt the category, but I’ll probably only target a middle-of-the-pack finish. I deploy my setup men when my starters don’t matchup well or my closers hit the skids.

Points leagues offer a slightly different use case for setup men. You can put more weight on the best available reliever rather than the guy with the best save opportunities. However, if a league is awarding points for saves and holds, you should still prefer closers. Glance back at the grid above. If saves and holds are each five points, you’re more likely to get bulk from a closer. Keep in mind, some of those top relievers contribute to both categories, so the raw numbers reported above can be slightly misleading. Many points leagues – like ottoneu – award more points for saves than holds, which should only further reinforce your preference for closers.

In most leagues, you should only pay for holds if you already would have paid for the reliever. Davis, Clippard, and Ken Giles are a few such examples. I might also target somebody like Clippard who has posted multiple seasons as a top holds guy. Such pitchers are rare – they’re usually either promoted to a closers job or their talent declines.

Who else has holds related strategies and tactics they’d like to share?

We hoped you liked reading Some Thoughts Regarding Holds by Brad Johnson!

Please support FanGraphs by becoming a member. We publish thousands of articles a year, host multiple podcasts, and have an ever growing database of baseball stats.

FanGraphs does not have a paywall. With your membership, we can continue to offer the content you've come to rely on and add to our unique baseball coverage.

Support FanGraphs




You can follow me on twitter @BaseballATeam

newest oldest most voted
mg
Guest
mg

I have played a number of h2h holds leagues. In my opinion, holds devalue all pitchers to a differing extent. Usually when a pitching stat is added, a hitting stat is added as well, so lets say the league is 6×6 and holds is the added pitching stat. All hitters can contribute to the added hitting stat. Starters and closers cannot contribute to the holds category. Therefore, they are devalued and hitters values are increased. And, as you brought up, holds are very inconsistent year to year, and also a very deep stat. Outside of the universally owned middle relievers (i.e. Wade Davis), i would wait until the very end of the draft to take a shot on potential breakouts and scan the waiver wire for the 2015 wade davis.

KC
Guest
KC

This exactly.

To counteract this and still expand the player universe, I play in a league that uses weighted saves/holds. So it’s still 5×5, but a hold equals 1/2 a save. In that format, elite closers get a slight bump, mediocre closers and strong holders are roughly equal, next valuable are bad closers/vultures, and then MR that aren’t really getting either. I think it’s a nice balance. We play on ESPN (the league is set up as 5×6 so we can easily track holds and saves separately) and I just have a spreadsheet that automatically updates so people can track the true standings. I wish the mainstream sites would incorporate the ability for something like this, but the spreadsheet works for now.

OT
Guest
OT

I think it would be interesting to have a 5×6 league with holds and saves as separate categories. Would keep pitchers from being so devalued

hebrew
Member
Member
hebrew

I play in a 20-man league with 5 SP, 3 RP and 1 P spot. We need holds.