I know, yet another reliever column. It seems like I’ve done a lot of these this season. Don’t worry, this is mostly about relievers but dives into other aspects of fantasy baseball, as well.
Earlier this week, Fred Zinkie of Rotowire posted this innocuous statement on Twitter:
After looking at the respective saves and holds leaders lists, I can’t figure out why there is so much resistance to SV+HD as a category. Let’s get to a place where we look at RP skills over managerial tendencies/preferences.
— Fred Zinkie (@FredZinkieMLB) September 20, 2022
I wouldn’t quit a league if it started using saves and holds as a combined fantasy category. However, for the most part I wouldn’t like it and don’t recommend it for quite a few reasons.
Holds is a terrible statistic
I agree with everyone who believes that saves are a terrible statistic in terms of measuring player value. However, adding another bad category to saves doesn’t solve this. It merely compounds the problem. This is somewhat different than swapping out wins for quality starts, where you are replacing a flawed statistic with a somewhat less flawed (but still flawed) stat but is part of the same problem-solving management theory. “This thing is bad! Let’s add something else to it that is bad!”
To offer context and show that I believe change is sometimes warranted, an example of a good change like this was when Tout Wars swapped out AVG in favor of OBP. On base percentage is a superior measurement of player value than AVG. Counting a hitters’ walks and HBP as part of his offensive value was an improvement to the fantasy game.
Adding holds would make gameplay worse
I presume the idea of adding holds as a category is to simply allow you to roster the best relievers and end the constant churning in the never-ending hunt for saves. While it is a noble reason to suggest adding the category, it would still lead to chasing relievers based on a pair of statistics that don’t truly measure reliever skills.
Unless you are only counting starting pitcher wins (or using quality starts), you’re also giving relievers even more categorical juice and diminishing starter pitcher value. If this is your goal, that’s fine, but before you add holds to your Roto league this is something you should be aware of before considering this change.
We’re not completely playing a skills-based game
I’m not going to hold Zinkie’s feet to the fire about the wording of his tweet. After all, tweets aren’t articles or books and we’ve all posted things on Twitter that aren’t as precise as they can be. But I don’t view either one of these statistics as a useful way to measure a relief pitcher’s skills any more than wins measure starting pitchers’ skills. This highlights a larger issue with category-based Roto leagues that goes beyond holds. If you want to reward player skill, I strongly recommend playing in a points-based league or perhaps even a WARP league if you can find one.
One thing Zinkie mentioned on a follow-up podcast on this subject is that we play fantasy baseball to determine who is the best at identifying player skills. I do not agree with this assessment. We all play for different reasons. I play fantasy baseball to see if I can win based on the rules of the contest. This might sound overly simplistic, but what this means is I know that 5×5 category Roto is a poor way to measure player ability. My goal as a fantasy player isn’t to participate in a contest that replicates real baseball.
We already use imperfect categories. ERA is a poor category for assessing pitcher ability. Few would argue it is an ideal category for measuring pitcher value. However, ERA does a better job of incorporating other aspects of real baseball into the fantasy game. A starting pitcher with a mediocre DRA and good ERA is frequently the beneficiary of a good defense, solid bullpen, and a manager who knows when to go to that bullpen. I could join a league that uses FIP or DRA, but I prefer ERA because it incorporates more into the game than a pitcher’s performance.
5×5 already accounts for non-closer, reliver value
If fantasy baseball only used wins and saves for pitchers, I could see why we would want to include holds as a separate category. But one thing that’s great about 5×5 is that relievers already have plenty of opportunities to add something to all five categories. In a deep league, a great middle reliever can elevate your ratios as much as almost any starting pitcher. There was a time when starting pitchers were the beneficiary of the win in most cases. In 2022, relievers are frequently entrusted to notch that all important W. While relievers don’t garner as many strikeouts as starting pitchers, their contributions in this category aren’t small potatoes.
Not every league is an overall contest
I suspect the consideration of using holds as a category is much more of an issue in overall contest formats than in standalone leagues. In a standalone league like LABR or Tout, the cost/benefit analysis when it comes to closers revolves around one category in one league. There are many instances where drafting only one closer or even eschewing the category entirely makes sense for the categorical reasons I outlined above. In a league with an overall contest component, dumping saves makes it extremely likely you won’t be able to win in the overall league of leagues, and makes churning for saves a necessity.
I don’t disagree with Zinkie that our game improperly values relievers. However, I don’t see the problem as a categorical one. Rather, I think of it as a roster composition issue. Even though nearly every major league team carries 12 hitters and 13 pitchers, most fantasy leagues carry 14 hitters and nine pitchers. Additionally, nearly every fantasy league team opts for 6-7 starting pitchers and 2-3 relievers. Even mono leagues typically see six starting pitchers and three relievers per team. The middle reliever is undervalued and underappreciated because we are using a roster model that is a period piece that belongs in a museum.
My two suggestions for ameliorating this problem would be to expand fantasy rosters to require teams to carry 10-11 pitchers and require teams to carry no more than six starting pitchers. If fantasy teams had to always carry five relievers this would reduce the endless chase for saves because most of the most skilled relievers would already be rostered. Relievers with poorer skill sets would be added for saves, but we’d do that full well knowing the risks. Pushing teams to roster more relievers would add more of their contributions to our fantasy teams in the preexisting five pitching categories while also pushing us to add more skill-based relievers to our teams, instead of chasing yet another stat that doesn’t truly fit this model.
Thank you for reading
This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.Subscribe now