The Fantasy Baseball Substitute for Wins?
In the past few weeks, I wrote about my and, I presumed, others’ distaste for certain categories in rotisserie and head-to-head baseball leagues. Some of us don’t care for wins, saves or other cats, or we think we have ideas for better ones. Reader response has been fantastic, I think. Some good discussion, some good points, some good concepts put forth.
Running Frog made a good point. Wins seem to correlate well enough with things like a great number of strikeouts, a good ERA and a good WHIP. He, like others, doesn’t see a reason to change.
I dislike the win in general because:
- If not for its somewhat subjective assignment to an individual, it’d be a team statistic, not just for fantasy purposes.
- Saves and holds are exclusive to relievers, whereas wins aren’t exclusive to starters.
The latter point may not be an issue for some. Relief pitchers generally perform better than starting pitchers in ERA and WHIP, but starters accumulate more strikeouts than relievers do. Some would like to see good non-closing relievers matter more, and somewhat random wins helps them to achieve that. I think that there may be other ways to do that, but I digress.
I really appreciate Quality Starts. It’s binary and therefore simple. In the absence of something better – and perhaps there isn’t something better – it works.
I really like the suggestion from Metsox: QS+CG. It’s a way to reward pitchers who come up with that pitching performance that’s of much greater value to his team than the quality start. Complete games are relatively rare, so in the instances that they’re earned, they can have some significant meaning in your standings.
Complaints about the QS are fair. It rewards a pitcher for having gone six innings and having accrued a 4.50 ERA, but not for having gone eight and doing the same thing. Is a 4.50 ERA really what we want to call “quality”? I think the quality of the start will have been inflicted upon our team ERA and team WHIP, as well, so the fact that three earned runs in six frames isn’t of great quality doesn’t concern me. The fact that four earned runs in eight innings doesn’t meet the criteria seems a bit unjust, though. In the instance in which a pitcher satisfies that otherwise unrewarded effort, perhaps he’d register a CG in Metsox’s league. At least there’d be some redemptive value, in that rare instance.
I stated that a modified version of Pure Quality Starts or Game Scores could be a good alternative to W’s. As I and commenters pointed out, the original versions of those rubrics incorporate many of the stats we already use for other fantasy categories. Game scores in particular become a lot like points. Perhaps that’s not a bad thing. But there seem to be some fundamental flaws with either of them, even if we remove some of the elements that count for them.
One idea intrigues me most, however. I’d begun to work on the development of it, jotted down some ideas and parameters. I wanted something that was like slugging percentage or ISO against a pitcher, but not a rate. Something like total bases allowed, but not that heavy on its own. Something that might even hit ground-ball and fly-ball pitchers a little more evenly, because the former tend to do worse in WHIP than the latter. It’d have to involve walks. Hit by pitch? Am I too finicky? The little things matter, but how much do I want them to matter in my game?
Then I took a break to go through comments and ponder them heavily. That’s when I came upon and contemplated the idea from MDL, one of the first commenters on the wins blog. That section of the thread pretty well captured my eventual destination, minus a couple of details. I’d include free passes issued. I’d also consider coming up with a threshold that the pitcher must reach, in each start, in order to record one of these – to make it binary, a trait that seems kind of an essential, at least for roto purposes. One responder noted that it’s quite similar to a points league, otherwise. Perhaps it becomes a points system, though, and we come up with some correlative cat for relievers. Kind of like Fantasy Game Scores for starters as well as relievers. Or perhaps they contribute to the same category, and we have only three! Heresy.
Kind of a kicker: Do away with WHIP to make four pitching categories and eliminate much of the duality. Use only four hitting categories to even things out.
To sum: O – TBA – BB = X, where X may be some number that must be equal to or greater than a level that we determine to be quality, after some more research and discussion, in order to register one of these things.
What could we call this category? Ideas, MDL? Perhaps the Quality Fantasy Start? Otherwise, Fantasy Game Score? I think it has promise.
Nicholas Minnix oversaw baseball content for six years at KFFL, where he held the loose title of Managing Editor for seven and a half before he joined FanGraphs. He played in both Tout Wars and LABR from 2010 through 2014. Follow him on Twitter @NicholasMinnix.
To me, you can choose almost literally ANY statistic to use, no matter how BS a statistic it is, and fantasy would still be fun. The reason is this: your league chooses how to value players; they don’t actually have to win games. So if you use pitcher W’s, it just makes you re-evaluate the value of a guy who will eat innings on a good team. The goal doesn’t have to be true talent. Heck, you could give points for blown save percentage and it would be super interesting — tons of value for a guy who is bad but will get lots of opportunities. You’d target over-the-hill guys with bigger names on bad teams with no obvious alternatives.
Scoring in fantasy is a moving target, and that’s part of what makes it cool. You win by understanding the format better and finding innovative ways to pad certain totals. We all know who the best real life players are, so it would be less fun if it was just a mad dash for the same guys. You have to be a little creative when it isn’t all SABR-perfect.
This is what I’ve said on every iteration of these articles, but much less eloquently. We use Ws because they dictate how we value players that would not otherwise be used in more strictly saber-defined leagues, and I don’t see why that’s a problem. Fantasy is fun because some categories are harder to predict.
To piggy-back on your ‘scoring being a moving target’ comment, stats other than Wins are dependent on things outside a player’s control. A stat like Runs is dependent on both a player getting on base and then someone knocking them in, which is outside of the guy getting the Run stat’s control. Ditto for RBI’s. You (usually) never look at a hitter in a vacuum. A lot of a hitter’s value depends on other players being on base when they’re up to bat, or players knocking them in when they’re on base. Having Wins as a stat leads to mediocre pitchers on really good teams having value. Why isn’t that ok?
I’m in the camp that think the situation (context) matters. Are two singles the same if one of them drives in a run? If we wanted to try and remove all context from players’ performance, I think we’d end up with only three fantasy baseball categories: maybe WAR for hitters, WAR for pitchers and then a fielding metric. It wouldn’t be nearly as fun.
I think you are overlooking the fact that there are many baseball fans who don’t read sites like Fangraphs and they DO use fantasy stats to define their opinions about players’ real life value.
And why does it matter why Joe Schmoe values a player highly because of wins? Does that take money out of your paycheck or something?
To make the game fun you need stats that managers can reasonably hope to project. If they can’t project then they’re going to ignore that stat entirely and it might as well not be included in your scoring. Imagine if you awarded points for “games in which the total of a starter’s walks and hits is evenly divisible by 3”.
You probably also want to avoid tracking stats that are inversely correlated for an individual player, since they’d tend to cancel each other out and lead managers to ignore both of them. Maybe losses and ERA, or ERA and negative-WHIP.
You can hope to project wins…it’s just not as simple as using Steamer, Oliver, or any other projection system as gospel. You have to consider team strength, outlook, injury risk, division/league…why is that bad?
I agree you can reasonably hope to project wins. That’s why I think it’s an acceptable stat to use from the point of view of “making fantasy baseball fun”.
I mentioned “projectability” as a caveat to Eminor3rd’s claim that “To me, you can choose almost literally ANY statistic to use, no matter how BS a statistic it is, and fantasy would still be fun.”
If all your league’s stats were completely unprojectable then fantasy would not, in fact, be fun. It would be equivalent to sitting around rolling dice and seeing who gets the highest total. Not really enjoyable.