Just wondering if there is any new thoughts, or past thoughts on bringing defense to the game and the thinking behind the possibility in the future. Thanks for the game!
Not really much to report on this front. Defensive stats are still high-variance and don’t have a huge sample size, especially in this TTO hitting environment. Defense is also affected by park effects, and park effects can change in-season, which would then retroactively change defensive performance. Those are the main factors that make me hesitant to add defensive stats into Ottoneu.
If anyone has any new research that shows evolution on defensive stats that I have not considered, please let me know.
The Outs Above Average (OAA) from Statcast data represent the latest development on defensive stats. Tango Tiger unveiled OAA for outfielders last year and released “iteration one” of the infield OAA a month or so ago.
They range from about -15 to +15 within a season (like other defensive metrics), so there is limited variation between most players. And the error bars are pretty large, just like UZR, etc.
IMHO, they would not make for a meaningful addition to Ottoneu.
I’ll read about OAA a bit. Thanks for pointing me in that direction.
Hey just adding to this. Fangraphs just announced they’re changing their calculation for WAR to incorporate Statcast’s Fielding Runs Prevented. Seems to imply that they’re more confident in its accuracy.
I found it particularly interesting that even though it didn’t change most players’ single season totals all that much, it did in a few cases alter it dramatically. The decidedly “glove first/bat last” Adam Engel in particular went from -0.8 WAR to 1.1 WAR in 2017.
Anyway, I know these stats are, as you say, still super high variance and defensive metrics in general are a bit of a crock, but hey, it would be awesome to be able to one day incorporate defensive acumen into Ottoneu. Perhaps that day will never come! But thought I’d share.
Here’s the story: A FanGraphs WAR Fielding Update | FanGraphs Baseball
This is just the OAA that @walt526 mentioned earlier.
I do not think this is compelling from a daily or weekly lineup fantasy standpoint to be honest but I’ll continue to read into it.
Oh yeah I know, I was sharing it because Fangraphs incorporating it means they have some more faith in it. I know it’s the same stat, and if you’re not comfortable incorporating it I totally get that. Again, high variance.
It’s definitely better than UZR.
What about a deduction/addition per actual error?