Quote:
Originally Posted by korme
As a guy that has fallen out of the nitty gritty of baseball day to day, I'm curious - is wRC+ and SIERA now the baseline stats you look at when generally determining a player's season's quality?
|
Hey man. Not sure if you were following when wOBA was the big thing, or OPS...OPS+ is a pretty popular (probably more popular/used than wRC+) as it takes OPS and then scales it so 100 is the league average. If you have a 110 OPS+, you're OPS is 10% better than average. It's great, but the flaw in OPS (and OPS+) is that it's not weighting baseball events properly.
That's where wOBA comes in, which wRC+ is using as a part of it's calculation. wRC+ is also normalizing park factors and such as well...no idea if this is replicated in OOTP! I just like it in real life so it's the most natural thing for me to use here. But generally, the idea is: 100 wRC+ is league average, 110 is 10% better than league average, 90 is 10% worse than league average, etc etc, and it values singles, doubles, homers, walks, etc more accurately than OPS+.
On the pitching side, FIP was the first (that I know of) stat that tried to show what a pitchers ERA should have been, removing everything outside of strikeouts, walks, and home runs. The argument was that only those three things were in the pitchers control. xFIP added some flavor to this to account for the randomness of fly ball outcomes.
There's been a change in that mindset a bit, in that maybe a pitcher has
some control over balls in play. If they can induce a shit-ton of pop-ups, or a shit-ton of grounders, they're usually better than guys who can't do that. Of course, there's better, smarter people out there explaining this stuff. I'd check this out if you're interested:
SIERA | Sabermetrics Library.
All of that said, I think wRC+ is a great stat to see what happened and it's surely got more predictive power year to year than batting average, OBP, or SLG on their own. But in reality, what most analysts and hardcore fantasy nerds like me are looking at is Statcast data, which is pitch/batted ball level data around exit velocities, launch angles, spray angles, etc.
On the pitching side, there's pitch modeling stats like Stuff+ which is looking ONLY at pitch movement, attack angles, release points, etc. and using machine learning to determine who's got the best "stuff". Location+ builds on that to determine strength of command, and Pitching+ puts some of that together but has it's own quirks as well.
Always happy to chat about that stuff! It's super interesting.