One study shows a slightly better correlation, and the new data is considered factual and the old data is thrown out?
No, that is not how science works in any field.
There's not much in the way of a rigorous peer review system in sabermetrics.
What stats do you think people should be looking at?
Because if "they" is message-board posters, they most certainly do. "LOL, you are still using FIP-? xFIP is what counts!"
The consistency is pretty loose, to be honest. FIP and xFIP usually have a pretty noticeable difference each year. Was Matt Garza a 3.59 pitcher last year or a 4.17?That's why you have several versions of advanced pitching stats to choose from and there is usually a consistency within those stats among pitchers who regularly effectively prevent runs.
At this point, I think we need to accept that a single ERA-substitute for predictively describing pitching performance is probably impossible. There are too many variables that you can't universally account for with a single-case statistic.What stats do you think people should be looking at?
Look at raw peripherals and stuff. Ks, BBs, HRs, GBs, FBs, velocity, swinging-strike percentage, things like that. Trying to distill it into a single stat seems futile.
Normalizing HR% to 12.5% just seems like a classic case of overfitting to me.
I'd so much rather one stop shop. It's just easier that way.Look at raw peripherals and stuff. Ks, BBs, HRs, GBs, FBs, velocity, swinging-strike percentage, things like that. Trying to distill it into a single stat seems futile.
I missed this a few pages back, but I love this offseason mock. Only change you could conceivably make would be the addition of Fujikawa, which would allow Vizcaino to be stretched out in the minors in preparation for a possible big league rotation spot in 2014 (or even 2nd half of 2013). Love the Garza to Red Sox trade, although I'm not sure we could get that package for only one season of him.
If you can choose between which of the six numbers you like best, then at that point the whole point of objectivity goes out the window and you can just fit the number to what story you want to tell. It basically becomes the eye test with numbers.
Sort of like how every new Cubs prospect is referred to by the highest ranking of any of the dozen or so prospect rankings out there.
Last edited by KyleJRM; 11-29-2012 at 01:07 PM.
I like looking at the peripherals more than I like looking at the ERA and DIP numbers myself. The stats Kyle mentioned all are more important for me projecting a guy moving forward than trying to work with one number that uses all (or some - velocity, swing strikes, GBs, etc are not in there) of those other numbers to make any guesses.
The fun part of FanGraphs is also one of the worst parts of it for me. It's nice and convenient to know those numbers, but getting your hands dirty and working with the basics (and while the data is only now become more available stuff like average velocity, pitch movement, swinging strike% all count as basic) is where the best guesses can be made.
Just in case the Kyuji Fujikawa thread goes overlooked:
Rodriguez I'm least interested in of the three, but I'd still be happy with the addition. He just lost out on a $17.5 million option, and instead received a $4 million buyout from the Brewers. Boras client; made $8 million in 2012.
Adams I like. He's far from a lock to re-sign with the Rangers, and there's history with Hoyer.