OOTP Developments Forums - View Single Post

NoOne · 03-20-2016, 12:19 PM

the important thing with more sophisticated stats is to remember they are not absolutely definitive. i love what dawn wrote because too often people think of these questions as yes/no answers... it is a breakeven analysis. the answer is wholey dependent on the relevant variables to the situation. simply put, the hr/power guy isn't always the best answer for a spot in the order that will get at least a good share of rbi opportunities.

a war of 7 is not 100% always better than a war of 6.9. it could be a poorly weighted component of the war equation causing something to be off. war is no where near that accurate while in its current form.

this applies to more than just WAR. e.g. something like normalizing for parks is not a perfect science (if normalizing for parks is actually down to +/- a minimal % error, i will happily retract that statement). with the advent of the new stats tracking technology, the picture will get much clearer in the coming years - especially defensive metrics. once that new data levels off, we are going to have some cool stuff leaked from the MLB clubs as each 'best' practice comes to light and is begrudgeonly shared over time.

i don't think of any one stat as the 'best.' that's not how they are meant to be used. some provide better information than others. some provide more specific info as opposed to more general info. my favorite one(s) are the ones that give the best information relevant to the question at hand.

03-20-2016, 12:19 PM	#17
NoOne Banned Join Date: Apr 2015 Posts: 7,273 Infractions: 0/1 (3)	the important thing with more sophisticated stats is to remember they are not absolutely definitive. i love what dawn wrote because too often people think of these questions as yes/no answers... it is a breakeven analysis. the answer is wholey dependent on the relevant variables to the situation. simply put, the hr/power guy isn't always the best answer for a spot in the order that will get at least a good share of rbi opportunities. a war of 7 is not 100% always better than a war of 6.9. it could be a poorly weighted component of the war equation causing something to be off. war is no where near that accurate while in its current form. this applies to more than just WAR. e.g. something like normalizing for parks is not a perfect science (if normalizing for parks is actually down to +/- a minimal % error, i will happily retract that statement). with the advent of the new stats tracking technology, the picture will get much clearer in the coming years - especially defensive metrics. once that new data levels off, we are going to have some cool stuff leaked from the MLB clubs as each 'best' practice comes to light and is begrudgeonly shared over time. i don't think of any one stat as the 'best.' that's not how they are meant to be used. some provide better information than others. some provide more specific info as opposed to more general info. my favorite one(s) are the ones that give the best information relevant to the question at hand. Last edited by NoOne; 03-20-2016 at 12:32 PM.