Posted by MikeT23 on 9/9/2012 9:36:00 PM (view original):
I'm 99% positive that you're wrong. You're assuming our fake players can take bad angles and be in a poor throwing position. That's seems ridiculously advanced for a sim game that costs $25 every three months.
You've lost me on your last post. I've made no assumptions regarding angles or positioning or footwoork. There are no angles, positioning or footwork in HBD. There is no geometry or physiology at all in HBD - just some algrebra and a hell of a lot a lot of arithmetic.
In your quiz question RF A has a range of 65 and RF B has a range of nine. We assume they both have two feet, two arms, and we know, because WIS tells us, that A is from Kansas and B is from the Dominican Republic. But all that is irrelevant. All that is relevant (in our controled experiment) is that A's range is 56 points, or 722%, greater than B's. This could mean that A is capable of getting to 722% more balls than B (probably not), or that he gets to balls 722% quicker than B (perhaps a little more likely, but probably not). Such speculation is academic. All we need to know is that A's range is 56 points better than B's. That enables us to know (within the context of a valid statistical sample with all other variables held equal) that A is "rangeier" that B, that in a 162 game season he is likely to get to more balls and achieve more putouts and assists than B.
You've constructed your quiz question so as to leave no doubt that A is a better defensive RF than B. Why are we arguing about that? I thought your hypothesis/arguement was that, although A is a better defensive RF than B, it's all right to go ahead and play B (the superior offensive player) in RF because in HBD RF defense is not statistically relevant enough to make it rational to play a defensivelly superior RF over an offensively superior C/RF. (Or something like that.)
You may very well be correct on this. But, for the reasons discussed above (and some others I've left undiscussed) I find your theory statisically dubious. I mean no offense. So little is known about HBD's programming, decision tree, and the interplay of the multitude of variables in HBD, it's doubtfual any theory such as yours can be statistically "proven."
In RL managers play hard hitting, low range slugs in RF all the time. (Remember when Jose Canseco got hit on top of the head with a fly ball?) Some days they get away with it, some days they don't. But the smarter managers know they're a lot likelier to get away with it with a lefthander on the mound than with a righthander. Baseball management has always been based on a tangle of tradeoffs, compromises and hunches levened with an overlay of statistics. Let's hope HBD is like that too.