Quote:
Originally Posted by chumdawg
That was a very well considered post, and I enjoyed reading it. But I'm afraid you left one thing out. Your null hypothesis has all the games being independent (but allows for a home-field advantage, the way I read it). In fact, that is *exactly* the model that gives the road team the edge in a series that does not go seven games. As in, there are three ways that a series can go something besides seven games:
sweep: two games at each site, net advantage zero
4-1: three games at road site, net advantage to road team
4-2: three games at each site, net advantage zero
Get it?
|
No, I don't. See my response to xrobx. Consider the Giants in last year's WS. Are you really going to argue that they lost home field advantage simply because they got it done in five games? I would say they had a *strong* grip on home field because they had two home games in their back pocket that they didn't even need.
Quote:
|
Now, I actually do believe that the games are not purely independent. But, as you say, this is complex and not at all easy to measure. But, in the case of baseball as opposed to basketball, there are indeed some tangible differences we can identify (outside of the "mojo" kind of things that you, probably rightfully, do not want to include). How you deploy your closer is one thing. That can vary, home or away, in certain game situations. Whether you use potential future starting pitchers, when/if necessary. Point is, the desperate team might have more bullets to fire, if it comes to it.
|
Yes. Baseball is unique because pitching matchups are the dominating factor over the course of a series. However, that would lead me to think all the more that the games are largely independent. The win probabilities won't be the same from game to game, but they would be independent nonetheless. For example, the Rangers won't have to worry about Carpenter tomorrow, so their probability of taking that game should be better than the last one (at least I hope!). Their chances would be determined by the matchup in that game, not in the fact they lost the previous one.
(And just because I have to say it... Leyland didn't have many bullets in Game Five.)
Quote:
And of course, if anything that runs counter to the "momentum" theory. But there's probably also something to the idea that a pitcher who takes the mound up 3-1 or up 3-2 has just a little more confidence in his pitches than a guy who is down 1-3 or down 2-3, and that can also be made manifest in the outcome.
I like to leave room for all these variables.
|
Which is another reason why I don't think one can decisively say it's always better to have Game Five at home. Maybe you'd rather have your ace pitch both Games One and Five on the road and let your #2 guy pitch his second game in front of the home crowd.
I will concede that certain events in one game can affect the outlook of another. But these are tangible things like pitcher availability as you said. I was arguing in abstraction that the timing of the third home game does not affect overall win probability due to "momentum".