[stupac2]

[quote author=stupac2 link=topic=13494.msg139410#msg139410 date=1446665441]

Well, that's only sort of right. At this point in the season you sure wouldn't have nailed 3/4. I have no problem with guessing what's going to happen after the CCGs. Also, we have a lot more historical data on the weather and we have a solid understanding of the physics that underlies it. If this were year 20 of the CFP playoff, I'd agree with you, but it's not. It's all too new. Silver even says so himself, they made significant changes to the model after last year, and will almost certainly be doing the same thing next year.

But, again, that's not really a thing I have a problem with. If people here want to build a complicated, comprehensive statistical model that's based on what little historical data we have, then go ahead. But most of what's happening in this thread is more along the lines of "red sky at morning, sailors take warning". Which, sure, maybe more accurate than flipping a coin (though I doubt it), but not particularly useful. The simple fact is that we don't really any idea if a 2-loss SEC champion Alabama trumps us, or if Clemson winning out helps us, or whatever the hell else people are trying to game out. We just don't know because it's too sensitive to other results and we don't understand the committee's thinking well enough.

OK, then you tell me what the weather is going to be on December 7th, when the final CFB playoffs are announced.  With years of historical data, you can probably narrow it down a fair bit.  And we can probably narrow it down, given what we know, to about 20 or so that have a reasonable likelihood, and nail three out of the four in a list of 10 teams.

Is it possible that something very unseasonable happens (like snow in the Bay Area on December 7th) and a team comes out of nowhere to nab a playoff spot?  Sure.  But just because we can't say with certainty doesn't mean the whole exercise is a waste of time. 

BC
[/quote]

Again, that's not what I'm saying. What I'm saying is that gaming out specific, unlikely scenarios is a waste of time. For two reasons, first because they're not that likely, and second because we don't really know what would happen around the edges of this new system.

I don't really get why this is at all comparable to predicting distant weather events, where our uncertainty is mostly because of our inability to measure the atmosphere well enough. That's not even remotely the problem here!

[stupac2]

[quote author=stupac2 link=topic=13494.msg139410#msg139410 date=1446665441]

Well, that's only sort of right. At this point in the season you sure wouldn't have nailed 3/4. I have no problem with guessing what's going to happen after the CCGs. Also, we have a lot more historical data on the weather and we have a solid understanding of the physics that underlies it. If this were year 20 of the CFP playoff, I'd agree with you, but it's not. It's all too new. Silver even says so himself, they made significant changes to the model after last year, and will almost certainly be doing the same thing next year.

But, again, that's not really a thing I have a problem with. If people here want to build a complicated, comprehensive statistical model that's based on what little historical data we have, then go ahead. But most of what's happening in this thread is more along the lines of "red sky at morning, sailors take warning". Which, sure, maybe more accurate than flipping a coin (though I doubt it), but not particularly useful. The simple fact is that we don't really any idea if a 2-loss SEC champion Alabama trumps us, or if Clemson winning out helps us, or whatever the hell else people are trying to game out. We just don't know because it's too sensitive to other results and we don't understand the committee's thinking well enough.

OK, then you tell me what the weather is going to be on December 7th, when the final CFB playoffs are announced.  With years of historical data, you can probably narrow it down a fair bit.  And we can probably narrow it down, given what we know, to about 20 or so that have a reasonable likelihood, and nail three out of the four in a list of 10 teams.

Is it possible that something very unseasonable happens (like snow in the Bay Area on December 7th) and a team comes out of nowhere to nab a playoff spot?  Sure.  But just because we can't say with certainty doesn't mean the whole exercise is a waste of time. 

BC
[/quote]

Again, that's not what I'm saying. What I'm saying is that gaming out specific, unlikely scenarios is a waste of time. For two reasons, first because they're not that likely, and second because we don't really know what would happen around the edges of this new system.

I don't really get why this is at all comparable to predicting distant weather events, where our uncertainty is mostly because of our inability to measure the atmosphere well enough. That's not even remotely the problem here!