(11-14-2017 03:34 PM)Frog in the Kitchen Sink Wrote:  

(11-14-2017 01:49 PM)ken d Wrote:  

(11-14-2017 10:59 AM)stever20 Wrote:  

(11-14-2017 10:47 AM)ken d Wrote:  

(11-14-2017 09:38 AM)stever20 Wrote:  Only a 14.5% chance that Texas beats WV and then loses to Tech.
more likely 18.6% chance that Tech beats TCU and then loses to Texas.

Kansas St has a 58% chance because they still do have a 16% chance to beat Oklahoma St, and then the Iowa St game is basically 50/50. So their chances have to be greater than 50%.

I'm not a statistician, and I understand that probability analysis is not intuitive. But to me, saying that K State has a 58% chance suggests that it is "more likely than not" that they will win at least one of two games in which they will be underdogs. To me, that conclusion is counterintuitive. It certainly isn't a bet I would make.

Well, one of the 2 games is basically a pick'em type of game. I don't know that they will be underdogs in the 2nd game at all.

And it really does make sense. If you believe that Kansas St has a 16% chance of beating Oklahoma St- and the 2nd game is a 50/50 game- You have to add in the chance that Kansas St beats Oklahoma St to the 50/50 chances. The 16% gets cut in half and added to the 50% for the 2nd game.

So you look at it-
7.9% to go 2-0
49.5% to go 1-1
42.6% to go 0-2

Actually, I don't believe that K State has a 16% chance of beating OK State. Three touchdown underdogs don't win at nearly that rate. The data I've collected suggest it's closer to 5%. And, I don't view their chances against ISU at 50-50 either. If the game were being played this week, I would expect ISU to be a 3-5 point favorite. That's not much, but underdogs in that point spread range won less than 30% of the time.

Underdogs do win sometimes, and I'd guess they win more often in November than they do in October. So there's a chance. I might even pick them in a Pickem Contest if I were behind and trying to catch the leaders. I just wouldn't bet any real money on their chances.

3 point underdogs win about 45% of the time. 5 point underdogs win about 35%. 1 TD underdogs win 30% of the time. 10 point underdogs win 25% of the time. 2 TD underdogs win about 15% of the time. 17 point underdogs 10 % of the time

https://www.boydsbets.com/how-often-the-...-football/

In the big 12, TCU is a 1 TD favorite over Tech, WVU is 2.5 over UT, OSU a 17.5 over KSU. So those percentages are pretty similar to the ESPN FPI probabilities.

So if you use those percentages, the chance all three lose this weekend is only 34%, with UT most likely to win.

And the problem for that is that Texas is also the most likely of the 3 to win next week...

Very possible it winds up at Texas at 7-5 and KSU and Tech both at 5-7.