Around the A10

[3star_recruit]

Here's how the top mid-majors did through the first weekend.  Solid showing by the A-10 so far:

Team    Record    W%    SOS
Colonial 17-3 0.850    -0.49
AAC 13-3    0.813    -16.22
A-10 16-5    0.762    -1.01
MWC 13-6    0.684    4.59
WCC 12-7    0.632    2.86
MVC 10-6    0.625    -6.75
Ivy 8-5        0.615    -3.89

[Old guy]

11 hours ago, Quality Is Job 1 said:

I give up.  You're obviously not understanding what I'm saying.  I think "Bayesian" is throwing you off.

Do not give up,  you are correct in what you are saying. What I am doing is pure estimation of probabilities without using formulas of any kind. From my point of view, which is reinforced by my long career in corporate, models are intrinsically flawed (some more flawed than others). They are approximations of  reality which may indeed come very close to reality, as the Wiz's models are. Those models that come closest to reality tend to have a built in adjustment capability that allows them to stay close to reality. The Wiz's model appears to be particularly good in this regard.

When I point out that a major betting service (Teamrankings) that is in the business of selling advice to bettors, shows a set of statistical results (it is indeed early on in the season and as we all know and the stats are far from set in stone) that favors LSU heavily against VCU, and then says in print that the odds are close to even for their coming game, that should be surprising, but should it really be surprising?

Quality, do you happen to work with statistical models in your job?

[billikenfan05]

17 hours ago, 3star_recruit said:

Bona has had a key injury three straight years in the beginning of the season.  They're snakebitten in November.

It could be possible to blame Schmidt’s preseason preparation for their injuries? 

[Quality Is Job 1]

18 minutes ago, Old guy said:

Do not give up,  you are correct in what you are saying. What I am doing is pure estimation of probabilities without using formulas of any kind. From my point of view, which is reinforced by my long career in corporate, models are intrinsically flawed (some more flawed than others). They are approximations of  reality which may indeed come very close to reality, as the Wiz's models are. Those models that come closest to reality tend to have a built in adjustment capability that allows them to stay close to reality. The Wiz's model appears to be particularly good in this regard.

When I point out that a major betting service (Teamrankings) that is in the business of selling advice to bettors, shows a set of statistical results (it is indeed early on in the season and as we all know and the stats are far from set in stone) that favors LSU heavily against VCU, and then says in print that the odds are close to even for their coming game, that should be surprising, but should it really be surprising?

Quality, do you happen to work with statistical models in your job?

The answer to your question is that I do not.

[Old guy]

OK, I understand better. I do not intend this as a put down in any way or manner. If things are not the way they look or the way the numbers indicate, that is because that is the way it is. Reality is not logical, or rational, and does not follow statistics, not all the time anyways. The Wiz is a wonderful resource to have in this blog, and so is Taj, and Clock and yourself.

Let's enjoy the surprising twists and turns we are about to get from this season. Life is good, at least so far, for the Bills. Let's hope and pray that we have enough horsepower in the team to continue this trend. Go Bills!