Started By
Message
re: Ole Miss has the widest potential W-L variance.
Posted on 7/15/13 at 12:28 pm to KaiserSoze99
Posted on 7/15/13 at 12:28 pm to KaiserSoze99
quote:
Ole Miss has the widest potential W-L variance.
Typically we think of variance in terms of all the cases, not just one case. In this example, Ole Miss would be a single case, and we would have an observed number of wins and losses for them. If we had all of the other cases (all of the other teams) we could calculate the win-loss variance. Obviously, this way of thinking about variance is much different from the way you are using it.
Another way of thinking about variance would be to imagine that we could simulate Ole Miss' season. In this scenario, each sim would be a "case" and we could calculate a variance for those sims.
Of course, that can't happen, as Ole Miss will only play the 2013 season once.
Anyways...it seems a trifling point, but I think what you are REALLY trying to say is that the probability that Ole Miss wins several important games is close to .50. Of course, these probabilities are dependent on various factors including the personnel, coaches, game locations, injuries, and prior performance, and they will change over the course of the season.
In all, I find your claim intriguing, although I think it is a bit more complicated than you make it out to be.
This post was edited on 7/15/13 at 12:32 pm
3
Posted on 7/15/13 at 12:30 pm to TxTiger82
You get his point obviously.
I've said it MANY times; there are 6 tossups that will define the 2013 season:
Vandy
Texas
Auburn
A&M
LSU
State
.500 or above is a good year
I've said it MANY times; there are 6 tossups that will define the 2013 season:
Vandy
Texas
Auburn
A&M
LSU
State
.500 or above is a good year
Posted on 7/15/13 at 12:34 pm to TxTiger82
quote:
Typically we think of variance in terms of all the cases, not just one case. In this example, Ole Miss would be a single case, and we would have an observed number of wins and losses for them. If we had all of the other cases (all of the other teams) we could calculate the win-loss variance. Obviously, this way of thinking about variance is much different from the way you are using it.
Another way of thinking about variance would be to imagine that we could simulate Ole Miss' season. In this scenario, each sim would be a "case" and we could calculate a variance for those sims.
Of course, that can't happen, as Ole Miss will only play the 2013 season once.
Anyways...it seems a trifling point, but I think what you are REALLY trying to say is that the probability that Ole Miss wins several important games is close to .50. Of course, these probabilities are dependent on various factors including the personnel, coaches, game locations, injuries, and prior performance, and they will change over the course of the season.
In all, I find your claim intriguing, although I think it is a bit more complicated than you make it out to be.
I enjoy a good academic discussion, but holy shite.
I couldn't think of a better way to state it, so "variance" was the word choice. Forgive my misuse.
Posted on 7/16/13 at 10:19 am to TxTiger82
quote:
Typically we think of variance in terms of all the cases, not just one case. In this example, Ole Miss would be a single case, and we would have an observed number of wins and losses for them. If we had all of the other cases (all of the other teams) we could calculate the win-loss variance. Obviously, this way of thinking about variance is much different from the way you are using it.
Another way of thinking about variance would be to imagine that we could simulate Ole Miss' season. In this scenario, each sim would be a "case" and we could calculate a variance for those sims.
Of course, that can't happen, as Ole Miss will only play the 2013 season once.
Anyways...it seems a trifling point, but I think what you are REALLY trying to say is that the probability that Ole Miss wins several important games is close to .50. Of course, these probabilities are dependent on various factors including the personnel, coaches, game locations, injuries, and prior performance, and they will change over the course of the season.
In all, I find your claim intriguing, although I think it is a bit more complicated than you make it out to be.
The most interesting approach is to set a confidence interval for the probabilities in each game and then run a simple Monte Carlo simulation. That would take a fairly straight-forward series of intuitive guesses (in the form of "I am 95% confident that the probability of Ole Miss beating State is between .65 and .8") and produce a probability for each potential record, from 0-12 to 12-0.
Page 1 of 1
Back to top
Follow SECRant for SEC Football News