Tweet
The simplest way to gauge who is doing best in the conference is to look at current standings. But given the overall challenges of winning on the road, some have tried to improve this by giving road wins a +1 and home losses a -1, and using the cumulative total of these marks as an “improved” method of gauging current standings. I have developed a simple model that tries to improve this a bit more by adding a quality of opponent component to this method.
For each of the past three years, I classified the teams into groups based on the final conference standings – the top 3 teams (“contenders”), the next 3 (“middle of the pack”), and the bottom 3 (“cellar-dwellers”). I then evaluated how each group fared against the other groups at home and on the road.
Thus, when contenders play other contenders, the home team wins about 2 out of 3. Contenders beat the middle of the pack about 8 out of 9 times at home, and about 2 out of 3 on the road. Contenders beat cellar-dwellers 100% of the time at home, and nearly that on the road (1 loss in three years).
The middle of the pack beat cellar-dwellers about 8 out of 9 at home, and about 2 out of 3 on the road. When they play each other (and when cellar-dwellers play each other), the home team wins 2 out of 3.
I applied this analysis to this year’s season by assigning three teams to each group, based on their preseason performance. I put BYU, SDSU, and UNLV as contenders. UNM and CSU were placed in the middle of the pack, and Wyoming as a cellar-dweller. That left Utah, TCU, and AFA, and in the end, I decided to put TCU in the middle of the pack, with Utah and AFA in the cellar with Wyoming.
As the conference season begins, the results of the analysis predicts that each contender will end the season with 12.5 wins, those in the middle of the pack will get 8 wins, and cellar-dwellers will have 3.5 wins. Then as the season progresses, these forecasts are adjusted to include actual results. Basically, if a team wins a game they had a low probability of winning, such as BYU at UNLV (1/3 probability), their expected wins goes up by the difference between 1 and that probability.
After last night’s games, BYU is now at 13.3 wins, and SDSU at 13.4. Here is the detail on how BYU’s is calculated:
Home games:
Vs contenders 0-0 2 games remaining 1.33 exp. wins
Vs middle of the pack 1-0 2 games remaining 1.78 exp. wins
(TCU)
Vs cellar dwellers 1-0 2 games remaining 2.00 exp. wins
(AFA)
Away games:
At contenders 1-0 1 game remaining 0.33 exp. wins
(UNLV)
At middle of the pack 0-0 3 games remaining 2.00 exp. wins
At cellar dwellers 1-0 2 games remaining 1.90 exp. wins
(Utah)
4 actual wins + 9.3 expected wins = 13.3 forecast.
This model currently forecasts the final standings to be:
SDSU 13.4 wins 13-3
BYU 13.3 wins 13-3
UNLV 10.8 wins 11-5
CSU 9.0 wins 9-7
TCU 7.2 wins 7-9
UNM 6.7 wins 7-9
Utah 4.8 wins 5-11
AFA 3.8 wins 4-12
Wyo 3.3 wins 3-13
For each of the past three years, I classified the teams into groups based on the final conference standings – the top 3 teams (“contenders”), the next 3 (“middle of the pack”), and the bottom 3 (“cellar-dwellers”). I then evaluated how each group fared against the other groups at home and on the road.
Thus, when contenders play other contenders, the home team wins about 2 out of 3. Contenders beat the middle of the pack about 8 out of 9 times at home, and about 2 out of 3 on the road. Contenders beat cellar-dwellers 100% of the time at home, and nearly that on the road (1 loss in three years).
The middle of the pack beat cellar-dwellers about 8 out of 9 at home, and about 2 out of 3 on the road. When they play each other (and when cellar-dwellers play each other), the home team wins 2 out of 3.
I applied this analysis to this year’s season by assigning three teams to each group, based on their preseason performance. I put BYU, SDSU, and UNLV as contenders. UNM and CSU were placed in the middle of the pack, and Wyoming as a cellar-dweller. That left Utah, TCU, and AFA, and in the end, I decided to put TCU in the middle of the pack, with Utah and AFA in the cellar with Wyoming.
As the conference season begins, the results of the analysis predicts that each contender will end the season with 12.5 wins, those in the middle of the pack will get 8 wins, and cellar-dwellers will have 3.5 wins. Then as the season progresses, these forecasts are adjusted to include actual results. Basically, if a team wins a game they had a low probability of winning, such as BYU at UNLV (1/3 probability), their expected wins goes up by the difference between 1 and that probability.
After last night’s games, BYU is now at 13.3 wins, and SDSU at 13.4. Here is the detail on how BYU’s is calculated:
Home games:
Vs contenders 0-0 2 games remaining 1.33 exp. wins
Vs middle of the pack 1-0 2 games remaining 1.78 exp. wins
(TCU)
Vs cellar dwellers 1-0 2 games remaining 2.00 exp. wins
(AFA)
Away games:
At contenders 1-0 1 game remaining 0.33 exp. wins
(UNLV)
At middle of the pack 0-0 3 games remaining 2.00 exp. wins
At cellar dwellers 1-0 2 games remaining 1.90 exp. wins
(Utah)
4 actual wins + 9.3 expected wins = 13.3 forecast.
This model currently forecasts the final standings to be:
SDSU 13.4 wins 13-3
BYU 13.3 wins 13-3
UNLV 10.8 wins 11-5
CSU 9.0 wins 9-7
TCU 7.2 wins 7-9
UNM 6.7 wins 7-9
Utah 4.8 wins 5-11
AFA 3.8 wins 4-12
Wyo 3.3 wins 3-13