- Joined
- Apr 1, 2013
- Messages
- 390
- Reaction Score
- 2,489
I compared UConn's schedule to its schedule over the last seven years using three different metrics. In each case I found that it was a bit easier than previous years, but pretty comparable to the 2008-09 campaign. Even though it was a bit easier, I don't think the schedule was as different from prior years as some would imagine based on all the complaining about our conference schedule.
For each metric the basic data is the Massey ranking at the end of the year, and I assumed a ranking (the # 5 team for example) was the same quality of team in each year. The first metric is how many top 5, top 10, or top 25 teams did we play? I don't care about teams ranked below 25 as we haven't lost to such a creature in 20 years. AS you can see, this year was easier.
The second metric is expected number of wins. For this I used the assumption that we would win 50% of our games against the #2 team, 52% against the #3 team... up to 100% against teams ranked 25 and below. This odds are roughly equal to our historical averages since 1994. As shown in the table below, the expected wins ranges from 26.5 to 28.4. I dropped one game from 2014 to keep the length of the seasons the dame at 30. Of course we had better records than that because we had better than average UConn teams.
The third metric was expected point spread where I estimated that with a regression of the log ranking against actual margins of victory over the past seven years (empirically this generates a point spread of 1 point against #1, 16 points against #5, 30 points against #25 and 48 points against #200). Again our margin of victory would have been expected to be larger this past season than in previous years.
For each metric the basic data is the Massey ranking at the end of the year, and I assumed a ranking (the # 5 team for example) was the same quality of team in each year. The first metric is how many top 5, top 10, or top 25 teams did we play? I don't care about teams ranked below 25 as we haven't lost to such a creature in 20 years. AS you can see, this year was easier.
The second metric is expected number of wins. For this I used the assumption that we would win 50% of our games against the #2 team, 52% against the #3 team... up to 100% against teams ranked 25 and below. This odds are roughly equal to our historical averages since 1994. As shown in the table below, the expected wins ranges from 26.5 to 28.4. I dropped one game from 2014 to keep the length of the seasons the dame at 30. Of course we had better records than that because we had better than average UConn teams.
The third metric was expected point spread where I estimated that with a regression of the log ranking against actual margins of victory over the past seven years (empirically this generates a point spread of 1 point against #1, 16 points against #5, 30 points against #25 and 48 points against #200). Again our margin of victory would have been expected to be larger this past season than in previous years.