Joined: 02 Nov 2009
Location: Carson City
Back to top
|Posted: 03/11/18 11:25 pm ::: Non Conference Strength of Schedule
|I realize I am just venting because of the release of the 8 bubble teams which seems to have Dayton safely in and Minnesota on the Bubble, but it is truly hard to believe how the Committee could come to that conclusion. So lets compare. W-L vs top 25 (Minn, Day) (3-3, 0-2) vs 26-50 (1-1, 1-1) vs 51-100 (6-2, 7-2) vs 101-200 (3-2, 5-1) vs 201+ (10-0, 10-0). This is a clear advantage for the Gophers. The overall RPI (41, 36) SOS (82, 92) and non-conference RPI (78, 69) are all comparable. But the one big difference is the Non conference Strength of Schedule (293, 53). The questions are, "Does it really mean anything?" and "Should it matter? My answers are, "Not much" and "No."
I have heard the argument that these are the games that coaches can control because they make their non-conference schedules. That is somewhat true, but there are many other reasons that go into how non-conference schedules are constructed that go to building a team to become better as the season goes on. There are also monetary considerations, local rivalries, tournaments where the schedules are set by others and team availability. You can also schedule what you believe to be a tough schedule, but if those teams underperform the schedule may wind up weaker than expected.
There are also vagaries of RPI that make it much better for SOS purposes to schedule top teams from bad conferences than bottom teams from good conferences, even though the teams may be roughly equivalent. Had Minnesota played George Mason, Texas St and Norfolk St instead of Wake Forest, North Carolina and Boston College their actual schedule strength wouldn't have been any harder but their NCSOS would have likely moved up into the top 250, maybe higher.
To show the true absurdity of the measure, here are the teams actually played by one tournament team:
To summarize, that is 1 Top 25 team, 2 teams 26-50. neither of whom would have been considered as an at-large, 2 teams 51-100, 3 teams 101-200 and 3 teams 200+. So where does this schedule rank in terms of NCSOS. Answer - 27. This is not a tough schedule, but because there is only 1 game against a power 5 conference and 3 teams won or tied for first in their mid-major conference, the NCSOS is stupid high. The team is Villanova, and it is probably worth at least one seed above what they deserve.
Using this metric is a double whammy. I have always believed that how a team is playing at the end of the year should count more than what they did at the beginning. Whether this is measured by last 10, conference record (at least in conferences with even schedules like the Big 12) or other ways, teams grow throughout the course of a season. The NCSOS is a dumb metric because it is already reflected in the record by the different stratifications. Forther using it makes the non-conference worth nore than the conference and that is just wrong.
If I were to calculate it I would value the first game at 1 and the final game at 1.5, with a straight line multiplier for the games in between. This may be different for non-majors where there best chances to impress are in the non-conference. Western Kentucky beat Missouri in the first game of the season on a neutral court, and while that is a big win I would not hold that loss against the Tigers as much as if it happened in February.
OK rant over.