(12-22-2021 04:22 PM)Lester Wrote: I'm still baffled by Kent State's NET ranking. Kent is a very good team that has had an excellent start to its season. It's hard for me to understand how the Rockets to date have outperformed Kent State. I would not be at all surprised to see Kent win the conference.
Yeah, NET (like any ranking system) doesn't necessarily accurately reflect which team is better. I agree Kent looks like the best team right now, with Buffalo close behind.
But again, NET weights efficiency highly, and only includes DI games, so Kent's 2 blowouts of John Carrol and Clarion don't count. Similarly, Ohio's win over Mercyhurst. If you look at unadjusted efficiency rates for a few MAC teams right now, you see:
Toledo 0.998-0.838 = +0.161
Kent 1.041-.927 = +0.113
Buffalo 0.990-0.887 = +0.104
Ohio 1.014-.932 = +0.082
CMU 0.810-0.988 = -0.179
Now again, NCAA supposedly adjusts those based on strength of opponents, but doesn't say how. I would expect that Kent, UB and OU's scores would be adjusted up relatively to UT's based on opponents, but can't say how much.
The other variable is team value index, which is weighted lower (unknown how much) and takes into account the strength of the opponent, the location (home/away/neutral) and winner. Supposedly does not factor in margin of victory. But who knows what the formula is.
UT has wins over teams 200, 235,261, 280, 295, 349 and losses to teams 19, 84, 120
Kent has wins over teams 62, 125, 147, 234, 249, 349 and losses to team 52.
One assumes Kent scores a lot better on this metric, but no way to know how much.
Do I think UT is a better team than Kent - no, not at all. But the NCAA's formula does, at least right now.