I have a strong quantitative analysis game. The Net ranking is pure garbage IMHO. It seems to over-engineer the predictive process and therefore getting too far away from the sole objective - Wins and losses. The proof will be in the pudding though. I have the Net rankings for all NCAA participants and intend to follow how well it does in the tourney.
Also, conference tournament success is different from regular season success and will be different than NCAA tourney success. The key differentiating factor - ability to play on just 24 hr rest. It's the same as why an MLB team can have a record breaking regular season yet falter in a quick Best of 5-7 in the playoffs. Different levels of rest materially impact how well a team plays vis a vis another.
I get the sense the predictive process is actually
under-engineered, data-starved and backward-looking, and would do better with richer data, particularly these days with the flow of transfers all over the map.
And later games should most definitely have more weight than earlier games. If it matters to the selection committee, then it should matter in the NET, otherwise what is the point of the NET?
Still, if the records vs. Q1 and Q2 mean anything, the Gamecocks being behind so many teams with far less impressive numbers is a mystery.
I definitely agree about downplaying conference tournament performance. Fatigue matters. Risk of injury matters. Having something/nothing to win/lose matters. Look at the teams that overperformed in any power-conference tourney. Underperformers didn't have much to lose except maybe a key player to injury. Hanging on to a 5 seed isn't worth gassing yourself. Nor is going from a 2 to a 1, for that matter.
As for some of the unfairly labeled "traditional one-bid" conferences, every team's whole season was on the line. James Madison's 31st win of the season was in the Sun Belt championship, and they had to get it. But that's another discussion.
www.ncaa.com
