Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Aug 22, 2018 10:28:50 GMT -5
|
|
|
Post by ptctitan on Aug 23, 2018 6:48:34 GMT -5
Looks like NET will use KenPom-like stats along with other stats. But the NCAA is keeping secret the algorithms used in NET. So, at-large tournament selections and seeding may become even less transparent with the committee being able to point to the final NET rankings as the reason for its decision.
For mid-majors in one-bid conferences like the HL, this increases the need of winning the conference tournament.
It would be interesting to compare how NET would have ranked last year's teams compared to RPI in order to see what effect, if any, NET would have on the tournament.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Aug 23, 2018 7:00:37 GMT -5
It would be interesting to compare how NET would have ranked last year's teams compared to RPI in order to see what effect, if any, NET would have on the tournament. That's exactly what I was thinking - give me a side by side listing to see how the tourney teams and seedings would differ
|
|
|
Post by Commissioner on Aug 23, 2018 7:13:56 GMT -5
I liked RPI. For all its flaws, and for the extent to which many people didn't understand it, it actually was quite simple (and hence understandable to non-statisticians) who wanted to take a few minutes to figure it out; it was easily replicable. Transparency and simplicity are two tremendous pluses for systems like this, in which people are being asked to accept the system as legit. And in the end, when I looked at RPI results, by the end of the year it was more or less as good as any of the more sophisticated systems for determining who was going to win games in the tournament. Because it was a transparent system, it had also enabled others to develop some neat features--I'm thinking mainly of RPI Forecast, which allowed you to enter hypothetical results for upcoming games and get a prediction of where your RPI might end up. Here's a thread that lists the Titans' RPI history. udtitanbasketball.freeforums.net/thread/306/titan-rpi-history. Unfortunately, RPI started just after some of our best years and as we began to fade from national relevance. I'd be very interested in someone someday recreating RPI rankings for, say, 1949--1979. Here is an early RPI printout, to see what it looked like in those early days. extra.ncaa.org/solutions/rpi/Stats%20Library/1980%20Final%20RPI.pdf
|
|
|
Post by Commissioner on Feb 24, 2021 9:22:14 GMT -5
Is NET superior to RPI? The number 1 team in the nation this year using the old RPI formula is ... Navy!
The Midshipmen are having a darn good season. They're 13-2, with a win at Georgetown (which is way down this year); their losses are at Maryland (by a lot) and to a decent Army squad, in overtime. #1? I don't think so. Navy is 85th by NET.
RPI has Clemson at #5, which leaps out at ya. Nothing in the NET rankings jump out quite that way.
There was a general sense that mid-majors would not do quite as well under NET as under RPI. Just eyeballing it, that looks true, but not dramatically so. Among those teams that might be fighting for NCAA bids, there are only a couple mid-major schools doing noticeably better by NET than RPI: Loyola: Loyola is #32 by RPI, but #11 by NET. With that NET ranking, you have to figure they're an at-large lock if they don't win the MVC tournament.
Drake: Loyola's MVC running mate is also doing a bit better by NET: 34 NET vs. 44 RPI. Those extra 10 spots could help a lot if they're fighting for an at-large bid.
But they're generally lower. That said, NET "feels" more correct. Mid-Majors in the top 50 RPI: BYU: 11 RPI, 20 NET (I don't really think of BYU as a Mid-Major, but they do play in the WCC) Colorado State: 14 RPI, 45 NET St. Bonaventure: 15 RPI, 39 NET Boise State: 18 RPI, 32 NET Wichita State: 20 RPI, 69 NET UC Santa Barbara: 23 RPI, 42 NET (we haven't checked in on the Gauchos this year, even in the scoreboard. They're 14-3 and atop the Big West). VCU: 25 RPI, 35 NET Western Kentucky: 34 RPI, 74 NET Louisiana Tech: 38 RPI, 87 NET.
Gonzaga, by the way--still a mid-major--is #1 in NET, #2 in RPI.
The Horizon's top team, Wright State, is benefiting slightly by NET--65 vs. a 69 RPI. Neither will get them an at-large bid, but if they push that NET up further with a conference tournament run, they could get a decent seed.
One team I left off above is really strange: Colgate. Colgate? The Red Raiders are 11-1 on the year, and riding a 10 game winning streak. That has their RPI at #10 but--get this--their NET is also a very sold #13, between Florida State and Southern Cal. But here's the real oddity--Colgate has only played 3 schools all year. They're 4-0 vs. Holy Cross and Boston University, and 3-1 vs. Army. Talk about an unbalanced schedule.
|
|