UConn 57th in First NET Rankings
- Thread starter jkfer98
- Start date
zls44
Your #icebus Tour Director
- Joined
- Aug 24, 2011
- Messages
- 9,137
- Reaction Score
- 24,629
He also thinks Memphis has, literally, ever mattered on the court as a program.
- Joined
- Aug 2, 2016
- Messages
- 4,706
- Reaction Score
- 59,878
ESPN's BPI has UConn at 91, I don't see RPI posted anywhere this season
- Joined
- Aug 26, 2011
- Messages
- 3,601
- Reaction Score
- 9,754
Yeah, it's just unlike all these other things (even the NET) RPI had no "secret sauce" baked in that made it impossible to replicate. So we could actually see what it looks like in comparison.
I have a feeling it might be a new worst ranking Nate Silver has ever seen.
I have a feeling it might be a new worst ranking Nate Silver has ever seen.
- Joined
- Aug 2, 2016
- Messages
- 4,706
- Reaction Score
- 59,878
It's early, but if this really doesn't account for strength opponent, this could really open the door for mid-majors to take more at-large bids. But we will have to see how well this correlates in March.
I've read that it's early so there isn't much confidence to adjust so the system is probably designed to not adjust much. There will probably be less adjustment than KenPom in general though thanks to the other criteria.
The biggest issue long term would seem to be that the system uses raw rather than adjusted efficiency. Having an offensive efficiency of 1.2 against Virginia should not be treated the same as having an offensive efficiency of 1.2 against Incarnate Word.
The reason it's so crazy right now is that unlike KenPom & others it doesn't have a baseline rating to prevent the early blowouts from affecting its initial ratings. Over time that will correct. The non-adjusted numbers do make me wonder how it will fare over the year, though.
The reason it's so crazy right now is that unlike KenPom & others it doesn't have a baseline rating to prevent the early blowouts from affecting its initial ratings. Over time that will correct. The non-adjusted numbers do make me wonder how it will fare over the year, though.
- Joined
- Aug 2, 2016
- Messages
- 4,706
- Reaction Score
- 59,878
OK I found UConn's ratings across the board:
KenPom: 87
Sagarin: 84
Massey: 88
BPI: 91
RPI: 128
NET: 57
KenPom: 87
Sagarin: 84
Massey: 88
BPI: 91
RPI: 128
NET: 57
- Joined
- Dec 29, 2011
- Messages
- 2,489
- Reaction Score
- 30,460
What we do know for certain is this: Unlike ratings such as those at KenPom.com, which use a proprietary formula to establish a preseason projection that is factored into early season numbers, every team’s ranking in the NET begins at 0. So the limited data set after just a few weeks of games inevitably will produce odd results based more on how a team’s schedule has broken to date than on the team’s actual value.
“I think we very much anticipated there would be some head-scratchers here, but we’re also confident that the more games are played, the picture will become clearer,” Gavitt said. “This tool is used in Selection Week in March. Ultimately, it’s not a tool that has full value until an entire schedule has been played out. Obviously, we’re a long way from that.
“Because it’s so new and so different from the RPI, we just think that getting out information, sharing it on a regular basis, letting coaches, student-athletes, fans and the media know what the implications are would be important.”
Hysterical overreaction to NCAA ratings can be NET positive for the sport
“I think we very much anticipated there would be some head-scratchers here, but we’re also confident that the more games are played, the picture will become clearer,” Gavitt said. “This tool is used in Selection Week in March. Ultimately, it’s not a tool that has full value until an entire schedule has been played out. Obviously, we’re a long way from that.
“Because it’s so new and so different from the RPI, we just think that getting out information, sharing it on a regular basis, letting coaches, student-athletes, fans and the media know what the implications are would be important.”
Hysterical overreaction to NCAA ratings can be NET positive for the sport
- Joined
- Dec 29, 2011
- Messages
- 2,489
- Reaction Score
- 30,460
Think it had more to do with Syracuse beating Ohio State.
True, both probably, but Oregon's jump helps your argument.
UConn doesn't have the money to move up in a point of sale ranking system.
- Joined
- Aug 2, 2016
- Messages
- 4,706
- Reaction Score
- 59,878
Up to 32 today
Husky25
Dink & Dunk beat the Greatest Show on Turf.
- Joined
- Sep 10, 2012
- Messages
- 18,701
- Reaction Score
- 19,783
25 place jump without playing a game...what a country?
- Joined
- Mar 21, 2018
- Messages
- 671
- Reaction Score
- 2,195
To be honest, the system doesn’t seem half bad. It’s at least seems less biased to pre existing opinions. The thing that bugs me the most is when teams that have multiple losses or a loss to a mediocre team being ranked over undefeated teams that have been blowing their competition out. Is it fair that a team like UNC is ranked 11th?
Obviously exceptions like Duke, nova (since they are defending champs), but why should UNC be ranked above Iowa? Because they’re a historical program who people pay more attention to (and resultantly look at their talent a lot more).
Obviously exceptions like Duke, nova (since they are defending champs), but why should UNC be ranked above Iowa? Because they’re a historical program who people pay more attention to (and resultantly look at their talent a lot more).
Because North Carolina's loss at Michigan is incredibly more difficult than any game Iowa has won. And Iowa will probably lose that same game later this year. It makes just as much sense to penalize for unequal opportunity this early in the season as it does to rank based on pre-existing opinions. Iowa's most difficult game was their win on a neutral court against Oregon, who then lost to Texas Southern at home. North Carolina's other loss, neutral court by 3 to top 20 ranked Texas, is also more difficult than that Oregon win. Who's to say Iowa wins either of those games? UNC has faced the 50th hardest schedule. Iowa the 250th. How do you reconcile that?
The truth of the matter is that systems and models use pre-existing "infrastructure" and player projections in the early season because it makes the model more accurate than starting from 0 and relying only on results. These guys test these things. Giving Roy Williams and Jay Wright the benefit of the doubt in the early season has been proven to be more accurate than not doing so and just relying on the small sample sizes at the beginning of the season.
Are we having this discussion if UNC hits 1 more 3 and wins against Texas in OT and their only loss is at Michigan? So we're not even quibbling about 1 game, but maybe 1 or a handful of possessions.
