• Welcome to OGBoards 10.0, keep in mind that we will be making LOTS of changes to smooth out the experience here and make it as close as possible functionally to the old software, but feel free to drop suggestions or requests in the Tech Support subforum!

Steve Forbes credibility watch

CONeill said:
If last night is the indication of what Reid will do all season, along with hitting 3s -- he and Forbes sound confident he can make those -- he's going to be a safer first-round projection than Sallis.
Click to expand...

Sure, but Sallis is already popping up on NBA draft boards. Reid is not.

If Sallis continues to shoot above 40% from 3 (currently an absurd 42.9%), he's gonna be a first round pick. His athleticism and measurables have never been in doubt. He was a plus defender and efficient scorer at Gonzaga inside the arc, but struggled to hit shots from the outside. And now he also gets to showcase how well he can handle the ball.

Sallis (probably) won't win ACC POY (largely because this Wake team around him is better), but he feels like a better pro than anyone else Forbes has had at Wake (LaRavia included).
 
Forbes credibility pretty darn high right now.

He took that call from the head coach (previous head coach?) of the Steward School in Richmond, VA.

"Hey, coach, I talked to Efton Reid and he's interested in moving back to the east coast. Are you interested?"

Forbes: "well, yeah."
 
Bones McKinney said:
You mean like Virginia did last year to Furman. Of course who wouldn’t take their natty but the year before that they lost to UMBC as the first #1 to go down to a 16.
Click to expand...
I would guess they were probably referring to 1 seed Purdue losing to Farleigh Dickinson in the first round. But they may have meant Maryland, Northwestern, Illinois, Penn State, Indiana, or Iowa not making it out of the first weekend.
 
re: The Big Ten - I don't think the NCAA committee is trying to predict who will do the best in the tourney are they? It's to put together the best at-large teams based on resume. That may be parsing definitions a bit, but I think there's a difference.
 
Deacsfan27 said:
re: The Big Ten - I don't think the NCAA committee is trying to predict who will do the best in the tourney are they? It's to put together the best at-large teams based on resume. That may be parsing definitions a bit, but I think there's a difference.
Click to expand...
This is similar to the discussion about the CFP - the 4 best teams or the 4 most deserving teams? In basketball, the committee is charged with selecting the 36 best teams. The best teams should also do the best in the tourney - I can't think of how that would differ.
 
They're not tasked with selecting the best 36 teams, they're picking the top 36 teams based on resumes and how they played. If they were just picking the best 36 teams, you'd obviously just use the metrics.
 
Again, the NET is by far the best system that the NCAA uses (with perhaps the Pairwise for hockey in the discussion where there is no committee at all - they just take the top X teams in the ratings after tournament winners). Is there room for tweaks? Of course. Is there much serious discourse in the analytics world about the KP/Torvik/Vegas models not doing a good job of evaluating/rating the top teams over the course of the season? No. Otherwise people would be making an incredible amount of money betting college basketball.
 
Rafi said:
This is similar to the discussion about the CFP - the 4 best teams or the 4 most deserving teams? In basketball, the committee is charged with selecting the 36 best teams. The best teams should also do the best in the tourney - I can't think of how that would differ.
Click to expand...
If they were trying to prognosticate the entire tournament there is no way anybody would have seen FAU, San Diego State and UConn to the Final Four. It is a crapshoot put together by the teams and their entire seasons. Two years ago in baseball the last team in, Ole Miss ends up winning it all at the CWS.
 
But hey I'm excited to have this conversation the next four months...take an eight month delay...and then have it again for the next four months on repeat as long as the NCAA keeps using the NET - which will be for a long time.

TLDR; Just please add more sections than "quads" so there isn't such a steep arbitrary cutoff between groups
 
Wakeforest22890 said:
Again, the NET is by far the best system that the NCAA uses (with perhaps the Pairwise for hockey in the discussion where there is no committee at all - they just take the top X teams in the ratings after tournament winners). Is there room for tweaks? Of course. Is there much serious discourse in the analytics world about the KP/Torvik/Vegas models not doing a good job of evaluating/rating the top teams over the course of the season? No. Otherwise people would be making an incredible amount of money betting college basketball.
Click to expand...
NCAA tennis also uses a metrics system to choose the participants, with no committee involved. But the model is published and can be checked by the public.
 
Wakeforest22890 said:
But hey I'm excited to have this conversation the next four months...take an eight month delay...and then have it again for the next four months on repeat as long as the NCAA keeps using the NET - which will be for a long time.

TLDR; Just please add more sections than "quads" so there isn't such a steep arbitrary cutoff between groups
Click to expand...
What's the argument for using quads, eighths, or any double-dipping at all?
 
Wakeforest22890 said:
They're not tasked with selecting the best 36 teams, they're picking the top 36 teams based on resumes and how they played. If they were just picking the best 36 teams, you'd obviously just use the metrics.
Click to expand...
The NCAA website explicitly states they are choosing the 36 best teams. "The committee selects the 36 best teams not otherwise automatic qualifiers for their conference to fill the at-large berths." https://www.ncaa.com/news/basketbal...i-mens-teams-picked-march-madness-each-season
 
I mean I'm right there with you - don't even use the Quads at all - just use the metrics and fire up the at-large teams from there. I'd imagine they want to use the Quad system to identify "quality wins"
 
Rafi said:
The NCAA website explicitly states they are choosing the 36 best teams. "The committee selects the 36 best teams not otherwise automatic qualifiers for their conference to fill the at-large berths." https://www.ncaa.com/news/basketbal...i-mens-teams-picked-march-madness-each-season
Click to expand...
Oh okay well that settles it! Again, if they were doing this then the 15-15 team every year that plays a top 3 schedule, finishes top 25-30 in all metrics, and goes like 8-10 against Q1 teams would get in. They don't, not because they're not one of the best 25-30 teams, but because they didn't put together one of the top 25-30 resumes.
 
Wakeforest22890 said:
I mean I'm right there with you - don't even use the Quads at all - just use the metrics and fire up the at-large teams from there. I'd imagine they want to use the Quad system to identify "quality wins"
Click to expand...
OK, now we're getting somewhere. Now, what if the chosen metric (the NET in this situation) continually overestimates a conference? Should that metric still be used? Should the model at least be made public? And what if a team is clearly different than they were in November, should we use a metric that does not differentiate a win in March vs November?
 
Rafi said:
OK, now we're getting somewhere. Now, what if the chosen metric (the NET in this situation) continually overestimates a conference? Should that metric still be used? Should the model at least be made public? And what if a team is clearly different than they were in November, should we use a metric that does not differentiate a win in March vs November?
Click to expand...
Overestimates relative to what? A sample size 7 games lol?

Let's see do I trust the models (like NET which is pretty spot on to KP and Torvik by the end of the year, albeit with the weird quality win boost added in) which are dealing with a sample size of thousands of games over the course of the season, or do I trust sample size 5 from the first two days of a single-elimination basketball tournament?

Or even going further...are these models biased over tens of thousands of games over the last 5 years where the Big 10 has been rated as a top 3 conference compared to the Big 10 faring poorly in the now...what....40-50 total games over those five years in the NCAAT?
 
Maybe it's a grand conspiracy in favor of the Big 10 and Mountain West Conference at the exclusion to the ACC. That's far more rational. Sure.
 
You must log in or register to reply here.
Back
Top