NCAA Updating NET Formula

[rjsuperfly66]

When I see a team like Minnesota with a Net of 42 with a record they have, I call BS.


Net Record Road Neutral Home
42 42 Minnesota Big Ten 15-16 2-10 2-1 11-5

I think the point is that no one thought the NET was a system without flaw, they just felt it had less imperfections and was more comprehensive than the RPI. NC State was an obviously issue two years ago and the committee made the fantastic decision to throw them in the NIT. The strength of the Big Ten led to some questionable placements this past season -- Minnesota and Purdue among them, as teams with .500 records but good efficiency metrics because they didn't frequently get blown out.

We can only hope that the new updates are a fix for the better.

There really were no updates or tweaks - they are no longer using the old Men's NET. The Men are adopting the NET that has been developed and approved by the NCAA Women's Development Group for 2020-2021.

To me, the women committee took the NET in it's original form, broke down what they didn't like, and fixed what they felt needed to be fixed. The TVI sounds to be the same as it was, and efficiency is still a metric, except it went to adjusted. I think we are parsing words, whether a tweak, update, complete rebuild, it's still a system in place that is supposed to analyze who, where, and how you play, and will hopefully do a better job than it did.

[ramster]

I think the point is that no one thought the NET was a system without flaw, they just felt it had less imperfections and was more comprehensive than the RPI. NC State was an obviously issue two years ago and the committee made the fantastic decision to throw them in the NIT. The strength of the Big Ten led to some questionable placements this past season -- Minnesota and Purdue among them, as teams with .500 records but good efficiency metrics because they didn't frequently get blown out.

We can only hope that the new updates are a fix for the better.

There really were no updates or tweaks - they are no longer using the old Men's NET. The Men are adopting the NET that has been developed and approved by the NCAA Women's Development Group for 2020-2021.

To me, the women committee took the NET in it's original form, broke down what they didn't like, and fixed what they felt needed to be fixed. The TVI sounds to be the same as it was, and efficiency is still a metric, except it went to adjusted. I think we are parsing words, whether a tweak, update, complete rebuild, it's still a system in place that is supposed to analyze who, where, and how you play, and will hopefully do a better job than it did.

To me, as I read about the Women's NET Team's Work on developing the NET they only reviewed the Men's NET and looked at the strengths and weaknesses. They developed their OWN Algorithm looking back on 10 years of data, 10 years of RPI Rankings and other data. They fed data into the Algorithm that was developed working with the Google Cloud Team. This is a new Algorithm, vey little is the same except for the name NET. It appears much more thought and analysis went into the development of this Women's NET Algorithm.

The Men just adopted the Women's NET exactly as it is.

One of the added features was recognizing away games in the formula. As I always suspected, the Men's NET did not recognize the Away Road Game Disadvantage. This is fixed in the Women's NET.

Key Question in the Q & A's provided by the Women's NET Development Team:

How was the current women’s basketball NET algorithm determined?

The Division I Women’s Basketball Committee decided to move from the RPI to the NET for the 2020-21 season after a lengthy evaluation process.

Included was a comprehensive assessment of the strengths and weaknesses of the men’s basketball NET that has been used by the Division I Men’s Basketball Committee as a sorting tool since the 2018-19 season.

Following an analysis of women’s basketball statistical data over a 10-year period by a team from Google Cloud Professional Services, the Division I Women’s Basketball Committee concluded that this algorithm was an optimal sorting tool and should be used beginning with the 2020-21 season.


https://www.ncaa.com/news/basketball-wo ... etball-net

[rjsuperfly66]

Yes of course, by dropping from 5 items to 2, they created their own algorithm, and they studied their data to make sure they liked the output. But of the two components, it does not appear that any tweaks were done to the TVI, which was created by the NCAA for the original NET. To keep that the same, I would have a hard time saying “very little is the same except for the name NET.” And they clearly liked the efficiency aspect, just knew that net efficiency really would not work for the women’s game and that’s one place they really deserve credit is the build-out of adjusted efficiency. But to me to have the top two components of the old NET be TVI and efficiency, and the two components of the women’s NET be TVI and efficiency (albeit different efficiency), that’s why I feel they broke down the walls and addressed them in a way to make it better. Had they rebuilt TVI, I’d feel differently.

[theblueram]

I just hope the NCAA starts using Net as a tool to punish teams. You get caught paying for players, a NET 50 reduction automatically going forward for a time.. I could buy into that. 1, no one will want to play them and 2 they won't make the NCAAT.

[Rhode_Island_Red]

exactly. Statisticians can always design an algorithm with any simple goal in mind.

In this case the goal is: How can we keep the non P5 teams out of contention.

Remember: Obvious nepotism hire Dan Gavitt has one job -- pacify The Cartel.

[URI2006_Andy]

The NCAA’s growing emphasis on efficiency ratings puts a lot of pressure on kids to be “on” every possession of every game. It’s exhausting for a fan, never mind a college-aged kid.

Your girlfriend breaks up with you, too bad. You still need to beat Fordham by 20, not by 6.

You got a paper due tomorrow, deal with it. You still need to keep your game against Kentucky close.

You want to be creative and innovative on the court when you got a comfortable lead. That’s frowned upon.

These efficiency metrics are turning basketball players into robots. Once players start to realize transferring is not the answer, you will see more talented players leave for leagues where wins and losses matter.

[Rhody15]

Let’s face it, our fan base will complain until the year 3000 about the way the NCAA tournament is selected.

And nothing is ever going to be in the mid majors favor.

It just is what it is.

[ramster]

And never win

The cinderella stories of non-P5's making the Sweet 16, Elite 8, Final 4 are wonderful stories over the years. But with reducing Mid Majors in the Tournament and now allowing P5s to recruit from Mid Majors and Low Majors it's not a good trend line for non P5 Schools

The fewer Mid Majors in the NCAA Tournament and the more P5's decreases fan interest in the Tournament. It is actually worse for the NCAA Tournament in the long run.

[rhodyfan3000]

Let’s face it, our fan base will complain until the year 3000 about the way the NCAA tournament is selected.

With good reason. Let's face it, anyone coming on here telling us how the NCAA just tweaked the process by which they decide how they select teams for the tournament is essentially pissing on us and calling it rain. Although, I don't blame superfly for that, he's just the messenger.

If the A-10 has three teams with 22 wins or more, and even own the tie breakers with other teams on the bubble, like we did with PC this past year, then the A-10 should get all three teams in, and not have to worry about only getting two, because the Big 10 team with the 15-17 record without the Big 10 autobid needs to get in instead.

I didn't mention this during the actual season because I don't believe in leaning on crutches and I think URI should have been constantly fighting to be in the top 3 in the A-10. But the fact remains, any process that weeds out that "one more" team from a conference like ours, isn't good for college basketball in the long run.

Edit:

And if we didn't make it into the top 3, then it should have been whomever else did, ie Richmond, St Louis, etc

[rambone 78]

The deck will always be stacked against schools at URI's level, no matter what metrics are used.

The question is, is the school willing to spend what it will take to at least have a chance to make the NCAA tournament somewhat consistently?

It's going to cost more and more going forward.

We are currently falling behind the best programs at our level. If that doesn't change...well we know the answer to that.

With revenue taking a huge hit these days, it's going to get much tougher to even maintain current budget levels, much less increase them. Donors will have to step up even more, to have any chance.

And compared to programs that have many more and bigger donors...that's another issue.

[rjsuperfly66]

If the A-10 has three teams with 22 wins or more, and even own the tie breakers with other teams on the bubble, like we did with PC this past year, then the A-10 should get all three teams in, and not have to worry about only getting two, because the Big 10 team with the 15-17 record without the Big 10 autobid needs to get in instead.

Not for anything, but what tiebreaker did you own over PC? Yes, you owned the head-to-head tiebreaker, and if PC and URI were neck and neck, that would be an acceptable time to say that we have two teams that we can't decide between, but one beat the other so we'll use that. But by the end of the season, PC was not on the bubble, more in the 8-9 seed range, so I don't see why that automatically makes URI deserving of a bid, or why using a hollow win total without analysis of strength matter in determining who should make the tournament? If a team goes 31-0, but beats the worst 31 teams in D1 basketball, does that suggest to us that they are a powerful team, or just beat up on bad competition? That is why college basketball needs a way to assess how impressive a team was in their 31 games of performance, not just looking at who they played and if they won or not. If the system does not adequately do that, I will join your on your parade. But I'm not going to participate in celebrating teams who play crap opponents and often play down to this opponents, while not being able to sufficiently beat good opponents.

Your resume is what you do over the entirety of the season ... You have to beat good/great teams over 31 games. It doesn't matter what you do one night, it doesn't matter what you do two nights, it really doesn't matter if you "owned" a team on any given night, the committee has the task of putting the best teams in the tournament, not by granting conferences an arbitrary number of bids before the season starts but because when they play the game of basketball on-the-court, they put the ball in the basket more often than not against their opponents, while doing it against a high-level of competition. That is why I enjoy college basketball. I love the underdog story, but I love it more when they do so because they deserve it. The underdog story wouldn't be cute if we just let any Joe Schmo in the tournament, and then they lost by 25 points but "That conference deserved three bids." No, did the team deserve it's bid? Did they beat good teams? Did they perform over 31+ games? If they did, they should be in. If they didn't, sorry but not sorry.

I'm not defending the NCAA. I think there have been some omissions in the last 5 years I didn't personally like (maybe 2 or 3 off the top of my head). But I also don't feel like there is this great conspiracy list of teams who got totally shafted who deserved a bid and didn't get one. For the most part, the teams who should have gotten a bid got one.

[rhodyruckus]

The deck will always be stacked against schools at URI's level, no matter what metrics are used.

The question is, is the school willing to spend what it will take to at least have a chance to make the NCAA tournament somewhat consistently?

Based on the apparent lack of increased spending in light of our increased revenue from NCAA credits, unfortunately I think we know the answer to that. And if the program is spending significantly more than previous I think it should be publicized a little more.