NCAA NET 23-24

[13th&Jackson]

After the Morehead St game, IU remained unchanged at 110. MS moved up two to 122. 

[rogue3542]

21 hours ago, Kdug said:

The committee doesn’t look at the version of NET on 12/19. They use the version at the end of the year after 30+ games have been played. Any metric is going to be incredibly noisy with only 10 or so games played by each team, a lot of them being against bad competition. The whole reason all other advanced metrics have a preseason component that’s still baked in is to help reduce that noise and make the metric more accurate. The NET doesn’t do that because it’s not used until the end of the season anyway.

The issue for me is this:

Just a couple years ago, blowing out a ~300 ranked team by 40 did nothing to help you in this metric; in fact, if you weren't playing your cupcakes in the 100-200 range, it significantly hurt your sos, and at the end of the season, it hurt your NET ranking along with the perception that you purposely played a significantly easier schedule than teams that beat better mid-majors by 10.

Now, we have a fair number of those better mid-majors scheduled to play that game, and it turns out we should have been scheduling the 200-300 teams this season so we could mercilessly beat them into the ground to artificially inflate our NET ranking.

Even by the end of the year, whatever changes they've made to it for this season are proving to render the NET nearly meaningless.

[FKIM01]

As of today, IU has zero business in the tournament.  Not sure when the lightbulb comes on for some of these guys, but you can't play down to the level of your competition and they consistently do that.  They almost got burned by Morehead State at home, for Christ's sake.

I wish I didn't care at all, but with each crappy performance, I care a little less.

[BruceDouglas]

2 hours ago, rogue3542 said:

The issue for me is this:

Just a couple years ago, blowing out a ~300 ranked team by 40 did nothing to help you in this metric; in fact, if you weren't playing your cupcakes in the 100-200 range, it significantly hurt your sos, and at the end of the season, it hurt your NET ranking along with the perception that you purposely played a significantly easier schedule than teams that beat better mid-majors by 10.

Now, we have a fair number of those better mid-majors scheduled to play that game, and it turns out we should have been scheduling the 200-300 teams this season so we could mercilessly beat them into the ground to artificially inflate our NET ranking.

Even by the end of the year, whatever changes they've made to it for this season are proving to render the NET nearly meaningless.

This is exactly the opposite of what happened. In 2020, the NCAA stopped using margin of victory:

https://www.cbssports.com/college-basketball/news/ncaa-announces-changes-to-simplify-formula-for-college-basketballs-net-ratings/

Winning percentage, adjusted winning percentage and scoring margin will no longer be components in the NCAA Evaluation Tool (NET) used to judge Division I basketball teams, the NCAA announced Monday. The changes will "increase accuracy and simplify" the NET system, according to the NCAA announcement.

The NET will now use just two factors in its evaluation: team value index (TVI) and adjusted net efficiency rating. Team value index is "a result-based feature that rewards teams for beating quality opponents, particularly away from home," while adjusted net efficiency rating accounts for "strength of opponent and location across all games played."

Now the "net efficiency rating" will most certainly be impacted by blowouts because the analytics around this suggest there is predictive value even in that "garbage time" result.

The stuff about playing teams in the 100-200 range sounds like the way programs used to "game" the old RPI formula and it did work. But fortunately RPI is no longer used.

[Kdug]

2 hours ago, rogue3542 said:

The issue for me is this:

Just a couple years ago, blowing out a ~300 ranked team by 40 did nothing to help you in this metric; in fact, if you weren't playing your cupcakes in the 100-200 range, it significantly hurt your sos, and at the end of the season, it hurt your NET ranking along with the perception that you purposely played a significantly easier schedule than teams that beat better mid-majors by 10.

Now, we have a fair number of those better mid-majors scheduled to play that game, and it turns out we should have been scheduling the 200-300 teams this season so we could mercilessly beat them into the ground to artificially inflate our NET ranking.

Even by the end of the year, whatever changes they've made to it for this season are proving to render the NET nearly meaningless.

We tried scheduling a 300+ team in Army, and barely snuck by them, which is one of the games holding our NET ranking down. It’s not as easy to get 30+ point wins as a lot of people make it seem. Also, getting a 10-15 point win vs 100-200 ranked teams is the same efficiency wise as 30+ point wins against 300+ ranked teams. We haven’t been able to do either, and have multiple blow out losses. The reason we’re ranked so low in the NET is because we’ve been a bad team outside of a few games.

[KoB2011]

5 hours ago, Kdug said:

we’ve been a bad team outside of a few games.

I appreciate your generous holiday spirit 

[rogue3542]

7 hours ago, BruceDouglas said:

This is exactly the opposite of what happened. In 2020, the NCAA stopped using margin of victory:

https://www.cbssports.com/college-basketball/news/ncaa-announces-changes-to-simplify-formula-for-college-basketballs-net-ratings/

Winning percentage, adjusted winning percentage and scoring margin will no longer be components in the NCAA Evaluation Tool (NET) used to judge Division I basketball teams, the NCAA announced Monday. The changes will "increase accuracy and simplify" the NET system, according to the NCAA announcement.

The NET will now use just two factors in its evaluation: team value index (TVI) and adjusted net efficiency rating. Team value index is "a result-based feature that rewards teams for beating quality opponents, particularly away from home," while adjusted net efficiency rating accounts for "strength of opponent and location across all games played."

Now the "net efficiency rating" will most certainly be impacted by blowouts because the analytics around this suggest there is predictive value even in that "garbage time" result.

The stuff about playing teams in the 100-200 range sounds like the way programs used to "game" the old RPI formula and it did work. But fortunately RPI is no longer used.

My point I guess is that even though they removed the formal scoring margin factor, there’s even more incentive to beat the other team by as much as possible, because in this current iteration of the metric that seems to correlate to the biggest boost in net rankings. Hence, schedule a bunch of cupcakes and beat them by 30, and you’re sitting pretty.

I’m not necessarily saying IU is a lot better than their net (they are at least somewhat better, though), but there’s teams in the top 50 that have no business being there, even ~12 games into the season.

[KoB2011]

2 hours ago, rogue3542 said:

My point I guess is that even though they removed the formal scoring margin factor, there’s even more incentive to beat the other team by as much as possible, because in this current iteration of the metric that seems to correlate to the biggest boost in net rankings. Hence, schedule a bunch of cupcakes and beat them by 30, and you’re sitting pretty.

I’m not necessarily saying IU is a lot better than their net (they are at least somewhat better, though), but there’s teams in the top 50 that have no business being there, even ~12 games into the season.

Have we been demonstrably better than our NET at this point?

I think before we complain other teams are “gaming” the system by blowing out inferior competition we should ask why we are getting blown out of games.

[NCHoosier32]

10 hours ago, rogue3542 said:

My point I guess is that even though they removed the formal scoring margin factor, there’s even more incentive to beat the other team by as much as possible, because in this current iteration of the metric that seems to correlate to the biggest boost in net rankings. Hence, schedule a bunch of cupcakes and beat them by 30, and you’re sitting pretty.

I’m not necessarily saying IU is a lot better than their net (they are at least somewhat better, though), but there’s teams in the top 50 that have no business being there, even ~12 games into the season.

isn't it kind of mind blowing that we can't beat anyone easily?  i mean spurts vs Auburn and UConn along with the Kansas game show we can be good right?  

[rogue3542]

27 minutes ago, NCHoosier32 said:

isn't it kind of mind blowing that we can't beat anyone easily?  i mean spurts vs Auburn and UConn along with the Kansas game show we can be good right?  

Yeah, they've been really bad about playing to perceived level of competition.  It's ridiculous that they can get up on Kansas by double digits for nearly the whole game, but also get down double digits to Morehead State for most of the game.

[tdhoosier]

13 minutes ago, NCHoosier32 said:

isn't it kind of mind blowing that we can't beat anyone easily?  i mean spurts vs Auburn and UConn along with the Kansas game show we can be good right?  

Not necessarily saying we are bad, but even bad teams have spurts. Look at Army. 😎

We are who we are. There’s a reason why we can’t put away bad teams and get blown out by good teams. No team in the top 25 is flawless, but for the most part, their consistencies in effort and execution outweigh the inconsistencies, which makes them good teams. Right now, this team is far from being consistent in these areas. We are average at best, with the talent to potentially turn it around. But polls don’t use ‘potential’ as a metric the way fans would. 

 

[KoB2011]

23 minutes ago, rogue3542 said:

Yeah, they've been really bad about playing to perceived level of competition.  It's ridiculous that they can get up on Kansas by double digits for nearly the whole game, but also get down double digits to Morehead State for most of the game.

I actually think it’s a pretty big stretch to say we play to the level of our competition based on the first 11 games. We have pretty consistently across 11 games shown we are far removed from the echelon of good teams.

We haven’t blown a single team out all year. We have been boat raced in two of our three losses. The loss that was close involved us being outscored by 17 points in the last 15 minutes of the game, at home. 

[BruceDouglas]

12 hours ago, rogue3542 said:

My point I guess is that even though they removed the formal scoring margin factor, there’s even more incentive to beat the other team by as much as possible, because in this current iteration of the metric that seems to correlate to the biggest boost in net rankings. Hence, schedule a bunch of cupcakes and beat them by 30, and you’re sitting pretty.

I’m not necessarily saying IU is a lot better than their net (they are at least somewhat better, though), but there’s teams in the top 50 that have no business being there, even ~12 games into the season.

Sure, because better teams win by more points? That's pretty common when looking at analytics for sports. And there's no "current iteration", it's the same formula they've used for the past several seasons since making those changes announced in 2020.

I've seen a lot of this on social media this season, i.e. "you can game the NET rankings by scheduling bad teams and beating them by a lot". I mean...OK, sure. Go ahead and do that. It's not quite as easy as people think. And it also misses the point of how NET is used, it's not really so much your ranking but the ranking (quadrants) of the team you beat. So yeah, beating up a bunch of bad Q4 teams might boost your NET ranking,  but it does nothing for your resume. And likely kills your non-con SOS, which the committee frowns upon.

[KoB2011]

34 minutes ago, BruceDouglas said:

Sure, because better teams win by more points? That's pretty common when looking at analytics for sports. And there's no "current iteration", it's the same formula they've used for the past several seasons since making those changes announced in 2020.

I've seen a lot of this on social media this season, i.e. "you can game the NET rankings by scheduling bad teams and beating them by a lot". I mean...OK, sure. Go ahead and do that. It's not quite as easy as people think. And it also misses the point of how NET is used, it's not really so much your ranking but the ranking (quadrants) of the team you beat. So yeah, beating up a bunch of bad Q4 teams might boost your NET ranking,  but it does nothing for your resume. And likely kills your non-con SOS, which the committee frowns upon.

Yup, and team sheets (what the committee is looking at when they compare teams) also shows the scores of the games in each quadrant.

They're going to see a team gamed their NET by blowing out Q3 and Q4 teams. In our case, they're going to see we got blown out in half of our Q1 games.

Metrics typically backup what people see with their own eyes. There are exceptions to that, but IU isn't one of them at this point in time. 

[13th&Jackson]

4 hours ago, rogue3542 said:

Yeah, they've been really bad about playing to perceived level of competition.  It's ridiculous that they can get up on Kansas by double digits for nearly the whole game, but also get down double digits to Morehead State for most of the game.

I'm not sure it's playing down to competition as much as it is matchups. Most of these mid-majors don't have size to match up, so they shoot a lot of threes. We don't defend the three well, so we get outscored by 20+ and the game is closer that it should be. Morehead St had 65 shot attempts, 34 from three. Even though they only made 10/34, they were +21

[13th&Jackson]

Without looking, where are the following teams ranked in the NET?

Team 1:  overall record 8-5, 0-1 road, 2-3 neutral, Quad 1 0-5 Quad 2 3-0

Team 2: overall record 10-3, 1-0 road, 1-2 neutral, Quad 1 0-3, Quad 2 1-0

[KoB2011]

31 minutes ago, 13th&Jackson said:

Without looking, where are the following teams ranked in the NET?

Team 1:  overall record 8-5, 0-1 road, 2-3 neutral, Quad 1 0-5 Quad 2 3-0

Team 2: overall record 10-3, 1-0 road, 1-2 neutral, Quad 1 0-3, Quad 2 1-0

5 and 99. And no I didn’t look. 

[Kdug]

1 hour ago, 13th&Jackson said:

Without looking, where are the following teams ranked in the NET?

Team 1:  overall record 8-5, 0-1 road, 2-3 neutral, Quad 1 0-5 Quad 2 3-0

Team 2: overall record 10-3, 1-0 road, 1-2 neutral, Quad 1 0-3, Quad 2 1-0

Team 1 has the best offense in the country. I think they’re a national title contender if they can just make marginal improvements on defense.

[kyhoosier29]

8 hours ago, Kdug said:

Team 1 has the best offense in the country. I think they’re a national title contender if they can just make marginal improvements on defense.

Alabama? 

[KoB2011]

1 hour ago, kyhoosier29 said:

Alabama? 

Yessir

[Stlboiler23]

11 hours ago, Kdug said:

Team 1 has the best offense in the country. I think they’re a national title contender if they can just make marginal improvements on defense.

I’d say more than marginal. They’re at 73 right now. They’re not that far off from what Purdue was (2nd on offense and 93rd on defense) back in ‘22. 

Edited January 2 by Stlboiler23

[13th&Jackson]

On 1/1/2024 at 1:46 PM, Kdug said:

Team 1 has the best offense in the country. I think they’re a national title contender if they can just make marginal improvements on defense.

But this demonstrates how much offensive efficiency really drives the NET when the NET #5 is 0-5 in Quad 1 and 0-1 in it's only true road game. 

[Kdug]

2 hours ago, 13th&Jackson said:

But this demonstrates how much offensive efficiency really drives the NET when the NET #5 is 0-5 in Quad 1 and 0-1 in its only true road game. 

The NET is an efficiency metric, so yeah adjusted offensive and defensive efficiency are the only things that drive it.

You’re referencing wins and losses, which are better reflected in resume based metrics. In KPI Alabama is 43rd and SOR they’re 58th

[KoB2011]

9 minutes ago, Kdug said:

The NET is an efficiency metric, so yeah adjusted offensive and defensive efficiency are the only things that drive it.

You’re referencing wins and losses, which are better reflected in resume based metrics. In KPI Alabama is 43rd and SOR they’re 58th

Right. We look a lot at records within Quads from NET, but the records don’t actually impact your own team’s ranking.

[BruceDouglas]

Underwood was asked about not subbing earlier in his post game presser and if it was impacted by NET's use of efficiency margins. He was pretty direct in acknowledging it's something he's very aware of.