RIP to the RPI

[OrangeXtreme]

 

[Eric15]

"The dates of games and their order are not included in order to give equal importance to both early and late-season games."

I'm very happy about this. A game played in November should count the exact same as a game played in early March.

 

[Toga]

Division I Men’s Basketball Committee adopts new ranking system

 

[dasher]

"The dates of games and their order are not included in order to give equal importance to both early and late-season games."

I'm very happy about this. A game played in November should count the exact same as a game played in early March.

I don't agree with that. A game in March is a much more reliable indicator of who you are as a team.

 

[Eric15]

A game in March is a much more reliable indicator of who you are as a team.

I agree with you here, but I still don't care. I think the season-long body of work should be equally weighted regardless of whether a team is trending up or down.

 

[Fat tire 13]

I agree with you here, but I still don't care. I think the season-long body of work should be equally weighted regardless of whether a team is trending up or down.

Agree too. In pro sports a team that gets hot near the end will be a playoff threat, but you still need to win enough games to get into the playoffs in the first place.

 

[Cusefan0307]

Long overdue.

 

[Col. Bleep]

"The dates of games and their order are not included in order to give equal importance to both early and late-season games."

I'm very happy about this. A game played in November should count the exact same as a game played in early March.

I disagree on that.

Teams often jell later in the season.
Teams that do so and get hot are going to be much bigger threats in the tournament.
Late season wins generally should be given more weight so that the tournament has the best current teams in it.

 

[Cusefan0307]

Schedules are unbalanced. Shouldn’t matter when games are played.

Also you guys are wrong. End of season performance never correlates to how we have done in the tournament.

We won the BET in 05 and 06 and lost in the first round.

Duke won the ACC in 2017 and lost early.

We stunk in 2013 and 2016 and made the FF.

You guys need to show me some metrics on end of season performance leading to good tournament performance because it seems like a myth to me.

 

[orange79]

I wonder if they are going to release the actual formula of the NET...

 

[Waltdods]

I don't agree with that. A game in March is a much more reliable indicator of who you are as a team.

People say this but I don't think there's any particular reason to think it's true.

Syracuse in 2013 was awful, terrible, could almost literally not score at the end of the season. It ended in the Final Four, which wouldn't have been much of a surprise if you looked at the whole body of work but would have been a huge surprise if you confined your view to March.

To take an unfair opposite example, Syracuse in 2006 was awesome in pre-NCAA March, ended up with an unearned 5-seed as a result, and then was out in one game.

My guess is that if someone ran the numbers, there would be little, and maybe zero, relationship between when a team wins in the regular season and in the tournament.

 

[Toga]

Has Lunardi's Mom called down to the basement yet?

 

[dasher]

People say this but I don't think there's any particular reason to think it's true.

Syracuse in 2013 was awful, terrible, could almost literally not score at the end of the season. It ended in the Final Four, which wouldn't have been much of a surprise if you looked at the whole body of work but would have been a huge surprise if you confined your view to March.

To take an unfair opposite example, Syracuse in 2006 was awesome in pre-NCAA March, ended up with an unearned 5-seed as a result, and then was out in one game.

My guess is that if someone ran the numbers, there would be little, and maybe zero, relationship between when a team wins in the regular season and in the tournament.

I think it's pretty obvious that a team that wins 8 out of 10 to end a season is more than likely playing better than a team that lost 8 out of 10 to finish the season. And that makes them a more deserving tournament team. Are there exceptions? Of course.

 

[RandygoCuse]

I think it's pretty obvious that a team that wins 8 out of 10 to end a season is more than likely playing better than a team that lost 8 out of 10 to finish the season. And that makes them a more deserving tournament team. Are there exceptions? Of course.

I think these are slightly different ideas:

1) A team that wins more toward the end of the season is more of a threat in the tournament - I would need to see data before I accepted this
2) A team that wins more toward the end of the season deserves a bid more than one that wins more toward the beginning - Would this be because at the beginning of the season teams are still figuring out what works best? Or just that you like the idea of a team finishing strong?

 

[orangemass]

We finally figured out how to game the RPI and now they change the metric... Dang !!!

 

[AustinCuse]

Schedules are unbalanced. Shouldn’t matter when games are played.

If the intention of the new Metric is to predict the performance of the teams in the tourney, it makes sense to weight late games higher. This is especially so where injuries to key players at the end of the season change expectations significantly.

 

[Eric15]

In my opinion, if you conducted an exhaustive and comprehensive analysis of every single team's season in the history of college basketball, it would almost certainly show that, on average, teams that have better records late in the season would have more tournament success than teams who have the majority of their success early in the season.

But obviously since we're talking about a sample size in the tens of thousands, there will be thousands of exceptions to this (e.g. 2006 SU, 2013 SU, etc.).

 

[Orangezoo]

If the intention of the new Metric is to predict the performance of the teams in the tourney, it makes sense to weight late games higher. This is especially so where injuries to key players at the end of the season change expectations significantly.

I think a point is being missed here. It's not like the selection committee is just abandoning the subjective components of picking the field. This new metric is one to provide a more comprehensive tool to base quadrant analysis on and then allow other factors like injuries and the eye test to be things for discussion when comparing two very similar teams mathematically.

 

[Cusefan0307]

If the intention of the new Metric is to predict the performance of the teams in the tourney, it makes sense to weight late games higher. This is especially so where injuries to key players at the end of the season change expectations significantly.

Even so if we weigh the later games more it's beyond stupid. It would be the only sport where games at the end mean more than games at the beginning. Now that we have settled that I am going to start an analysis.

 

[Orangezoo]

The other issue with March records is that you start looking outside the power 5 and everything gets skewed. You would have to specifically score those games in terms of SOS as well if you are going to do that.

 

[OttoMets]

I agree with you here, but I still don't care. I think the season-long body of work should be equally weighted regardless of whether a team is trending up or down.

Any bias at play here?

My gut agrees with you, but my brain can't rationalize this. And I'm guessing it's because SU's gotten boned with seeding so often in the past - 19-0 and 20-1 starts that earned us #4 seeds because we tend to fade in February really stick in my craw.

 

[dasher]

I think these are slightly different ideas:

1) A team that wins more toward the end of the season is more of a threat in the tournament - I would need to see data before I accepted this
2) A team that wins more toward the end of the season deserves a bid more than one that wins more toward the beginning - Would this be because at the beginning of the season teams are still figuring out what works best? Or just that you like the idea of a team finishing strong?

A team that is playing well at the end of the season would seem to be a better team and more deserving team than one that played well earlier in the season. My sr year at Syracuse, 1978, we were great at the start of the season. We had a vet team and we beat New Mexico at new Mexico and Mich St with Magic, both early in the season. At the end of the year, teams had caught up with us. Who are you really as a team? I think the team at the end of a season is who you are. I'm not discounting early seasons wins. If I had to decide between two teams and they were really close, I would take the team playing better at the end of a season over a team that wasn't.

 

[Cusefan0307]

A team that is playing well at the end of the season would seem to be a better team and more deserving team than one that played well earlier in the season. My sr year at Syracuse, 1978, we were great at the start of the season. We had a vet team and we beat New Mexico at new Mexico and Mich St with Magic, both early in the season. At the end of the year, teams had caught up with us. Who are you really as a team? I think the team at the end of a season is who you are. I'm not discounting early seasons wins. If I had to decide between two teams and they were really close, I would take the team playing better at the end of a season over a team that wasn't.

Maybe you are right, maybe you aren't.

Maybe my mind is clouded with about 10 Pitt and Georgetown teams that went on huge winning streaks in February and March only to lose the first weekend.

 

[Orangezoo]

A team that is playing well at the end of the season would seem to be a better team and more deserving team than one that played well earlier in the season. My sr year at Syracuse, 1978, we were great at the start of the season. We had a vet team and we beat New Mexico at new Mexico and Mich St with Magic, both early in the season. At the end of the year, teams had caught up with us. Who are you really as a team? I think the team at the end of a season is who you are. I'm not discounting early seasons wins. If I had to decide between two teams and they were really close, I would take the team playing better at the end of a season over a team that wasn't.

I think generally speaking this holds up but it's not something you can really measure fairly across the whole landscape without data being skewed. I think it's still very fair as something to look at if it's a decision down to two teams for a bid or in seeding for sure. But as a variable of NET I just don't think it works.

 

[SoBeCuse]

Any bias at play here?

My gut agrees with you, but my brain can't rationalize this. And I'm guessing it's because SU's gotten boned with seeding so often in the past - 19-0 and 20-1 starts that earned us #4 seeds because we tend to fade in February really stick in my craw.

I’m not really remembering getting boned with seeds. In those starts I think we loaded up on cupcakes then started a tailspin against good teams. Short BET as well? Boeheim has changed his tune with scheduling the past few years for the better.

This thread is really difficult to evaluate. UVA and Sparty come to mind. I’m not sure how much better we were in March. I mean TCU shot 39% and 17% from 3. They shouldn’t have even been in the game. Our defense was good but it’s a ‘gimmick’ D unfamiliar to many teams in the Tourney especially on short notice. Our offense was what it was...terrible. We were very lucky to advance. These games just should not have been that close when those teams are turning it over like they were, shooting in the 30s from the field, and in the teens from 3. I mean Sparty shot 26%(!!) for the GAME and had 14 turnovers yet it was a one possession game, lol. We weren’t a good team.