Serious question about the NCAA's "Net Rankings"

[tykurez]

We’ll see if the luck lasts because so far it has. As much as we like to chide that program they keep figuring out a way to win. If they luck their way to the E8 it’s still a team that got to the E8.

 

[Tenspro2002]

But they made all those 10 shots and are not at .500. They have one of the best records in the country and lead the Big East. At some point in the season, reality has to trump predictive.

The "reality" is they've won a ton of close games.

Is it because they're just so plucky and clutch? Or have they gotten lucky? Probably some of both.

The big question that nobody is addressing head on is:

Should NCAA Tournament seeding be based on your resume (W/L and SOS) or should it be based on how good the Committee thinks a team really is?

 

[auror]

Providence fans want you to believe those wins were, in fact, wins. That is a sound position.

We are not trying to seed alternate-universe Providence with a .500 record.

Deprioritizing single-game outcomes is a bad idea.

The final minutes of a close regular season game are exciting based on the premise that the outcome matters.

If the NCAA decides that the outcome doesn’t matter - that a 1-point loss is really the same thing as a 1-point win - then those final minutes are meaningless. The regular season just becomes one long, continuous scoring exhibition for each team. However efficient the computers may find that, it’s boring as heck. There is a good reason every major sports league seeds teams by wins rather than “season-long scoring differential.”

Remember that the NET ranking is not what the committee uses to select or seed the teams. It is not the NCAA sanctioning anything.

It's just one metric.

They use it to sort the wins on the sheet. When sorting the wins to determine how strong a win is, would you want to sort the opponent by our best guess of how good a team is (so how tough the win was) or how good a season the opponent has had? The former makes a lot more sense.

When actually selecting the field, you'd want to weigh a combination of the two. Reward those that are the most deserving while also selecting the teams most likely to strongly compete for the championship. This is why the committee considers both predictive and resume metrics when actually selecting teams. It's why despite being in the 40s in KenPom, Providence was #15 in the committee top 16 reveal.

College basketball is different than most sports (especially pro sports) because the schedules are vastly different. In MLB, a win is a win because all the teams are professionals, and there are only 30, and they are roughly on par with each other in talent and schedule. In college basketball, a win over the #350 team at home is much different than a win against #6 on the road. Some teams only play teams worse than #200 in conference play, and some only play teams better than #100. We need better tools than simply just number of wins.

 

[HuskyHawk]

The "reality" is they've won a ton of close games.

Is it because they're just so plucky and clutch? Or have they gotten lucky? Probably some of both.

The big question that nobody is addressing head on is:

Should NCAA Tournament seeding be based on your resume (W/L and SOS) or should it be based on how good the Committee thinks a team really is?

It needs to be both. Take UConn women for example. Seeding on straight results is unfair to whoever has to play them with Paige. NCAA has said they do look at how a team is playing and returning players, so "how good you are in March" matters. But they can't simply ignore W/L SOS where a team like PC looks like they'd perhaps be a 1-2 seed right now.

Honestly, if PC wins out and wins the BET they need to be and deserve a #1 seed. Even if the games are close. If they lose early on, then the Committee could decide more luck was at play.

 

[tzznandrew]

We’ll see if the luck lasts because so far it has. As much as we like to chide that program they keep figuring out a way to win. If they luck their way to the E8 it’s still a team that got to the E8.

You can't really luck your way to an E8. Honestly, if they make the second weekend you just have to tip your cap to them and throw out the "luck" talk. Outliers exist.

 

[VAHuskyFan]

Remember that the NET ranking is not what the committee uses to select or seed the teams. It is not the NCAA sanctioning anything.

It's just one metric.

They use it to sort the wins on the sheet. When sorting the wins to determine how strong a win is, would you want to sort the opponent by our best guess of how good a team is (so how tough the win was) or how good a season the opponent has had? The former makes a lot more sense.

When actually selecting the field, you'd want to weigh a combination of the two. Reward those that are the most deserving while also selecting the teams most likely to strongly compete for the championship. This is why the committee considers both predictive and resume metrics when actually selecting teams. It's why despite being in the 40s in KenPom, Providence was #15 in the committee top 16 reveal.

College basketball is different than most sports (especially pro sports) because the schedules are vastly different. In MLB, a win is a win because all the teams are professionals, and there are only 30, and they are roughly on par with each other in talent and schedule. In college basketball, a win over the #350 team at home is much different than a win against #6 on the road. Some teams only play teams worse than #200 in conference play, and some only play teams better than #100. We need better tools than simply just number of wins.

Agree with just about all of this. Obviously who you play and SOS matters. That part of the NET is fine. No one is suggesting a win against the #350 team at home should be treated the same as a win against #6 on the road.

The issue is how much you weight the fact of a win or loss vs. margin of victory (or loss). I don’t even have a problem including margin and efficiency stats in the calculus to be honest. You just need to value and reward binary outcomes over margins and other metrics.

My point is that when the primary NCAA-created metric has Houston as the #4 team in the nation despite being 0-3 in Quad 1 games — and Providence as the #28 team despite being 6-2 in Quad 1 games — then you have designed a shoddy metric that is clearly not valuing outcomes enough. (Both teams are 7-1 in Quad 2 games.)

Your point about schedule strength being different only highlights how off of the mark the NET rankings are. Providence has a better record against a tougher SOS. And yet NET says that warrants a 7 seed vs a 1 for Houston.

The reason the NET is spitting out such foolishness is that it is underweighting outcomes and overweighting margin of victory/loss. Again, the NCAA can keep margin as part of the equation, they just need to dial it way (way) back from wherever they have it set currently.

 

[navery12]

Agree with just about all of this. Obviously who you play and SOS matters. That part of the NET is fine. No one is suggesting a win against the #350 team at home should be treated the same as a win against #6 on the road.

The issue is how much you weight the fact of a win or loss vs. margin of victory (or loss). I don’t even have a problem including margin and efficiency stats in the calculus to be honest. You just need to value and reward binary outcomes over margins and other metrics.

My point is that when the primary NCAA-created metric has Houston as the #4 team in the nation despite being 0-3 in Quad 1 games — and Providence as the #28 team despite being 6-2 in Quad 1 games — then you have designed a shoddy metric that is clearly not valuing outcomes enough. (Both teams are 7-1 in Quad 2 games.)

Your point about schedule strength being different only highlights how off of the mark the NET rankings are. Providence has a better record against a tougher SOS. And yet NET says that warrants a 7 seed vs a 1 for Houston.

The reason the NET is spitting out such foolishness is that it is underweighting outcomes and overweighting margin of victory/loss. Again, the NCAA can keep margin as part of the equation, they just need to dial it way (way) back from wherever they have it set currently.

This part in bold is where you keep missing the point. You keep harping on the results and questioning how PC can be so low with their record. That's not the purpose of the NET rankings, it's not about rewarding teams for their results. It's not even about seeding, which is largely based on wins and losses and already has plenty of metrics that exist to judge.

It's a way for you and me to look at a team and have a rough starting point of is this a good win/bad loss. Just because the NET isn't giving out the results you want doesn't mean it's wrong

 

[Watchdog]

These are college men (and women). Blowing a team out is not poor sportsmanship IMO.

When I was growing up, I was on bad soccer teams. We lost every game one year. We made a choice to get better and work hard, and guess what happened? My HS team won the state championship back-to-back years, with a core group of the same kids who got annihilated when we were in 5th/6th grade. We ran right through some of the same towns that beat up on us growing up. I am for blow-outs. Sometimes in life you have to lose to learn how to win.

"Sometimes in life you have to lose to learn how to win."
What a dinosaur. You need to learn that no one should ever suffer defeat today. How do they get their self esteem back?

 

[XLCenterFan]

"Sometimes in life you have to lose to learn how to win."
What a dinosaur. You need to learn that no one should ever suffer defeat today. How do they get their self esteem back?

We learn from our losses, not our wins.

 

[XLCenterFan]

@husky429 do you have a problem with the winning margins of the past 2 women’s games?

 

[businesslawyer]

Providence fans want you to believe those wins were, in fact, wins. That is a sound position.

We are not trying to seed alternate-universe Providence with a .500 record.

Deprioritizing single-game outcomes is a bad idea.

The final minutes of a close regular season game are exciting based on the premise that the outcome matters.

If the NCAA decides that the outcome doesn’t matter - that a 1-point loss is really the same thing as a 1-point win - then those final minutes are meaningless. The regular season just becomes one long, continuous scoring exhibition for each team. However efficient the computers may find that, it’s boring as heck. There is a good reason every major sports league seeds teams by wins rather than “season-long scoring differential.”

Thank you.

Predictive models are good for just that -- predicting. They are not particularly meaningful for telling you who has accomplished more for the simple fact that you're not accomplishing more whether you're winning by 30 or 3 -- you're winning and moving on to the next game.

 

[husky429]

@husky429 do you have a problem with the winning margins of the past 2 women’s games?

I don't watch the women, so I have no idea.

When coaches keep in the starters until the last two minutes so they can win by 35 instead of 15, I have a problem. It's really pretty simple.

 

[storrsroars]

I don't watch the women, so I have no idea.

When coaches keep in the starters until the last two minutes so they can win by 35 instead of 15, I have a problem. It's really pretty simple.

I have a problem with it when it's a mismatch of a major vs a mediocre to poor mid-major or lower early in the season. But, at the end of the season, when you're getting ready for tournament play and you're clicking... go for it. I might also make an exception for the last game before conference play starts.

And that would apply to the UConn women today. Last game before BET and 2nd game to get Bueckers firing again. As well as Senior Day.