UConn's KenPom ranking 2-02

[auror]

And before anyone says we should schedule better cupcakes, please realize we have played all the winning record Am East teams in recent years, with games against UMass Lowell, Maine, Bryant, Vermont.

We haven't played Vermont in over a decade. Haven't played them with Becker as their coach at all. Put another way, Vermont has won 15 conference trophies since we played them last. We played Bryant last over a decade ago as well.

I don't think there is much upside to playing Vermont. But yeah it has been a bit.

 

[upstater]

We haven't played Vermont in over a decade. Haven't played them with Becker as their coach at all. Put another way, Vermont has won 15 conference trophies since we played them last. We played Bryant last over a decade ago as well.

I don't think there is much upside to playing Vermont. But yeah it has been a bit.

OK, man, sorry. I can't keep track. I just remember thinking that playing Bryant was sure to damage our record. I know we played Maine last year, and UMass Lowell recently. Regardless, I'm pretty sure some of these NET rankings are tracking whether you're playing UMass-Lowell or New Hampshire, as if that matters.

 

[pepband99]

This has to be a gaslight, right? It’s our league, but you will never admit it.

It's a complete intentionally stupid gaslight.

How many times has this guy railed on about metrics and blowouts dictating ratings? And now he asks this vague nothingburger?

Our kenpom ranking is low because our defensive ranking is low. That's the only correct answer.

 

[jibsey]

How has Ken Pom‘s ranking matched up with the Final 8 the last 2 years? Or the Sweet Sixteen? Injuries are not part of this analysis so the last 5 games stats on our team hurt the ranking assuming we get to and stay at full strength?

 

[nelsonmuntz]

It's a complete intentionally stupid gaslight.

How many times has this guy railed on about metrics and blowouts dictating ratings? And now he asks this vague nothingburger?

Our kenpom ranking is low because our defensive ranking is low. That's the only correct answer.

If you don’t know, then shut up. There are posters that really understand this stuff, and there are posters like you, who just just like to take cheap shots at other posters.

 

[pepband99]

If you don’t know, then shut up. There are posters that really understand this stuff, and there are posters like you, who just just like to take cheap shots at other posters.

1) Pot/kettle
2) Refute my answer, in that case.

Our defensive efficiency rating (115) is the worst of any team ranked in the top 54 of Kenpom, and even the number of triple-digit defensive rankings in the top 70 or so overall is very low.

Unless you think our defense has been anything above middling all year ...

 

[7774]

1) Pot/kettle
2) Refute my answer, in that case.

Our defensive efficiency rating (115) is the worst of any team ranked in the top 54 of Kenpom, and even the number of triple-digit defensive rankings in the top 70 or so overall is very low.

Unless you think our defense has been anything above middling all year ...

You don't get it, his brain has decided the metrics are bad, so then whatever the metrics say is meaningless, becasue the metrics are bad. Unfortunately for him, it's actually his brain that is bad, he just doesn't realize it because the thing that's in charge of figuring that out is his brain.

 

[pepband99]

You have no idea what you are talking about, which should be so incredibly simple that I will never understand how you can't grasp it.

...except his answer is exactly correct, but I'm the one calling names, in your mind.

Both KenPom and NET are used by the Selection Committee for seeding and field selection. I have always had a problem with the "black box" nature of these models, and it looks like there are problems with them. 2 that come to mind are:

1) There is some kind of implicit bias in the model based on the opening published results that never gets fully washed out as the season goes on. While it is normal to have a model mute new outlier information to keep the model from getting too volatile in the beginning of its use, the projected starting point should be completely removed by some point, and it is not clear that is happening.

2) The model over-weights immaterial outlier events. This is a flaw in many models. A 50 point win over a really bad team should not move a model over a 25 point win over the same team.. The people running these models claim this is not happening, but they are black boxes, so who knows?

In this gem - you clearly demonstrate that YOU have no idea what you're talking about.

Kenpom and the NET are 2 completely different things. You and I agree (and have agreed in past) on the NET having black box stuff. It's not a good look, and breeds questions like the ones you have here.

Kenpom, however, is much much more open. It's basically 2 things:
1) Measuring offensive and defensive efficiency on per-possession basis
2) Measuring those measures against other teams measures, adjusting for factors like home/away, tempo (so number of possessions), etc.

It's also a PREDICTIVE measure, not a reactive one like polls (and probably NET, at least the black box piece of it). If Kenpom says that we are lower in overall ranking, it means that it's predicting a score and result where that team does better, all being relative.

Upcoming St John's game - perfect example. St John's is 16, and we're 34 right now. The NetRtg, which is what determines the ranking order, puts St John's about 5 net points ahead of us. Accounting for the fact we're playing at home, which eats into that 5 point margin, KenPom actually predicts a win for us.

Will outliers dictate swings in Kenpom? Sure, but nothing i would deem nefarious. Your "50 point win" scenario only really matters if it moves our per-possession numbers materially, which it might not at all.

 

[pepband99]

It's why Gonzaga is ranked 11th currently when they have no good reason to be ranked that high. They pulverize the crap teams of the WCC which prettys-up their KenPom even though it adjusts for "SOS". I made a post about it in a different KenPom thread a week or so ago.

Right answer, but wrong reason. 2 outlier monsters for Gonzaga

1) Beating Baylor by 38 at home in game 1.
2) Beating a not terrible Lowell team by 59 in game 3.

The rest kind of fall into line.

 

[RuffRuff]

...except his answer is exactly correct, but I'm the one calling names, in your mind.



In this gem - you clearly demonstrate that YOU have no idea what you're talking about.

Kenpom and the NET are 2 completely different things. You and I agree (and have agreed in past) on the NET having black box stuff. It's not a good look, and breeds questions like the ones you have here.

Kenpom, however, is much much more open. It's basically 2 things:
1) Measuring offensive and defensive efficiency on per-possession basis
2) Measuring those measures against other teams measures, adjusting for factors like home/away, tempo (so number of possessions), etc.

It's also a PREDICTIVE measure, not a reactive one like polls (and probably NET, at least the black box piece of it). If Kenpom says that we are lower in overall ranking, it means that it's predicting a score and result where that team does better, all being relative.

Upcoming St John's game - perfect example. St John's is 16, and we're 34 right now. The NetRtg, which is what determines the ranking order, puts St John's about 5 net points ahead of us. Accounting for the fact we're playing at home, which eats into that 5 point margin, KenPom actually predicts a win for us.

Will outliers dictate swings in Kenpom? Sure, but nothing i would deem nefarious. Your "50 point win" scenario only really matters if it moves our per-possession numbers materially, which it might not at all.

On average what I’ve seen in terms of KP dictating line is often the spread plus/minus the 3-4 for home court, give or take .5-1. That would suggest St John’s being a 1-1.5 point favorite on Friday, barring any major swings after their game tomorrow.

For instance Kansas is a 2.5 favorite over Iowa St tonight. Iowa St is has a 1 point KP advantage over Kansas, Kansas is getting 3.5 for home court.

That said I’m not sure I’d call KP predictive as it’s basing its metric on historical data. It’s simply a measurement that can imply trend, and quality of play against a variable schedule, if you want to call that predictive. NET is simply measuring quality of resume more than quality of play, which is then used to evaluate body of work for seeding purposes. The difference between the two is valuing how a team wins and loses.

 

[Iluvsuebird]

Geez this is nerdy stuff. I am with the dude that narrows it down to one thing: defense. You improve the D and when taking into account our offense, we will be really, really hard to beat, nerds notwithstanding.

 

[nelsonmuntz]

You don't get it, his brain has decided the metrics are bad, so then whatever the metrics say is meaningless, becasue the metrics are bad. Unfortunately for him, it's actually his brain that is bad, he just doesn't realize it because the thing that's in charge of figuring that out is his brain.

What is your problem? If you don't like what I say, then move on to the next post. The stalking thing is getting weird.

 

[nelsonmuntz]

...except his answer is exactly correct, but I'm the one calling names, in your mind.



In this gem - you clearly demonstrate that YOU have no idea what you're talking about.

Kenpom and the NET are 2 completely different things. You and I agree (and have agreed in past) on the NET having black box stuff. It's not a good look, and breeds questions like the ones you have here.

Kenpom, however, is much much more open. It's basically 2 things:
1) Measuring offensive and defensive efficiency on per-possession basis
2) Measuring those measures against other teams measures, adjusting for factors like home/away, tempo (so number of possessions), etc.

It's also a PREDICTIVE measure, not a reactive one like polls (and probably NET, at least the black box piece of it). If Kenpom says that we are lower in overall ranking, it means that it's predicting a score and result where that team does better, all being relative.

Upcoming St John's game - perfect example. St John's is 16, and we're 34 right now. The NetRtg, which is what determines the ranking order, puts St John's about 5 net points ahead of us. Accounting for the fact we're playing at home, which eats into that 5 point margin, KenPom actually predicts a win for us.

Will outliers dictate swings in Kenpom? Sure, but nothing i would deem nefarious. Your "50 point win" scenario only really matters if it moves our per-possession numbers materially, which it might not at all.

It doesn't have to be nefarious to be a glitch in the model.

I work with a lot of models in my job, and most of them have implicit guardrails that tend to influence the model's interpretation of new data, i.e. the model's starting point never gets completely removed from the model. I also find that models frequently equally weight data, including outliers. I imagine this is unavoidable with a model based on per possession efficiency, which is fine for measuring offense and defense, but can skew outcomes when looking at the team as a whole.

By focusing on efficiency, KenPom has to be leaving the margin uncapped, even if the model is adjusting for quality of opponents. There is no other way to incorporate efficiency with a capped margin because how would the model know which possessions to throw out?

 

[RuffRuff]

It doesn't have to be nefarious to be a glitch in the model.

I work with a lot of models in my job, and most of them have implicit guardrails that tend to influence the model's interpretation of new data, i.e. the model's starting point never gets completely removed from the model. I also find that models frequently equally weight data, including outliers. I imagine this is unavoidable with a model based on per possession efficiency, which is fine for measuring offense and defense, but can skew outcomes when looking at the team as a whole.

By focusing on efficiency, KenPom has to be leaving the margin uncapped, even if the model is adjusting for quality of opponents. There is no other way to incorporate efficiency with a capped margin because how would the model know which possessions to throw out?

Any chance you work with Heidi Klum or Emily Ratjkowski?

 

[7774]

What is your problem? If you don't like what I say, then move on to the next post. The stalking thing is getting weird.

The only reason we’re interacting in this thread is you replied to one of my posts which was addressed to a completely different individual. Even the post you replied to here was not addressed to you, so maybe you should stay out of the conversations I'm having with other posters. You should stick to going into the “other NCAA games” threads and complaining about how people are discussing other NCAA games, that’s more your strength. Things involving math just aren’t for you in this life.

 

[pepband99]

That said I’m not sure I’d call KP predictive as it’s basing its metric on historical data.

Errr - what would be an example of a predictive metric, given the above. Voodoo? Tarot cards?

 

[RuffRuff]

Errr - what would be an example of a predictive metric, given the above. Voodoo? Tarot cards?

Personally prefer fortune tellers as the crystal ball has a great track record in college hoops.

How many different metrics are there these days based on their own presumptive algorithm? KP, Torvik, Evan M, others. Torvik has us ranked 13th, KP has us 34th. Big delta. None of these are dead center, just directional, but I guess you can tag all of them with the term predictive. I prefer directional data. Still far too many human factors involved (injury, portal gelling, imbalanced schedules, travel, 1 and one tourney, etc), in a short season. Case in point, NC State last year. College hoops isn't MLB & moneyball, much more art involved, coaching impact, and circumstantial. But I get what you're saying.

 

[jibsey]

I saw one poll today in which UConn was not in the top 25 AFTER beating Marquette #9. You can take the polls, the Ken Pom and whatever else you have and stick them where……..

 

[UCann]

I saw one poll today in which UConn was not in the top 25 AFTER beating Marquette #9. You can take the polls, the Ken Pom and whatever else you have and stick them where……..

Was it this one, if so I agree what a joke (also somehow Marquette only drops one spot):

College basketball rankings: North Carolina joins Top 25 And 1 after late addition of international prospect

Tar Heels coach Hubert Davis continues to improve his roster by landing Luka Bogavac, a 6-foot-5 guard from Montenegro

www.cbssports.com

 

[navery12]

Personally prefer fortune tellers as the crystal ball has a great track record in college hoops.

How many different metrics are there these days based on their own presumptive algorithm? KP, Torvik, Evan M, others. Torvik has us ranked 13th, KP has us 34th. Big delta. None of these are dead center, just directional, but I guess you can tag all of them with the term predictive. I prefer directional data. Still far too many human factors involved (injury, portal gelling, imbalanced schedules, travel, 1 and one tourney, etc), in a short season. Case in point, NC State last year. College hoops isn't MLB & moneyball, much more art involved, coaching impact, and circumstantial. But I get what you're saying.

Not sure where you're looking but we're 33 on Bart Torvik and 34 on KenPom. There is not a large delta

 

[DRBackup]

It doesn't have to be nefarious to be a glitch in the model.

I work with a lot of models in my job, and most of them have implicit guardrails that tend to influence the model's interpretation of new data, i.e. the model's starting point never gets completely removed from the model. I also find that models frequently equally weight data, including outliers. I imagine this is unavoidable with a model based on per possession efficiency, which is fine for measuring offense and defense, but can skew outcomes when looking at the team as a whole.

By focusing on efficiency, KenPom has to be leaving the margin uncapped, even if the model is adjusting for quality of opponents. There is no other way to incorporate efficiency with a capped margin because how would the model know which possessions to throw out?

I'm not sure I'd call what KenPom does a "model". The only thing "subjective" that he does is adjust for opponent and tempo. You could call that a "model", I suppose, but I imagine he has a literal ton of empirical data to back that up. I'd be curious how much these adjustments even affect the overall ratings, honestly.

Our defense has not been impressive this year. Combine that with a lot of teams that played very well on offense, keyed up to play the 2 time defending champs.

 

[Get a Job]

Some dude once said, "All models are wrong, but some are useful."

 

[auror]

That said I’m not sure I’d call KP predictive as it’s basing its metric on historical data. It’s simply a measurement that can imply trend, and quality of play against a variable schedule, if you want to call that predictive. NET is simply measuring quality of resume more than quality of play, which is then used to evaluate body of work for seeding purposes. The difference between the two is valuing how a team wins and loses.

There are two types of college basketball team metrics and they are usually referred to as "predictive", aka future facing. Or "resume", backwards facing. They both look at what a team has done, but 1 uses that to model how the team will perform in the future, and the other just looks at how the team has done relative to others. So obviously the models will be variously good at doing the predicting, but that's the umbrella term they fall into.

Popular metrics:

Predictive:
KenPom, T-Rank, Miya Relative Rankings, BPI, Haslametrics

Resume:
SOR, KPI, Torvik and NET WAB, the new Miya Resume Quality, Haslametrics "Deserves"

The NET is a blend of both, but primarily predictive. Based on comparing it to other models, I look at it as predictive with a bonus for teams that have a lot of raw wins.

 

[Orange]

UConn is only 35 as of right now, with Gonzaga still playing. There are 14 teams with 6 or more losses ahead of UConn in KenPom, including 3 that UConn beat head to head. What is driving UConn's bad ranking? Is it the 3 games against teams ranked in the bottom 10 of D1?

Unlike BPI, Pomeroy completely removes preseason projection stuff from his rankings around the time that conference play begins. That's part of why BPI always rates us considerably better than KenPom does. We're better than how we've actually played so far this year.

 

[nelsonmuntz]

The only reason we’re interacting in this thread is you replied to one of my posts which was addressed to a completely different individual. Even the post you replied to here was not addressed to you, so maybe you should stay out of the conversations I'm having with other posters. You should stick to going into the “other NCAA games” threads and complaining about how people are discussing other NCAA games, that’s more your strength. Things involving math just aren’t for you in this life.

You are the neediest poster on this site.

 

[BigErnMcCracken]

You are the neediest poster on this site.

If you don’t know, then shut up. There are posters that really understand this stuff, and there are posters like you, who just just like to take cheap shots at other posters.

 

[superjohn]

This thread is worth it to watch Waylon dunk all over himself, it never gets old.

We get the Waylon trifecta here.

Dunks all over himself while being a clueless jerk, plays the victim, and then accuses others of exhibiting behavior he exhibits here daily.

 

[pepband99]

Geez this is nerdy stuff. I am with the dude that narrows it down to one thing: defense. You improve the D and when taking into account our offense, we will be really, really hard to beat, nerds notwithstanding.

Feeding the nerd mojo seems to work on occasion - I know what my role is

 

[6Nattys4Us]

Torvik has a metric called "Game Score" (G-Sc) that you can see on his team page. It's essentially the margin of victory (including mean margin during the course of the game) adjusted for opponent and location. You can consider this "how the result of the game impacts the team's overall adjusted efficiency". This is specific to Torvik, but other than the in-game mean, it should apply pretty similarly to KenPom.

You can see which games really affected our rankings. Dayton, bad. Marquette and Texas on the road, good. Since the start of January (and missing McNeeley), we've had 6 white or light green-colored results, which are not up to standard for a potential top 25 team. At Georgetown and at Marquette were good results.

Seems pretty basic - the more you score the better your O score; the less you give up the better your D score. Win/Loss is barely a factor.

 

[UCTitleTown]

That loss to Colorado isn't helping us now and definitely not when NCAA seedings come out. They're currently 9-12.....and 0-10 in the Big 12. Yuck.