NET and KenPom Tracker 23-24 (SU = 84 3/9/24)

[CardiacCuse]

A lock?

14 more games to go, minimum

a 13-5 team going only 7-7 to close the year would be rather unlikely, imo

theyve already played the meat of their schedule, too

theyll be favorites to win more than 7 times from here on out

 

[jncuse]

And further to my "further to this", Kenpom isn't as good of a predictor anymore due to the instant transfer portal and going pro early. Some of the baseline assumptions are no longer reliable since so many teams change drastically from year to year. I don't think Kenpom has come up with a good way to deal with the player movement. It would be a lot more accurate if players stayed at one college for 4 years.

I agree that it is not a good predictor for the first 5-8 games. I don't think anybody is particularly a good predictor in this new environment of crazy transfers. Be it AP, Coaches Poll, KP whatever. Everything is a major grain of salt.

But his preseason weights are all gone at this point. The ADJ EM is all based on current year connections and current year results.

 

[EnviroSciGuy]

Will be interesting to see how the weekend games impact our score.

 

[Sbucketball]

That's not correct.

The luck factor at least by this point of the season has nothing to do with our preseason ranking.
It is driven by not having proportional levels of blowouts, or winning or losing an abnormal amount of close games. We can see that in our results.

At this point of the season the opening KP is probably irrelevant in his system (which we started around #100) Around Christmas, his system has drawn down the "preseason weight" to around 20% of his score based on his prior comments. So at that point we were probably weighed down by 5-7 spots. I would imagine by January 20 his preseason weights are either worth 0% or very close to it.

The luck factor is simply how Kenpom explains a team beating better teams. My point is we've been undervalued since Day 1. We started with a bad metric, and add in this team giving up in blowouts, we haven't been able to climb out to a proper ranking. Not saying we're still dragged way down by the preseason start, but it doesn't help when you start with a bad metric. If we started at #10 and had the same season to date, I'm guessing we'd have a better metric right now. All of the teams move in relation to each other, like NET, so our wins/loses have a different effect on our opponents when we're not ranked well. Look at teams with similar resumes but are much higher in NET and Kenpom because they started way higher. Maybe I'm wrong, but it certainly feels that way. It's tough to climb in these rankings.

 

[Sbucketball]

I agree that it is not a good predictor for the first 5-8 games. I don't think anybody is particularly a good predictor in this new environment of crazy transfers. Be it AP, Coaches Poll, KP whatever. Everything is a major grain of salt.

But his preseason weights are all gone at this point. The ADJ EM is all based on current year connections and current year results.

The AP is always wild. UNC pre-season #1 last year because they got hot for 5 games. Miami pre-season top 15 due to Final Four. It's always so horrible and there's always a load of teams that are ranked purely based on tournament performance the year before, notwithstanding how they performed during the regular season or if the team changed over a bit.

 

[jordoo]

The problem with math is that we try to use it to explain the universe and the events that unfolded in it, however, the universe has no need of math to operate.

 

[jncuse]

I don’t think there is an algorithm that accounts for a sophomore laden team with many new pieces and a first year coach growing over the course of a season. So, this one calls it “luck”, because it defies his logic. I see growth. I’ve been watching sports long before these geeks started trying to ruin sports. It is simply a young team improving. It’s the best thing about college basketball, in my opinion. It’s even more fun when it’s your team improving. I’m not going to sweat the analytics, because they could never account for the things that really matter in sports.

At this point your just making stuff up to what you think it is, and then attacking it. None of what you claim has anything to do with this analytic.

I'm not even defending the "Luck" term, in facts its often just a point of general interest - I already said it should have a different name at minimum. But at least I am going to tell people what it stands for
a) has nothing to do with pre-season ranking.
b) has to do with blowouts (more one way than the other)
c) performance in close games.(more one way than the other)

 

[orangefog]

At this point your just making stuff up to what you think it is, and then attacking it. None of what you claim has anything to do with this analytic.

I'm not even defending the "Luck" term, in facts its often just a point of general interest - I already said it should have a different name at minimum. But at least I am going to tell people what it stands for
a) has nothing to do with pre-season ranking.
b) has to do with blowouts (more one way than the other)
c) performance in close games.(more one way than the other)

I’m not making anything up. I don’t give a crap about this statistical BS. People are trying to create algorithms to predict games. When the game doesn’t turn out as predicted, in this case, they throw it in to “luck”. That says to me that the algorithm is useless. Either way, I don’t care. Our guys are improving. And, it’s “you’re”.

 

[JustPherU]

14 more games to go, minimum

a 13-5 team going only 7-7 to close the year would be rather unlikely, imo

theyve already played the meat of their schedule, too

theyll be favorites to win more than 7 times from here on out

A lock to me means 7 games left where we would be huge favorites, it’s not like when we win we are blowing people out. I think we have a chance , but by no means are we a lock

 

[jncuse]

The luck factor is simply how Kenpom explains a team beating better teams. My point is we've been undervalued since Day 1. We started with a bad metric, and add in this team giving up in blowouts, we haven't been able to climb out to a proper ranking. Not saying we're still dragged way down by the preseason start, but it doesn't help when you start with a bad metric. If we started at #10 and had the same season to date, I'm guessing we'd have a better metric right now. All of the teams move in relation to each other, like NET, so our wins/loses have a different effect on our opponents when we're not ranked well. Look at teams with similar resumes but are much higher in NET and Kenpom because they started way higher. Maybe I'm wrong, but it certainly feels that way. It's tough to climb in these rankings.

#1. I don't think the bolded is accurate. His results are connected based on how the team has played all year. There is no longer a preseason weight overlaying his connected data, to adjust his ranking (if it is, it has to be quite small factor at this point) I posted a link earlier in this thread from one of his articles some years back, that

#2. You do raise a valid point though. We are being punished in KP and NET because of our early season blowouts -- even our recent ones unfortunately. Should our team be viewed worse for that? I think that depends on mood.
After all, we see points in all the threads here bemoaning the fact that we aren't really good because of those 5 losses -- and then we also see the opposite especially after we win by one point!

 

[UpstateSM]

Don't look now but we are projected to be tied for the last double bye spot in the ACCT per Kenpom right now. The next two games at home are sooooo important.

 

[jncuse]

I’m not making anything up. I don’t give a crap about this statistical BS. People are trying to create algorithms to predict games. When the game doesn’t turn out as predicted, in this case, they throw it in to “luck”. That says to me that the algorithm is useless. Either way, I don’t care. Our guys are improving. And, it’s “you’re”.

Great point. His algorithm predicted we would win by 2 points today.
It should have known we would win by 3.

I do agree we are improving. At least we agree on that and the fact that it was a huge great win.

You're welcome.

 

[721Comstock]

I know I speak for OiG when I say -

I’ll see Ken Pomeroy in Hell.

 

[Sbucketball]

#1. I don't think the bolded is accurate. His results are connected based on how the team has played all year. There is no longer a preseason weight overlaying his connected data, to adjust his ranking (if it is, it has to be quite small factor at this point) I posted a link earlier in this thread from one of his articles some years back, that

#2. You do raise a valid point though. We are being punished in KP and NET because of our early season blowouts -- even our recent ones unfortunately. Should our team be viewed worse for that? I think that depends on mood.
After all, we see points in all the threads here bemoaning the fact that we aren't really good because of those 5 losses -- and then we also see the opposite especially after we win by one point!

I haven't specifically calculated anything so again maybe I'm off here, but my gut tells me the system makes it tough to climb up the rankings when you are undervalued. Take NET for example (since that's more important than Kenpom) - if we are undervalued at #100 and we beat #60, the #60 team might drop 10 spots making our win worth less. If we're overvalued at #30 and beat #60, the #60 team might actually rise up a few spots for a close game making our win more valuable. It's like quicksand. Starting out undervalued makes your wins worth less. And once conference play starts, we only play conference teams so if a bunch of teams started slow and are undervalued, it's a conference full of quick sand. For example, VT a year or two ago started hot then lost a bunch straight when Catoor was injured, and didn't really drop as much as you would expect. Had they dropped those games early in the year, my gut tells me it would have made them worse off. IDK, hopefully I'm wrong but that's my instinct on it.

 

[jncuse]

Here is the important thing though.

KenPom isn't used to select teams. Or standalone NET for the most point.
Nor should they. When I start selecting teams out of P6, the KP goes in the garbage, and the NET os merely a look at me number. I look at it a bit more for trying to measure mid-majors because this is one of the few ways to understand them

The most important factor is your W-L, and your Q1 and Q2 wins. And that is what we did today. Winning by one doesn't really help your KP, but it really helped our team to make the tournament.

If we finish the ACC at 8-5, our NET will not be a factor to push us in, but it will not be a hindrance either... I suspect it would get somewhere in the high 50's as a guess by that point.

 

[CardiacCuse]

A lock to me means 7 games left where we would be huge favorites, it’s not like when we win we are blowing people out. I think we have a chance , but by no means are we a lock

its a lock that of the remaining at leat 14 games...that 7 of them will be wins...based on everything we've seen so far.

how many wins do you think the team will end the regualr season with?

go look at the schedule and count them off

 

[storange]

its a lock that of the remaining at leat 14 games...that 7 of them will be wins...based on everything we've seen so far.

how many wins do you think the team will end the regualr season with?

go look at the schedule and count them off

26

 

[JustPherU]

its a lock that of the remaining at leat 14 games...that 7 of them will be wins...based on everything we've seen so far.

how many wins do you think the team will end the regualr season with?

go look at the schedule and count them off

By definition lock means an almost certainty that you will win.Like I said, we definitely can win 20, but no way are we a lock to do it.

 

[baggerbob]

Why would he give them credit for that? They haven’t been able to compete with a quality team yet, let alone beat them. If you take off the orange colored glasses, this team is not anywhere close to the tournament.

It doesn't make any difference what board or topic you never have anything positive to say.
Won't put you on ignore because then it shows I didn't read all the messages, but won't read anymore from you.

 

[OttoinGrotto]

it’s not like when we win we are blowing people out.

Other than today, that's absolutely how it's been tho.

 

[storange]

Why would he give them credit for that? They haven’t been able to compete with a quality team yet, let alone beat them. If you take off the orange colored glasses, this team is not anywhere close to the tournament.

Saying they’re not close is ignorant

 

[jncuse]

I haven't specifically calculated anything so again maybe I'm off here, but my gut tells me the system makes it tough to climb up the rankings when you are undervalued. Take NET for example (since that's more important than Kenpom) - if we are undervalued at #100 and we beat #60, the #60 team might drop 10 spots making our win worth less. If we're overvalued at #30 and beat #60, the #60 team might actually rise up a few spots for a close game making our win more valuable. It's like quicksand. Starting out undervalued makes your wins worth less. And once conference play starts, we only play conference teams so if a bunch of teams started slow and are undervalued, it's a conference full of quick sand. For example, VT a year or two ago started hot then lost a bunch straight when Catoor was injured, and didn't really drop as much as you would expect. Had they dropped those games early in the year, my gut tells me it would have made them worse off. IDK, hopefully I'm wrong but that's my instinct on it.

I believe you are describing a system that would have some "ELO" type mechanics. I'm fairly certain KP doesn't do that, but I could be wrong.

 

[storange]

I believe you are describing a system that would have some "ELO" type mechanics. I'm fairly certain KP doesn't do that, but I could be wrong.

Electric Light Orchestra is involved ? Great song writers and Jeff Lynn part of the “Traveling Willburries “. Awesome

 

[orangefog]

Electric Light Orchestra is involved ? Great song writers and Jeff Lynn part of the “Traveling Willburries “. Awesome

It’s used all over the world.

 

[PoppyHart]

There’s certainly some strange magic involved with all these algos, but I would turn to stone if this team doesn’t make the dance.