sorry, not sorry

[Frybry02]

If they bring the hammer to AZ, then the NCAA better bring a fucking Jackhammer to a few other programs

[UAEebs86]

Sean Miller is a witch.
Dick Vitale is a bitch.

[Chicat]

[Longhorned]

You know what this sports media coverage is like? The Italian media coverage of Amanda Knox.

[Alieberman]

My morning routine:

Get out of bed
Pee
Make coffee
Open laptop- refresh BearDownWildcats tab. Check to see if Arizona Basketball and Sean Miller are still a thing.
Drink coffee

I don't know what would happen to my mornings if Arizona Basketball was on stable ground.

[UAEebs86]

You know what this sports media coverage is like? The Italian media coverage of Amanda Knox.

Well, I know for a fact Sean Miller tried to kill a British Royal AND Jack Ryan.

[YoDeFoe]

KenPom ranked every basketball program starting with the 1997 season... Arizona comes in at number five.

https://kenpom.com/programs.php

1. Duke
2. Kentucky
3. Kansas
4. UNC
5. Arizona
6. MSU
7. Nova
8. Lousiville
9. Florida
10. Texas

[Spaceman Spiff]

KenPom ranked every basketball program starting with the 1997 season... Arizona comes in at number five.

https://kenpom.com/programs.php

1. Duke
2. Kentucky
3. Kansas
4. UNC
5. Arizona
6. MSU
7. Nova
8. Lousiville
9. Florida
10. Texas

We need to start paying more. Also, Texas in the top ten is crazy.

[azcat49]

Didn’t know where to put this and I thought it might fit here given where we rank. The four blue bloods then us. Something to fight for

https://www.google.com/amp/s/www.cbsspo ... grams/amp/" target="_blank

[ghostwhitehorse]

Didn’t know where to put this and I thought it might fit here given where we rank. The four blue bloods then us. Something to fight for

https://www.google.com/amp/s/www.cbsspo ... grams/amp/" target="_blank

3 things. . .

Interesting that the data starts to be unreliable past '97

Top 5 YEAH!

Autoplay vids should be chucked into the wood shredder.

[U.P. Zona Fan]

Man I miss the days of the kenpom blog, I loved nerding out to that stuff.

[Captain Obvious]

KenPom ranked every basketball program starting with the 1997 season... Arizona comes in at number five.

https://kenpom.com/programs.php

1. Duke
2. Kentucky
3. Kansas
4. UNC
5. Arizona
6. MSU
7. Nova
8. Lousiville
9. Florida
10. Texas

Never put much stock in these types of rankings. Most of these teams have been to the Final Four multiple times since 1997. The team with the longest Final Four drought on the list is......drum roll.....Arizona. Texas made it in 2003. That's exactly why I could care less about rankings and statistics. They can be manipulated rather easily to make things seem better than they actually are. We are nowhere near being a blue blood program. Associating us in any capacity with any of those programs is ridiculous to the point of being offensive.

[Longhorned]

Even if your algorithm is neutral and transparent, you're doing fuzzy math if Arizona ranks highly. That Ken Pomeroy can't stop lying just to try to make us look better than we are, so I'm here to remind you that we suck. And if you get a promotion tomorrow and your supervisor shakes your hand, just remember his hand is covered in fecal matter, and so are you. And that you're going to hell. I hate you. I hope you fall into a trough-urinal naked.

[CatFanOneMil]

Kenpom has made his money selling the idea that his algorithm is solid...but it is solid speculation and he ignores a huge swath of the data that is unpredictable...never give your money to an internet carny unless you simply want to be entertained, you're not winning the big stuffed trophy at Kenpoms "toss-a-ring" booth...its rigged to ONLY take your money as are ALL the prediction models, (We see you nate silver 538)...

Prediction models of both the future and the past are charlatan games if they worked these guys would predict stocks and be billionaires.

Falls into the same category as ufo's and bigfoot...now that everyone has a high quality cam recorder in their pocket the videos of these elusive mythological creatures from both alien worlds and the hollow earth have mysteriously dried up...go figure...data science will make us all a little more depressed but reality is boring as heck.

[dovecanyoncat]

never give your money to an internet carny

Ceezarneezies eezare eeza neezateezioneezal treezsure eezand deezon't yeezou eezeveezer feezorgeezet eezit!

[YoDeFoe]

KenPom ranked every basketball program starting with the 1997 season... Arizona comes in at number five.

https://kenpom.com/programs.php

1. Duke
2. Kentucky
3. Kansas
4. UNC
5. Arizona
6. MSU
7. Nova
8. Lousiville
9. Florida
10. Texas

Never put much stock in these types of rankings. Most of these teams have been to the Final Four multiple times since 1997. The team with the longest Final Four drought on the list is......drum roll.....Arizona. Texas made it in 2003. That's exactly why I could care less about rankings and statistics. They can be manipulated rather easily to make things seem better than they actually are. We are nowhere near being a blue blood program. Associating us in any capacity with any of those programs is ridiculous to the point of being offensive.

I think if you asked someone if Arizona was in the same grouping as MSU, Nova, and Louisville... they'd say yes. Our FF drought is a problem, but the program has consistently won games, conference titles, and recruiting battles. I'm not sure I'd rank us fifth but somewhere in the top ten seems right.

[SabinoDrifter]

Kenpom has made his money selling the idea that his algorithm is solid...but it is solid speculation and he ignores a huge swath of the data that is unpredictable...never give your money to an internet carny unless you simply want to be entertained, you're not winning the big stuffed trophy at Kenpoms "toss-a-ring" booth...its rigged to ONLY take your money as are ALL the prediction models, (We see you nate silver 538)...

Prediction models of both the future and the past are charlatan games if they worked these guys would predict stocks and be billionaires.

Falls into the same category as ufo's and bigfoot...now that everyone has a high quality cam recorder in their pocket the videos of these elusive mythological creatures from both alien worlds and the hollow earth have mysteriously dried up...go figure...data science will make us all a little more depressed but reality is boring as heck.

Hate to break it to you, but every division 1 team has a kenpom account and they often reach out to him when trying to beef up their non-conference schedules.

Also, not being a billionaire doesn't mean you can be smart and successful.

[CatFanOneMil]

Kenpom has made his money selling the idea that his algorithm is solid...but it is solid speculation and he ignores a huge swath of the data that is unpredictable...never give your money to an internet carny unless you simply want to be entertained, you're not winning the big stuffed trophy at Kenpoms "toss-a-ring" booth...its rigged to ONLY take your money as are ALL the prediction models, (We see you nate silver 538)...

Prediction models of both the future and the past are charlatan games if they worked these guys would predict stocks and be billionaires.

Falls into the same category as ufo's and bigfoot...now that everyone has a high quality cam recorder in their pocket the videos of these elusive mythological creatures from both alien worlds and the hollow earth have mysteriously dried up...go figure...data science will make us all a little more depressed but reality is boring as heck.

Hate to break it to you, but every division 1 team has a kenpom account and they often reach out to him when trying to beef up their non-conference schedules.

Also, not being a billionaire doesn't mean you can be smart and successful.

Not sure what point you are making, but my point was simple Kenpoms prediction model is crap.

It is.

Does not mean people won't use it, hell people still think body fat is based purely on calories in (which is just really bad science)...lots of diets out there that people sign onto because they're popular...

Just because Kenpom is popular does not make him right, it just means the crowd believes him.

And we know where that leads.

[Spaceman Spiff]

Kenpom has made his money selling the idea that his algorithm is solid...but it is solid speculation and he ignores a huge swath of the data that is unpredictable...never give your money to an internet carny unless you simply want to be entertained, you're not winning the big stuffed trophy at Kenpoms "toss-a-ring" booth...its rigged to ONLY take your money as are ALL the prediction models, (We see you nate silver 538)...

Prediction models of both the future and the past are charlatan games if they worked these guys would predict stocks and be billionaires.

Falls into the same category as ufo's and bigfoot...now that everyone has a high quality cam recorder in their pocket the videos of these elusive mythological creatures from both alien worlds and the hollow earth have mysteriously dried up...go figure...data science will make us all a little more depressed but reality is boring as heck.

Hate to break it to you, but every division 1 team has a kenpom account and they often reach out to him when trying to beef up their non-conference schedules.

Also, not being a billionaire doesn't mean you can be smart and successful.

Not sure what point you are making, but my point was simple Kenpoms prediction model is crap.

It is.

Does not mean people won't use it, hell people still think body fat is based purely on calories in (which is just really bad science)...lots of diets out there that people sign onto because they're popular...

Just because Kenpom is popular does not make him right, it just means the crowd believes him.

And we know where that leads.

It's more about college basketball being the least predictable sport and less about Kenpom being bad. Kenpom's model is fine, but predicting college basketball is as hard as it gets.

CBB experts regularly do terribly in predicting the tournament. CBB has several factors that lends itself to being really hard to accurately predict.

[Beachcat97]

It's more about college basketball being the least predictable sport

Which is why Vegas LOVES it.

[CatFanOneMil]

Kenpom has made his money selling the idea that his algorithm is solid...but it is solid speculation and he ignores a huge swath of the data that is unpredictable...never give your money to an internet carny unless you simply want to be entertained, you're not winning the big stuffed trophy at Kenpoms "toss-a-ring" booth...its rigged to ONLY take your money as are ALL the prediction models, (We see you nate silver 538)...

Prediction models of both the future and the past are charlatan games if they worked these guys would predict stocks and be billionaires.

Falls into the same category as ufo's and bigfoot...now that everyone has a high quality cam recorder in their pocket the videos of these elusive mythological creatures from both alien worlds and the hollow earth have mysteriously dried up...go figure...data science will make us all a little more depressed but reality is boring as heck.

Hate to break it to you, but every division 1 team has a kenpom account and they often reach out to him when trying to beef up their non-conference schedules.

Also, not being a billionaire doesn't mean you can be smart and successful.

Not sure what point you are making, but my point was simple Kenpoms prediction model is crap.

It is.

Does not mean people won't use it, hell people still think body fat is based purely on calories in (which is just really bad science)...lots of diets out there that people sign onto because they're popular...

Just because Kenpom is popular does not make him right, it just means the crowd believes him.

And we know where that leads.

It's more about college basketball being the least predictable sport and less about Kenpom being bad. Kenpom's model is fine, but predicting college basketball is as hard as it gets.

CBB experts regularly do terribly in predicting the tournament. CBB has several factors that lends itself to being really hard to accurately predict.

My point has been stated more than once that Kenpoms model is NOT fine...the fact that CBB experts (who depend deeply on Kenpom) do terribly is the proof...Kenpoms model leaves out a LOT of the intangibles that make it hard...Kenpom started as a decent project, but lets be very clear here, his prediction models are complete bunk, it is NOT science, it is completely subjective opinion disguised as math...its not even good math. I mean for crying out loud he recently changed one of the metrics from "multiplying" to "adding"...(regarding teams competing and defensive vs offensive ratings)

Here's the money quote from his own site "I’m not sure how basketball really works, but my hunch is that it’s probably closer to additive than multiplicative"

His fucking HUNCH...no data scientist uses the word "hunch" when giving you his formula...

Drop Kenpoms formula into a stats class...see how well it floats...it does not float at all.

I get it, we all love some math when we want to look at teams/etc, hell I enjoy a good Kenpom rating, it gives me a warm fuzzy feeling that snuggles my hopes...but its not REAL math...its hopeful math. He might as well include a stat on shoe sizes it has just as much bearing as some of his other factors.

If I had the data I could construct a model that was based purely on shoe sizes at each position and it would be just as accurate.

And while we're at it we should b just as honest about player ratings...what does a "Five Star" even mean? How has that translated in the real world?

I'm sure there might be some metric that allows us to track it, but for crying out loud we still think it matters more than it should...where are all those top classes of recruits in the grand scheme of things?

Only one team wins the natty...you can load up on carbs and five star recruits and get the same result.

We've ALL witnessed second rate teams dismantle our Five Star loaded crew...its painful to watch...Bahamas anyone?

[Longhorned]

Drop Kenpoms formula into a stats class...see how well it floats...it does not float at all.

CFOM, it sounds like you know a lot about the granular aspects of Kenpom's algorithms, and I don't. But my wife's requirements for a doctoral degree in information science -- in the premier information science program in the country -- included a stats class taught by one of the most widely respected statisticians in the world, and he taught the class using examples from Kenpom, which he thinks very highly of. Maybe you're a stats guru, but this is the first time I've ever heard about fundamental problems with Kenpom.

[CatFanOneMil]

CFOM, it sounds like you know a lot about the granular aspects of Kenpom's algorithms, and I don't. But my wife's requirements for a doctoral degree in information science -- in the premier information science program in the country -- included a stats class taught by one of the most widely respected statisticians in the world, and he taught the class using examples from Kenpom, which he thinks very highly of. Maybe you're a stats guru, but this is the first time I've ever heard about fundamental problems with Kenpom.

As far as COLLECTING and organizing stats, Kenpom is fine, but for use as a prediction model it is not...

Statistics isn't about discovering correlations, it's about eliminating coincidence, and coincidence is the bedrock of Kenpoms model.

Don't really care who quotes him, his model is based on coincidence and when you get into predicting things ascertaining causation is extremely hard while uncovering correlations is easy which is what Kenpoms model does...but in order to do that it ignores a ton of secondary effects which is incredibly wrong in predicting things.

As far as the granular aspects of Kenpoms model, no I don't need to know, all I need is his basic premise which he uncovers for you in his own page...I simply know enough about math and prediction models to know that ignoring fat tails and secondary effects is bad science no matter who does it.

[97cats]

some fun facts:

did you know, over the last 18 years, only three national champions (UCONN twice) had an adjusted offensive ranking outside the top 10, and eleven of those champions had an adjusted offense in the top 4.

further, every national champion finished in the Kenpom top ten with one exception, UCONN in 2014 at #15, and fifteen of those champions being in the top 3.

also interesting, outside of UCONN in 2011 & 2014 - every Nation Champion has come from the PreSeason AP Top 10.

[ChooChooCat]

some fun facts:

did you know, over the last 18 years, only three national champions (UCONN twice) had an adjusted offensive ranking outside the top 10, and eleven of those champions had an adjusted offense in the top 4.

further, every national champion finished in the Kenpom top ten with one exception, UCONN in 2014 at #15, and fifteen of those champions being in the top 3.

also interesting, outside of UCONN in 2011 & 2014 - every Nation Champion has come from the PreSeason AP Top 10.

Whoever the UCONN fan is that sold his soul to the devil for his team's success must have been a really important person, because good lord no team deserves as much bullshit luck as they have gotten.

[CatFanOneMil]

some fun facts:

did you know, over the last 18 years, only three national champions (UCONN twice) had an adjusted offensive ranking outside the top 10, and eleven of those champions had an adjusted offense in the top 4.

further, every national champion finished in the Kenpom top ten with one exception, UCONN in 2014 at #15, and fifteen of those champions being in the top 3.

also interesting, outside of UCONN in 2011 & 2014 - every Nation Champion has come from the PreSeason AP Top 10.

UCONN proves the equation is a coincidence not a correlation...this was my point...Kenpom does not allow you to dig down and eliminate coincidence he simply demonstrates it exists.

You could produce the same stats by using the color Blue in uniforms as a metric...but it'd probably be way higher.

[97cats]

UCONN proves the equation is a coincidence not a correlation...this was my point...Kenpom does not allow you to dig down and eliminate coincidence he simply demonstrates it exists.

You could produce the same stats by using the color Blue in uniforms as a metric...but it'd probably be way higher.

UCONN as a case study is proving that it is a correlation, and they and their results is the anomaly.

fifteen out of eighteen seasons is 83.3% of the time.

is that a coincidence?

im no mathematician, but it seems something occurring more than eighty percent of the time over an 18 year period would classify as a correlation.

[CatFanOneMil]

UCONN proves the equation is a coincidence not a correlation...this was my point...Kenpom does not allow you to dig down and eliminate coincidence he simply demonstrates it exists.

You could produce the same stats by using the color Blue in uniforms as a metric...but it'd probably be way higher.

UCONN as a case study is proving that it is a correlation, and they and their results is the anomaly.

fifteen out of eighteen seasons is 83.3% of the time.

is that a coincidence?

im no mathematician, but it seems something occurring more than eighty percent of the time over an 18 year period would classify as a correlation.

Try it with the color blue in uniforms and get back to me.

[Spaceman Spiff]

UCONN proves the equation is a coincidence not a correlation...this was my point...Kenpom does not allow you to dig down and eliminate coincidence he simply demonstrates it exists.

You could produce the same stats by using the color Blue in uniforms as a metric...but it'd probably be way higher.

UCONN as a case study is proving that it is a correlation, and they and their results is the anomaly.

fifteen out of eighteen seasons is 83.3% of the time.

is that a coincidence?

im no mathematician, but it seems something occurring more than eighty percent of the time over an 18 year period would classify as a correlation.

80% of the time, it works every time.

[KaibabKat]

National Champions (35) since 1985 have the following medians:

33 wins
5 losses
4 Preseason AP Ranking
3 Final AP Ranking
3 RPI
1 Seed

[PennZona20]

It's more about college basketball being the least predictable sport

Which is why Vegas LOVES it.

Actually there are plenty of algorithms out there that consistently beat cbb lines. Now granted it’s mostly due to pure volume of lines w 360 or whatever D1 teams and most of the algorithms attack non power conference totals but cbb totals can get killed because there are a lot of weak lines out there if you do your homework diligently enough.