ADVERTISEMENT

Top 3 in each conference

Again, merely an example, that's certainly not the only thing I base my own opinions of teams on. Not even close. It's simply one of the only "real" things we can base teams on. Teams from all conferences actually playing each other on the court. The results are definitive, not determined by metrics.

So why not use OOC results, or OOC results against other major conferences? That would be a much bigger sample size than a random one and done tourney.
 
  • Like
Reactions: KisteK
What advance metrics cannot incorporate into an analytical formula is the true quality of a team or opponent. It can factor wins and losses, opponents wins and losses, etc, etc.... but it can't rate quality. Team A that is 20-0 against teams in a low tier conference that only plays teams in said conference would look better analytically than team B that is 15-5 but has played obviously stonger opponents.

That is absolutely false.
 
SOS is based 100% exclusively on wins and losses, period. Nothing else. 2/3 is opponent's record and 1/3 is their opponent's record.

Dude...what in the hell are you talking about? Seriously? I'm not trying to be rude, that's just so inaccurate it's nuts.

It sounds like you're referring exclusively to the RPI--which isn't even an "advanced metric." It's a crappy rating system that can be easily manipulated and even the creator says shouldn't be used to determine NCAAT teams.
 
Advanced metrics say this:
Team A: 20-0 in the Greater Southeastern Alaskan Peninsula League, whose player of the year last season was a 5'3" legally blind one-legged Eskimo. Opponent's combined record: 220-43. Opponent's opponents' combined record: 854-317

Team B: 20-0 in the ACC. Opponent's combined record: 220-43. Opponent's opponents' combined record: 854-317.

Advanced metrics say these teams are equally as good. So, based on that we should expect a great game between 2017 UNC and the Lower Talkeetna College Fightin' Seal Clubbers.

100% incorrect.
 
I'm not either, then again I haven't ever delved that deep into it.

Okay that explains a lot. You have zero clue what you're talking about.

At least I know what the heck is going on now.

KenPom uses efficiency and adjusts for tempo, strength of the opponents D, strength of the opponents O, and the four factors. It doesn't even care if you win or lose. You could win while playing poorly against a weak team and go down. You could lose while playing well against a great team and go up.
 
Okay that explains a lot. You have zero clue what you're talking about.

At least I know what the heck is going on now.

KenPom uses efficiency and adjusts for tempo, strength of the opponents D, strength of the opponents O, and the four factors. It doesn't even care if you win or lose. You could win while playing poorly against a weak team and go down. You could lose while playing well against a great team and go up.
I wasn't really talking about KenPom, but my point does not change. That is fine, KenPom can factor all those things all he wants. What I am saying is that those factors will be altered by the quality of opponent you are playing!!! High major teams are playing high major teams all the time. Low tier teams are playing low tier teams all the time. That is what is not and can not be factored into analytics. Do you think Duke is going to have better offensive and defensive efficiency numbers playing in the ACC or in the Southland Conference?
 
Last edited:
I wasn't really talking about KenPom, but my point does not change. That is fine, KenPom can factor all those things all he wants. What I am saying is that those factors will be altered by the quality of opponent you are playing!!! High major teams are playing high major teams all the time. Low tier teams are playing low tier teams all the time. That is what is not and can not be factored into analytics. It simply can't.

No, that IS what's factored in. It's not factored into the RPI because the RPI is a freaking joke. But it IS factored into kenpom.

That's why 9-1 Georgetown is 102nd and 7-4 UF is 30th.
 
No, that IS what's factored in. It's not factored into the RPI because the RPI is a freaking joke. But it IS factored into kenpom.

That's why 9-1 Georgetown is 102nd and 7-4 UF is 30th.
Let me put it this way, do you think Duke would put up better offensive and defensive efficiency numbers playing in the ACC or playing in the Southland Conference?
 
Let me put it this way, do you think Duke would put up better offensive and defensive efficiency numbers playing in the ACC or playing in the Southland Conference?

Their adjusted efficiency numbers would likely be the same as whatever they are now.

Duke's adjusted efficiency against Southern was 112.5, their adjusted efficiency against Michigan State was 117.9.
 
Their adjusted efficiency numbers would likely be the same as whatever they are now.

Duke's adjusted efficiency against Southern was 112.5, their adjusted efficiency against Michigan State was 117.9.
I guess I am not understanding what this adjustment is based on then.
 
"The adjusted efficiency margin shows the number of points a team would be expected to outscore the average Division-I team over 100 possessions without adjusting for location of the game."

That certainly doesn't sound very precise to me. What exactly is "the average Division-I team" and who gets to set that metric? This is what I am talking about....
 
Okay that explains a lot. You have zero clue what you're talking about.

At least I know what the heck is going on now.

KenPom uses efficiency and adjusts for tempo, strength of the opponents D, strength of the opponents O, and the four factors. It doesn't even care if you win or lose. You could win while playing poorly against a weak team and go down. You could lose while playing well against a great team and go up.

And what if the bench is cleared with 10 min left in the game and a 25 point lead is whittled down to single digits? What does it do then?

Or how does it account for injuries or absent players?
 
Last edited:
  • Like
Reactions: Kevin Bryan
"The adjusted efficiency margin shows the number of points a team would be expected to outscore the average Division-I team over 100 possessions without adjusting for location of the game."

That certainly doesn't sound very precise to me. What exactly is "the average Division-I team"? This is what I am talking about....

The average division 1 team is the mathematical mean of all 351 teams' efficiencies. It's literally the average team. So it would change every year and even within the year.

But of course they aren't playing the average team. So for each individual game it's adjusted based on their actual performance vs the predicted performance against that quality of opponent. This is done for both offense (against that specific opponents' defense) and for defense (against that specific opponents' offense).
 
Someone other than KU mightvtake the Big 12 this year


And what if the bench is cleared with 10 min left in the game and a 25 point lead is whittled down to single digits? What does it do then?

Or how does it account for injuries or absent players?

KenPom doesn't account for injuries in a single game. (Though I believe BPI does...)

But over time, if a key guy misses a bunch of games, that team's performance would likely suffer which would in turn make them less efficient.
 
  • Like
Reactions: Boilermaker03
I wasn't really talking about KenPom, but my point does not change. That is fine, KenPom can factor all those things all he wants. What I am saying is that those factors will be altered by the quality of opponent you are playing!!! High major teams are playing high major teams all the time. Low tier teams are playing low tier teams all the time. That is what is not and can not be factored into analytics. Do you think Duke is going to have better offensive and defensive efficiency numbers playing in the ACC or in the Southland Conference?

Why is it so hard for you to understabd SOS is huge in advanced metrics like KenPom?

You keep saying all of these things that are completely false as if youre sure of what youre talking about.
 
The average division 1 team is the mathematical mean of all 351 teams' efficiencies. It's literally the average team. So it would change every year and even within the year.

But of course they aren't playing the average team. So for each individual game it's adjusted based on their actual performance vs the predicted performance against that quality of opponent. This is done for both offense (against that specific opponents' defense) and for defense (against that specific opponents' offense).
I understand all that and it makes sense, but still isn't a legitimate estimation of the true quality of a team, IMO. Like I said, over the span of a season a team like Duke is developing their efficiency ratings based on playing far more superior and talented teams than Stephen F Austin is when they are developing their rating, for example. Basically, I am saying if Stephen F Austin is playing against Duke, VT, ND (all top 10 in offensive efficiency) and the rest of the ACC as opposed to playing in the Southland Conference where (I shit you not) 6 of the 13 teams are below #300 in offensive efficiency obviously their defensive efficiency rating would be much, much lower than it is playing against those scrubs all the time.
 
Why is it so hard for you to understabd SOS is huge in advanced metrics like KenPom?

You keep saying all of these things that are completely false as if youre sure of what youre talking about.
Because SOS is based solely on W-L records, literally...that's it. Does not matter who you beat or lost to, or who your opponents beat or lost to, etc. etc. You could've beaten the sisters of the blind 20 times. Those 20 wins count the same as knocking off the #1 team in the nation 20 times.
 
Because SOS is basely solely on W-L records, literally...that's it. Does not matter who you beat or lost to, or who your opponents beat or lost to, etc. etc. You could've beaten the sisters of the blind 20 times. Those 20 wins count the same as knocking off the #1 team in the nation 20 times.

Jfc this isnt even true. Im like 99% sure this isnt true for KenPom or Sagarin.

Why cant you just admit you dont have the slightest clue what youre talking about rather than making things up to continue this absurd argument?
 
Come on, dude. I never said freakin' Wofford is a better team than UNC (although obviously they were last night). However, when half your conference gets bounced from a single-elimination tourney in the 1st weekend it is indicative that maybe they were overrated. I don't see why that seems so far-fetched to some people. You talk about the "entire body of work". Well, 65% or so of that body of work for the entire season is against the conference in question.....

Maybe that's an indication of how tough that conference is? I just can't buy what you're selling here. I feel like the full body of work trumps a team getting bounced in a 1 game elimination tourney. There are just too many factors that can play into why a team or teams from a particular conference can get beaten on any given night.
 
Last edited:
  • Like
Reactions: KisteK
Jfc this isnt even true. Im like 99% sure this isnt true for KenPom or Sagarin.

Why cant you just admit you dont have the slightest clue what youre talking about rather than making things up to continue this absurd argument?
Why you mad that Purdue isn't the 6th best team in the country? They are still ok.
 
Maybe that's an indication of how tough that conference is? I just can't buy what you're selling here. I feel like the full body of work trumps a team getting bounced in a 1 game elimination tourney. Their just too many factors that can play into why a team or teams from a particular conference can get beaten on any given night.
Exactly! It could be an indication of how tough that conference is, but it could also be an indication that it isn't as tough as it is perceived to be. How do you know when they are playing each other 65% of the time? All you can go by are the regular season non-conference games and NCAA tourney non-conference games.
 
Exactly! It could be an indication of how tough that conference is, but it could also be an indication that it isn't as tough as it is perceived to be. How do you know when they are playing each other 65% of the time? All you can go by are the regular season non-conference games and NCAA tourney non-conference games.

While I can agree with part of this, my point remains. Plus I think most of us who know basketball go by the proverbial "eye test". Advanced metrics, RPI, BPI etc. are good measuring sticks and most of the time it aligns with what we see. But it is definitely not the be-all end-all.
 
  • Like
Reactions: Kevin Bryan
While I can agree with part of this, my point remains. Plus I think most of us who know basketball go by the proverbial "eye test". Advanced metrics, RPI, BPI etc. are good measuring sticks and most of the time it aligns with what we see. But it is definitely not the be-all end-all.
Yes, exactly. I am with you on that. Advanced metrics have their place, but basketball isn't an algorithm. There are way, way, waaaaaaayyyyy too many variables that you can't just plug into some mathematical equation. It's freakin' basketball, people. GET A GRIP!
giphy.gif
 
Because SOS is based solely on W-L records, literally...that's it. Does not matter who you beat or lost to, or who your opponents beat or lost to, etc. etc. You could've beaten the sisters of the blind 20 times. Those 20 wins count the same as knocking off the #1 team in the nation 20 times.

NO IT IS NOT.

Good lord, man. How are you just ignoring what is true and repeating the same falsity over and over again? It's like arguing with a flat earther.
 
  • Like
Reactions: Mgkcbb
NO IT IS NOT.

Good lord, man. How are you just ignoring what is true and repeating the same falsity over and over again? It's like arguing with a flat earther.
Good lord, man! I am not talking about Ken Pom or Sagarin or any other dude you guys suck off. I am talking about the regular SOS, the basis of the RPI that so many people took as the gospel before Pomeroy and Sagarin sucked you all into their cults. Those two incorporate their adjusted efficiency numbers into the SOS, big deal! It still is not an accurate depiction of the true quality of a team, ie how good they really are. THEY ARE JUST NUMBERS!
 
Yes, exactly. I am with you on that. Advanced metrics have their place, but basketball isn't an algorithm. There are way, way, waaaaaaayyyyy too many variables that you can't just plug into some mathematical equation. It's freakin' basketball, people. GET A GRIP!
giphy.gif

There is absolutely a place for both the advanced metrics and the eye test (now, there is no place for the piece of garbage that is the RPI, but I digress...). Anyone who relies on only one thing is likely missing out on some valuable piece of info.

However, your continued descriptions of how advanced metrics work are stunningly off base.
 
  • Like
Reactions: Boilermaker03
Good lord, man! I am not talking about Ken Pom or Sagarin or any other dude you guys suck off. I am talking about the regular SOS, the basis of the RPI that so many people took as the gospel before Pomeroy and Sagarin sucked you all into their cults.

The RPI hasn't been cited in a legitimate conversation on hoops, other than by the committee, in years. Honestly. People stopped taking that seriously more than a decade ago.

You keep saying "advanced metrics" when talking about your wins and losses SOS. The RPI isn't an advanced metric. It's not a metric at all. It's an index.
 
There is absolutely a place for both the advanced metrics and the eye test (now, there is no place for the piece of garbage that is the RPI, but I digress...). Anyone who relies on only one thing is likely missing out on some valuable piece of info.

However, your continued descriptions of how advanced metrics work are stunningly off base.
Now we slightly agree. But, my descriptions of advanced metrics are not "stunningly off base". You guys just aren't listening to what I am saying. Read my post about Duke-Stephen F Austin again.
 
The RPI hasn't been cited in a legitimate conversation on hoops, other than by the committee, in years. Honestly. People stopped taking that seriously more than a decade ago.

You keep saying "advanced metrics" when talking about your wins and losses SOS. The RPI isn't an advanced metric. It's not a metric at all. It's an index.
I am talking about offensive and defensive efficiency and "adjusted" offensive and defensive efficiency. The RPI has certainly been used by the selection committee much more recently than a decade ago too, as worthless as it is.
 
Good lord, man! I am not talking about Ken Pom or Sagarin or any other dude you guys suck off. I am talking about the regular SOS, the basis of the RPI that so many people took as the gospel before Pomeroy and Sagarin sucked you all into their cults. Those two incorporate their adjusted efficiency numbers into the SOS, big deal! It still is not an accurate depiction of the true quality of a team, ie how good they really are. THEY ARE JUST NUMBERS!

Youre either next level dumb or an absolute troll. Youre polarized to those two possibilities.
 
Youre either next level dumb or an absolute troll. Youre polarized to those two possibilities.
Sagarin has Purdue at 4. KenPom has them at 5. I guess the real question here is, when are you going to kick KenPom's arse for trolling your boys with his blasphemous algorithm that is obviously out of whack?
 
I am talking about offensive and defensive efficiency and "adjusted" offensive and defensive efficiency. The RPI has certainly been used by the selection committee much more recently than a decade ago too, as worthless as it is.

Literally two posts ago you said you were only talking about the RPI. The RPI doesn't have any kind of efficiency built in.

So you're either changing your story or you're WAY WAY WAY off about how real advanced metrics work.
 
Literally two posts ago you said you were only talking about the RPI. The RPI doesn't have any kind of efficiency built in.

So you're either changing your story or you're WAY WAY WAY off about how real advanced metrics work.
In regard to SOS I was talking about the "old" formula if you will, what the RPI is based on (2/3 opponent's winning % and 1/3 their opponent's winning %).... only reason I brought up the RPI. I don't know how Sagarin and KenPom calculate their SOS. I assume they use the same basic formula and calculate it based on their efficiency rankings intertwined as well. It's still just about as worthless as I tried to explain in my post about Duke vs. Stephen F Austin re: efficiency ratings.
 
Funny thing about this thread is.....it's all speculation at this point. So however good your arguement is....it's moot right now. Come February, the top 3 in each conference will be defined by actual conference play records. Can we put our little johns back in our pants, and just wait and see what the actual records dictate? It's just too early to know what the best 3 teams in each conference are. I go off conference records. Pretty cut and dry.
 
In regard to SOS I was talking about the "old" formula if you will, what the RPI is based on (2/3 opponent's winning % and 1/3 their opponent's winning %).... only reason I brought up the RPI. I don't know how Sagarin and KenPom calculate their SOS. I assume they use the same basic formula and calculate it based on their efficiency rankings intertwined as well. It's still just about as worthless as I tried to explain in my post about Duke vs. Stephen F Austin re: efficiency ratings.

That's just not true. You can try to explain it all you want--It doesn't make it true.
 
Funny thing about this thread is.....it's all speculation at this point. So however good your arguement is....it's moot right now. Come February, the top 3 in each conference will be defined by actual conference play records. Can we put our little johns back in our pants, and just wait and see what the actual records dictate? It's just too early to know what the best 3 teams in each conference are. I go off conference records. Pretty cut and dry.

NO SPECULATING ON A COLLEGE BASKETBALL MESSAGE BOARD YOU GUYS!!!
 
  • Like
Reactions: KisteK and wesr
ADVERTISEMENT