Synergy

[chum1]

high kenpom ratings make me very happy

That's not a good test case because you can't have them independent of winning.

you can't win if both your offensive and defensive efficiencies suck.
 

Call it what you want, it's still the same arbitrary event.

[michigancat]

high kenpom ratings make me very happy

That's not a good test case because you can't have them independent of winning.

you can't win if both your offensive and defensive efficiencies suck.
 

Call it what you want, it's still the same arbitrary event.

What is "it"?

[chum1]

high kenpom ratings make me very happy

That's not a good test case because you can't have them independent of winning.

you can't win if both your offensive and defensive efficiencies suck.
 

Call it what you want, it's still the same arbitrary event.

What is "it"?

What I would call "winning." Here's my question:  Isn't being very happy with a high kenpom ranking ultimately the same is being very happy with lots of wins?  Is it possible for you to be happy with a high ranking in a stat based ranking system that had no correlation with wins whatsoever?

[michigancat]

What I would call "winning." Here's my question:  Isn't being very happy with a high kenpom ranking ultimately the same is being very happy with lots of wins?

No, they aren't perfectly analogous. Is your enjoyment of wins over the SWAC equal to your enjoyment of wins over teams like Florida?

Is it possible for you to be happy with a high ranking in a stat based ranking system that had no correlation with wins whatsoever?

well, no. Because it definitely wouldn't use extremely useful and always superior possession-based stats.

[chum1]

What I would call "winning." Here's my question:  Isn't being very happy with a high kenpom ranking ultimately the same is being very happy with lots of wins?

No, they aren't perfectly analogous. Is your enjoyment of wins over the SWAC equal to your enjoyment of wins over teams like Florida?

Is it possible for you to be happy with a high ranking in a stat based ranking system that had no correlation with wins whatsoever?

well, no. Because it definitely wouldn't use extremely useful and always superior possession-based stats.

Useful for what?

[michigancat]

Useful for what?

comparing how well teams score
comparing how well teams prevent scoring
combining the two to predict how well a team will perform against another team

Lots of other things, too. What would you like to do?

[nicname]

i wanna coach right now
i'm BW and i came to get down
i'm not internationally known
but i'm known to go and recruit alone
cause i fly a lot and i mean a lot
stay away from me if you don't have a good shot
cause i've been a winna and sometimes a losea
head coach of da cats is what i choosea

nicname thoroughly enjoyed this rickdaris post.

[chum1]

Useful for what?

comparing how well teams score
comparing how well teams prevent scoring
combining the two to predict how well a team will perform against another team

Lots of other things, too. What would you like to do?

So, knowing those things is an end in itself, having meaning independent of wins?  That's a new concept to me and I'm still skeptical of it.  It no doubt originated in an attempt to predict wins.  Even if you're just talking about which team is better, it seems nearly everyone would be thinking about which team would win more in one way or another.

[michigancat]

Useful for what?

comparing how well teams score
comparing how well teams prevent scoring
combining the two to predict how well a team will perform against another team

Lots of other things, too. What would you like to do?

So, knowing those things is an end in itself, having meaning independent of wins?  That's a new concept to me and I'm still skeptical of it.  It no doubt originated in an attempt to predict wins.  Even if you're just talking about which team is better, it seems nearly everyone would be thinking about which team would win more in one way or another.

They definitely exist (in part) to compare the strength of teams and help predict wins and losses. And they are always superior to per-game stats for predicting wins and losses.

[chum1]

Useful for what?

comparing how well teams score
comparing how well teams prevent scoring
combining the two to predict how well a team will perform against another team

Lots of other things, too. What would you like to do?

So, knowing those things is an end in itself, having meaning independent of wins?  That's a new concept to me and I'm still skeptical of it.  It no doubt originated in an attempt to predict wins.  Even if you're just talking about which team is better, it seems nearly everyone would be thinking about which team would win more in one way or another.

They definitely exist (in part) to compare the strength of teams and help predict wins and losses. And they are always superior to per-game stats for predicting wins and losses.

Well, for one, you don't need to adjust for tempo when tempos are similar enough.  I can also imagine cases in which margin of victory is as successful a predictor as kenpom.  My point wasn't about the best system in general.  It was about the dogmatic insistence that everyone must use that system in every situation and they're stupid if they don't.

[kso_FAN]

It was about the dogmatic insistence that everyone must use that system in every situation and they're stupid if they don't.

I don't believe anyone said this in this thread.

I will admit I may imply it sometimes (I did say I hate regular stats, and most of the time I do), but it was never explicit in this thread.

[chum1]

It was about the dogmatic insistence that everyone must use that system in every situation and they're stupid if they don't.

I don't believe anyone said this in this thread.

I will admit I may imply it sometimes (I did say I hate regular stats, and most of the time I do), but it was never explicit in this thread.

Someone did.  It happens a lot, though, and not just from that person.  The assumption made very often here is that conclusions drawn from stats are completely invalid if the stats aren't adjusted for tempo.  That's a false assumption.

[CHONGS]

nothing means anything guys

especially if it is basic statistics stuff I don't understand

quit talking about anything guys

[chum1]

especially if it is basic statistics stuff I don't understand

Ooooh!  Please explain statistics stuff to me!  Do it!!!

[CHONGS]

especially if it is basic statistics stuff I don't understand

Ooooh!  Please explain statistics stuff to me!  Do it!!!

gather ye rosebuds while ye may

[felix rex]

No stat means itself. Widening gyre. Infinite play.

[kougar24]

[michigancat]

Well, for one, you don't need to adjust for tempo when tempos are similar enough.  I can also imagine cases in which margin of victory is as successful a predictor as kenpom.  My point wasn't about the best system in general.  It was about the dogmatic insistence that everyone must use that system in every situation and they're stupid if they don't.

they are also often ignorant. But when it's a sportswriter, I'm pretty much okay with thinking they're stupid. I'll try to make my tone nicer next time so you're not offended.