• Welcome to OGBoards 10.0, keep in mind that we will be making LOTS of changes to smooth out the experience here and make it as close as possible functionally to the old software, but feel free to drop suggestions or requests in the Tech Support subforum!

Wake opens +9

RJKarl said:
I can't stop laughing that first someone would have the balls to say, "Hey I've got a formula to figure how lucky your team has been, is and will be for the rest of the season."

That's batshit crazy.

What's even crazier is to believe someone who is saying such nonsense.
Click to expand...

It's a pretty good metric actually. It has been statistically proven that Kenpom efficiencies are a better future predictor than a tool like RPI that looks at wins and losses.

Luck might be a bad term...it is basically just how different your record is than what your offensive and defensive outputs say that it should be.
 
GW (Lonergan) is smoking LaSalle, which is good for those who have to wait until this game is over before Wake comes on.
 
VaDeac said:
One of the big KenPom guys on here (sorry whomever, can't remember who) explained this to me recently and in doing so admitted that "Luck" is a really poor name for the differential between expectation and actual occurrence, which usually comes about by winning close games.
Winning the close games you're in is undoubtedly a skill or set of skills that is overwhelmingly positive and anything but luck. The term "luck" really does teams a disservice IMO because it sounds flukey and dismissive.
Click to expand...

It's probably a combination of skill and luck, but mostly luck. That is, treating winning close games as pure luck makes the model a better predictor of future games than otherwise. Or, winning close games doesn't appear to be a repeatable skill in most cases.

Kenpom's ORTG (not a substitute for metrics like PER) is useful when used in concert with usage rate, but isn't very good as a one-off metric. People like numbers really need to stop using it as a metric that measures overall offensive performance/value.
 
Last edited:
bobknightfan said:
Yeah, we beat a #13 & #12 team by 1 & 4 points. Then the next year we blew a 13-point lead and lost to a #7 seed in the 2nd round....after having to overcome an 11-point deficit to beat a #15 seed in the first round.

And talking about rebuilding... Only one coach in ACC history has ever had a team that was ranked #1 in the country one year, then finished dead last in the conference the next year.
Click to expand...

Kryzasd;fljski 93-95?
 
IAppreciateIt said:
It's a combination of skill and luck, but mostly luck. That is, treating winning close games as pure luck makes the model a better predictor of future games than otherwise.
Click to expand...

I get that... just think it shouldn't be called luck. Call it "Differential" or something.

If I were cynical, I might accuse Ken of being a bit self-serving by dismissing the differential between his expectations and actual game outcomes as sheer luck. But I'm not, so...
 
RJKarl said:
I didn't say better offensive player. I said better offensive numbers.

Carson also had better numbers in spite of scoring and boarding less per minute; having a much lower FG%. There is no rational way to create a formula that would make Carson be more efficient than Devin this year.
Click to expand...

They score the same number of points per possession they use. Carson turns it over less and shoots better FTs. But Carson never shoots and is a clearly a weak offensive player.

But yeah, Devin is 10X the offensive player Carson is which could be seen by a 2 second glance and their advanced stats.
 
If you really want to use a TO stat per possession, you should really count touches not just the times a player throws it away or commits another TO.

The fact that KenPom postulates Devin is rated lower shows a gross problem with his entire system.
 
RJKarl said:
If you really want to use a TO stat per possession, you should really count touches not just the times a player throws it away or commits another TO.

The fact that KenPom postulates Devin is rated lower shows a gross problem with his entire system.
Click to expand...

It doesn't postulate that at all. The efficiency metric used alone doesn't rate players in any way whatsoever.

In fact, CJ Fair in tonight's game is ranked as the 6th best player in college basketball per Kenpom's purely statistical model. But his offensive efficiency is only 102 (remember Tony C is 110!!)
 
RJKarl said:
Which shows the uselessness of that metric.
Click to expand...

Just one one piece of the puzzle...Although if you have a player that is high efficiency and high usage then you can be pretty damn sure that's an excellent offensive player.
 
You must log in or register to reply here.
Back
Top