Ken-Pom Silliness | Syracusefan.com

Ken-Pom Silliness

orange2win

orange2win

2nd String
Joined
Aug 30, 2011
Messages
920
Like
540
1. Arizona 18-0
2. Villanova 16-1
3. Syracuse 18-0

If Nova has one loss and we beat them ...

It is why our own observations and common sense are still superior to these computer geeks, their algorithms and game leverage calcs. No doubt the computer numbers are correct; but don't always yield the best team.
 
Not only did we beat Nova, we spanked them. After starting out red-hot shooting 3's to start the game, Nova then proceeded to get steamrolled on a 71-37 run the rest of the way by Syracuse.

We won the game by 16, after being down by 18. That's amazing.

#3? ...

riiiiiight.jpg
 
From Doctor Po-Po himself:

The first thing you should know about this system is that it is designed to be purely predictive. If you’re looking for a system that rates teams on how “good” their season has been, you’ve come to the wrong place. There are enough systems out there that rank teams based on what is “good” by just about any definition you can think of. So I’d encourage you to google college basketball ratings or even try the opinion polls for something that is more your style.

Ken Pomeroy Ratings Explaination
 
SUMBA said:
From Doctor Po-Po himself:

The first thing you should know about this system is that it is designed to be purely predictive. If you’re looking for a system that rates teams on how “good” their season has been, you’ve come to the wrong place. There are enough systems out there that rank teams based on what is “good” by just about any definition you can think of. So I’d encourage you to google college basketball ratings or even try the opinion polls for something that is more your style.

Ken Pomeroy Ratings Explaination
Expand...

Slippery words. "Purely predictive" Hmmm. He's predicting that Nova is stronger than SU as of today? Recent results show to the contrary.
 
orange2win said:
1. Arizona 18-0
2. Villanova 16-1
3. Syracuse 18-0

If Nova has one loss and we beat them ...

It is why our own observations and common sense are still superior to these computer geeks, their algorithms and game leverage calcs. No doubt the computer numbers are correct; but don't always yield the best team.
Expand...



I want to bash my head on the table when I see posts like this. Are you trying to say the human polls do yield the best team?

And I can see the point of criticism if it ranks Cuse 10th and Nova 2nd, but its one spot. Its like when people had Ohio St and Wisc ranked 1 whole spot ahead of us. Is it really so outrageous that its worth complaining about?
 
I honestly don't know why people get so bent out of shape about these computer rankings. Pomeroy's algorithms take into account every possession from every game, so of course there are going to be some weird results. As we all know, SU was not all that impressive in some of their games against cupcakes, and I'm sure that hurts them with Pomeroy's system. I personally don't care that SU didn't demolish High Point or Cornell, but a computer does. To any human with a brain, the demolition of Nova vaults SU above them, but Pomeroy's system doesn't work that way. Even Pomeroy has said that his system isn't perfect, and I'm sure that he would vote SU ahead of Nova if he had an AP vote.
 
SUMBA said:
From Doctor Po-Po himself:

The first thing you should know about this system is that it is designed to be purely predictive. If you’re looking for a system that rates teams on how “good” their season has been, you’ve come to the wrong place. There are enough systems out there that rank teams based on what is “good” by just about any definition you can think of. So I’d encourage you to google college basketball ratings or even try the opinion polls for something that is more your style.

Ken Pomeroy Ratings Explaination
Expand...

One thing I know for sure is we have too many fancy schmancy computer stuff nowadays and not enough good ol' fashion"sort by record + head to head = eye test" rankings. Those are always right.
Now get off my lawn!
 
orange2win said:
Slippery words. "Purely predictive" Hmmm. He's predicting that Nova is stronger than SU as of today? Recent results show to the contrary.
Expand...

How do you create a ranking system with a rule that no team can be ranked higher than a team that beat it?

That's like saying Jamie Dixon has to be considered a better coach than Boeheim because he has a winning record head-to-head even though JB is better against everybody else than Dixon.
 
BuffOrange said:
One thing I know for sure is we have too many fancy schmancy computer stuff nowadays and not enough good ol' fashion"sort by record + head to head = eye test" rankings. Those are always right.
Now get off my lawn!
Expand...

lol...like you, and Digger/Dukie V, I also believe the 'eye test' is the most effective way to rank teams.
 
it's just a temporary fluke that will even out as more games are played. Pomeroy ratings are essentially Offensive Efficiency minus Defensive Efficiency.

On the one hand, Syracuse had a terrific defensive game vs. Pitt on Saturday. Pitt came in averaging 119.1 points per 100 offensive possessions; we held them to 101.9 points per 100.

However,although that was a great defensive job vs. a highly efficient team, if you look at our own defensive efficiency in a vacuum, it looks as if we underperformed - we came into the game allowing 91.7 points per 100. That great defensive effort actually hurt our overall defensive efficiency numbers. It raised our defensive rating by 7 tenths of a point, just enough to allow Nova to pass us by 15 ten thousandths of a point.

it's weird because, given the efficiency of the Pitt offense, it was arguably SU's best defensive performance of the year, even though from a pure numbers perspective it was one of the poorest. But as with all good statistical analyses, these flukes will even out as more data are added.
 
I gotta say I am enjoying how how much are getting bent out of shape about this. The two teams are basically identical in the rating. Nova is ahead of us by an indistinguishable amount; there is simply no way the numbers are accurate enough to be significant to the 4th decimal place. We did beat them pretty handily at our place; that's one game out of 18 or whatever it's been. In the other 17, Nova has outperformed us a little, I guess. They have beaten two top 10 teams on a neutral court; we have 2 top 10 wins at home.

I also love when people trash Ken Pom and imply they know better. The Vegas lines are basically always right in line with KP; if you know so much better than KP you should be able to clean up at the books. I've got some money in my betus account; I'll put it up against anyone on here in a picks contest for the rest of the year; I'll take the KP picks and you make your own picks.

However,although that was a great defensive job vs. a highly efficient team, if you look at our own defensive efficiency in a vacuum, it looks as if we underperformed - we came into the game allowing 91.7 points per 100. That great defensive effort actually hurt our overall defensive efficiency numbers. It raised our defensive rating by 7 tenths of a point, just enough to allow Nova to pass us by 15 ten thousandths of a point.
Expand...

The ratings are opponent adjusted, so shouldn't this wash out?

Slippery words. "Purely predictive" Hmmm. He's predicting that Nova is stronger than SU as of today? Recent results show to the contrary.
Expand...

He's predicting that if Nova and SU played on a neutral court tomorrow the game would basically be a pick em.
 
Welcome moqui to the annual kick ken-com string. :) I know you appreciate his work and yes the results get better as the season goes on.

He is into the high-leverage basketball minutes thing and claims we had 21 mins of high-leverage against Minnesota in November and that was our finest moment (written before the Pitt game). As for the Nova game, he says because of the incredibly lopsided runs both ways, only 7 mins in that game were considered high-leverage. This explains how we get less credit in his system for that win.

Some of his stats however baffle me. He says "Syracuse has played the weakest schedule thus far of the undefeated squads." Really??? We have played 3 ranked teams while Wichita State has played none. Must be by his more sophisticated calcs he gets different results.

Anyway, I prefer the Colley Matrix. Check out Colley's roots. Ken-pom seems more geared to the big money in predicting the spread, available only to subscribers. I also see he has quit his day job. Must be doing alright.

http://www.colleyrankings.com/hcurrank.html
 
He is into the high-leverage basketball minutes thing and claims we had 21 mins of high-leverage against Minnesota in November and that was our finest moment (written before the Pitt game). As for the Nova game, he says because of the incredibly lopsided runs both ways, only 7 mins in that game were considered high-leverage. This explains how we get less credit in his system for that win.
Expand...

Got a link for that? I didnt realize he was weighting things by high leverage minutes. I know he had broken the minutes out like that, but I didn't realize they had an impact on how the ratings came out.

Some of his stats however baffle me. He says "Syracuse has played the weakest schedule thus far of the undefeated squads." Really??? We have played 3 ranked teams while Wichita State has played none. Must be by his more sophisticated calcs he gets different results.
Expand...

Well rankings isn't the way he's going to look at things. He's probably going to use his own rankings. Also what is the date on that? Because currently, we have a tougher SOS rating on his website than Wichita. Was that statement made before the Pitt game?
 
Yeah, I am about 99% sure the distribution of high leverage minutes doesn't have anything to do with his ratings.

And the article was from last week; their SOS has passed Wichita with the Pitt game being included.
 
orange2win said:
Correct. It is from a Ken-pom contributor on his site.
Expand...
then why are you attributing the high-leverage argument and the direct quote about the schedule to Pomeroy?
 
moqui said:
then why are you attributing the high-leverage argument and the direct quote about the schedule to Pomeroy?
Expand...

Until you pointed it out, I thought it was written by him.

The quotes should be removed, but he appears to agree with the writer as he posted it without comment.
 
I don't believe leverage is used to compute OE or DE. It is just a measure of output in the in game win probability chart he produces for each game.
 
You must log in or register to reply here.

Similar threads

SWC75
    • Like
History of the Pesci Bowl
Replies
3
Views
419
dollarbill44
dollarbill44
SWC75
Against Ranked Teams - Basketball - after Week 1
Replies
0
Views
416
SWC75
SWC75
SWC75
    • Like
    • Wow
SU Doubleheaders
Replies
1
Views
613
JonB
JonB
SWC75
Against Ranked Teams - Basketball - after week 3
Replies
0
Views
237
SWC75
SWC75
SWC75
Against Ranked Teams - Basketball - after week 6
Replies
0
Views
216
SWC75
SWC75

Forum statistics

Threads
171,265
Messages
4,941,601
Members
6,018
Latest member
CnyTarheel

Share this page

Online statistics

Members online
230
Guests online
1,496
Total visitors
1,726
Totals may include hidden visitors.


...
Top Bottom