Untitled Document
Brown Columbia Cornell Dartmouth Harvard Penn Princeton Yale



Username Post: Pomeroy        (Topic#20750)
SomeGuy 
Professor
Posts: 6391

Reg: 11-22-04
11-27-17 10:52 PM - Post#238165    

This may not last, with Harvard's whole team presumably recovering from a nasty illness, and Nova looming for Penn, but . . . for the first time in a long time, Penn is ahead of Harvard in Pomeroy.

And that's with the preseason stuff baked in. If we could isolate what has happened in 2017-18 thus far, I assume Penn would be further ahead (though in a much smaller sample size).

 
mbaprof 
Senior
Posts: 342

Age: 66
Reg: 12-24-11
11-29-17 07:41 AM - Post#238261    
    In response to SomeGuy

In case you want to see the whole list
https://kenpom.com/
note you need a paid sub to search, but from your browser bar search works on the page
You can also save to PDF and do whatever you want to do with it
Interestingly they have Harvard and Monmouth next to one another


 
Mike Porter 
Postdoc
Posts: 3615
Mike Porter
Loc: Los Angeles, CA
Reg: 11-21-04
12-03-17 03:20 AM - Post#238863    
    In response to mbaprof

Now into second in Kenpom about 21 spots below Yale and 3 spots ahead of Princeton after their respective rough nights.

 
SomeGuy 
Professor
Posts: 6391

Reg: 11-22-04
12-03-17 11:47 AM - Post#238883    
    In response to Mike Porter

Tough night for the league. Amazing that Penn is 2nd in kenpom. And within the margin of error where winning the league regular season looks perfectly reasonable. And, as I said when we passed Harvard, even more remarkable that the baked in info from last year is still propping the other guys up -- I assume that if you isolate this season's numbers, we are more significantly ahead of Princeton and Harvard, and quite possibly ahead of Yale. Still lots of noise in e performances so far, and lots of excuses for HYP, but I'm encouraged.

 
penn nation 
Professor
Posts: 21086

Reg: 12-02-04
12-03-17 11:56 AM - Post#238884    
    In response to SomeGuy

It just means that the other Ivy teams haven't (yet) performed up to expectations. And if we're 141 in Pomeroy and rank 2nd among the Ivies, this is a down year.

We only have one bad loss--the Fairfield game.

 
Mike Porter 
Postdoc
Posts: 3615
Mike Porter
Loc: Los Angeles, CA
Reg: 11-21-04
12-04-17 03:14 AM - Post#238934    
    In response to penn nation

Oh yeah, I was just highlighting where things stood. To be clear, we are basically where we were expected to be... other folks have come down to our level for various reasons. I'd love to see us around 120 end of season and then let's see what happens in tournament.

 
mbaprof 
Senior
Posts: 342

Age: 66
Reg: 12-24-11
12-04-17 08:40 AM - Post#238936    
    In response to Mike Porter

As of this am,
Yale 121
Penn 141 (but behind monmouth)
Princeton 144
Harvard 157
Columbia 182

NOVA #1!
how does kenpom use road losses? I thought Harvard had a pretty good loss to #13 Kentucky sat

 
mrjames 
Professor
Posts: 6062

Loc: Montclair, NJ
Reg: 11-21-04
12-04-17 09:02 AM - Post#238937    
    In response to mbaprof

The way that the Pomeroy rankings move is essentially as follows:

Take a team's efficiency margin (pts/poss) in a game on offense and defense. Take the opponent's season-long efficiency margins, adjusted for the site of the game (+~5pts/100 poss for home games, opposite for road games). That yields the adjusted efficiency margin for a team in a game. Sum those up across the entire season and you generally get a team's adjusted offensive efficiency and defensive efficiency.

Wins and losses, in and of themselves, do not matter, KP is a margin-based system, since margin is far more descriptive of a team's true ability than simple Ws and Ls.

There are some more nuances that just that. KP uses preseason weighting to keep the numbers from bouncing around crazily to start the season - those preseason weights diminish throughout the year and are pretty negligible by late December. He also uses some recency weighting in the calculation.

Most of all, though, it's just adjusted efficiency margins driving the system.

 
section110 
Masters Student
Posts: 847

Loc: south jersey
Reg: 11-22-04
12-04-17 03:06 PM - Post#238993    
    In response to mrjames

Does the calculation have any safety valve for meaningless end of clear victory scoring--i.e, running up the score? If not it helps explain why Jay Wright doesn't go to his deep bench until the very last minute in 25 to 45 point wins or how a 10-4 team with a victory margin of 15 to 20 points over clearly inferior opponents gets rated equal to or above a 14-0 team with a 10-15 point margin of victory with the starters watching the last 4 or so minutes.

 
mrjames 
Professor
Posts: 6062

Loc: Montclair, NJ
Reg: 11-21-04
12-04-17 04:34 PM - Post#238999    
    In response to section110

Despite the fact that higher margin is more predictive, KP has put a governor on the margin, such that you stop getting “credit” at a certain point. One thing that I use here is game script, which is the average lead held at any time. I find that to be quite predictive and it controls for expanding your lead wildly at the end (or letting a team chip away at a huge lead at the end).

 
penn nation 
Professor
Posts: 21086

Reg: 12-02-04
Pomeroy
12-04-17 04:36 PM - Post#239000    
    In response to mrjames

So technically this is more like a median, then, as opposed to an arithmetic average or mean.

Edited by penn nation on 12-04-17 04:37 PM. Reason for edit: No reason given.

 
mrjames 
Professor
Posts: 6062

Loc: Montclair, NJ
Reg: 11-21-04
12-04-17 05:00 PM - Post#239003    
    In response to penn nation

It’s a mean, but best to think of it as the notion “area under the curve” from calculus. Like, if I add up the margin at every second in a game and then divide by the 2400 seconds in a game, what is that mean margin - that’s the game script.

 
section110 
Masters Student
Posts: 847

Loc: south jersey
Reg: 11-22-04
12-04-17 05:32 PM - Post#239005    
    In response to mrjames

Addresses the problem. Guess Jay just likes to roll up the score.

 
penn nation 
Professor
Posts: 21086

Reg: 12-02-04
12-04-17 05:59 PM - Post#239007    
    In response to mrjames

Got it. Thanks.

 
Old Bear 
Postdoc
Posts: 3988

Reg: 11-23-04
12-04-17 08:05 PM - Post#239017    
    In response to penn nation

I have rarely heard a team fan complain about running up the score. Nor have I heard of a winning team's fan bitch about the refs. Unfortunately, I have been guilty of bitching about both.

 
SomeGuy 
Professor
Posts: 6391

Reg: 11-22-04
12-09-17 07:02 PM - Post#239849    
    In response to mbaprof

Up to 131 after today's win. That's the highest we've been since 2007. Still trailing Yale, but putting some space between us and Harvard/Princeton.

Go Quakers!

 
Icon Legend Permissions Topic Options
Report Post

Quote Post

Quick Reply

Print Topic

Email Topic

1718 Views





Copyright © 2004-2012 Basketball U. Terms of Use for our Site and Privacy Policy are applicable to you. All rights reserved.
Basketball U. and its subsidiaries are not affiliated in any way with any NCAA athletic conference or member institution.
FusionBB™ Version 2.1 | ©2003-2007 InteractivePHP, Inc.
Execution time: 0.244 seconds.   Total Queries: 16   Zlib Compression is on.
All times are (GMT -0500) Eastern. Current time is 06:28 AM
Top