Untitled Document
Brown Columbia Cornell Dartmouth Harvard Penn Princeton Yale



 Page 3 of 3 ALL« First<123
Username Post: Nolan Cressler        (Topic#12981)
Silver Maple 
Postdoc
Posts: 3765

Loc: Westfield, New Jersey
Reg: 11-23-04
02-09-12 11:06 AM - Post#119844    
    In response to mrjames

Did you read what I wrote? I said nothing whatsoever about your skills.

 
Howard Gensler 
Postdoc
Posts: 4141

Reg: 11-21-04
02-09-12 02:12 PM - Post#119873    
    In response to Silver Maple

Hey Silver Maple, on another thread didn't you tell me to drop it.

While none of my questions regarding Mike's fine work are meant to be taken as a personal affront, his Adjusted Sensitivity Rating on this issue is something like a .9643, which is far greater than it was a year ago when he was more patient with us numbskulls.

My issue is the notion that a league can improve relative to itself (its own quantitative number can be higher from year to year) but yet fall back in the ratings due to the relative strength of other leagues and be called "better." To me, that makes the league worse - maybe because I wasn't a math guy.

Teams do not exist in a vacuum, they exist to compete against other teams. So in my linear brain, if the Ivy League is ranked 21st one year and 26th a few years later, when they're ranked 26th, the League is not as good as it was when it was rankled 21st. I can see how one can make the argument that is not the case (although it's a much more logical argument to make with individual teams whose strength of schedules can change drastically) but over the course of an entire league schedule I would think enough scheduling anomalies would even themselves out that the league ranking would be a truer representation. If I'm wrong, I'm wrong, I'm okay with being ignorant in some areas.



 
Silver Maple 
Postdoc
Posts: 3765

Loc: Westfield, New Jersey
Reg: 11-23-04
02-09-12 02:25 PM - Post#119874    
    In response to Howard Gensler

I think Mike's point there, which I agree with, is that the team ranking measure is simplistic. It's simply not an adequate way of evaluating the strength (relative or absolute) of a system as complex as an eight-team athletic conference.

And you're right. I'm banging my head against the wall. But I'm expecting a message from Mike in 20 years acknowledging that I was correct. I guess it's appropriate that the abbreviation for my screen name is SM.

 
Old Bear 
Postdoc
Posts: 3988

Reg: 11-23-04
02-09-12 05:20 PM - Post#119897    
    In response to Silver Maple

Whatever happened to Nolan Cressler?

 
Jeff2sf 
Postdoc
Posts: 4466

Reg: 11-22-04
02-09-12 05:39 PM - Post#119900    
    In response to Old Bear

Speaking from a baseball point of view, I know exactly how tiring it is to have these conversations. It's like "we've been over this before, I'm right and am backed up by data that's been rigorously analyzed, you're wrong and are letting your eyes fool you".

I admit that I'm a little puzzled by how the pythag can be SOOO much better yet the relative ranking has fallen. It's not very intuitive. Is it zero sum? In other words, if the absolute pythag has risen while relative rank has gone down, does that imply that the tails are much fatter. That conferences 27 and below just suck REALLY bad? Or could, theoretically, every single conference have improved?

 
pchrystie 
Masters Student
Posts: 673

Reg: 03-14-06
02-09-12 10:07 PM - Post#119922    
    In response to mrjames

  • mrjames Said:
It wasn't pchrystie's point that was stupid, it was his example.



So you're saying that if Vegas determined that it could make the most money in a Kentucky-Binghamton game at Kentucky +7, they would set the line at Kentucky -37 just out of some sense of honor in spread setting? I understand that Vegas would never set the line at Kentucky +7, but it has nothing to do with that being the inaccurate prediction; it has to do with the inaccurate prediction losing money.

And Vegas does set lines where an objective observer and/or computer would not set them. While we may not agree on much on this board, I assume we can all agree that Cowboy fans are mor0ns. My understanding is that Vegas sets the line to compensate for their mor0nitude, i.e. not where an objective observer or a computer would tell them to set it.

It's not about accuracy. It's about money. If the two coincide, fine (and they probably usually do). But they don't always.

 
mrjames 
Professor
Posts: 6062

Loc: Montclair, NJ
Reg: 11-21-04
02-09-12 11:16 PM - Post#119924    
    In response to pchrystie

No, no - you're point is valid, but it's just the magnitude that's off. The New York effect and, sure, the Cowboys effect exists and can cause spreads to have to be adjusted off of actuarially fair number. We can talk about the inefficiencies of pricing the college game - HCA being the biggest among them - but this fan base adjustment doesn't really prove out at the college level, when you look at the data. And the adjustment isn't 40 points, and it isn't even four points. It's a point here or there to entice people to the other side.

So, the difference between accuracy and money is often zero, but even when it exists it is tiny. If it ever got more than a couple points, arb money would flow in and push the spread back.

As for the Pythag issue... It is zero sum. But if there's more parity, as you suggest, then you're going to have a tight middle and ranks can vary wildly with small changes in Pythag. If there's less parity, the middle might be wide open so it takes very large changes in Pythag to move in conference rank. That's the difference between 2011 and 2012. It's also why I like looking at Pythag and not rank. Pythag is based on how my conference has performed on an SOS-adjusted basis. Rank depends on things like whether there is parity or not. I might want to set a target of being a .500 conference. That could be a No. 15 conference, a No. 12 conference, a No. 17 conference - but that just depends on the year.

If anyone wants to talk to me in private about my research, I'm more than happy to do so. My e-mail is mrjames2006@gmail.com. People that have bothered to ask me questions that way have found out that I'm aware of the limitations of my analysis. It's just hard to discuss anything openly here, because the tone is far too hostile for intellectual honesty.

 
 Page 3 of 3 ALL« First<123
Icon Legend Permissions Topic Options
Report Post

Quote Post

Quick Reply

Print Topic

Email Topic

56530 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: 0.231 seconds.   Total Queries: 16   Zlib Compression is on.
All times are (GMT -0500) Eastern. Current time is 05:59 PM
Top