Pool C

Started by Pat Coleman, January 20, 2006, 02:35:54 PM

Previous topic - Next topic

0 Members and 2 Guests are viewing this topic.

sac

Quote from: sac on February 11, 2009, 10:58:32 AM
Quote from: Dave "d-mac" McHugh on February 11, 2009, 06:36:29 AM
Well sac... the OWP tells you how many games your opponents won, while the OOWP balances out if a team that is 20-1 is beating nothing but sub-.500 teams while a 16-5 team is battling a tough schedule with maybe half of its opponents at better than .500. So... there is some math that keeps a team with a great record (Chapman) from getting too much credit for beating up an easy schedule.

Team A plays Haverford (10-1) ..................Haverford played Ursinas(20-1)

Team B plays Augustana.(10-1).................Augustana played Washington(16-5)

Guess which one is going to have a higher OOWP Dave?  It really doesn't balance anything out, it just perpetuates the notion that wins is the way to rate a teams strength.

Guess I'll explain it better

Haverford and Augustana have played the exact same schedule, except for the games with Washington/Ursinas and teams A/B.

KnightSlappy

To get the most ideal SOS rating, one may wish to include OOOWP, OOOOWP, and OOOOOWP etc.  Many people feel that OWP and OOWP will get you 'close enough' to true SOS.

If this was included, the games against Wash U and Ursinus may be more accurately reflected.

Hugenerd

Quote from: sac on February 11, 2009, 12:31:21 AM

QuoteI need a valued opinion...

Mr Ypsi and/or others, do you read Strength of Schedule to be the OWP/OOWP or does Strength of Schedule include the numerical rank as well as the raw data from OWP and OOWP?

I don't really consider either a strength of schedule.  But it gets tossed around that way.


Example from last year Wash U's inregion record 16-5, Ursinas' inregion record 22-1.........if you played both you get more 'credit' for playing Ursinas but clearly WashU was the much better team.

OWP and OOWP just tell you how many games your opponents won, nothing more.

Both OWP and OOWP are heavily tied to the conference you are in.  Since most teams have a majority of their games in conference (NESCAC is an exception, they only play 9 +tourney), the OWP and OOWP of that team will approach the average OWP and OOWP of that conference. For example, in the case where all 25 of your games are against conference opponents, your OWP and OOWP will be 0.500.  That is why it is important to play very strong out of conference teams if you want a Pool C, because if you play cupcakes your ceiling is around 0.500 for both those numbers.  It also doesnt help if you are in a conference that plays 18 or more conference games so it is basically impossible to get very high OWP and OOWP numbers. That is also why an OWP or OOWP of 0.500 should not be considered a very good value.  In my opinion, anything over 0.6 is exceptional (A), 0.55 and above is good (B), 0.525 and above is ok (C), 0.50 and above is barely passing (D) and below 0.5 means that your out of conference schedule was, for the most part, filled with bad teams (teams with losing records).

The main issue, for me, with OWP and OOWP, is that the metric relies too heavily on the number of conference games your league has.  You could be in the best conference, but if your schedule has 2/3 conference games, you are essentially weighting your OWP by 2/3*(0.5)+1/3(OWP of out of conference teams).  It is not an easy problem, but it would be nice if they could come up with a metric that is independent of the number of conference games your team plays (and other factors that are mandated by the conference) and only focused on the strength of your league and the out of conference games that your team schedules. If you play in a strong conference, you should be rewarded and if you schedule a tough out of conference schedule, you also should be rewarded.  However, I dont think you should be punished because your larger conference plays double-round robin and therefore 18 of your games are in conference.

Ralph Turner

Quote from: sac on February 11, 2009, 10:58:32 AM
Quote from: Dave "d-mac" McHugh on February 11, 2009, 06:36:29 AM
Well sac... the OWP tells you how many games your opponents won, while the OOWP balances out if a team that is 20-1 is beating nothing but sub-.500 teams while a 16-5 team is battling a tough schedule with maybe half of its opponents at better than .500. So... there is some math that keeps a team with a great record (Chapman) from getting too much credit for beating up an easy schedule.

Team A plays Haverford (10-1) ..................Haverford played Ursinas(20-1)

Team B plays Augustana.(10-1).................Augustana played Washington(16-5)

Guess which one is going to have a higher OOWP Dave?  It really doesn't balance anything out, it just perpetuates the notion that wins is the way to rate a teams strength.
Sac, if you look at the ASC teams, they are all around .500.

McMurry gets into the North Central group with that game.
UT-Dallas gets into the Guilford group with their game.

The best place to look at these OWP/OOWP values are with the final 2007-08 group on these sites, both men and women.

Ralph Turner


Dave 'd-mac' McHugh

Quote from: sac on February 11, 2009, 11:46:16 AM
Quote from: sac on February 11, 2009, 10:58:32 AM
Quote from: Dave "d-mac" McHugh on February 11, 2009, 06:36:29 AM
Well sac... the OWP tells you how many games your opponents won, while the OOWP balances out if a team that is 20-1 is beating nothing but sub-.500 teams while a 16-5 team is battling a tough schedule with maybe half of its opponents at better than .500. So... there is some math that keeps a team with a great record (Chapman) from getting too much credit for beating up an easy schedule.

Team A plays Haverford (10-1) ..................Haverford played Ursinas(20-1)

Team B plays Augustana.(10-1).................Augustana played Washington(16-5)

Guess which one is going to have a higher OOWP Dave?  It really doesn't balance anything out, it just perpetuates the notion that wins is the way to rate a teams strength.

Guess I'll explain it better

Haverford and Augustana have played the exact same schedule, except for the games with Washington/Ursinas and teams A/B.

I guess that is why MANY teams play more than one game outside of conference and usually do NOT have the exact same schedule as other opponents. Heck, if everyone played the exact same schedules, I am sure it would be easier to pick the tourney teams AND do the Top 25!

It comes down to scheduling and best preparing yourself for the season and possible post-season selection. The rules haven't changed.
Host of Hoopsville. USBWA Executive Board member. Broadcast Director for D3sports.com. Broadcaster for NCAA.com & several colleges. PA Announcer for Gophers & Brigade. Follow me on Twitter: @davemchugh or @d3hoopsville.

sac

Quote from: Dave "d-mac" McHugh on February 11, 2009, 12:50:46 PM
Quote from: sac on February 11, 2009, 11:46:16 AM
Quote from: sac on February 11, 2009, 10:58:32 AM
Quote from: Dave "d-mac" McHugh on February 11, 2009, 06:36:29 AM
Well sac... the OWP tells you how many games your opponents won, while the OOWP balances out if a team that is 20-1 is beating nothing but sub-.500 teams while a 16-5 team is battling a tough schedule with maybe half of its opponents at better than .500. So... there is some math that keeps a team with a great record (Chapman) from getting too much credit for beating up an easy schedule.

Team A plays Haverford (10-1) ..................Haverford played Ursinas(20-1)

Team B plays Augustana.(10-1).................Augustana played Washington(16-5)

Guess which one is going to have a higher OOWP Dave?  It really doesn't balance anything out, it just perpetuates the notion that wins is the way to rate a teams strength.

Guess I'll explain it better

Haverford and Augustana have played the exact same schedule, except for the games with Washington/Ursinas and teams A/B.

I guess that is why MANY teams play more than one game outside of conference and usually do NOT have the exact same schedule as other opponents. Heck, if everyone played the exact same schedules, I am sure it would be easier to pick the tourney teams AND do the Top 25!

It comes down to scheduling and best preparing yourself for the season and possible post-season selection. The rules haven't changed.

This is a hypothetical, only to demonstrate OWP, OOWP is not a strength of schedule measure.  I thought I'd use smaller numbers so it would be easier to understand.

I made no arguments about how many games anyone should play.


John Gleich

This system is decidedly better than the previous system, at the very least because it allows teams who are "deserving" to get a bid.  The previous system allowed only the most "deserving" under certain very stringent criteria to make it. 

Ultimately, if a team is relying on a Pool C, then they're not as good as they could be or they haven't done what they needed to do to get a bid.  This is a rather broad statement, I understand, but each conference (that has an auto-bid) has provisions to get a bid to the NCAA tournament.  Now, decidedly, the tournament has come a long way to being a championship where the best teams are competing for the title instead of just being a championship based off of  involvement.  We're pretty darn close to getting a championship where all of the "championship caliber" teams are competing for the championship.  Now, "championship caliber" is a subjective term... and the NCAA tournament selection tries to be as objective as possible. 

In terms of the number of games a conference has, that is set by the conference.  They can change the number if they so desire, so as to better prepare themselves for the NCAA tournament... insomuch as playing more non-con games could prepare you... I think that generally, the conference is the equalizer... but it is certainly true that a team from a stronger conference will likely be more prepared to play in tough games... they just might have a tougher time getting to the NCAA tournament.
UWSP Men's Basketball

National Champions: 2015, 2010, 2005, 2004

NCAA appearances: 2018, '15, '14, '13, '12, '11, '10, '09, '08, '07, '05, '04, '03, '00, 1997

WIAC/WSUC Champs: 2015, '14, '13, '11, '09, '07, '05, '03, '02, '01, '00, 1993, '92, '87, '86, '85, '84, '83, '82, '69, '61, '57, '48, '42, '37, '36, '35, '33, '18

Twitter: @JohnGleich

Greek Tragedy

The MIAC plays 20 of their 25 games in conference...ummm.  Advantage NESCAC?
Pointers
Breed of a Champion
2004, 2005, 2010 and 2015 National Champions

Fantasy Leagues Commissioner

TGHIJGSTO!!!

Ralph Turner

Quote from: Old School.... (Tom Doebler) on February 11, 2009, 02:58:12 PM
The MIAC plays 20 of their 25 games in conference...ummm.  Advantage NESCAC?
The ASC plays 20/21 (East Division/West Division) games in-conference!

Pat Coleman

Publisher. Questions? Check our FAQ for D3f, D3h.
Quote from: old 40 on September 25, 2007, 08:23:57 PMLet's discuss (sports) in a positive way, sometimes kidding each other with no disrespect.

Hugenerd

Just to clarify this point, I am going to use the ASC teams an ideal example of how not to use the system to your advantage.  The ASC appears to be the conference with the most "in-bred" schedules, meaning nearly all of their opponents and opponents' opponents are from the same group of teams (the other teams in the ASC).  As comparison, I will use the NESCAC and UAA as counter examples (each has advantages due to different reasons)

OWP and OOWP for ASC teams (with 4-6 conference games to go, these numbers will likely approach 0.500 even more):

Taxas-Dallas 0.521 0.515
Letourneau 0.521 0.511
Miss. College 0.520 0.501
Texas-Tyler  0.537 0.508
East Texas Baptist 0.503 0.509
Ozarks 0.503 0.507
LA College 0.520 0.503

McMurry 0.511 0.514
MHB 0.550 0.509
Concordia 0.473 0.494
Hardin Simmons 0.503 0.510
Sul Ross State 0.466 0.504
Texas Lutheran  0.467 0.504
Schreiner 0.499 0.501
Howard Payne 0.466 0.504

Even though each of these teams has about 20% of their games left (and all of them in conference), all of the OWP and OOWP numbers are clustered very close to 0.500.  In fact, only one team, MHB, has an OWP of 0.55 or higher and only 2 have OWP have OWP's of 0.525 or higher. In terms of OOWP, there is no team with a value greater than 0.515.

The NESCAC, on the other hand:

Middlebury 0.515 0.516
Amehrst 0.491 0.552
Williams 0.522 0.534
Colby 0.582 0.526
Bowdoin 0.582 0.518
Trinity 0.573 0.533
Bates 0.573 0.514
Conn College 0.506 0.532
Wesleyan 0.536 0.501
Tufts 0.555 0.530

We see here that 5 of 10 teams have OWPs over 0.55.  In the NESCAC, the OWP is more of a reflection of the schedule that the teams scheduled out of conference.  You can also see that Amherst's OWP is the lowest in the conference, while their OOWP is the highest.  This is likely due to the fact that their out-of-conference schedule was filled with cupcakes while most of their other opponents, including those in the conference, played tougher schedules so Amherst's OOWP is high.

Here is another example, the UAA, that you would expect to have a range of OWP and OOWP because there will likely be very little "in-breeding" in their out of conference schedules beause they are in geographically different regions:

WashU 0.560 0.545
CMU 0.520 0.537
Rochester  0.543 0.538
Brandeis 0.605 0.540
Case 0.553 0.520
Chicago 0.594 0.533
NYU 0.532 0.536
Emory  0.530 0.503

We see that although the UAA plays more conference games (given the season is not over yet and each team has 5 more conference games) than the NESCAC, they are also at an advantage because most teams play in different regions of the country and therefore have less common opponents in their out of conference schedules (or less "in-breeding" within their schedules).  All teams but one have OWP greater than 0.525 and 4 of 8 have higher than 0.55.

Ryan Scott (Hoops Fan)


You also see the NESCAC advantage of playing in a region with lots of teams.  The so called "second tier" schools in New England (essentially some of the best non-con competition a NESCAC school can schedule) all have better winning percentages simply by virtue of how many lower level schools are out there.

For example, the University of New England, this season, is sitting at 19-3.  They've played virtually no one of great significance.  In past years, this is the kind of team (solid, but with an inflated record) that Amherst would beat and greatly improve their SOSI).  With the new system, a win against UNE wouldn't count for as much.  However, the problem is just moved down the line.  UNE played Colby, Tufts, and Bates this season, so the inflated 19-3 is still helping Amherst and Middlebury.

The sheer size of the NE region means that there is a slightly more padded schedule for everyone, unless a school chooses to avoid the cupcakes.  We've seen in the past several season what avoiding the cupcakes gets a school (ten losses and no shot at the post-season, usually).

I do think it's a better system, but not by much.  And, of course, so long as the criteria continues to be regional while choosing nationally, being in a tough region will hurt rather than help a team.
Lead Columnist for D3hoops.com
@ryanalanscott just about anywhere

Ralph Turner

#1993
Quote from: hugenerd on February 11, 2009, 08:24:30 PM
Just to clarify this point, I am going to use the ASC teams an ideal example of how not to use the system to your advantage.  The ASC appears to be the conference with the most "in-bred" schedules, meaning nearly all of their opponents and opponents' opponents are from the same group of teams (the other teams in the ASC).  As comparison, I will use the NESCAC and UAA as counter examples (each has advantages due to different reasons)

OWP and OOWP for ASC teams (with 4-6 conference games to go, these numbers will likely approach 0.500 even more):  (In-region, non-conference games are listed)

Taxas-Dallas 0.521 0.515  (Austin College - SCAC x2, Guilford - ODAC)
Letourneau 0.521 0.511   (Lake Forest - MWC and UWW - WIAC at UMHB)
Miss. College 0.520 0.501  (Rust - Indep, Sewanee - SCAC)
Texas-Tyler  0.537 0.508   (Guilford - ODAC at UT-Dallas)
East Texas Baptist 0.503 0.509  (Austin College - SCAC, UDallas NEAC)
Ozarks 0.503 0.507  (Hendrix- SCAC)
LA College 0.520 0.503 (Millsaps - SCAC)

McMurry 0.511 0.514 (North Central - CCIW; and St Vincent PA in Las Vegas)
MHB 0.550 0.509  (Lake Forest- MWC, UWW - WIAC, Trinity TX - SCAC, Southwestern - SCAC)
Concordia 0.473 0.494 (none; beat WPI 86-85)
Hardin Simmons 0.503 0.510 (UDallas - NEAC, Austin - SCAC, Southwestern - SCAC)
Sul Ross State 0.466 0.504 (none)
Texas Lutheran  0.467 0.504 (none; lost to Anderson 79-84)
Schreiner 0.499 0.501 (Trinity - SCAC, Southwestern - SCAC)
Howard Payne 0.466 0.504 (none; lost to Anderson 93-78)

Even though each of these teams has about 20% of their games left (and all of them in conference), all of the OWP and OOWP numbers are clustered very close to 0.500.  In fact, only one team, MHB, has an OWP of 0.55 or higher and only 2 have OWP have OWP's of 0.525 or higher. In terms of OOWP, there is no team with a value greater than 0.515.

The NESCAC, on the other hand:

Middlebury 0.515 0.516
Amehrst 0.491 0.552
Williams 0.522 0.534
Colby 0.582 0.526
Bowdoin 0.582 0.518
Trinity 0.573 0.533
Bates 0.573 0.514
Conn College 0.506 0.532
Wesleyan 0.536 0.501
Tufts 0.555 0.530

We see here that 5 of 10 teams have OWPs over 0.55.  In the NESCAC, the OWP is more of a reflection of the schedule that the teams scheduled out of conference.  You can also see that Amherst's OWP is the lowest in the conference, while their OOWP is the highest.  This is likely due to the fact that their out-of-conference schedule was filled with cupcakes while most of their other opponents, including those in the conference, played tougher schedules so Amherst's OOWP is high.

Here is another example, the UAA, that you would expect to have a range of OWP and OOWP because there will likely be very little "in-breeding" in their out of conference schedules beause they are in geographically different regions:

WashU 0.560 0.545
CMU 0.520 0.537
Rochester  0.543 0.538
Brandeis 0.605 0.540
Case 0.553 0.520
Chicago 0.594 0.533
NYU 0.532 0.536
Emory  0.530 0.503

We see that although the UAA plays more conference games (given the season is not over yet and each team has 5 more conference games) than the NESCAC, they are also at an advantage because most teams play in different regions of the country and therefore have less common opponents in their out of conference schedules (or less "in-breeding" within their schedules).  All teams but one have OWP greater than 0.525 and 4 of 8 have higher than 0.55.


hugenerd, I have been applauding your posts for a couple of weeks now!

This post is worth dozens of karma!  There are several of us in the ASC who have been fighting this.  I will cut-and-paste this on the ASC board.
Here is another analysis (from the Pool C board) of what we are not doing to help ASC teams to get NCAA Pool C bids.  We have been used as an example of doing it the wrong way!


The non-conference D-III games that you see in the ASC are the only D-III teams in the area!

Hugenerd

Thanks.  It would be nice if the ASC adapted to the system by changing their format, but then again you never know if the NCAA will change the system again.  However, in any system if you are playing 80% (or more, non-d3 and out-of-region games dont count) of your region games in conference it is hard to get a handle how good a team is with respect the rest of the country.  With such a small sample size against non conference opponents (5 games or less), it is hard to think of a system where you would get a good idea of the relative strength of that team against the rest of the country.  Massey only has the ASC as having 67  games against out-of conference opponents (I dont know how many are non d3 or non region), which comes out to about 4.5 per team.  I think the system could use reform regardless.  However, I guess there is some added bonus to the schools, because I am sure they dont have to worry as much about filling a schedule and they probably have more consistent travel costs.