Great Lakes Region

[almcguirejr]

Sorry for the delay, here's the data through Sunday's games.

REG   #   WP      OWP     OOWP    SOS     RPI     RPI50   NAT   Pool        REG     OVR   CONF     Team
GL   01   0.833   0.511   0.501   0.508   0.589   0.671   022   A   C      20-4    20-5   NCAC     Wooster
GL   02   0.813   0.530   0.488   0.516   0.590   0.664   028   A   C      13-3    18-7   MIAA     Hope
GL   03   0.813   0.508   0.492   0.503   0.580   0.658   031   C   007    13-3    17-8   MIAA     Calvin
GL   04   0.773   0.555   0.509   0.539   0.598   0.656   035   A   C      17-5    19-5   OAC      John Carroll
GL   05   0.739   0.524   0.520   0.523   0.577   0.631   049   C   016    17-6    18-7   OAC      Wilmington
GL   06   0.760   0.474   0.504   0.484   0.553   0.622   064   A   C-2    19-6    19-6   PrAC     Thomas More
GL   07   0.739   0.476   0.497   0.483   0.547   0.611   074   C   035    17-6    17-7   AMCC     Penn State-Behrend
GL   08   0.714   0.506   0.486   0.499   0.553   0.607   076   C   037    15-6    19-6   NCAC     Wittenberg
GL   09   0.739   0.464   0.494   0.474   0.540   0.606   077   C   038    17-6    17-7   PrAC     Bethany

Thanks for all the efort you put into this.

[sac]

Today's Great Lakes Region ranking

1. John Carroll
2. Wooster
3. Hope
4. Calvin
5. Thomas More
6. Wilmington

.....except for Thomas More, it mirrors RPI

[jcu_fan]

Knightslappy,

Is there a post where you explain the abbrieviations/calculations?  I'd love to know more about your method...

[KnightSlappy]

Knightslappy,

Is there a post where you explain the abbrieviations/calculations?  I'd love to know more about your method...

Basically, I collect data (from the d3hoops.com team schedules) on every team in the country. I then calculate the in-region winning percentage (WP), opponents winning percentage (OWP) and opponents-opponents winning percentage (OOWP).

I take these values and compute in-region RPI (25% WP and 75% SOS) and a value that I call RPI50 (50% WP and 50%SOS)

Strength of Schedule (SOS) is defined by the NCAA as 2/3 OWP and 1/3 OOWP

I currently rank the regions according to RPI50 (that seemed to work well in the early rankings), but now, as sac points out above, the standard RPI seems to be correlating better with the NCAA's official rankings.

I have awarded the team with the top RPI50 in each Pool A conference with bids.

I then rank all the Pool B and C teams according to RPI50.

That's really all there is to it. It's just a hard calculation based on numbers only (no head to head adjustments or anything).

[Knight81]

I apologize in advance for asking a question that's probably been answered before on this board, but I'm a newby, and it is not clear to me how at large teams are selected for the DIII tournament. Last year Calvin won the MIAA league championship but didn't get the automatic bid because it lost in the league tournament, and I can envision a similar scenario this year. How does this infamous "Pool C" work?

[ziggy]

I apologize in advance for asking a question that's probably been answered before on this board, but I'm a newby, and it is not clear to me how at large teams are selected for the DIII tournament. Last year Calvin won the MIAA league championship but didn't get the automatic bid because it lost in the league tournament, and I can envision a similar scenario this year. How does this infamous "Pool C" work?

Pool C is the NCAA's designation for At-Large bids.

The NCAA prepares a final set of regional rankings like the set released yesterday. After removing all the automatic qualifiers the committe will put the highest remaining team from each region "on the board". Since there are 8 regions, 8 teams are considered per round and one will be awarded a pool c based on the criterion in the handbook (winning percentage, strength of schedule, wins versus regionally ranked opponents, etc.). When that team leaves the board they are replaced by the next highest ranked team in that region. The process continues until all 19 pool c bids have been awarded.

[KnightSlappy]

Pool A is made up of conference automatic qualifiers. With the exception of the UAA, all conferences award this to their tournament winner. (40 bids this year).

Pool B is made up of teams that either play independent of a conference, or in a conference that has not yet qualified for an automatic bid. (2 bids this year).

After the Pool A and B bids are distributed all remaining teams are eligible for Pool C selection (at-large). (19 bids this year).

The NCAA uses specific criteria to select Pool B and C teams: http://d3hoops.com/faq.php?question=45

The d3hoops.com Tournament FAQ is very good: http://d3hoops.com/faq.php?category=NCAA%20Tournament

[Dark Knight]

After removing all the automatic qualifiers the committe will put the highest remaining team from each region "on the board". Since there are 8 regions, 8 teams are considered per round and one will be awarded a pool c based on the criterion in the handbook (winning percentage, strength of schedule, wins versus regionally ranked opponents, etc.). When that team leaves the board they are replaced by the next highest ranked team in that region. The process continues until all 19 pool c bids have been awarded.

Is this different from simply selecting the highest ranking of the remaining teams from all regions?

[KnightSlappy]

After removing all the automatic qualifiers the committe will put the highest remaining team from each region "on the board". Since there are 8 regions, 8 teams are considered per round and one will be awarded a pool c based on the criterion in the handbook (winning percentage, strength of schedule, wins versus regionally ranked opponents, etc.). When that team leaves the board they are replaced by the next highest ranked team in that region. The process continues until all 19 pool c bids have been awarded.

Is this different from simply selecting the highest ranking of the remaining teams from all regions?

Only 8 teams are considered at once (one per region), so head to head results and the like are not used in comparison to teams not yet on the discussion table.

That being said, all 19 could theoretically come from one region (but this would never actually happen).

[ziggy]

After removing all the automatic qualifiers the committe will put the highest remaining team from each region "on the board". Since there are 8 regions, 8 teams are considered per round and one will be awarded a pool c based on the criterion in the handbook (winning percentage, strength of schedule, wins versus regionally ranked opponents, etc.). When that team leaves the board they are replaced by the next highest ranked team in that region. The process continues until all 19 pool c bids have been awarded.

Is this different from simply selecting the highest ranking of the remaining teams from all regions?

I'm not sure I fully understand the question but the process is set up as it is so that the second, third, fourth, ect best non-auto qualifying team from a region can't get passed up by a less deserving team in another region simply because the latter was highly ranked in their particular region. Does that answer your question?

[Dark Knight]

After removing all the automatic qualifiers the committe will put the highest remaining team from each region "on the board". Since there are 8 regions, 8 teams are considered per round and one will be awarded a pool c based on the criterion in the handbook (winning percentage, strength of schedule, wins versus regionally ranked opponents, etc.). When that team leaves the board they are replaced by the next highest ranked team in that region. The process continues until all 19 pool c bids have been awarded.

Is this different from simply selecting the highest ranking of the remaining teams from all regions?

I'm not sure I fully understand the question but the process is set up as it is so that the second, third, fourth, ect best non-auto qualifying team from a region can't get passed up by a less deserving team in another region simply because the latter was highly ranked in their particular region. Does that answer your question?

I think so. Your response implies that there is a difference between the way the regional raking is done, to get on the board, and the way the national ranking is done between teams of different regions.

I had been under the impression that teams were selected based on regional rankings from the different regions, which seems less than ideal since regions differ. But if there is some correction factor based on the strength of the region or something, that would help.

Edit: KnightSlappy says above that head-to-head comparisons are done. That might be hard for teams in different regions, but it's something, I suppose.

[Pat Coleman]

But if there is some correction factor based on the strength of the region or something, that would help.

Something of that kind might occur, but it would be subjective and in the minds of the individual committee member. (Some might term it "bias," in fact.)

[ziggy]

But if there is some correction factor based on the strength of the region or something, that would help.

Something of that kind might occur, but it would be subjective and in the minds of the individual committee member. (Some might term it "bias," in fact.)

Our problem is there is no way to predict how the committee will value each number. How good does a winning percentage need to be to overcome a weak strength of schedule? How strong of a schedule is needed to make a lesser winning percentage look competitive for a Pool C?

[KnightSlappy]

After removing all the automatic qualifiers the committe will put the highest remaining team from each region "on the board". Since there are 8 regions, 8 teams are considered per round and one will be awarded a pool c based on the criterion in the handbook (winning percentage, strength of schedule, wins versus regionally ranked opponents, etc.). When that team leaves the board they are replaced by the next highest ranked team in that region. The process continues until all 19 pool c bids have been awarded.

Is this different from simply selecting the highest ranking of the remaining teams from all regions?

I'm not sure I fully understand the question but the process is set up as it is so that the second, third, fourth, ect best non-auto qualifying team from a region can't get passed up by a less deserving team in another region simply because the latter was highly ranked in their particular region. Does that answer your question?

I think so. Your response implies that there is a difference between the way the regional raking is done, to get on the board, and the way the national ranking is done between teams of different regions.

I had been under the impression that teams were selected based on regional rankings from the different regions, which seems less than ideal since regions differ. But if there is some correction factor based on the strength of the region or something, that would help.

The #2 Mid-Atlantic team is not always selected before the #6 Midwest team, just because they were a #2 vs. a #6.

On Hoopsville, they had a committee member on, and he said they do consider the validity of strength of schedule numbers. For example, a team that plays 20 conference games will have a SOS that trends toward 0.500. A team that plays only 10 or 12 conference games has more room for strength of schedule variance due to the increased number of non-conference games. So in that respect, they try to "look inside the numbers".

[David Collinge]

JCU gets upended in the OAC semifinals, putting themselves (I guess) on the Pool C bubble and opening the door for Wooster or the MIAA champ to be the region's top seed.