Pool C

[Ralph Turner]

+1 hugenerd!

The OWP works in the parts of the country with numerous teams, but name the D-III teams that are immediately to the south of Millsaps, Mississippi College, Louisiana College, Texas Lutheran, Trinity, Schreiner, McMurry, Sul Ross State, etc. 

I maintain that the NESCAC only plays single round robin to boost its OWP, and therefore its members can avoid inflicting another loss on its "bottom feeders".  Those NESCAC "bottom feeders" then in turn can defeat a "top rung" team from the CCC or the MASCAC or NAC or "future NECC" that has boosted its OWP on their own double round robin bottom feeders.

(I proved this back on the old message board in 2005.  Currently, Tufts is 11-13 on the season, 10-5 in non-conference and 1-8 in conference.  Assume that they would be 2-16 in a double round-robin format, and that 10-5 non-conference record becomes no better than 2-5, and the Tufts season record becomes 4-21!  Put that and the season records of the other NESCAC teams reflecting a "double round robin" into the Amherst OWP!  I will concede the autonomy to the NESCAC in scheduling the conference games to their needs, but let's not unfairly award Amherst the benefits that accrue thereto, or what some might consider "gaming the system".)

I heard Mr Grace talking about the way that the committee considers these factors in its deliberations, especially with the West and South Region (beginning about 8 minutes into the interview).  I am glad that that consideration is on the table.

If no rules are broken why should they be penalized. It is the rules that should be changed.

I agree. No one should be penalized, but also the integrity and accuracy of the statistical tool should be understood.

I have used the example of Rogers Hornsby and George Brett last year in a Daily Dose blog when we were considering the validity of the QOWI.  We may need to consider that in the OWP/OOWP's that we are seeing.

http://www.d3hoops.com/dailydose/2007/02/25/waiting-for-the-rules-to-change/

#  Ralph Turner Says:
February 27th, 2007 at 9:31 pm

I will appreciate some real statisticians to post, but I remember a stat about the best baseball Batting Average ever, Rogers Hornsby 1924 .424 or George Brett's 1980 .390.

It seems that Brett's .390 was higher in terms of Standard Deviations above the mean than Hornsby's.

Do we have a sufficiently large "n" to determine whether Amherst's Opponents opponents record is higher above the mean than UW-SP's or Mississippi College's or Wooster's? If we have a regionally based statistic, doesn't the standard deviations above the mean consider that?

Thanks.

An understanding of the "regional natures" [my quotations, not his] was one of Mr Grace's specific points in his interview last night.  It is different in the West and the South than in other regions.

[golden_dome]

I attempted to make this point to the ASC when I addressed their coaches and conference administrators a few years back. ASC teams do not help themselves, either, by scheduling a large portion of their non-conference games against scholarship schools, which does not weigh into this important metric whatsoever.

If the ASC teams are irate, they need only look at their own non-conference scheduling practices, not at other conferences' scheduling.

Pat, I understand what you are saying and that point is valid, but I wouldn't put all of the blame on the ASC coaches. The scarcity of Division III schools in this part of the country is a huge problem that other regions of the country, outside of the West, do not deal with. I would expect that the travel budgets for other teams already pale in comparison to the ASC shools. Most teams are forced to use the other 4-5 nonconference games to play guaranteed games, or stay at home. 99% of Division III administrations would do the same thing under similiar circumstances.

There was a map posted a few weeks back showing the location of the all DIII institutions. I think most people would be surprised how few DIII schools there actually are in the South and West regions compared to other parts of the country.

I know there are some ways the schedules could be better, but scheduling alone will never make up the difference. And I don't know if nonconference scheduling will ever be much better considering the already very large travel budgets for conference games.

[Cards7580]

A breakdown of Division III schools locations

131 = Northeast  Maine to New York
  98 = Atlantic      Pennsylvania to South Carolina
120 = Central      MN, WI, MI, OU, IN, IL, KY, MO, IA
  17 = SouthEast AR, TN, LA, MS, AL, GA, FL
  19 = Midwest     ND, SD, NE, KS, OK, TX, MY, WY, CO
  22 = West          WA, OR, CA, ID, NV, UT, AZ, NM

These numbers may have changed because I think the numbers are a couple of years old, but the general idea is most of the Division III schools are in three basic areas.

[LogShow]

A further break down of the west:

WA = 4
OR = 5  The NWC

Southern Cal = 8   The SCIAC

Throw in UC Santa Cruz and thats 18 of the 22

[Ralph Turner]

A further break down of the west:

WA = 4
OR = 5  The NWC

Southern Cal = 8   The SCIAC

Throw in UC Santa Cruz and thats 18 of the 22

Cal State East Bay, Menlo, Chapman and provisional LaSierra.

[Pat Coleman]

I attempted to make this point to the ASC when I addressed their coaches and conference administrators a few years back. ASC teams do not help themselves, either, by scheduling a large portion of their non-conference games against scholarship schools, which does not weigh into this important metric whatsoever.

If the ASC teams are irate, they need only look at their own non-conference scheduling practices, not at other conferences' scheduling.

Pat, I understand what you are saying and that point is valid, but I wouldn't put all of the blame on the ASC coaches.

I'm sorry -- if ASC coaches do not make out their own schedules then I will shift my focus to the ADs.

However, as long as you have control over your schedule, then you have no reason to complain if, as a league, you hover at .500.

Hugenerd -- I don't think I agree with your assessment that scheduling more D-III non-conference opponents would have little effect. I did a really quick count and it looks like the ASC played 13 non-ASC games against regional opponents and 42 against non-regional or non-Division III opponents. (Plus, there were a couple non-conference games against conference teams, which are essentially the same as conference games to this metric.)

Are you sure that if the ASC had played three times more non-conference regional games that there wouldn't be an effect?

I hope your table, however, finds its way into the conference meetings this offseason. This is precisely the symptoms you find with an inbred schedule, and it is what I warned of.

[LogShow]

A further break down of the west:

WA = 4
OR = 5  The NWC

Southern Cal = 8   The SCIAC

Throw in UC Santa Cruz and thats 18 of the 22

Cal State East Bay, Menlo, Chapman and provisional LaSierra.

Oh yeah thanks!  So basically, ID, NV, UT, AZ, NM are included in the west region but there aren't any schools.

[Pat Coleman]

In basketball.

[pbrooks3]

+1 hugenerd!

The OWP works in the parts of the country with numerous teams, but name the D-III teams that are immediately to the south of Millsaps, Mississippi College, Louisiana College, Texas Lutheran, Trinity, Schreiner, McMurry, Sul Ross State, etc. 

I maintain that the NESCAC only plays single round robin to boost its OWP, and therefore its members can avoid inflicting another loss on its "bottom feeders".  Those NESCAC "bottom feeders" then in turn can defeat a "top rung" team from the CCC or the MASCAC or NAC or "future NECC" that has boosted its OWP on their own double round robin bottom feeders.

(I proved this back on the old message board in 2005.  Currently, Tufts is 11-13 on the season, 10-5 in non-conference and 1-8 in conference.  Assume that they would be 2-16 in a double round-robin format, and that 10-5 non-conference record becomes no better than 2-5, and the Tufts season record becomes 4-21!  Put that and the season records of the other NESCAC teams reflecting a "double round robin" into the Amherst OWP!  I will concede the autonomy to the NESCAC in scheduling the conference games to their needs, but let's not unfairly award Amherst the benefits that accrue thereto, or what some might consider "gaming the system".)

I heard Mr Grace talking about the way that the committee considers these factors in its deliberations, especially with the West and South Region (beginning about 8 minutes into the interview).  I am glad that that consideration is on the table.

Excellent argument explaining things, Ralph.

[golden_dome]

I'm sorry -- if ASC coaches do not make out their own schedules then I will shift my focus to the ADs.

However, as long as you have control over your schedule, then you have no reason to complain if, as a league, you hover at .500.

Hugenerd -- I don't think I agree with your assessment that scheduling more D-III non-conference opponents would have little effect. I did a really quick count and it looks like the ASC played 13 non-ASC games against regional opponents and 42 against non-regional or non-Division III opponents. (Plus, there were a couple non-conference games against conference teams, which are essentially the same as conference games to this metric.)

Are you sure that if the ASC had played three times more non-conference regional games that there wouldn't be an effect?

I hope your table, however, finds its way into the conference meetings this offseason. This is precisely the symptoms you find with an inbred schedule, and it is what I warned of.

First of all, this has less to do about scheduling and more to do about money. Maybe coaches have total control over who they play in your region of the country where travel budgets don't matter with a Division III opponent at every corner, but options are limited here with so few DIII members. Coaches still answer to administrators and have to work within the framework of budgets.

I'm not saying it's impossible, you are 100% correct to say it could be done better. But  even if there were plenty of DIII opponents and travel was not a consideration, it's still easy to see that 4 nonconference games have little affect on a team's RPI when compared to 12-15 nonconference games for others. The 20-21 conference games were instituted because there are so few DIII opponents and teams needed games. Lowering that number will probably only make the problem worse.

There isn't a perfect system and really there's no one to blame for it, DIII is bigger in certain parts of the country but hopefully it will continue to expand to our regions. I really don't think the OWP and OOWP are that big of a deal because regional winning % is still most important and most deserving teams from the South will get an opportunity in the NCAA Tourney.

[Pat Coleman]

I think it's also clear that playing your four or five non-conference games against regional opponents would have more impact than playing one of them that way.

If the ASC doesn't like the hand its dealt, that's fine. But it needs to play its hand better than it is.

[Pat Coleman]

The 20-21 conference games were instituted because there are so few DIII opponents and teams needed games. Lowering that number will probably only make the problem worse.

It would at least give the schools that want to make the effort the opportunity to do so, and if the ASC is that strong, then it will benefit everyone in the OOWP, even the schools who play Wiley and Belhaven and LSU-Shreveport.

[Hugenerd]

Are you sure that if the ASC had played three times more non-conference regional games that there wouldn't be an effect?

Obviously it could make a slight difference but it depends on the quality of the opponents you play.  It will still not make as big a difference as having 16 non-conference games compared to 4.

What it comes down to is this:  If OWP and OOWP are only used for tiebreakers to rank teams in-region (and is pretty much not used in the south or west), then I guess it is better than using a coinflip. 

However, if there is a team from the south, like Mississippi College (19-5),  and they are being compared to a team from the northeast, midwest, etc., and the committee cant differentiate on any other metric and work their way down the criteria list to OWP and OOWP and say, "Eureka! The 4th team from the NESCAC (19-6) has an OWP of 0.615 and MC only has an OWP of 0.507, thats a difference of 0.108, that is clearly significant." Then I have a big problem of using OWP as a criteria because that 0.108 means nothing.  It just means that the 4th NESCAC team didnt have as in-bred a schedule.

[Mr. Ypsi]

Chris, I agree with your larger point (even with 'wiser' scheduling, the ASC is at a disadvantage in playing the OWP card), but your specific numbers are off.  As far as I know, only the NESCAC has a single round-robin format, providing 16 non-con games.  Otherwise, very few conferences would have 12-15 non-con games: an 8-team league doing a double round-robin has 11 such games; a 9-team has but 9; etc.

Pat, if it would be readily available for you, could you list the # of conference games by conference?

[golden_dome]

The 20-21 conference games were instituted because there are so few DIII opponents and teams needed games. Lowering that number will probably only make the problem worse.

It would at least give the schools that want to make the effort the opportunity to do so, and if the ASC is that strong, then it will benefit everyone in the OOWP, even the schools who play Wiley and Belhaven and LSU-Shreveport.

The problem is that there will always be far fewer teams wanting that opportunity so I doubt the coaches would vote to lower conference games for financial concerns. ASC member schools are just like DIII's throughout the country who don't have unlimited resources, and most would not spend even more to travel all around the country to improve the ASC's OWP and OOWP.  I doubt there is a conference anywhere in the country who would handle the situation any differently under the same cirumstances. Conferences like the NESCAC work the system well, but it is also convenient in many instances.

But even if travel wasn't considered, how many teams in this area would actually help the RPI that much. Trinity is already playing two ASC teams, and there really isn't another team in this area within 500 miles of the Texas schools. The only logical game not being played would be a MS College vs Millsaps game, but that's just one game. LC is already playing Millsaps. There are few DIII teams here, and even fewer who could actually help the numbers.

Having said that, I still don't think this is a big deal as long as the regional winning percentage is the primary criteria. The ASC has usually managed to get two teams in the tourney most years which is better than most conferences. That is plenty of opportunity to prove yourself in the NCAA Tourney.