Pool C

[pabegg]

Just to make sure I made myself clear. Losing always hurts the numbers. It has no mathematical impact on the OWP/OOWP and harms the RWP.

I would agree, if you put the word "significant" in front of "mathematical", because it does have a miniscule impact (you are giving your opponent a victory so their WP will increase), but, in reality, the result of a single game on your OWP is pretty much negligible (and even less so on your OOWP).

No so. Your opponent doesn't get credit for its win over you. As I described up-post, the games against you don't count in the calculation.

Again, using the Hope-Alma example: Alma starts the day at 3-15 in the region, with 2 of those losses to Hope. When it comes time to calculate Hope's OWP, Alma's record is treated as 3-13 (.1875) and the OWP average is calculated using the .1875 number (twice because of the two games). After tonight, a third copy of .1875 will be entered into Hope's calculations. So it doesn't matter whether Alma is 4-15 by beating Hope or 3-16 by losing, the 3-13 record is all that is used in the OWP.

[pabegg]

I think what the NJAC did splitting a ten team conference into two divisions is the best of both worlds.  While it is not gaming the system like the NESCAC does, it also eliminates 5 conference games and the teams don't beat each other up as much.  They also did a nice job of splitting the traditional powers between the two, although the north is the stronger division overall with NJCU, Ramapo, WPU and MSU in it.  I would think more conferences would be looking at a solution like this.

Actually, this is quite common, playing two games against your own division and one against the other division.

The NAC, CUNYAC, SUNYAC, NJAC, ASC, and SCAC all have this kind of scheduling. In some leagues, there are actual divisions, and in others there are not.

The NESCAC actually has a bit of this already, as two old "mini-conferences" already have a second game scheduled: Colby, Bates, and Bowdoin each play each other a second time for the "CBB" crown and Amherst, Williams, and Wesleyan each play each other a second time.

So going to full 2 and 1 scheduling would only add 2 games to the schedules of these six, which will not have enough impact on the numbers to change the objections of those who claim the NESCAC is "gaming" the system.

In practical terms, the NESCAC would have a hard time splitting into two divisions, because the weekend scheduling is built around travel partners and one set of travel partners would have to be in separate divisions (Wesleyan and Conn College). I think the league would have to go to 12 schools before this is possible.

[sac]

So going to full 2 and 1 scheduling would only add 2 games to the schedules of these six, which will not have enough impact on the numbers to change the objections of those who claim the NESCAC is "gaming" the system.

The potential to rack up 2 more "in-region" losses would be big.

[Hugenerd]

Just to make sure I made myself clear. Losing always hurts the numbers. It has no mathematical impact on the OWP/OOWP and harms the RWP.

I would agree, if you put the word "significant" in front of "mathematical", because it does have a miniscule impact (you are giving your opponent a victory so their WP will increase), but, in reality, the result of a single game on your OWP is pretty much negligible (and even less so on your OOWP).

No so. Your opponent doesn't get credit for its win over you. As I described up-post, the games against you don't count in the calculation.

Again, using the Hope-Alma example: Alma starts the day at 3-15 in the region, with 2 of those losses to Hope. When it comes time to calculate Hope's OWP, Alma's record is treated as 3-13 (.1875) and the OWP average is calculated using the .1875 number (twice because of the two games). After tonight, a third copy of .1875 will be entered into Hope's calculations. So it doesn't matter whether Alma is 4-15 by beating Hope or 3-16 by losing, the 3-13 record is all that is used in the OWP.

So then for OOWP, do they not count games in which the teams in question are involved.  For example, If team A has beaten team B twice, and team B has beaten team C twice.  If team A has also beaten team C twice,  do these games not count in the OOWP?  (To further clarify, lets say team A is 16-4, team B is 10-10, and team C is 5-15, when you are calculating team A's OOWP, is team C 5-15, 5-13, or 5-11?)

[David Collinge]

The NESCAC actually has a bit of this already, as two old "mini-conferences" already have a second game scheduled: Colby, Bates, and Bowdoin each play each other a second time for the "CBB" crown and Amherst, Williams, and Wesleyan each play each other a second time.

They have a name, too; Amherst, Wesleyan, and Williams constitute the "Little Three."

[pabegg]

So then for OOWP, do they not count games in which the teams in question are involved.  For example, If team A has beaten team B twice, and team B has beaten team C twice.  If team A has also beaten team C twice,  do these games not count in the OOWP?  (To further clarify, lets say team A is 16-4, team B is 10-10, and team C is 5-15, when you are calculating team A's OOWP, is team C 5-15, 5-13, or 5-11?)

OOWP is actually simpler. Once you've done the OWP calculation for every team, you simply average the OWP of all your opponents (weighted by the number of time that you've played them).

So in the example, once you've calculated B's OWP (treating C's record as 5-13) and all the other OWP's on A's schedule, then average all of those OOP's (again, weighted by the number of games against each team).

[pabegg]

Rankings are out

http://ncaasports.com/basketball/mens/polls/rankings/diviii

[pabegg]

Regional correction: Lycoming has a 13-8 regional record. Upon further review, their game with Shenandoah is not regional. Williamsport, PA to Winchester, VA is 198 miles, but from the campus addresses, it's 200.3.

Thoughts on the ranking, from west to east

Whitewater #1 in the west? Seems generous but consistent with their #4 last week.
Platteville is at #6, but has already dropped to virtual #8 with their loss.

Great Lakes I got exactly, with Heidelberg moving ahead of Albion due to their win over Capital.

DePauw and Miss College are a coin flip at the bottom of the south, with the same record and statistically identical OWP and OOWP values.

I'm a little surprised that Albright is not higher; then again, I thought they deserved to be in the poll at #7 last week, so their gain from probable #9 to #6 makes sense to me.

The RIC-Trinity-Middlebury order must be due to the fact that RIC beat Trinity, and Trinity beat Middlebury. I would think that Middlebury jumps to the top of the group if they beat Trinity in the NESCAC SF.

[Greek Tragedy]

Thoughts on the ranking, from west to east

Whitewater #1 in the west? Seems generous but consistent with their #4 last week.
Platteville is at #6, but has already dropped to virtual #8 with their loss.

Why is that, everyone else lost in that region!

[Ralph Turner]

Pabegg, I think that we can now look at your calculations with the understanding that the last 4-6 Pool C bid teams will be the last ones at the table, and that the last minute outcomes in the tourneys will juggle the final results.

If you are "C-10" or better, then win your tourney!

[pabegg]

Thoughts on the ranking, from west to east

Whitewater #1 in the west? Seems generous but consistent with their #4 last week.
Platteville is at #6, but has already dropped to virtual #8 with their loss.

Why is that, everyone else lost in that region!

I'm assuming your question is about Whitewater's #1. I just don't think that Whitewater has the best numbers in the region. It's also possible that they're in first based on the non-mathematical components used to rank teams.

[scooterman]

pabegg--In your opinion what does Brockport St. got to do to get Pool C bid if they dont get AQ by winning SUNY tourney? I thought getting to finals with Plattsburgh would do it, am I right?

[Hugenerd]

So then for OOWP, do they not count games in which the teams in question are involved.  For example, If team A has beaten team B twice, and team B has beaten team C twice.  If team A has also beaten team C twice,  do these games not count in the OOWP?  (To further clarify, lets say team A is 16-4, team B is 10-10, and team C is 5-15, when you are calculating team A's OOWP, is team C 5-15, 5-13, or 5-11?)

OOWP is actually simpler. Once you've done the OWP calculation for every team, you simply average the OWP of all your opponents (weighted by the number of time that you've played them).

So in the example, once you've calculated B's OWP (treating C's record as 5-13) and all the other OWP's on A's schedule, then average all of those OOP's (again, weighted by the number of games against each team).

So they only account for your wins in the first order, they dont worry about your victories against opponents opponents.

[Hugenerd]

What??? Chicago beats Brandeis and NYU and drops a spot in the rankings? How do you explain that?

[monsoon]

All of this math is making my head hurt - and that's just from reading about it, not doing the work!  Thanks to those of you trying to make sense out of all of this.