MBB: USA South Conference

[notamensa]

compliments of david collinge over on the top-25 board -


Top 25 News and Notes, Vol. VIII, Issue 111

Two weeks into the season, and we still have 53 teams receiving votes.  This week the top 25 saw a shake-up, including at the top, but the inevitable shake-out is still on the horizon.

Congratulations to the new #1 team, a familiar name to readers of this column: the College of Wooster.  Although Wooster is almost a permanent member of the top 10 (having been ranked there 81 times in 111 polls), it is actually only the 6th week that the Fighting Scots have reached the top step of the ladder.  In the 2000-01 season, they reached the top in Week 12, and lost to arch-rival Wittenberg that weekend.  In the 2004-05 season, the Scots became #1 on two separate occasions (Weeks 7 and 13), and both times lost to Wittenberg that week.  Then last season, when Wittenberg was #1, #2 Wooster knocked them off to take over the top spot.  So who is on the Scots' schedule this week?  I mean, besides Oberlin?  That's right, #6 Wittenberg.  Stay tuned to this channel for further developments. 

Debutantes:

#8 Whitworth has reached the top 10 for the first time in poll history; and
#20 Averett broke into the top 25 for the first time ever, in just their second week receiving votes.
Congratulations to the Pirates and Cougars!

Streakers:
In nabbing the poll's top two spots, Wooster and Amherst extended their record streaks:

#1 Wooster is in the top 10 for the 41st consecutive week, and received votes for the 111th straight time, and
#2 Amherst is in the top 25 for the 66th consecutive week.
Elsewhere,

#2 Amherst is in the top 10 for the 10th straight week;
#4 Virginia Wesleyan and #10 Lincoln are in the the top 25 for the 10th consecutive time each;
#25 UW-Whitewater received votes for the 80th consecutive week; 
Baldwin-Wallace received votes for the 20th consecutive week, but had their streak of top 25 appearances snapped at 15 straight.
Congratulations to these great programs!

High-Water Marks:

#8 Whitworth, #17 Bates, and #20 Averett each set new highest-ever-ranked marks in this poll;
#3 Ohio Northern, #5 St. Thomas, and #10 Lincoln each tied their highest ranking ever (ONU, Week 14 of 2000-01; UST, 4 weeks in the 2001-02 season; and Lincoln, last year's final poll.)
Congratulations on these achievements!

Milestones:

#3 Ohio Northern is a top 10 team for the 10th time;
#5 St. Thomas is a top 25 team for the 40th time;
#13 Mississippi College received votes for the 60th time;
#9 Lawrence received votes for the 50th time; and
Keene St. received votes for the 30th time.
Congratulations to these teams, and to all who continue to make this such a great game! 

[PrideSportBBallGuy]

These "Power Rankings" are like the BCS, in a sense that it isn't perfect.  i just compared it to the national level I took d3hoops top 25 (the top 10) and ran my numbers against it and squeezed out my top 10.  Again it is still early.  This is probably something that should be used until the end of the season.  There is a slight emphasis on games the teams play.  Losing conference games hurts alot in my little system.  I am not giving away my formula but, expected wins and wins games played and conference losses all are part of it.

MY TOP 10 

Whitworth
UW-Stevens Point
Amherst
Wooster
VA Wesleyan
Augustana
St. Thomas
Ohio Northern
Wittenberg
Lawrence

D3hoops

Wooster (22)
Amherst (3)
Virginia Wesleyan
St. Thomas
UW-Stevens Point
Wittenberg
Ohio Northern
Whitworth
Lawrence
Augustana

As you can see my top team is 10-0, 8-0 in second spot.  Once more games are played it all should balance out.  Unless ya'all feel the top 10 isn't that bad. If the team is undefeated well they are higher on the list.  Again I will say it is still too early.

Losing a conference game is probably the worst thing a team can do to affect the position.

[PrideSportBBallGuy]

I will keep it to the conference for now and release a national one each week when the regional rankings come out (I will do a top 25).  I will even use my system in the pick'em contest to pick my games enroute to the National Championship and I will let you know how well I will do.

[notamensa]

I am not giving away my formula

i dont know how everyone else feels, but if i cant see how you derived your rankings, you might as well be posting a personal opinion. why the secrecy with the formula? do you think it wont stand up to scrutiny?

[old_lion]

I am not giving away my formula but, expected wins and wins games played and conference losses all are part of it.

Translation ... it's a WAG.

For those of you unfamiliar with the acronym ...
W - wild
G - guess
I'm sure you can come up with the "A". 

[PrideSportBBallGuy]

I said what the formula uses in prior post. what else to you want. If you really must have it then I guess I will give it out.
Simplifed version.

((EW+W)/(2))-(Conference)L

EW wins uses PF and PA against and GP

First part uses statistical wins and actual wins and finds the avg.  You take out any conference losses from that average.  I think if you can't dominate your conference how do you expect to dominate on the national scene.

I will put numbers out with the rankings once the season gets more underway.  You may look at it and say it doesn't take in account of SOS outside of the conference. In fact it very well may.  If you play some close games out of conference against good teams with one or two points either way you may win 2 out 4 or 4 out of 4. But if you dominate your conference it looks pretty good.  If you don't then it looks bad.  Whatever the situation PF and PA it should equal out. We will see what it looks like much later in the season.

It almost predicted the conference finish of last year.  Only difference was the SU FC spots.  When in fact though SU got most of thier wins out of conference.

[narch]

I said what the formula uses in prior post. what else to you want. If you really must have it then I guess I will give it out.
Simplifed version.

((EW+W)/(2))-(Conference)L

EW wins uses PF and PA against and GP

First part uses statistical wins and actual wins and finds the avg.  You take out any conference losses from that average.  I think if you can't dominate your conference how do you expect to dominate on the national scene.

I will put numbers out with the rankings once the season gets more underway.  You may look at it and say it doesn't take in account of SOS outside of the conference. In fact it very well may.  If you play some close games out of conference against good teams with one or two points either way you may win 2 out 4 or 4 out of 4. But if you dominate your conference it looks pretty good.  If you don't then it looks bad.  Whatever the situation PF and PA it should equal out. We will see what it looks like much later in the season.

It almost predicted the conference finish of last year.  Only difference was the SU FC spots.  When in fact though SU got most of thier wins out of conference.

this really clarifies things for me...

how about the un-simplified version for those of us that can understand simple mathematical equations?

by the way...i looked at your last 75 posts and the only one i saw which even MENTIONED the formula, was the post quoted by mensa...

[narch]

pridesportbasicballguy - please understand i'm not attacking you here...simply trying to get an understanding of this statistical system you are using to predict the outcome of games and give teams "power rankings" - if this is truly as accurate as you claim, i might take it to vegas with me and cash in

[PrideSportBBallGuy]

I really don't know how accurate it is.  I mean it mixed up FC and SU's spot. (Last year's final standings) I don't know of the top of my head how close all SU games were in conference play compared to FC games.  They may have played closer games then FC.  The reasons why the system put them ahead of Ferrum College.

The formula used is what it is.

Only EW has a more complex system.  Its used already and the guy who used/came up with that formula says there is a about +/-  3 games.  Thats why I averaged that number with the teams actual wins.

EW calculation is GP*(((PF)^14)/(((PF)^14)+((PA)^14))). 

Thats how it is listed in my spread sheet.

Its interesting at least.  I want to see the effectiveness compared outside.  I don't know how accurate it is early on.  My guess is it is just like the massey ratings as of right now.  Averett is currently #2 in that.

[CNU85]

too complicated for this simpleton. I'll just watch the games and the team with the highest number under their nameon the scoreboard gets a W...the other team gets the L. The team with the most W's at the end of the season can be in first place.

[PrideSportBBallGuy]

although I think it may still be to early in the season.  I am going to make 3 game picks tonight to see how it works.  I don't know if this formula can be worked like this.  I guess we will see

4-4 NC Wesylean Beats 4-3 Hampden-Sydney (It looks like most people on the ODAC board have picked HSC)

4-3  Juniata beats 4-3 York(PA)  (I actually think this game will be really close. The formula has Juniata as the better team by .04 points.  This game is played at York. So that might push them over the edge) I am sticking to the formula though.

0-7 Fitchburg State beats 1-8 Becker.

If I go 0-3 tonight,  I know and I should hear it tomorrow.  (I think the system still might be missing something.)  This is used to to deterime expected wins not wins against other teams. I just can't get over how well it worked out and got the standings fairly accurate from last year.

[notamensa]

I actually think this game will be really close. The formula has Juniata as the better team by .04 points.  This game is played at York. So that might push them over the edge

i think youve hit it on the head right here. your formula should have a home/away variable.

[SU97]

Regardless of the formula, I still don't think Greensboro should be "ranked" that high among conference teams - unless their lineup of transfers are all like the Berg guy from last year.  That guy could shoot.

Still backing my formula until conference games start:  Dragan > Gboro 

[narch]

I just can't get over how well it worked out and got the standings fairly accurate from last year.

question - did you apply the formula to last season before or after the games were played?

and what is the carrot thingy? i'm not familiar with that symbol

what is the significance of the number 14?

how is this applied?  do you calculate the formula for each of the teams then the team with the higher number is your winner?

details, my man, details

i like su97's formula better, although i'm not sure it's correct

[PrideSportBBallGuy]

went 1-2 with my picks.  The bottom feeder game.  I just don't think there are enough games yet.  Since there is an emphasis on conference games.  NCWC of course has no conference games under thier belt threw off my numbers.  Yeah maybe I am pushing it to be right.  I don't think it can be used to determine games since it is used to figure out wins  .  I will pick games every day and see what happens.  Work it out I guess to get it right.

narch-
When I did it of course it is after games from last year.  I orginally set it up just as EW (that crazy formula)  It had SU in 4th in the conference.  I said that would be impossible since they only won one conference game.  I had to factor in a way for conference losses.  I tried subtracting conference losses as my first step.  In order to get all wins in the conference.  It ended up making the order alot better with the exception of SU and FC.

Carrot thing is "power"

14 is based on the EW- Source:(http://basketball-reference.com/about/glossary.html) ; Ask this guy.

I set out and apply that formula stated earlier and the higher number is my winner.  I think there are still quite a few bugs to work out.  The current order if you play more games you are going to be on top.  Again though with more wins it makes it easier to be up top.  I thought about it all day.  I might want to set up a ratio based on total number games.  500 points in 5 games means 2700 points in 27 games.  Catch my drift applying it on a larger scale. Something along those lines.  I might like to add a QOWI to it somwhere.  Right I say it is good for Power Rankings not for predicting games.

SU97-Look at like opponets and tell me who beat who in out of conference schedule.  I just glanced at and saw my power rankings could reflect it.