Re: January 21st Poll Speculations
Actually the poll thing is a good illustration of some of the flaws of the use of the Borda Count method for determining rankings of candidates for anything. The problem is Arrow's impossibility theorem, which says that there is no fair way to rank a list that is longer than two choices.
On another issue, Utica fans might be interested in the rankings that will appear on another site later today...
You are exactly right. It is counter-intuitive for sure, but if you think about it, the poll rankings are just a list of best to worst of average ranking by the voters. If you break down the voting points here is a breakdown of the average ranking from last week to this week:
1.Norwich 2.05-->1.55
2.St. Norbert 1.65-->2.25
3.Bowdoin 4.20-->3.00
4.Oswego 4.90-->5.05
5.Hobart 2.80-->5.40
6.Wisc - EC 6.3-->5.70
7.Adrian 7.25-->6.75
8.Utica 8.85-->6.90
9.Plattsburgh 9.05-->9.40
10.Amherst 9.80-->10.20
11.M'ville 11.50-->12.55
12.Williams 12.75-->13.45
13.Mass-Boston 14.85-->13.90
14.Babson 14.10-->13.95
15.MSOE 15.15-->14.40
15.Neumann 15.30-->14.40
As you can see, Utica actually has an average ranking of 6.9 (up from 8.85 last week) but there are/were 7 teams with a better average ranking both weeks. Very rarely does it work out that a teams average ranking is also their actual ranking like it did with Bowdoin this week.
You will have to excuse the in depth analysis on such a (NCAA wise) meaningless poll, I'm a bit of a poll geek.
Actually the poll thing is a good illustration of some of the flaws of the use of the Borda Count method for determining rankings of candidates for anything. The problem is Arrow's impossibility theorem, which says that there is no fair way to rank a list that is longer than two choices.
On another issue, Utica fans might be interested in the rankings that will appear on another site later today...