Voting Notes PDF Print E-mail

Point Scoring:

How the totals were calculated : Based on a total of 31 potential votes, a person's 1st place choice was given 31 points, 2nd place 30 points, 3rd place 29 points and so on. 30th place received 2 points, 31st place received 1 point. This worked fine as long as the person had replied for all 31 films.

For those submitting less than 31 choices, some adjusting, factoring, inflation or deflation had to be done. The dilemma was this: if a person has only voted on 10 films, should their 1st choice still get 31 points, their 2nd 30, and their 10th and last 22 points. No, that seemed to overvalue the movie they liked least. But if you went in the other direction, giving their 10th-rated film 1 point, then their favorite would only get 10 points. That seemed to undervalue the movie which they liked the most.

In the end, I made up sort of a sliding scale, assigning points (somewhat arbitrarily) differently, depending on how many films the person had actually voted on. Confusing enough? I'll provide the precise details to anyone who requests them; hopefully for your sake you aren't that desperate! So, for the person having voted on 10 films, the point allocation was:

1st place -- 30 points
2nd -- 27
3rd -- 24
4th -- 21
5th -- 18
6th -- 15
7th -- 12
8th -- 9
9th -- 6
10th -- 3

An explanation of Report #4 (Averaging point total breakdown):

Reports #1, 2 and 3 dealt with "raw" totals. Some of you may have noticed that some of your favorites fared more poorly than you might have expected. Part of this could be explained by you having favorites which are outside of the "mainstream" : ). Another factor is the number of respondents who ranked the movie.

There are some films which got a fairly high score from each person who voted on it, but not very many people voted (i.e., had seen it). There are other films whose higher point totals had more to do with it being seen by more people, even if the individual ranking was not so high.

For the Group of 10 (those who saw all 31) this was not a factor....all 10 people saw all 31 films, so there is no "averaging" effect. The averaging does come into play for the Group of 13 however.

Report #4 contains the standings, Group of 13 only, of the 31 films, adjusted for averaging. To arrive at these, I took the point total and divided by the number of people who had ranked the particular film.

Further notes on the averaging report (Report #4):

Dancing Lady makes a spectacular rise out of 31st place, for perhaps the first and only time! Which illustrates that if you fiddle around with numbers long enough, you can "prove" anything! (Much to my personal chagrin.....I was the one who artificially raised Yolanda's score in the first place by ranking it 7th. What was I thinking !? There truly is no accounting for taste; mine that is!)

Also noteworthy: The biggest gainers and losers in this exercise?

Gainers :

You Were Never Lovelier -- up 22 places, from 26th to 4th
The Sky's The Limit -- up 11, from 25th to 14th
Broadway Melody of 1940 -- up 8, from 17th to 9th

Losers :

Royal Wedding -- down 11, from 9th to 20th
Daddy Long Legs -- down 7, from 14th to 21st
Blue Skies -- down 7, from 19th to 26th

Averaging across the entire group:

Continuing to flog the averaging concept:

Even though averaging the Group of 10 on its own does not provide us with anything new, we can still average the combined scores which were contained in Report #1.

Report #5 contains the standings, Group of 13 and Group of 10 combined, of the 31 films, adjusted for averaging.

To arrive at these, I took the point total and divided by the number of people who had ranked the particular film.

Further notes on the second averaging report (Report #5):

Dancing Lady returns to its rightful place --- 31st place. Several of the scores are so close they are essentially ties. Does that mean that if we get enough people seeing all of the films, it would all sort of, er, average out....to a 10-way tie for 1st, a 10-way tie for 2nd, a 10-way tie for 3rd, and Dancing Lady alone at 4th?

Perhaps noteworthy : The biggest gainers and losers in this exercise? (Comparing the standing in Report #1 to the one presented here.)

Gainers :

The Sky's The Limit -- up 9 places, from 16th to 7th
You Were Never Lovelier -- up 8 places, from 22nd to 14th

Losers :

Royal Wedding -- down 7, from 15th to 22nd
Several others -- down 3 places

Have we learned anything else from all of this?

  1. Gary should not be left unattended with a loaded calculator!
  2. Those of you out there who are looking for a suggested Fred film to see next, it would seem that The Sky's The Limit and the 2 films with Rita Hayworth would be highly recommended.
  3. Royal Wedding --- very accessible, seen by many; but as you see more and more Fred films, it tends to lose some of its appeal in comparison to others.

Positional voting:

One note of explanation on the appearance of some of the standard favorites (Easter Parade, The Band Wagon) amongst the second-last and last-place votes. It's not that some people were making unexpectedly contrary choices, rather that they had not seen all that many of the Astaire musicals (say 4 or 6), so some of the lower-ranked ones turn out to be some of the traditionally highest-rated ones. It's awfully difficult to choose a meaningful least-favourite from Top Hat, Swing Time, Easter Parade and The Band Wagon!

-- Gary Shupak

Last Updated on Wednesday, 11 November 2009 10:47