Dan Connelly wrote:

> I find this very discouraging, as by so doing you are unfairly weighting your
> vote.  I will struggle on choosing a correct distribution of points,
> but by tossing around 20's, you thus make my votes, in which contestants
> differ by a few points, of trivial relative significance.
>

There's probably good counsel in your post, Dan.

In the Stills, I usually give 20-20-20's to my favorite, and probably >16-16-16
to six or eight of my other favorites. And I do differentiate, especially among
the top 12 or so, between excellent scores in the different categories: one might
be 20-15-14 and another 14-15-20.  In that Animations round,  I think I gave Pool
Shark close to 20-20-20,  but I don't feel that I "stuffed the ballot box."  I'm
sure I gave >16-16-16 to two or three others, notably Rusty's Late.I've never
done the math, but I suspect that I give out a uniform distribution of rankings,
from near 0-0-0 to near 20-20-20.  I'm wondering if from your post, you typically
give out a normal distribution centered around 10-10-10, with just one or two
20-20-20's  per year.

I quote Dan again:

> Votes need to justify themselves.  If A scores higher than B in category
> C, it is because A is better than B in C, not because one had
> a generally good feeling for A overall.  Was Pool Shark an all-time
> ground-breaking masterpiece in all three areas?

Well, for example, there was once an entry that looked just like a photograph of
a group of people in bearksin working on building a temple.  The problem is that
these humans were all bald, pale, hairless, and identical septuplets.  That
problem  wrecked the image for me, both in its artistic appeal,  its attention to
technical detail, and its theme. I did not rate it highly in any category.  I
felt guilty about my vote, as I was sure some voters would be swept away by its
"looking just like a photograph."

Perhaps one solution to "rogue voters" is for the IRTC staff to put more
"experts" in the voting list for Animations, especially if we're only getting
10-15 entries.

Post a reply to this message

Attachments:
Download 'us-ascii' (3 KB)