>>There is just something wrong with the competition that is hard
>>to put a finger on.

My observations:

1) One of the "problems" is that people are too nice to each other when
voting.

If a picture is off-topic, it should get a 1.

If a picture is poorly composed, it should get a 1.

If a picture is a Bryce Render with 3 objects in it, it should get a 1.

If a picture is so dark you can't see anything in it, it should get a 1.

By the same token...

If a picture is so on-topic it knocks your socks off, it should get a 20.

If a picture is so well composed you say "wow", it should get a 20.

If a picture is realistic, even of imaginary objects, it should get a 20.

Basically, everyone voting in the range of 5-15 for pictures is saying
nothing about the pictures they are voting for.  The average scores tend to
center around the 9-12 point range.  The really lousy pictures tend to get
about the same scores as the really good pictures.

I vote in two passes.  The first vote is for good or bad, basically
eliminating all the pictures that either didn't impress me, or that simply
were off topic.  The ones that do impress me get some vote based on how well
they impress me.

The voting program then gives you a screen with the pictures in order of
your voting score for each of them.  I then perform a second pass through
the top 10-15 pictures revoting on just those pictures for a final vote.

I've found this technique gets rid of the pictures that are really lousy to
begin with and allows me to focus in on the good pictures for a better study
and more accurate vote.  I've taken what used to take me a week to do and
shortened it down to about 3-4 hours.

2)  I also find that the written text document submitted with a work is very
important in determining the artists intent with regard to their work.  If
when looking at a picture you are unsure of what it is, or how it fits the
topic, the text description can be very instrumental in turning a picture
that appears off-topic into a picture that actually is on-topic.  If a
picture is not documented or it lacks a description on how a particular
effect was achieved, I've even gone so far as to download the source to
render it for myself before giving a vote.  Naturally, if the doc is lacking
and there is no zip file included, you can't hope to resolve questions like
this, you have to doc the artist in your vote.

3)  Comments cannot be retracted once posted.  While not a flaw in the
voting process, people must be careful in what they post.  Constructive
critism about the picture is good feedback and should be taken as such.
Judges should keep away from making statements that are not in that vein.

4) The competition has no way to take into the hours one picture may have
taken to develop versus the 20 minutes another artist spent in a last-ditch,
whim submission.  Usually this is quite apparent and can be handled with the
voting technique listed above.  However, as a competition, artists should
respect the time that judges are giving their work and this type of
submission should be frowned upon.  As a judge, you will spend more time
judging the submissions than some people spend on putting their entire
submission together... including render time.  (Most of my pictures have in
excess of 100 hours of work in them by the time they are submitted.)

Post a reply to this message