4a343141$1@news.povray.org...
> The thought regarding using the ICA-style scoring is that you add up each
> judge's scores for each category and then get an overall score that's
> used for the final result.  That gives a chance (not a big one, but a
> chance) that someone who isn't artistically competent a shot at winning
> based on their concept and technical aspects.
>
> That also makes it less clear who the winner is likely to be just based
> on looking at the image.

Some years ago I ran correlation statistics on IRTC votes, and the 3 notes 
were extremely correlated. Basically it all comes down to a single "wow or 
not" rating. There were outliers (the occasional 
bad-concept-but-good-execution-image and the other way round) but they are 
just that, outliers, and they didn't affect the output of the competition. 
Having several scores is more intellectually satisfying than really 
significant and useful.

G.

Post a reply to this message