Can the voting process include the average time a voter spent viewing a photograph? Understandably the software can't determine if a person is actually viewing the photograph, but it should be able to capture both the time of url request for the photo and the time the vote was received.
I think it might be another guage of the sucess, or failure of a photograph.
Except for the times I vote and advance to the next photo, and meanwhile go make lunch while it loads. You'd also have to measure and compensate for each person's (current) connection speed.
Most often, I think you already get the stat you want in the number of comments garnered.
[personal opinion]
It would be fun to look at a histogram of "viewing times", I agree it would be a very interesting way to judge success. You'd have to throw out the outliers, where the voter went for a beer, LOL.
[/personal opinion]
It would probably be difficult or impossible to really implement this, since variability in server response and download times could introduce as much (or more?) variation than the actual voting patterns.
Oh wow. this is one statistic that would look awesome for me. It takes, on average, between 2 and 5 minutes per picture, just to download (depending on what speed my molasses.net dialup chooses to be), then look at it for a second or two (hahahaha), score, comment perhaps. Like kirbic mentioned, the outliers would have to be thrown out, but with dialup, the actual viewing time probably is an outlying number. if that makes any sense.
It is, in theory, a good idea. But it is one of those things that has so many variables that I don't know that it would be a particularly accurate stat. I think number of views and number of comments gives a better idea of how much the picture was looked at.
