A statistic that would be useful in the challenge results would be the Standard Deviation of the Population result. A smaller number means that folks had general agreement in their voting and a larger number implies that there was a lot of spread in the voting. It is just an interesting number.
I suspect my Broken II entry will have a large spread do to the range of comments I'm getting.
what about photographer age, elsapo and joeylawrence auto DQ hahaha
Okay, that was funny.
But seriously. There have been a number of discussions of scoring and the STDEVP would provide some indication of how controversial your picture is. It is something that can be calculated with Excel but...
I'm sorry, I must be missing something. What part of this information is missing from the way it is currently implemented where you see the actual number of each vote you got?
If I want to see if a lot of people agreed, I look at the curve on the bar graph. It's easy and very visual.
But ah - dleach is one of us who subscribes to "Your eyes can decieve you; don't trust them" (Obi-Wan Kenobi, Star Wars IV). Quantification is required for direct scientific comparison.
Only problem is you are assuming a normal curve, and I suspect we can get quite bimodal at times, among other things. Interesting idea, though.
interesting idea. Perhaps I shall calculate it as a service to the community. omfg imagine trying to find the Sum of Squares for 600 entries, each with 250 votes >_<
*brain explodes and splatters all over my room
"Standard deviation (SD). A measure of the variability of a distribution of scores. The more the scores cluster around the mean, the smaller the standard deviation. In a normal distribution, 68% of the scores fall within one standard deviation above and one standard deviation below the mean." (reference: //www.wrightslaw.com/links/glossary.assessment.htm )
I'm only curious about it because I suspect my broken entry is going to have a larger standard deviation. I think it would be interesting to see if you could create an image that is controversion and gets a higher standard deviation.
I calculated one of my entries using excel but need to work on the formula some so I don't have to enter all 600 votes and instead just enter the count for each number.
Correlating the various comments I believe there is sufficient variability in interest to conclude that an export of the votes and entrant id for each photo could be made available for those outlying participants who care to deviate from the standard curve available, whereby one of us math types could provide an Excel or other means of indulging the search for bimodal behavior, which is nevertheless premised on the significance of the coefficients i.e. the voters interest in the subject challenge would require factoring. But we should be decent and say nothing about it to those less open to deviation, especially as we are dealing with what the sum of those squares might mean!
