Can't the lawyer still DQ one of the final videos?VicBond007 wrote:Final DQ list is out. See first page.
Even with the standard distributions... Not really. There are an infinite number of moments above the first and second moments (also known as average and standard deviation). There is skew and kurtosis (3rd and 4th moment)... and all the moments above them. Without making an assumption about the distribution, you can't really make a conclusion even if you did have the standard deviation. Although... I guess there'd only be 34 moments in the worst case (the conversion to a moment is equivalent to a Fourier Transform. So it'd be a Discrete Fourier transform and it'd have no more than 34 data points).Code wrote:Somebody that understands statistics! Now that's rare.chui101 wrote:Well, you can't statistically draw that conclusion without comparing the standard deviations of the lengths of the videos as well. Just because the means are similar doesn't mean the distribution of times is the same.irriadin wrote:Based on the number of videos, it doesn't look like many longer videos made the cut. There were 34 finalists last year and that was with the six minute time restriction in place.
Take the 33 one-minute videos and a 1 hour 27 minute videos.... this has (approximately) the same standard deviation as a 16 videos -13 minutes long and 16 vids 16 minutes long. (yes, negative minutes :-p).
Basically, just because the means AND the standard deviations are the same... doesn't mean that the skew and the kurtosis is the same. The only thing we have is an average time of ~3.5 minutes and a minimum time of 1 minute.