Top 10% List "Score" Modified

[Phade]

Hey All,

The "score" value for the Top 10% List has been modified by using a Bayesian Average. This curves the scores based on the number of opinions as well as the average value of the score. Here's how it works:

score = Bayesian Average(avg(compscore+overall+review))

Bayesian Average = (v ÷ (v+m)) × R + (m ÷ (v+m)) × C

where:
R = average for the video (mean) = (Rating)
v = number of votes for the video = (votes)
m = minimum votes required to be listed in the top 10% (7)
C = the mean score across all videos (average all vids)

This means that the more opinions a video has, the more likely it is that the average score is correct. This will get rid of the "A bunch of DBZ fans gave this vid all 10's and now it sits on the top of the list" thing that unfortunatly seems to happen.

I hope this helps. Good luck and have fun!! ^_^

Phade.

[dokidoki]

Evolution > Glory of Love

[dwchang]

Hey All,

The "score" value for the Top 10% List has been modified by using a Bayesian Average. This curves the scores based on the number of opinions as well as the average value of the score. Here's how it works:

score = Bayesian Average(avg(compscore+overall+review))

Bayesian Average = (v ÷ (v+m)) × R + (m ÷ (v+m)) × C

where:
R = average for the video (mean) = (Rating)
v = number of votes for the video = (votes)
m = minimum votes required to be listed in the top 10% (7)
C = the mean score across all videos (average all vids)

This means that the more opinions a video has, the more likely it is that the average score is correct. This will get rid of the "A bunch of DBZ fans gave this vid all 10's and now it sits on the top of the list" thing that unfortunatly seems to happen.

I hope this helps. Good luck and have fun!! ^_^

Phade.

I'm very glad to see you finally implement this. I don't mean to assume too much, but I think this list makes it more apparent what the "top videos" are across the years and that have been "validated" by more opinions.

I guess I'm assuming too much, but I don't think many people could argue that the top 20 or so vids don't deserve to be there. Almost all have more than 100 people validating their scores and are regarded as "classics."

I for one thought it was pretty lame that "Failed Experiments" or "Shameless Rock Video" were near the bottom/middle of the list when a vid with maybe 8 - 10 opinions had a higher score, but less people agreeing.

Anyway nice work Phade! . Anyone who disagrees can go disagree with IMDB.com and well, sheer statistics/mathematics.

Evolution > Glory of Love

I hate you...this is the only bad thing to come of this

[Otohiko]

Wow. Holy... it just eliminated all but a couple of vids with >20 ops out of the list.

But, I guess that's only fair. Good work!

[Otohiko]

Wow. Holy... it just eliminated all but a couple of vids with >20 ops out of the list.

But, I guess that's only fair. Good work!

...err, that's <20 ops.

Hm, I guess there are bad things that happen when you don't take math for 3 years

[trythil]

Anyway nice work Phade! :). Anyone who disagrees can go disagree with IMDB.com and well, sheer statistics/mathematics.

There's lies, damn lies, and statistics.

[angelx03]

Anyway nice work Phade! . Anyone who disagrees can go disagree with IMDB.com and well, sheer statistics/mathematics.

There's lies, damn lies, and statistics.

LIES!

Anyway, nice work there, Phade. Of course I think I downloaded like more than half the videos that's in the Top 10%. Now there won't be any "good" ones to download.

..for the time being of course

[fyrtenheimer]

Ohh, I wondered what was up when of my videos was back on the top ten.

[Char the second]

man its alwalys that koopiskeva vidoe i like it but i really want to see a new one

[Zarxrax]

Phade: how about an option to remove all videos from the list that you have downloaded locally?