Must've hurt you bad. 3 is like 2 hooks merged together.mckeed wrote:It makes sense, i was just using 30 as a random number I pulled out of my arse.
Top 10% List "Score" Modified
- Bote
- Joined: Sat Aug 09, 2003 8:20 am
- Location: Belgrade, Serbia
- Contact:
My Youtube channel: Bote Logos
NEW!!! One Piece AMV - "YUM YUM 2.0"
Berserk - Man of Sorrows (upscaled to 4k)
NEW!!! One Piece AMV - "YUM YUM 2.0"
Berserk - Man of Sorrows (upscaled to 4k)
Beowulf@RDS wrote:RECTANGLES AND AFTER EFFECTS WONT SAVE YOU NOW MOTHERFUCKERS
- mckeed
- Joined: Tue May 15, 2001 1:02 pm
- Location: Troy, NY
- Contact:
- genestarwind21122
- Samurai Master
- Joined: Sun Feb 23, 2003 4:52 pm
- Location: space the final frontier....
- Contact:
Okay everyone but take for instance Shrine of the Goddess that video was up there for 4 days or so on top above Euphoria I know you may say that's wrong but I was getting ops from various people. If a video is on the top 10% the old way the only thing you guys should have done is took a look at it and see why it got there. If you think it is a good video then op it honestly for the reason why it is good and the same thing goes for bad. If it is bad then op honestly why it is bad and not give it straight 1's. The old system had it's advantages for creators that aren't well known they could have got on there because they finally had a breakthrough and people could see it then if it was truely good.
Do you see where I'm going with it. I'm not saying I'm offended by the new policy but it doesn't mean that a 30 op video couldn't be proven worthy over a 200 op video or a video that is blown out of proportion.
Do you see where I'm going with it. I'm not saying I'm offended by the new policy but it doesn't mean that a 30 op video couldn't be proven worthy over a 200 op video or a video that is blown out of proportion.
- Zarxrax
- Joined: Sun Apr 01, 2001 6:37 pm
- Contact:
New creators can still have a breakthrough and get on the list. You think Koop was a major player before he released Euphoria? How bout Majin Kenshin? Nappy? Kurt?
New videos will make it to the list of they are worthy. I think you are worried about videos that are not *truely* worthy of being on the list getting up there just so they can have some extra exposure. This is wrong though. The weekly star lists can give these videos a better shot though.
New videos will make it to the list of they are worthy. I think you are worried about videos that are not *truely* worthy of being on the list getting up there just so they can have some extra exposure. This is wrong though. The weekly star lists can give these videos a better shot though.
- mckeed
- Joined: Tue May 15, 2001 1:02 pm
- Location: Troy, NY
- Contact:
Though I think the list is more accurate currently, I do agree with gene. My Uber depressing video started getting way more ops once it made the top 10 and continued to get solid revierws. Imust have checked the top 10% right before the switch as it had manged to move up to #17. Then after the switch it moved down to the bottom of the list. Personally I think it would be good to have two lists up, as they both tend to highlight different videos. People with a string of good reviews can get some much needed exposure. Maybe a supplemenatry list with videos that only apprear on the old list. Like a "up and comming video list" which is calaculated how the old top 10% was. Could make this shorter as well. Like only 20 videos or so. After a video establishes itself on the main 10% it would get removed or if the reviews fell down as a result of more scrutiny. Could be a nice way of helping lesser-established creators out. Just an Idea.
- mckeed
- Joined: Tue May 15, 2001 1:02 pm
- Location: Troy, NY
- Contact:
Also, maybe we should combine videos that are co-created on the list as well. For instance, This is the same video as they worked on it togeather and have different amounts of oppinions.
67 VegettoEX (Michael LaBrie) Alone in the World 9.09 30
67 MeriC (Meri Cantoni) Alone in the World 9.09 28
One thing is that they have different ammount of ops. They probally should be combined to give a more accurate score for the video and it should only take up one line.
67 VegettoEX (Michael LaBrie) Alone in the World 9.09 30
67 MeriC (Meri Cantoni) Alone in the World 9.09 28
One thing is that they have different ammount of ops. They probally should be combined to give a more accurate score for the video and it should only take up one line.
- Phade
- Site Admin
- Joined: Fri Oct 20, 2000 10:49 pm
- Location: Little cabin in the woods...
Hey,
With the old method of a strait average, it was possible, for example, for a group of DBZ fans to collaborate and put all 10's on a mediocre video to make it to the top of the list instantly. Should the video be on the #1 spot? No, not really. The new system helps take care of that. I'll try to explain how this works.
downwithpants has posted a mathematical technical definition of what is going on with the calculations, but I'll try to explain further. Here's the formulas:
old score = avg(compscore + overall + review)
new score = (votes ÷ (votes + min_votes)) × old_score + (min_votes ÷ (votes + min_votes)) × avg(all videos)
The new score formula has two parts which are added together:
(votes ÷ (votes + min_votes)) × old_score
and
(min_votes ÷ (votes + min_votes)) × avg(all videos).
Each part is a ratio multiplied by either the old_score or the average of all videos. With few opinions (ex. 1 opinion), the overall average has the greatest weight:
(1/(1+7))*old_score = 1/8 * old_score
while
(7/1+7)*overall_avg = 7/8 * overall_avg
If you'll notice that since we are adding the two parts together, 1/8 + 7/8 = 1, which means that we are getting a whole score.
As far as the average goes, this means that, mathematically, any new video is assumed be be an average video until proven otherwise. As the video gets more opinions (ex. 50 opinions), the average of the video has more weight than the overall average:
(50/(50+7))*old_score = 50/57 * old_score
while
(7/50+7)*overall_avg = 7/57 * overall_avg
By this time, the actual average of the video is dominating the value of the new score. If your average is above average, your video will march up the rankings. If your average is below average, your video will march down the rankings.
If we go back to the buttmunch DBZ fanboy example who are trying to artificially push a video to the top of the list, here's what happens at 7 opinions of all 10s:
current overall average = 7.406224807350
(7/(7+7))*10 + (7/(7+7))*7.406224807350 = 5.00000 + 3.70311 = 8.70311 = 8.70
So instead of the video taking the #1 spot, the video takes a much lower position. Say the video reaches 30 opinions and the average is a little more sane at 8:
(30/(30+7))*8 + (7/(30+7))*7.406224807350 = 6.48649 + 1.401178 = 7.887668 = 7.89
At 50 opinions and the average still being 8, the new score looks like this:
(50/(50+7))*8 + (7/(50+7))*7.406224807350 = 7.01754 + 0.909534 = 7.927074 = 7.93
This method has now prevented an artifical score inflation by a bunch of fanboys, but then yields a score closer to the actual average as the number of opinions grows. The more opinions a video has, the more valid the score becomes.
I hope this explanation helps.
Phade.
With the old method of a strait average, it was possible, for example, for a group of DBZ fans to collaborate and put all 10's on a mediocre video to make it to the top of the list instantly. Should the video be on the #1 spot? No, not really. The new system helps take care of that. I'll try to explain how this works.
downwithpants has posted a mathematical technical definition of what is going on with the calculations, but I'll try to explain further. Here's the formulas:
old score = avg(compscore + overall + review)
new score = (votes ÷ (votes + min_votes)) × old_score + (min_votes ÷ (votes + min_votes)) × avg(all videos)
The new score formula has two parts which are added together:
(votes ÷ (votes + min_votes)) × old_score
and
(min_votes ÷ (votes + min_votes)) × avg(all videos).
Each part is a ratio multiplied by either the old_score or the average of all videos. With few opinions (ex. 1 opinion), the overall average has the greatest weight:
(1/(1+7))*old_score = 1/8 * old_score
while
(7/1+7)*overall_avg = 7/8 * overall_avg
If you'll notice that since we are adding the two parts together, 1/8 + 7/8 = 1, which means that we are getting a whole score.
As far as the average goes, this means that, mathematically, any new video is assumed be be an average video until proven otherwise. As the video gets more opinions (ex. 50 opinions), the average of the video has more weight than the overall average:
(50/(50+7))*old_score = 50/57 * old_score
while
(7/50+7)*overall_avg = 7/57 * overall_avg
By this time, the actual average of the video is dominating the value of the new score. If your average is above average, your video will march up the rankings. If your average is below average, your video will march down the rankings.
If we go back to the buttmunch DBZ fanboy example who are trying to artificially push a video to the top of the list, here's what happens at 7 opinions of all 10s:
current overall average = 7.406224807350
(7/(7+7))*10 + (7/(7+7))*7.406224807350 = 5.00000 + 3.70311 = 8.70311 = 8.70
So instead of the video taking the #1 spot, the video takes a much lower position. Say the video reaches 30 opinions and the average is a little more sane at 8:
(30/(30+7))*8 + (7/(30+7))*7.406224807350 = 6.48649 + 1.401178 = 7.887668 = 7.89
At 50 opinions and the average still being 8, the new score looks like this:
(50/(50+7))*8 + (7/(50+7))*7.406224807350 = 7.01754 + 0.909534 = 7.927074 = 7.93
This method has now prevented an artifical score inflation by a bunch of fanboys, but then yields a score closer to the actual average as the number of opinions grows. The more opinions a video has, the more valid the score becomes.
I hope this explanation helps.
Phade.
- mckeed
- Joined: Tue May 15, 2001 1:02 pm
- Location: Troy, NY
- Contact:
- godix
- a disturbed member
- Joined: Sat Aug 03, 2002 12:13 am
A week or two ago I went through the top 10% and downloaded whatever I hadn't seen. A lot of what I downloaded was crap that I assume got up there just because eight of the creators friends were incredibly nice in their reviews. After Phade did this change I went and downloaded whatever I didn't have on the list again. I can't say I liked all of them but this time there wasn't one listed that made me think 'WTF? How did this get on this list?' I prefer the list to be what it claims to be, the top 10% of all videos made and this recent change seems to be doing a better job at it than before. Thanks Phade.
- Phade
- Site Admin
- Joined: Fri Oct 20, 2000 10:49 pm
- Location: Little cabin in the woods...


