Top 10% List "Score" Modified

[dwchang]

I for one understand the argument that lesser known vids won't get as much exposure since well...I took on the devil's advocate and debated the exact same thing with Phade earlier. I don't honestly believe in it, but I did present it.

It's a valid concern, BUT Phade has already provided an answer in the Star Scale. That will contain any vid with even 5 stars and as we have seen over the year that vids come and go near the top every week. There's your answer. If a video truly is good, it'll remain near the top of the star scale and eventually climb in the top 10% list as well.

Also as Phade and Zarxrax have implied, a good video WILL make the list in time. Alan provided plenty of new vids that are near the top and from (then) unknown creators. There videos were good and thus stood against the test of a multitude of opinions.

This method may seem a bit harsh, but think about it, shouldn't such a list be harsh and be open to such procedures (esp. with how well it has been explained)? I think people are more or less arguing about the possibility a newer vid may not make the list, but that's a lot less likely than a bunch of vids "undeserving" vids making it. We already KNOW that happens right? This would prevent such a thing and shouldn't a video need to "work" to get there?

My prediction is that the list will remain fairly stable near the top and a few times a year, a video that is spectacular will continually climb to the top spots. This makes perfect sense compared to the way the old system worked where a vid could start at the top and move up or down. I don't understand why people would find it offensive that a good video has to work it's way up. That makes sense if it's a good video yes?

Also I don't mean to sound like ass, but there's no way to say it....let's be honest, how many videos a year truly deserve top spots like that? Again, this list is meant to be strict and harsh. We want it to be accurate yes? If that's the case, how many Right Now's, Euphoria's, Tainted Donuts, etc. come out in one year (and no I don't mean the type of video, but a video that just captures everyone in the community)?

If you HONESTLY believe you've been "wronged" by this list or something, then prove it. Get more opinions and work your way up there. It sure as hell is more valid (to me) than just popping up near the top and frustrating people who download the vids and expect it to be good. I for one used to not download vids on the Top 10% list with less than 20 opinions.

I don't think it's too much to ask to make the list harsher and more accurate. You guys are basically saying "OMG my video with 8/10/20/XY opinions isn't on the list anymore" or as Snhknives parodied "OMG it's harder to be good!!11" He obviously didn't mean that, but he did voice the biggest complaint, which is in reality, the reason the list exists. To highlight good videos to viewers. Phade's changes coincide exactly with that purpose.

And in the end, you guys do realize we're arguing about an interweb list right?

[mckeed]

ok, this makes a whole lot more sense now. I still think there should be an "up and comming" list though. Yeah, people could still stack it, but they would get shot down pretty quick if they sucked since it would be a much smaller list. Just my 2 cents.

[dwchang]

ok, this makes a whole lot more sense now. I still think there should be an "up and comming" list though. Yeah, people could still stack it, but they would get shot down pretty quick if they sucked since it would be a much smaller list. Just my 2 cents.

Again, Star Scale and weekly Star Scale.

[mckeed]

Again, Star Scale and weekly Star Scale.

Again, You don't quite understand. Let me explain this since the logic escapes you.

Well, for one, not everyone puts their videos up on local. So that thorws your star scale out the window for people who don't put up their videos for local. Not everyone on the current top 10% has their video's up for local download. Most of them do, but not all. I even found one that isn't availiable for download at all. Using a method to identify videos that automatically throws videos out cause they aren't local isn't a very fair system.
Second, Weekly is just weekly and also relies on a local video. What if something slowly gains momentum over a period of months? The normal star scale isn't enough. For one its not really detailed enough to give any real worth of a video. The same "stacking" that could take place with the old scale can and would take place here. Since now they can't stack to get on the top 10%, i would imagine people will start doing it here as well. You could argue this gets rid of opinions that just try to give all 10's. That might be true. Or people will just realize they just need more people to get on the top 15% list. So they go on the internet or local DBZ board and find 16-30 people willing to do this for them. Harder? Yes, but still doable.

Thirdly, a system that relies on opinions encorages people to get opinions to someday make the number of opinions that gets them on the top 10%. The star method is an indirect method at best. If were going to change a system, why not address some issues for why some people liked the old system. It mainly was just exposure. I know my only video started very low on the list and as people saw it, i started getting more opinions on it, which caused it to slowly move up the list. So why not create a supplementary system to the top 10% that accomplishes just that. You argue that the star scales do this. I don't agree that it is the best way of acomplishing this.

I'm not saying the current system is bad. I think its very accurate. I just think there should be a better way of spotlighting videos. That is all.

[dwchang]

I'm not saying the current system is bad. I think its very accurate. I just think there should be a better way of spotlighting videos. That is all.

Then make a suggestion in the site suggestions, that's why they're there.

By continuing to argue within THIS thread, it just makes it look like you're adamantly against this change and I know you don't mean to, but it just makes you look like you're saying "OMG my video with XY opinions isn't ranked as highly as the videos everyone loves and have 200 opinions validating them." I know that isn't the case and you're trying to highlight that newer vids aren't favored, but it sure doesn't come across that way if you continually argue the same things.

In all honesty, if you agree that it's accurate, that ends this argument. If you feel that we should also have a system to highlight new videos, then by all means suggest it...I for one would welcome such a change, but don't know how to implement it and thus haven't suggested anything.

Sorry if this comes across incorrectly, but if you read the replies in the last two pages, it's just one giant circle...then again a lot of .org discussions are like that .

[Phade]

Hey,

Well, for one, not everyone puts their videos up on local. So that thorws your star scale out the window for people who don't put up their videos for local. Not everyone on the current top 10% has their video's up for local download. Most of them do, but not all. I even found one that isn't availiable for download at all.

But they were most likely downloadable at some point. Are you saying that a new video that is not downloadable should easily get onto the list?

Using a method to identify videos that automatically throws videos out cause they aren't local isn't a very fair system.

Nearly 40% of the videos listed in the catalog are locally downloadable. Of the video entries that are not, most of them are bogus entries. The rest of them are old entries of creators who don't come around anymore or just haven't bothered to put them up yet (usually due to "it was encoded in 1998 and a new encode would be much better, but I haven't done it yet")

Second, Weekly is just weekly and also relies on a local video. What if something slowly gains momentum over a period of months? The normal star scale isn't enough.

If a video gains momentum slowly, there must be a reson for it. Generally that reason is because the video is only mildy above average in quality. A video that is very good quality will have no trouble getting recognized because people will generally talk about that vid more. Unless the new video creator likes to spend lots of money on video hosting, they are very likely to put the video onto the local storage server so that people can get to it easily.

For one its not really detailed enough to give any real worth of a video. The same "stacking" that could take place with the old scale can and would take place here. Since now they can't stack to get on the top 10%, i would imagine people will start doing it here as well.

This is not likely to succeed. Videos that are rated high will get more downloads. As it gets more downloads, the average will reflect a more true value. This is the case in the 10% list as well as the star lists.

You could argue this gets rid of opinions that just try to give all 10's. That might be true.

This already happens. The top and bottom 10% scores are removed before calculating the average on the Top 10% list. For the star scale, people who give the same scores all the time (1's, 2's, 3's, 4's and 5's) are removed from the calculation.

Or people will just realize they just need more people to get on the top 15% list. So they go on the internet or local DBZ board and find 16-30 people willing to do this for them. Harder? Yes, but still doable.

Yes, much much harder now than it was before.

Thirdly, a system that relies on opinions encorages people to get opinions to someday make the number of opinions that gets them on the top 10%. The star method is an indirect method at best. If were going to change a system, why not address some issues for why some people liked the old system. It mainly was just exposure. I know my only video started very low on the list and as people saw it, i started getting more opinions on it, which caused it to slowly move up the list. So why not create a supplementary system to the top 10% that accomplishes just that. You argue that the star scales do this. I don't agree that it is the best way of acomplishing this.

So what is the best way to accomplish this?

I'm not saying the current system is bad. I think its very accurate. I just think there should be a better way of spotlighting videos. That is all.

If you can come up with something, show how to do it, and show that it is mathamatically sound, the system will be considereder.

Phade.

[mckeed]

Yeah, that is true. I figured this was the place to do it though since the thread was about the change as people who would complain about it while reading it would post here and wouldn't think to actually go to site suggestions. I was just trying to think ahead. I personally don't feel wronged by the change, but I can imagine certain people would. I was just excited that my video had climbed up to a relatively high position which meant even more people would op it and then the change happened. I personally just want people to see my videos which is why i was arguing for a spotlight section. The old top 10% was just a vehicle for that in my eyes, but by looking at some of the stuff on it is was clear that there were some problems with it so I didn't give it TOO much stock. The new list is excelent in terms of content and have been routinely downloading videos to add to my collection. Some I completely forgot about. Especially since now can track what i got if i use the local downloads.

[dwchang]

I was just excited that my video had climbed up to a relatively high position which meant even more people would op it and then the change happened. I personally just want people to see my videos which is why i was arguing for a spotlight section.

I can understand that since that's how probably most of us feel. The higher on the list generally equates to more downloads and exposure. This is why I agree with you in perhaps having a section, but my contention is that this isn't the forum for that and if as Phade says, if you can come up with a mathematically sound way to do this, I'm 100% down with it. It's not like I don't like watching new vids .

The old top 10% was just a vehicle for that in my eyes, but by looking at some of the stuff on it is was clear that there were some problems with it so I didn't give it TOO much stock. The new list is excelent in terms of content...

Exactly.

It may be more difficult to make the list and you have to "climb," (which I like since I can put faith in the results) but ultimately the list itself is very sound and I challenge anyone to disagree that the top 20 vids AREN'T top videos on the site. Sure they may not be your favorites, but as godix said, you can at least realize why they're there. And with ~100 people agreeing well...that's pretty sound imo .

I personally put very little faith in the list in the past and I imagine that probably contributed to lack of enthusiasm in watching vids lately. Now I can be fairly confident with vids on that list and with Phade's additions to the Super Search, perhaps I'll find some gems I missed out on.

All in all, Phade has been a coding monster lately and we're reaping the benefits. Thanks Phade!

[mckeed]

Hey,

Second, Weekly is just weekly and also relies on a local video. What if something slowly gains momentum over a period of months? The normal star scale isn't enough.

If a video gains momentum slowly, there must be a reson for it. Generally that reason is because the video is only mildy above average in quality. A video that is very good quality will have no trouble getting recognized because people will generally talk about that vid more.

I don't quite agree with this statement. Some people don't pimp their videos alot or it only showed at one con or something. It took over a year for one of my videos to make the old 10%. Once it got there there were more validating reviews and I got many more oppinions on it. I'm sure there are many creators who probally have similar stories.

You could argue this gets rid of opinions that just try to give all 10's. That might be true.

This already happens. The top and bottom 10% scores are removed before calculating the average on the Top 10% list. For the star scale, people who give the same scores all the time (1's, 2's, 3's, 4's and 5's) are removed from the calculation.[/quote]
Ok, I was playing devil's advocate to myself so this proves that using the old 10% system combined with an algorithm that would compare both lists and remove videos that are on the new 10% would be decently effective as far as a publicity vehicle. As the bogus opinions would be removed when calculating for the new top 10% anyway.

Thirdly, a system that relies on opinions encorages people to get opinions to someday make the number of opinions that gets them on the top 10%. The star method is an indirect method at best. If were going to change a system, why not address some issues for why some people liked the old system. It mainly was just exposure. I know my only video started very low on the list and as people saw it, i started getting more opinions on it, which caused it to slowly move up the list. So why not create a supplementary system to the top 10% that accomplishes just that. You argue that the star scales do this. I don't agree that it is the best way of acomplishing this.

So what is the best way to accomplish this?
[/quote]

In summary (Old 10% scale - videos on current top 10%) then take top 20 on that list. Maybe you could "retire videos" off the list that hadn't had an opinion in over 4 months or something. Or once it hits the "min votes" level it gets removed?

I'm not saying the current system is bad. I think its very accurate. I just think there should be a better way of spotlighting videos. That is all.

If you can come up with something, show how to do it, and show that it is mathamatically sound, the system will be considereder.
[/quote]
Math isn't my strong suit, and I don't know how to begin to program this. It seems like if the top 10% was good for so long, it might be fine for such a small list like this anyway since it required a decent amount of reviews to get on the list. Yeah, it could be stacked, but if people know that the list is just a exposure kind of thing. It might not be to bad.

[mckeed]

crap i screwed up the quoting....i'm sure you can figure it out though based on phade's previous post.