Search
 
 
 
 
 
Search Pages:           Advanced
VWAR Ratings
VWAR (Volume Weighted Average Rating) is a new method of calculating a submission's overall CSA rating by taking into account the number of assessments as well as their values. The VWAR was introduced in recognition of the fact that submissions with more assessments are generally better - that is to say, a shortage of assessments is an assessment in itself. The previous calculation method meant that an unpopular submission with five 10/10s (overall 10/10) could out-rate a popular submission with twenty-eight 10/10s and two 9/10s (overall 9.9/10).



a = Assessments Aggregate
The sum of all assessments made on the submission. If a submission has two 10/10s and three 8/10s, then the aggregate is 10+10+8+8+8 = 44.

b = Number of Assessments
The number of assessments the submission has received.

c = Category's Record Number of Assessments
The submission's category's average number of assessments per submission. For popular categories, this will be higher, and for less popular games it will be smaller. This variable ensures less popular games are not penalized.
Published by tom ::. 16 days ago
Promote Me
Category: Documentation
Comments
Page of 13   1 2 3 4 ... 11 12 13 Next Page
Darkstorn .
no. | Y2M
Posted 2 hours ago
Posted by Weighted Companion Snake
Laziness and Apathy. People are lazy, or in a hurry, or simply not sure what to say sometimes, therefore, they had the option to not assess a submission.


Exactly, i also made this point a few pages back or something.

I've personally, for the entire 4+ years i've been here, assessed a total of _28_ submissions.
Thesy .
The Weighted Snake. | Y2M
Posted 2 hours ago
I apologize to once again add to an what-seems to be a long-going debate, but I hope to point out the one actual flaw that I see.

Laziness and Apathy. People are lazy, or in a hurry, or simply not sure what to say sometimes, therefore, they had the option to not assess a submission.

I repeat, sometimes I've gone through an entire section, seen some horrible things and some excellent things, and not assessed more than a few. I know I'm not the only one.

Just because someone doesn't assess something doesn't mean they don't like it or dislike it, infact, usually when I don't assess something, it's because I'm not sure what to rate it.

You're 100% right though Tom, it's not an art gallery, it's a website. People who go into Galleries and critique generally have a life around that, people who get paid just to say something witty about some French painting.

This is a website where people go to customize their games, not critique pieces of art, how many guests do you get in a day? Do you get more guests than users? Well Guests can't assess things, this system is basicly saying that all guests think that all the submissions should be rated lower.

I'm aware that's a bad way to word it, but you can't deny the fact that this system effectively causes everything to be rated lower, simply because you allow Guests to view content.

Which actually made me think of something else, I'll leave a message in your profile about that though.
Darkstorn .
no. | Y2M
Posted 5 hours ago
Posted by tom
Why don't you go over to imdb.com or rottentomatoes.com and check the ratings of a handful of your favorite movies then tell me why every one of your submissions is so utterly perfect in every conceivable way that it's offensive to receive anything less than a 10 for them.


I don't watch movies much. But my submissions aren't perfect. I'd still say they aren't worth a 5, when the critic rating is already 10.

I'd like to point out to you that NONE of my submissions had a 10/10. TWO submissions had one even close, the majority were around 9.0 and a few were below that. Now i have two submissions that are almost 10/10, and pretty much ALL the rest have fallen below 6/10.


This is a website, not a contemporary art gallery. The observation that more ratings are conducive to a better submission is accurate in the majority of cases.


It is NOT accurate, and it is NOT better.

How is the average meaningless? Are you fucking serious(excuse my French)? USERS give the assessments. If you want accurate and better scores, you should try to change the people. But before that happens, the average _is the best and most accurate way to express user opinions about it_. It can be unfair, because people have objective opinions but hell, now it does the SAME but with the addition of a retarded variable... People can STILL abuse the system as much as they want to, now it just punishes those who DON'T.


That's funny, because I'd appreciate it if you'd pay a bit more attention to what I've been saying which you clearly haven't been.


I have, i just think your answers were conceived by a 3-year-old who didn't think it further...

You haven't made any comparisons to the old rating system


The average is the most fair, accurate and simply the best way. The new does not even NEED user input for this, EVEN if you would count only 10's rather than all the ratings, you'd still get a result VERY similar to how it's now... Broken.

You claim that it's unfair to see an submission with less assessments get a better score, i ask of you; HOW? You clearly have absolutely zero clue about art...

If a submission gets 10 tens and thus an average of 10, and another gets 28 tens and 2 nines, it'll have worse average. AND IT IS FOR A REASON; It got a few nines. If that submission with lesser assessments would even be given a SINGLE 9, its average would be WORSE.

If a model gets 2 nines even if their other ratings are 10, it deserves it. Did you see me whining about it? I had two models rated 9.9 because of this VERY same thing. And i didn't complain.

Tell me how this system is NOT meaningless? It totally and utterly removes actual user input, it cares only for the amount of those assessments. I've seen 6 average submissions get 10 several times, and 10 average submission get 6. Is that right to you?


Please do elaborate on all these problems...


Seriously, i suggest you either learn how to read, or stop putting words into my mouth.

Its major problem happens to be the fact that it ignores the average; Even schools use averages to determine final scores, not some retarded ill-thought system that takes only into account one thing when it should take them all, or none.

These are user assessments, for works of art. Sure, if you insist on that it has nothing to do with art, put this system to scripts and ignore it from the areas you have zero knowledge about.

I mean, I've only been developing this site for 5 years and I've only got one university degree in design so I'm sure you know a lot more about the artistic process, critiquing and peer assessment.


That only proves that i do... Because you make it sound like you need an university degree in order to know what rating system is the best. I'll raise you with my 12 years of modeling and texturing experience....

What do i win?

What most people don't understand is that running a resource website involves a constant struggle between the needs of artists and the needs of those downloading the art.


You don't seem to understand. There were MUCH less 10/10 submissions around than you think, they just tend to be the most visible ones... NONE of my stuff had 10/10 for example.

Also, you seem to ignore the fact that i'm not complaining because my scores fell from 10/10 to 8/10... No, only one did. The rest went from close to 10/10 to _6.0 or less_.

I accept the fact that you're constantly trying to tweak it to fit the needs of the users better, but this is one of those things you should just not tweak at all... Maybe some people have complained about the averages being 'unfair'(it is not), but i'd say a hell of a lot more people dislike this.

The top rated skins section is also utterly broken now...

This system is broken, and it is not any more fair, now it requires people to spam those 10's rather than people to earn it... This fixes nothing, people will still give the 10, now they just have to give MORE of them...

Not to mention you totally and utterly ignored my argument that this place is old, a VERY LARGE AMOUNT of submissions were released before the assessment system ever became reality. What do you think of those? Majority of my releases were like that, the only reason my AK got so many assessments is that it was / is the most downloaded AK on this site... I have _two_ submissions with more than 10 assessments... The rest have closer to 5.

The problem is that this system ignores reason. It gives newer models an advantage because they're more visible in the new skins section. Old skins are _very very very very_ hard to find. Especially now as their scores are much lower...

/E: Had to cut down the quotes a lot, not because i ignored it, rather there's a character limit for posts. Refer to the original posts for what he said.
Шipeout .
5th gear pinned! ^.^ | Y2M
Posted 5 hours ago
Well sounds like a great idea.. Im sure it will make the rating scheme more fair
lordshenlong .
Indeed | Y2M
Posted 13 hours ago
So the rating is based on how many ratings ye get as well as the scores of each one?
I think people may be a tad unhappy cos not everything gets rated anyway, and there's plenty of crap stuff that has been rated higher than some good stuff that has no rating at all, or a low one cos of few people rating it.
::.
? | Y2M
Posted 14 hours ago
Posted by Weighted Companion Snake

Honestly, I really hate this.

Anything that doesn't have atleast 5 10 out of 10 assessments is rated beginner, which not only makes it harder for skins to be seen at all, it also pushes down older subs that never got many assessments, as well as more unpopular games' sometimes quite good submissions.

Nice try, but honestly I think this just pushes down everything's rating except for people like Elfa/Snake have hundreds of people in their watch list.

Even if that doesn't convince, think about this; just because someone doesn't assess doesn't mean it's bad.

I think all of Millenia's work is amazing, am I going to go and assess every single submission of his? No. And there are people more lazy than me, I only actually assess about half the stuff I look at.


This is mostly generalization and exaggeration. It also demonstrates that you haven't read or haven't understood the documentation on this or understood the flaws of the old system. Provide specific examples to support your case next time.

Posted by Fuckles_The_Cat

Lesson to your users Tom. Most people hate this.


Really? "most" as in 51% of registered and unregistered users on FPSB, or "most" as in "everybody is going to this party dude, everybody". Please forward me the survey you conducted.

Posted by Darkstorn

I think it's quite a bit more extreme than that, one of my submissions has nine 10's and one 9 and its 'average' is suddenly 8.04, and one has three 10's and one nine and is rated _5.95_ and a critic rating of 10. Some stuff is plain in danger of being trashed regardless of their old good rating...


The trasher threshold has been changed to compensate for harder ratings. Why don't you go over to imdb.com or rottentomatoes.com and check the ratings of a handful of your favorite movies then tell me why every one of your submissions is so utterly perfect in every conceivable way that it's offensive to receive anything less than a 10 for them.

Posted by Darkstorn

Seriously Tom, again you can't expect to rate art by a retarded equation. Doesn't help that it takes the amount of assessments as more important as the assessments themselves... A release with an average of 6 but enough assessments can still get a score close to 10... And the top rated skins section is now seriously broken because of the system.


This is a website, not a contemporary art gallery. The observation that more ratings are conducive to a better submission is accurate in the majority of cases. For the tenth time though, this system is not perfect, but it's a hell of a lot better than the old rating system which was utterly meaningless.

Posted by Darkstorn

I'd at least appreciate if Tom would pay attention to what i've said, but it's obvious that he doesn't understand game customization. This is just another attempt at gaining more popularity for the place, but if it angers those who CREATE the content, then something's wrong, no?


That's funny, because I'd appreciate it if you'd pay a bit more attention to what I've been saying which you clearly haven't been. You haven't made any comparisons to the old rating system, which I should remind you was scrapped in response to community suggestions and my own observations of its meaninglessness.

Posted by Darkstorn

This system causes far more problems than it fixes. It is NOT a good way of reviewing pieces of art. And i know what i'm talking about, Tom doesn't. The moment he chose to try and defend this by proving people wrong with more equations, he lost. You. Cannot. Rate. Art. With. A. System. Like. This.


Please do elaborate on all these problems you have identified as a discerning art critic and expert on internet communities and user behavior. I mean, I've only been developing this site for 5 years and I've only got one university degree in design so I'm sure you know a lot more about the artistic process, critiquing and peer assessment.

What most people don't understand is that running a resource website involves a constant struggle between the needs of artists and the needs of those downloading the art. If every artist gets a 10, they may feel a false sense of happiness, but for those trying to appreciate the art (end-users) they are left with a meaningless rating system where they can't distinguish between good work and great work. Darkstorn, even if you are getting an 8 out of 10 it's still a decent rating and it'll still be distinguished from the pack better than it was before where it would be hidden amongst a raft of 10 rated recolors. The new rating system isn't perfect, but it has meaning - it factors in the observed internet behavior that better submissions get more ratings - not in all cases but in the vast majority of cases. Internet systems are not perfect, they never will be. You need to accept this, and accept the fact that I'm continually trying to tweak things and make them work better, and that changes don't always benefit all stakeholders in the system, because they naturally can't.
-
Fuckles_The_Cat .
Obama: the other white meat. | Y2M
Posted 14 hours ago
Lesson to your users Tom. Most people hate this.
Darkstorn .
no. | Y2M
Posted 1 day ago
Posted by Weighted Companion Snake

Anything that doesn't have atleast 5 10 out of 10 assessments is rated beginner


I think it's quite a bit more extreme than that, one of my submissions has nine 10's and one 9 and its 'average' is suddenly 8.04, and one has three 10's and one nine and is rated _5.95_ and a critic rating of 10. Some stuff is plain in danger of being trashed regardless of their old good rating...

Seriously Tom, again you can't expect to rate art by a retarded equation. Doesn't help that it takes the amount of assessments as more important as the assessments themselves... A release with an average of 6 but enough assessments can still get a score close to 10... And the top rated skins section is now seriously broken because of the system.

I'd at least appreciate if Tom would pay attention to what i've said, but it's obvious that he doesn't understand game customization. This is just another attempt at gaining more popularity for the place, but if it angers those who CREATE the content, then something's wrong, no?

This system causes far more problems than it fixes. It is NOT a good way of reviewing pieces of art. And i know what i'm talking about, Tom doesn't. The moment he chose to try and defend this by proving people wrong with more equations, he lost. You. Cannot. Rate. Art. With. A. System. Like. This.
Thesy .
The Weighted Snake. | Y2M
Posted 1 day ago
Honestly, I really hate this.

Anything that doesn't have atleast 5 10 out of 10 assessments is rated beginner, which not only makes it harder for skins to be seen at all, it also pushes down older subs that never got many assessments, as well as more unpopular games' sometimes quite good submissions.

Nice try, but honestly I think this just pushes down everything's rating except for people like Elfa/Snake have hundreds of people in their watch list.

Even if that doesn't convince, think about this; just because someone doesn't assess doesn't mean it's bad.

I think all of Millenia's work is amazing, am I going to go and assess every single submission of his? No. And there are people more lazy than me, I only actually assess about half the stuff I look at.
.
It's called feedback for a reason. | Y2M
Posted 2 days ago
I dont like this idea D:
Page of 13   1 2 3 4 ... 11 12 13 Next Page
Add Comment
You must login or register to access this!