Log inRegister an accountBrowse CSDbHelp & documentationFacts & StatisticsThe forumsAvailable RSS-feeds on CSDbSupport CSDb Commodore 64 Scene Database
You are not logged in - nap
CSDb User Forums


Forums > CSDb Discussions > The voting system is completely retarded
2018-03-31 19:37
jailbird

Registered: Dec 2001
Posts: 1576
The voting system is completely retarded

I just can't wrap my head around this unmitigated, pure crap we use for voting. Here's a situation where I actually have to vote waaaay lower than it deserves, in order to increase a release's rating.

This is simply ridiculous.



"Hey, this release is really cool, I'll give it a tenner."




"WTF, the rating got lower?!? Ok, maybe it counts as an illicit upvote and it is filtered with some clever algorythm, let's go with 9 then."




"Damn, now it's way lower than it would be without my vote. Maybe an 8 will help?"




"Nah, this is getting worse, but let's see what happens if I vote with a 7."




"Shit, still no good, I'll jump to an ever lower number just for the sake of experiencing more of this shit."




"Oh great, now we're getting somewhere, one more try."




"YEY! I love this release, so it will get a 6 from me."
 
... 66 posts hidden. Click here to view all posts....
 
2018-04-02 08:47
Rastah Bar

Registered: Oct 2012
Posts: 336
Quote: I just can't wrap my head around this unmitigated, pure crap we use for voting. Here's a situation where I actually have to vote waaaay lower than it deserves, in order to increase a release's rating.

This is simply ridiculous.



"Hey, this release is really cool, I'll give it a tenner."




"WTF, the rating got lower?!? Ok, maybe it counts as an illicit upvote and it is filtered with some clever algorythm, let's go with 9 then."




"Damn, now it's way lower than it would be without my vote. Maybe an 8 will help?"




"Nah, this is getting worse, but let's see what happens if I vote with a 7."




"Shit, still no good, I'll jump to an ever lower number just for the sake of experiencing more of this shit."




"Oh great, now we're getting somewhere, one more try."




"YEY! I love this release, so it will get a 6 from me."


My guess of what is happening: a reference average vote and a spread around the mean are calculated. Perhaps the median and standard deviation. The final average vote is a weighted average of all votes where the votes are progressively downweighted according to their distance from the reference average, where the "distance" is normalized by the spread.

This means that if you vote low, the spread increases and some of the high votes will get more weight. Resulting in an increase in the final weighted average vote. But if you look at what happens when the vote becomes lower and lower, at some point the weights of the highest votes have maxed out, and the decrease in weight of the low vote does not compensate for its actual lower value and then the final weighted average vote can go down again.
2018-04-02 09:27
jailbird

Registered: Dec 2001
Posts: 1576
Sooo... Magic? Got it :)
2018-04-02 09:52
ilesj

Registered: Jun 2012
Posts: 27
Mr. Ammo and Moloch raise a good point. The average rating calculation may be faulty, but so is peoples' voting habits. Any new release gets votes only in the 8-10 range routinely, with the few random "downvotes", unless it's somehow a crappy release. Anything below 8 is considered a downvote, apparently.

How well a release is received can in practice be judged by the number of nines and tens it receives, not by the (weighted) average rating.

I see biggest problem with compos held on CSDb with CSDb voting. In a compo voting I would like to use the full 1-10 scale to differentiate the entries from each other. But CSDb voting puts me off, because giving anything below 8 is "being an asshole".
2018-04-02 10:20
DKT

Registered: Aug 2004
Posts: 96
What ilesj said.
And... Compos should not use CSDB voting system, because we can't compare vote value 10 for new release with vote value 10 for Crest Avantgarde or Dawnfall, a classic productions. It's better to provide external simple vote-sheet and then full range of votes will be used by people (like on parties).
Beside this, people don't have time to vote (to provide good result with efficient number of votes) so who cares...
2018-04-02 10:41
jailbird

Registered: Dec 2001
Posts: 1576
Quote: Mr. Ammo and Moloch raise a good point. The average rating calculation may be faulty, but so is peoples' voting habits. Any new release gets votes only in the 8-10 range routinely, with the few random "downvotes", unless it's somehow a crappy release. Anything below 8 is considered a downvote, apparently.

How well a release is received can in practice be judged by the number of nines and tens it receives, not by the (weighted) average rating.

I see biggest problem with compos held on CSDb with CSDb voting. In a compo voting I would like to use the full 1-10 scale to differentiate the entries from each other. But CSDb voting puts me off, because giving anything below 8 is "being an asshole".


Seems that the main culprit of the annoyance is that there's a voting statistic for the releases, so people can see when someone votes undesirably.
Why is this list of number of votes per value necessary at all?
Why are the votes of anonymous voters public anyway?
2018-04-02 11:04
ilesj

Registered: Jun 2012
Posts: 27
IMO, seeing what kind of ratings are given, and how many, tells a lot more than the stupid weighted average. If there would be just the average rating without any details on the given ratings, that would tell absolutely nothing.

Why is the calculated average rating so important?
2018-04-02 11:20
jailbird

Registered: Dec 2001
Posts: 1576
Quoting ilesj
IMO, seeing what kind of ratings are given, and how many, tells a lot more than the stupid weighted average.
Is it? When the majority votes with 8-10 anyway? That doesn't seem too useful to me. But I am not sure I'd be able to draw any conclusions from the votes even if they'd be more impartial, really.

Quoting ilesj
If there would be just the average rating without any details on the given ratings, that would tell absolutely nothing.
This system as a whole tells absolutely nothing either, except from how many supporters/friends and/or haters you have on the scene.

Quoting ilesj
Why is the calculated average rating so important?
It isn't, that's the point.
2018-04-02 12:13
No-XS

Registered: Mar 2002
Posts: 78
Can I say something? I will anyway...

I never realized you could vote that way. I have always used the blue 'vote' button. And then it seems to work?!

So the 'rate this release' sidebar widget is broken.
2018-04-02 12:52
iAN CooG

Registered: May 2002
Posts: 3132
Quote: Can I say something? I will anyway...

I never realized you could vote that way. I have always used the blue 'vote' button. And then it seems to work?!

So the 'rate this release' sidebar widget is broken.


uh? no, they are the same, what makes you think they act differently? Have you noticed differences using one or another?
2018-04-02 13:11
No-XS

Registered: Mar 2002
Posts: 78
No. Like I said, I never used the voting from the sidebar. But by using the button, I never saw my voting go wrong. So I tested it. And When I vote 10, the number goes up. When I vote 1, the number goes down.

Now I tested it with the sidebar vote, and it also goes right for me.

So not sure then. Was trying to provide some more information.
Previous - 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 - Next
RefreshSubscribe to this thread:

You need to be logged in to post in the forum.

Search the forum:
Search   for   in  
All times are CET.
Search CSDb
Advanced
Users Online
Frostbyte/Artline De..
Didi/Laxity
Guests online: 129
Top Demos
1 Next Level  (9.8)
2 Mojo  (9.7)
3 Coma Light 13  (9.7)
4 Edge of Disgrace  (9.6)
5 Comaland 100%  (9.6)
6 No Bounds  (9.6)
7 Uncensored  (9.6)
8 The Ghost  (9.6)
9 Wonderland XIV  (9.6)
10 Bromance  (9.6)
Top onefile Demos
1 It's More Fun to Com..  (9.8)
2 Party Elk 2  (9.7)
3 Cubic Dream  (9.6)
4 Copper Booze  (9.5)
5 Rainbow Connection  (9.5)
6 TRSAC, Gabber & Pebe..  (9.5)
7 Onscreen 5k  (9.5)
8 Wafer Demo  (9.5)
9 Dawnfall V1.1  (9.5)
10 Quadrants  (9.5)
Top Groups
1 Oxyron  (9.3)
2 Nostalgia  (9.3)
3 Booze Design  (9.3)
4 Censor Design  (9.3)
5 Crest  (9.3)

Home - Disclaimer
Copyright © No Name 2001-2024
Page generated in: 0.064 sec.