| |
booker
Account closed
Registered: Jul 2003
Posts: 333 |
Unfair charts voting results
Hi, allow me just to quote the best idea I've read so far about voting system:
Quoting enthusiAnd Im not growing tired of suggesting to just count all votes > 8 accumulative for gfx,music,code.
6 * vote 10 = 60 points
20 * vote 8 = 160 points
etc.
Maybe someone who is (even ;-) more into database stuff than me can play with CSDb votes and give this approach a try?
Last time I checked that manually I was pretty content with the overall results.
Seems to be very fair option and as well relatively easy to implement using already existing data.
Even then bugs like 16th place of T.C. with 4 votes (36 pts. total) vs. 16th place of Hayes (46 pts. total) vs. Scortia 18th (138 pts. total) would be solved automatically ;)
|
|
... 36 posts hidden. Click here to view all posts....
|
| |
enthusi
Registered: May 2004
Posts: 675 |
Groepaz: read again please. You missed the point DRAMAtically ;-)
The suggestion asked for votes > 8.
Where 8 = $ANY_FUNNY_NUMBER_THAT_REFLECTS_A_POSITIVE_VOTE
Still many 8-votes outweight fewer 10-votes.
And I think that is a good thing. Apply it and see. |
| |
chatGPZ
Registered: Dec 2001
Posts: 11116 |
ah so that involves even more pointless magic then ?
so what about these 1 and 2 and 3 votes in your system then? why would only > 8 qualify?
really, this all yet again boils down to: more votes needed. period. trying to get meaningful results from 10 votes is futile, no matter how much magic tricks you apply. |
| |
Skate
Registered: Jul 2003
Posts: 490 |
solution is simple. announce a date to remove hidden vote option. down voters will automatically change or delete their votes before that date.
1) lamers won't be exposed (even if they should).
2) many products/groups/sceners will reach a closer value like they deserved.
of course there is nothing to do about the ones who are shameless to downvote publicly. let the evil take care about them later. ;)
forget about math. that's the solution. |
| |
STE'86
Registered: Jul 2009
Posts: 274 |
exactly why is it more valid to have 20x8 votes than it is to have 6x10?
It seems to me that such as system, would in effect only positively affect "known" current "faces" on here who could call upon lots of associates and fellow group members to advance their standing.
so "unknown" freshers whose work could well be streets ahead of the "faces" get penalised.
also it is a matter of historical fact that MANY 25 year old entries on here are also far ahead of many current ones. as one sees when a piece of art appears on the "radar" thats 20+ years old and you get the "my god thats amazing" comments. but only 5 people have ever voted on it.
a new fair system?
sorry. think again.
Steve |
| |
chatGPZ
Registered: Dec 2001
Posts: 11116 |
Quote:solution is simple. announce a date to remove hidden vote option. down voters will automatically change or delete their votes before that date.
and i will remove mine as well, and so will many other regular voters.
no, making the votes public does not solve anything. it's out of the question too. |
| |
Joe
Registered: Apr 2002
Posts: 224 |
The problem seems, to me atleast, still be that too few takes voting serious. There is around 100.000 releases here (?) and roughly only me and Ed holds a margin of 10% coverage of those (in voting), which of course has taken a serious amount of time to review and reflect upon.
Now if everyone took it equally as serious, instead of dismissing the number of amounts of votes per release as a possible guideline I'd guess that the levels would even out and hopefully be more accurate in the end.
/James |
| |
Perplex
Registered: Feb 2009
Posts: 254 |
The now retired Internet Top 100 SF/Fantasy List of books used something they called diffused average to rank the books. Something similar might work well here. It's mainly to avoid new entries shooting straight for the top spots, though.
|
| |
Perplex
Registered: Feb 2009
Posts: 254 |
How about ranking entries based on Bayesian estimates?
Obviously with a smaller value for m than the one they use on IMDb. |
| |
Joe
Registered: Apr 2002
Posts: 224 |
I don't know how the figures work over at Bitfellas Artcity, but it takes only 1 vote (instead of 5) to give a total figure, which of course changes during time. |
| |
Mermaid
Registered: Jan 2002
Posts: 335 |
Booker, here's the best solution to any worries people have about unfair charts:
http://noname.c64.org/csdb/userpage/rightbar.php
Uncheck all the charts, save changes, worries gone! |
| Previous - 1 | 2 | 3 | 4 | 5 - Next |
