| |
booker
Account closed
Registered: Jul 2003
Posts: 333 |
Mr. Rating Balancer - hello !
Mister :D Whoever (or whatever scripted) you are and scanning people's/prod's ratings and "voting" low, like 4 or 6 to keep those people/prods suddenly not appearing high on lists you are my hero and I love you. [ Just seen this "voting" again on Stinsen's rating page ].
C'mon, don't be shy, reveal yourself. :) |
|
... 56 posts hidden. Click here to view all posts....
|
| |
jailbird
Registered: Dec 2001
Posts: 1576 |
Quoting Perff from 2002:
"I will not describe exactly how it is calculated, but I'll try to come with a brief explanation here.
The weighted average is a modified mathematical average which take into account the spread and number of votes.
This results in that the weighted average will always be smaller (or the exact same) as the mathematical average.
The higher the spread and/or the fewer votes, the lower the weighted average is compared to the mathematical average.
Noone said that this is the correct way to do it, but it's better than simply using the mathematical average."
Also Perff, a year later:
"As I posted in another tread I have made a new 'weighted average' function that tries to find and eliminate the fake (extreme) votes."
Not sure how the algorithm works (I'm assuming it removes extreme downvotes and upvotes), but if it was up to me, I'd remove all those votes from the calculation which are unusually "lonely" and strike out from the average as well. |
| |
CreaMD
Registered: Dec 2001
Posts: 3035 |
Calculating charts only from public votes would show respect towards people who vote publicly. As far as anonymous votes are concerned, they can still be counted in the product detail page.
edit: people who don't care, can continue not caring, and those who do, can finally spend their time lobbying, accusing, arguing and disrespecting the shameful public downvoters. |
| |
JackAsser
Registered: Jun 2002
Posts: 1989 |
Again: more voters will make "fake"-votes insignificant to the end result. That is the -only- problem. |
| |
JackAsser
Registered: Jun 2002
Posts: 1989 |
-> solution. Require login. Let downloading cost pointa. Earn points by voting. :) |
| |
Yazoo
Registered: Nov 2006
Posts: 227 |
agree to jackasser... more voters... problem solved |
| |
booker
Account closed
Registered: Jul 2003
Posts: 333 |
Quote: Calculating charts only from public votes would show respect towards people who vote publicly. As far as anonymous votes are concerned, they can still be counted in the product detail page.
edit: people who don't care, can continue not caring, and those who do, can finally spend their time lobbying, accusing, arguing and disrespecting the shameful public downvoters.
Actually prods got public/ano ratings displayed, why not the same for ppl?
More voters needed yeah (not gonna happen).
Anyways there's few words about crowd sourced ratings in "How to predict unpredictable" by the guy who wrote "Are you smart enough to work for Google". Vaguelly remember ratings with average scores lie, you should look at those which are occuring most, because they tell how the most of the public is voting. If you see the example I gave above (which is why I acutally did this thread) a lone, one low vote is drastically changing the scoreboard position. However this one vote is only 5% of all votes.
Any bells ringing regarding proper display of how crowd is voting vs scoreboard mess caused by ONE voter? :) |
| |
CreaMD
Registered: Dec 2001
Posts: 3035 |
More votes doesn't completely prevent destructive effect of manipulative anonymous voting. Effect is visible in highly competitive categories. Even single vote can change the position of entry in chart. |
| |
HCL
Registered: Feb 2003
Posts: 717 |
Why is 6 downvoting? Only if you compare to all the 10s, but perhaps not compared to what that person voted for other entries.
Now, Joe/Wrath Designs already explained all this last year (or was it two years ago? probably 5 then :P).
My own opinion is that the 10s are actually the problem. The voting system degrades after a few years, and today the only sane vote you can give is a 10 for a good prod, 9 for an average, and 8 for some really ugly shit. Pehaps the one that voted 6 on Comaland actually took the time and reviewed all his/her votes and normalized them around 5, and the voting system became useful again.. |
| |
CreaMD
Registered: Dec 2001
Posts: 3035 |
When adjusting happens on single entry and kicks it out of first place in charts it can also be easily mistaken for downvoting. Adjusting can't happen selectively on just some entries.
And for votibg system becoming useful again... I think it needs some assistance from the developer of this site. He could make some kind of adjusting / decreease the value of points? / over the time... |
| |
booker
Account closed
Registered: Jul 2003
Posts: 333 |
Quote: Why is 6 downvoting? Only if you compare to all the 10s, but perhaps not compared to what that person voted for other entries.
Now, Joe/Wrath Designs already explained all this last year (or was it two years ago? probably 5 then :P).
My own opinion is that the 10s are actually the problem. The voting system degrades after a few years, and today the only sane vote you can give is a 10 for a good prod, 9 for an average, and 8 for some really ugly shit. Pehaps the one that voted 6 on Comaland actually took the time and reviewed all his/her votes and normalized them around 5, and the voting system became useful again..
There's something in what you say, perhaps voting system should get +5 points :)
Vote 6 is statistically best to ruin the rating, that's why it's easy to use it.
There's difference between down voting in purpose, and rating properly, though.
When you see most % of votes are around 7,6,5 and there are only like 5% of 10 or 9 then this looks like an attempt to up vote a shitty prod, simply, or someone might just have a mood to give it high, or just thinks differently.
Likewise 5% of 6 on something 95% of crowd voted 10,9,8 is simply either a deliberate try to control the scoreboard, or someone just has a different opinion.
THAT"S FINE FFS
Problem is when you have 5% of votes DRASTICALLY mangling on the scoreboard. (see my example earlier post 14). That's why the rating should ignore the lowest % of votes. Then the scoreboard will more accurately reflect the crowd rating. |
| Previous - 1 | 2 | 3 | 4 | 5 | 6 | 7 - Next |
