| | Merman
Registered: Dec 2002
Posts: 140 |
Transparent votes?
There's been a lot of "discussion" in the one-liner box about whether voting should be transparent, i.e. you can see who voted (and their vote's value).
I feel that things should stay as they are, just the number of votes and "weighted" vote shown for each entry. That would save all the arguments "why did you vote low for me?" or "he/she is a lamer, they voted low for me/my release"
What do others think? |
|
... 116 posts hidden. Click here to view all posts....
|
| | Steppe
Registered: Jan 2002
Posts: 1510 |
I bow down to the wise decision and will of course remove my 1 from Arcanum immediately. =) |
| | Cruzer
Registered: Dec 2001
Posts: 1048 |
Why isn't the survey transparent? :) |
| | CreaMD
Registered: Dec 2001
Posts: 3048 |
Cruzer I have to shout this with applause: *very good point* |
| | WVL
Registered: Mar 2002
Posts: 896 |
why not just calculate the average from the transparent votes only? everybody can give their vote, whether they want them public or not, but they just won't count in the average... |
| | CreaMD
Registered: Dec 2001
Posts: 3048 |
Quote: why not just calculate the average from the transparent votes only? everybody can give their vote, whether they want them public or not, but they just won't count in the average...
I was suggesting this solution too. Simply to disable non-transparent votes on default, and have possibility to enable them if anyone is interested.
I've got that *cheschire cat smile* on my face again.. I mean, yeah, it's not a very nice solution towards anonyms, but I don't see any reason in having any special affection or tolerance towards such type of voters. |
| | Board Rider
Registered: Jan 2002
Posts: 15 |
I think it would be good to see who voted for what and how the feel about the release too...
Board Rider/CSD! |
| | fade
Account closed
Registered: Mar 2002
Posts: 290 |
board rider: thats what comments are for.. :)
How about a good, bad or ugly icon, with text on_mouseover "fade hates this" "fade would let this one tickle his twig and berries" you catch my drift.. Just PLEASE enough whinging about it.. I hate hearing this shit and unfortunately it isnt Big Brother so nobody is getting evicted from the scene |
| | _V_
Account closed
Registered: Jan 2002
Posts: 124 |
Voting, imho, should be an honest, anonymous process done by many, many people. If there's no honesty, there's no point in interpreting the votes. |
| | fade
Account closed
Registered: Mar 2002
Posts: 290 |
maybe we can catch some fairies while we wait for any honesty. And maybe find a fakelabel that understands what being a fakelabel is all about. How about everyone fucks off this pointless topic and does some c64 datawork because this is as pointless as a conversation on religion. Someone put their foot down and end it all.. argh! :) |
| | _V_
Account closed
Registered: Jan 2002
Posts: 124 |
On the way home, I recalled something which might be interesting... if these apparent downvotes are a problem, why not switch to the much stabler median to calculate the central value instead of the mean?
If you have an ***ordered*** sequence of N numbers:
x.1, x.2, x.3, ..., x.N
Then the median is defined as:
x.(N+1)/2 (if N is odd)
(this is, simply, the value of the number in the middle)
or 1/2 * ( x.N/2 + x.N/2+1 ) (if N is even)
(this is the arithmetical mean of the values of the two middle numbers)
The median is statistically more robust than the arithmetical mean, especially when the amount of numbers N is small (which seems to be the case for most prods). It is the number which literally stands in the middle of the other numbers: there are as many numbers on the right as there are on the left.
So for example, let's say we have 10 ordered votes for a prod as follows:
6, 7, 7, 7, 7, 8, 9, 9, 9, 10
Then the two middle numbers are 7 and 8, and the median is 1/2*(7 + 8) = 7.5
Note that the arithmetical mean is 7.90 in this case.
However, now let's assume some hip kid decides to downvote and adds a 1 to the sequence:
1, 6, 7, 7, 7, 7, 8, 9, 9, 9, 10
The median now is 7, whereas the arithmetical mean is 7.27. So the median drops 0.5, while the mean drops 0.63, which is a bit more. It gets worse if we add another downvote, let's say a 2:
1, 2, 6, 7, 7, 7, 7, 8, 9, 9, 9, 10
Now the median is 7, while the arithmetical mean is 6.83. As you can see, the median holds its position for a much longer time than the arithmetical mean.
The disadvantage of the median is that, in the current voting system, it will always have values like 6, 6.5, 7, 7.5 etc... but you can remedy this by coupling it to the arithmetical mean and calculate:
1/2 * (median + arithmetical mean)
That way you can calculate a quantity which is fairly robust against to downvoting (and also upvoting, alas) while still being coupled to the arithmetical mean (and thus produce numbers other than just 7, 7.5, 8 etc).
Just a thought. And if you remarked that you can do a lot of weird things with statistics, you'd be right. You can twist and turn the results of a test any way you want with the right formulas. :) |
| Previous - 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 - Next | |
