maemo.org - Talk

maemo.org - Talk (https://talk.maemo.org/index.php)
-   Community (https://talk.maemo.org/forumdisplay.php?f=16)
-   -   MeeGo Coding Competition 2011 (https://talk.maemo.org/showthread.php?t=71561)

Helmuth 2011-11-30 12:32

Re: MeeGo Coding Competition 2011
 
I would have prefered a out of the box solution - counted, evaluated and analyzed directly at Nemain without any community influence - but if it is replicable by anyone I'm fine with the script solution as I would be fine with the Borda system.

For the protocol: neal has participated at the contest - but luckily his application is not affected by this decision - he is at the second place in his category, no matter if we select the ftstv script or the borda system.

Anyway, do we have cases with the exact same number of votes and hence ties at one of the counting procedures?

neal 2011-11-30 13:02

Re: MeeGo Coding Competition 2011
 
Borda is a bad solution. From wikipedia:

Quote:

under the simplest form of the Borda count, if there are five candidates in an election then a candidate will receive five points each time they are ranked first, four for being ranked second, and so on, with a candidate receiving 1 point for being ranked last (or left unranked).
Thus, strategically, I should only vote for my submission (or my favorite submission) and not vote for any others. Consider the case where there are 3 submissions: A, B and C and my submission is A. To maximize my chance of winning, I should vote: A, because I give A 2 points and B and C 0. If I were instead to vote A, B, I'd give A 2 points and B 1 point, which hurts A.

The point of methods like STV is to eliminate this strategic voting. If I place A first, A gets my full support and it doesn't matter how I rank the rest of the candidates. I think most people understood the election in this sense (as it is how the maemo council election was run and how Debian does elections). Therefore, I am strongly against using Borda.

mmlado 2011-11-30 13:35

Re: MeeGo Coding Competition 2011
 
As we "outsourced" the voting counting, and anyone can do it, I'm still for the idea, to use council voting style counting. There can be enough people to do the work and it wouldn't be too much for one person.
Use the voting with one seat. Do turns. Each turn remove the winner of the previous turn and we get a ranking.
The procedure should be done only 5 times to get winners for all prizes.
Running the script is easy, and removing the winner applications can be done with search/replace.
I would volunteer for a category or two.

Edit:

Sorry. Simple search/replace won't work, but... I can write a script that removes the winner from the file, and adjusts all other votes in the file.

I found this is the format of the file:
Quote:

4 2 # four candidates are competing for two seats
-2 # Bob has withdrawn (optional)
1 4 1 3 2 0 # first ballot
1 2 4 1 3 0
1 1 4 2 3 0 # The first number is the ballot weight (>= 1).
1 1 2 4 3 0 # The last 0 is an end of ballot marker.
1 1 4 3 0 # Numbers in between correspond to the candidates
1 3 2 4 1 0 # on the ballot.
1 3 4 1 2 0
1 3 4 1 2 0 # Chuck, Diane, Amy, Bob
1 4 3 2 0
1 2 3 4 1 0 # last ballot
0 # end of ballots marker
"Amy" # candidate 1
"Bob" # candidate 2
"Chuck" # candidate 3
"Diane" # candidate 4
"Gardening Club Election" # title
So we would need a script that would remove a number, for example 2 and decrement all bigger numbers in the votes.

Edit2: Unfortunately don't know if this would create ties. :(

Helmuth 2011-11-30 13:59

Re: MeeGo Coding Competition 2011
 
Quote:

Originally Posted by neal (Post 1131045)
Thus, strategically, I should only vote for my submission (or my favorite submission) and not vote for any others. Consider the case where there are 3 submissions: A, B and C and my submission is A. To maximize my chance of winning, I should vote: A, because I give A 2 points and B and C 0. If I were instead to vote A, B, I'd give A 2 points and B 1 point, which hurts A.

Mmmh... very good argument. :)

However, there is a small difference how the Borda system is implemented at openSTV compared to the description from wikipedia. The first place gets n-1 points. So at 5 submissions the first place gets 4 points and the last 0.

But your Argument is still valid. :rolleyes:

Quote:

Originally Posted by neal (Post 1131045)
The point of methods like STV is to eliminate this strategic voting. If I place A first, A gets my full support and it doesn't matter how I rank the rest of the candidates. I think most people understood the election in this sense (as it is how the maemo council election was run and how Debian does elections). Therefore, I am strongly against using Borda.

Yes, the STV system seems to me to be a little bit more fair to me - even if I have some problems to understand it. :D

We are never aware of strategical voting. Think about voting for your own project for number 1 and then only the underdogs, the worst working and always crashing projects and the top favourites at the last place. The STV system wouldn't cover such kind of strategical voting.
The best solution against all kind of strategical voting would be a large ballot and get many, many users to vote. But this was out of our control if only 1% of all email receivers used their chance to vote. :(

neal 2011-11-30 14:02

Re: MeeGo Coding Competition 2011
 
Quote:

Originally Posted by mmlado (Post 1131062)
Running the script is easy, and removing the winner applications can be done with search/replace.
I would volunteer for a category or two.

My script takes a simpler approach. What's the problem with my script?

mmlado 2011-11-30 14:12

Re: MeeGo Coding Competition 2011
 
Sorry. How stupid of me! :eek: :(
We already have a mechanism to remove candidates, by saying they have withdrawn. It's even easier to run my proposal.

Here are the results I got, in less then 5 minutes :) :


neal 2011-11-30 14:14

Re: MeeGo Coding Competition 2011
 
Quote:

Originally Posted by mmlado (Post 1131085)
Sorry. How stupid of me! :eek: :(
We already have a mechanism to remove candidates, by saying they have withdrawn. It's even easier to run my proposal.

Here are the results I got, in less then 5 minutes :) :

That's a nice idea. Does it yield different results from the approach that I took? (Can you generate a diff, please.)

mmlado 2011-11-30 14:17

Re: MeeGo Coding Competition 2011
 
Quote:

Originally Posted by neal (Post 1131078)
My script takes a simpler approach. What's the problem with my script?

There's no problem with the script, I think people were expecting the same voting process as for council voting.
With my solution I got the same result, in Business & Office Category, as when running the script with 5 seats, only now we have a ranking.

neal 2011-11-30 14:24

Re: MeeGo Coding Competition 2011
 
Quote:

Originally Posted by mmlado (Post 1131088)
There's no problem with the script, I think people were expecting the same voting process as for council voting.
With my solution I got the same result, in Business & Office Category, as when running the script with 5 seats, only now we have a ranking.

??? My script generates a ranking.

mmlado 2011-11-30 14:27

Re: MeeGo Coding Competition 2011
 
Quote:

Originally Posted by neal (Post 1131086)
That's a nice idea. Does it yield different results from the approach that I took? (Can you generate a diff, please.)

diff? Sorry, no.
I did it only on Business & Office Category, by hand. So run election for 1 seat, copy the winner. Add winner to the withdrawn list, run the election again.
I think it should give same results as frictional transfer, only we would have ranking.


All times are GMT. The time now is 13:05.

vBulletin® Version 3.8.8