EVE Information Portal

if(_currentuser.loggedInPosition.isDev)
{
    FilterOut(_currentThread.GetElements().Select(post => post.GetOwner().VotedForCSM));
}

How do we know you don't already have this?

They do, but right now it just filters out posts by current and former CSM members.

Right, they get ~proportional representation~, and meanwhile one bloc has taken 1/3 of the seats. How many are left after the other 2 major blocs (HBC, N3) get their seats (3 if the Russian community has their own candidate)? It'll be an accurate tally of who's voting, sure, but it sure as hell won't be a diverse CSM, which will likely put a big fat dent in CCP's efforts to increase voter turnout (whatever those efforts may be).

The old system, flawed as FPTP is, was moving towards a trend of blocs only putting forth a single official candidiate. This was in large part due to the fact that more than one candidate was completely unnecessary - there were no votes or anything of the sort, everything that needed to be decided (chair, who goes to Iceland) was decided before anyone took office, so "stacking the deck" was a pointless gesture. What this new system has done is introduce the CSM voting on things AND given a voting system that lends itself towards multiple candidates (or at the very least allowing nullsec to dictate who the majority of the council are), which undoes all of that. That's not really a good thing, especially when you're still trying to convince non-bloc-aligned people to actually care enough to participate.

After giving some thought to this, I'd say the simple-vote + multi-vote combo exemplifies how the thought process went here: they put a lot of effort raising every single issue they could find in the old system, then did their best to solve each one of them individually. What we got from this is a big patchwork, each piece on it's own supposedly addresses something, but the solution as a whole makes no sense.

The result is some nonsensical things, like making dozens of thousands of people to vote to eliminate five candidates, at most, out of a pool of some 50 to 60*, and then two weeks later making all these people vote again for the election result, now with a ranked ballot.

It really seems like they are so immersed in a problem-solving mentality that they are unable to make an objective analysis of the changes they are making as a whole. Instead of recognizing that, maybe, all these complains are not entirely unfounded, what we see in this thread are responses doubling down on their decisions. And with this they are failing to consider that, as the rules change, so do the incentives and thus how people will approach the subject of placing representatives in the CSM.

Seriously, there would be a lot more pleased comments in this thread if all they did was announce just two of these changes:
- Secretary and vice-secretary positions selected internally.
- A big effort to increase the awareness of the election, yet to be disclosed and the main problem here.
Instead of iterating on the process they are changing too much at once, and thus putting a lot more in jeopardy.


* I'm already expecting a larger pool of candidates as there will be an STV system in place.

Darius supporters voted, I would almost feel bad for CCP, if they were the only ones allowed to post feedback.

Otherwise, I didn't vote last year, so now I do understand the feedback threads now. Oh well, with great apathy comes great tribulation.

I see one potential problem with more than one vote and weighting votes at that.

Let’s say I like candidate 1 the best so I put a #1 vote on them, and I like candidate 2 the second best. If I have to chose one of those it is clearly candidate 1 and I would like candidate 2, but not at the expense of candidate 1.

So if candidate 1 has a certain value given him/her by all the ranked votes given and candidate 2 also has a close number of votes. If I vote #2 for candidate 2 can it push him to the point he has a higher rank overall than candidate 1 and thus in case of a close race, bump candidate 1 out of the running?

what code language are you using to code this?

I want to make sure I understand this correctly. Say there are 60k votes; the threshold is then 4001 votes. Now say I were to be marked as top preference on 5000 ballots. I would get my 4001 votes at full value and be declared provisionally elected, and then the remaining 999 votes are distributed evenly to the second choices on the 5000 ballots that ranked me as first, giving each slightly less than 0.2 votes apiece. This continues until all seats are filled. Because of the proportional transfer, it doesn't matter whose votes I receive "first", because a portion of all votes transfer.

Is that correct, at least in the basic details?

This system of voting will give Null sec blocs an unprecedented security to get at least 1 and most likely 2 candidates in
The only way that independents can run is through sheer name recognition or they need to ally themselves into political groups.

I have to congrat CCPgames by playing the meta game on a new level by rigging the rules, either convert or stay on the outside.

This only confirms my suspicion that from now on CCP wants to invest it time into nullsec (Sovereignty, sanctums, alliance mechanics, futher intergration of Dust514, bottom up income, technetium moons, capital fleet force projection, itterate on TiDi)

My only hope is that they dont forget to give good opportunities for small independent entities, Null sec needs to be a place where newer or smaller alliances can survive and larger ones can't easily crush them, add alot of diminishing returns as a prevention imo.

You can find complete details on the Wright-STV algorithm here. It's a bit laborious but pretty straightforward.

A simplified explanation of the algoritm is as follows:

If at any time there only enough candidates left to fill the needed number seats, or that many people are provisionally elected, the election ends.

Total up how many first-preference votes each candidate received.

Any candidates who get more than the threshold are declared provisionally elected. The threshold is 1 + int(number of active votes / (number of seats+1))

The provisionally elected candidate with the most votes distributes his overages. If for example the threshold is 4000 votes, and a candidate gets 5000, then he needs 80% of his voting power to get elected, and has 20% to give away. You can imagine that those ballots are cut into two parts, an 80% part that the candidate keeps, and a 20% part. The 20% part is transferred to the most-preferred candidate on it who has not already been provisionally elected. So if you had been provisionally elected and where transferring a ballot that read "mynnna-Trebor-Malcanis", and I had also already been provisionally elected, the 20% ballot would go to Malcanis.

Once the overages are distributed, a check is made to see if anyone else has been provisionally elected.

Overage distribution continues until there are no provisionally elected candidates with overages to distribute. Note that fractional ballots that went to a candidate who became provisionally elected will be fractionally distributed -- if a candidate got a 20% ballot and himself distributes 10%, then he keeps an 18% ballot and sends on a 2% one!

At this point, if we don't have enough provisionally elected candidates, the candidate with the fewest number of first-place votes is eliminated (if there is a tie, the loser is determined randomly). His name is crossed off all the ballots they appear on. Any ballots that don't have any names left are considered "exhausted" and are discarded (which reduces the threshold in the next round of the election).

The election is then restarted from scratch, with one less candidate. All those ballots are glued back together, and we go through the whole process again. If you need to get rid of 25 candidates, you're going to go through 25 rounds unless you get lucky. Better sharpen those scissors...

I am coding up my own election software as a cross-check (it's a fun little project). There are a couple of fine points I want to confirm with CCP Veritas, but that's basically it.