Jita Park Speakers Corner

Over the past few CSM elections, it has become increasingly clear that the simple "vote for one candidate" election system should be revisited in the hopes of finding a system that produces a CSM that better represents the electorate (ie: those members of the community who care enough to vote).

In the most recent election, for example, almost 25% of the votes were cast for candidates who did not win election to the CSM. These lost "undervotes", coupled with the fact that some candidates have increasingly overwhelming information and organizational advantages, threatens to effectively disenfranchise a significant portion of the electorate.

Since it is in everyone's long-term interest for the CSM to broadly represent the electorate, a strong case can be made that the election system should be reformed. CSM elections should not only be fair, but should be seen to be fair.

This posting is the first step in that process, and it is our hope that it will spark a serious discussion of the topic and provide the CSM with community guidance as we solidify our recommendations to CCP.

Goals of a Reformed System

It has been mathematically proven that there is no perfect voting system. However, "The perfect is the enemy of the good", and there are many election systems that are clearly better than the current one.

The CSM believes that any new CSM voting system should, at a minimum:

1) Be easy to implement (by CCP), use (by the voters) and permit public verification of the results given the published raw vote totals.

2) Reduce the need for tactical voting ("I like A better than B, but I have to vote for B because I think she's more electable"). This is typically done by reallocating votes cast for candidates that are eliminated from the election to other candidates, thus reducing or eliminating undervotes.

3) Reduce (but not eliminate) the advantages held by highly organized voting blocs. In the previous election, for example, one voting bloc did extremely sophisticated exit-polling; if they had chosen to use this information to efficiently split their votes, they could have won 3 of the top 7 positions on the CSM.

Below we present one possible system that attempts to meet the above goals. We caution readers not to assume that this is a system we have decided upon; rather, it is presented as an example for discussion and improvement.

Candidate-Designated Single Transferrable Vote

The most common "better" election system is Single Transferrable Vote (STV). Typically in STV systems, voters rank the candidates (or some subset of them) in order of preference, and if their first candidate is eliminated from the election, their vote transfers to their second choice, and so on.

However, standard STV systems do not meet our first goal. Implementing classic STV would require significant work by CCP to update the system, would require extra effort by voters (which would tend to favor highly organized groups), and may not be publicly verifiable.

Therefore, perhaps a simpler system might achieve most of our goals. Consider, for example, Candidate-Designated Single Transferrable Vote.

In this system, voters vote for a single candidate, just as they do now. However, each candidate publicly states which other candidates they want their votes to be transferred to if they are eliminated from the election.

CD-STV, while not providing the full spectrum of choice that STV does, does meet all the goals. The amount of work CCP will have to do to implement it is small (listing the candidates' preferences on their profile and voting page), it is no more difficult to use than at present, and anyone can run the raw vote totals through an election simulator.

The Nitty-Gritty:

Currently, once the final list of approved candidates is published by CSM, there is a 2 week campaign period before voting starts. In the new system, candidates will have to disclose to CCP and publish in the first post of their campaign thread and on any campaign websites who their preferred alternate candidates are before the end of the first week. After that, they are locked in and cannot be changed -- thus there will be a week for voters to comment on these choices and decide how best to allocate their votes (troll and feeder candidates will be fairly obvious).

Candidates should specify somewhere between 3 and 6 alternates in order of their preference (their "preferences"), so that if one or more of their preferred candidates has already been elected or knocked out, their votes will not be wasted.

On the actual voting page, there would be a list of the alternate candidates so that when actually voting, people will know where their vote will go.

The determination of the results is straightforward:

In each round:

* If the top vote-getter has more than 1/n of the remaining vote pool, where n is the number of CSM slots still available (14 at the start), she is directly elected; the vote pool is reduced by the number of votes she currently controls (and n goes down by 1 in the next round)

* If this does not happen, then the bottom candidate is eliminated, and her votes are reallocated to the topmost candidate in her preference list that is still in the race.

This process continues until everyone has either been elected or eliminated, at which point we have our rankings.

Reserved for replies

dont agree with the second one as that would tip the balance into a landslide if conspired with other players.

Ground floor of what should be a nice, calm discussion.

Let's see how badly this will get gamed, and how much whining will come off it when hisec still doesn't have "enough representation", even though people can't seem to be able to figure out what "enough representation" is, or what "hisec representation" should look like.

It seems like this is a transparent attempt to game the system so the CSM dictates the results rather than the voters.

this is the portion I'm talking about, the change from the STV system that the CSM hoped nobody would notice

the proper way to do a STV is transfer the "excess" votes of anyone who obtains a seat

I'm glad to see the CSM has spent its time discussing issues that affect EVE as a whole, as opposed to immediately jumping down the line of "how can we make sure Mittens doesn't get elected to the #1 spot on the CSM for the third consecutive term?"

To be fair it screams "how do we ensure that we stay on the CSM and not get mobbed out by bloc candidates next term" rather than "how do we keep the bloc candidates off entirely"

Mitanni would still have been elected under this, or probably any, voting system. Robert's post is more concerned at looking at the bottom end of the voting than the top.

And the CSM hasn't decided anything, we talked about it a bunch (hint, we can read and reply to several forum threads at a time, shocking i know) and felt the discussion reached an appropriately structured place that we could start a good conversation with the EVE player base.

Goons would probably be better served engaging in that discussion rather than inventing things to tinfoil hat about.