Also why did the election got moved to the later date. According to this dev blog: http://community.eveonline.com/devblog.asp?a=blog&nbid=3386 CSM8 should take office April 3rd 2013.

So wormholers will have a primary after all.

The summit attendees will use a new 2+5 system, with CCP and the CSM working together to pick the 5 hardest working and most feature relevant CSMs being flown to each summit and the final 2 attendees will be the "most preferred" candidates, chosen by reentering the election results into an STV election to select the top 2 candidates.

A back-of-the-envelope calculation indicates that a Schulze-STV election with 14 seats and 28 candidates would take over 9 years to compute on a decent PC.

Will there be increased efforts to raise awareness of the CSM season this time, which in many people's eyes would be a far more important step for the CSM, and do a far better job of making the process legitimate, than fiddling with the voting system? Asking us to vote twice and to make 14 choices on the second occasion isn't exactly going to appeal to the apathetic Joe Random Eve Player.

Is it 14 votes exactly, or up to 14? If I only see 4 candidates I like the look of, do I have to cast the remaining 10 or can they be discarded?

Is it 14 votes exactly, or up to 14? If I only see 4 candidates I like the look of, do I have to cast the remaining 10 or can they be discarded?

The 'most feature relevant' part of the summit attendee section bothers me somewhat, since the summits should be covering a wide spectrum of topics, not just "lets talk about our new lowsec (for example) revamp expansion for three days". How will this be decided?

As mentioned in the devblog, CCP has not settled on the exact STV counting method to be used in the election. This is basically a tradeoff between systems that are simple and ones that are more complex but generate slightly better proportional results.

In all cases, the actual casting of ballots is the same -- you specify one or more candidates, in your order of preference. It's just the counting method (and the reallocation of votes as candidates are elected and eliminated) that changes.

One system that isn't feasible is Schulze-STV. While it is a very good counting method, its computational complexity explodes as the number of candidates and seats goes up. A back-of-the-envelope calculation indicates that a Schulze-STV election with 14 seats and 28 candidates would take over 9 years to compute on a decent PC.

The following links may be useful to those who are interested in suggesting counting systems:

* Wikipedia: Single Transferable Vote
* Wikipedia: Counting single transferable votes
* OpenSTV (Software)

I was trying to look it up and having no luck. What is the Big O notation you got for that algorithm?

I'm probably misunderstanding this, because I'm a little dopey.

Will you have to stack rank all 14 candidates? That seems somewhat onerous, wouldn't be better to say you can select up to 14 people who may ore may not get your vote. Making the voting system harder work seems counter intuitive.

I was trying to look it up and having no luck. What is the Big O notation you got for that algorithm?

Not sure if this should be asked somewhere else, but is there a specific reason for cms's needing to be over 21? Just one of those old rules that we're sticking to, or is there something i'm not getting at first glance?

Will there be increased efforts to raise awareness of the CSM season this time, which in many people's eyes would be a far more important step for the CSM, and do a far better job of making the process legitimate, than fiddling with the voting system? Asking us to vote twice and to make 14 choices on the second occasion isn't exactly going to appeal to the apathetic Joe Random Eve Player.

Will you have to stack rank all 14 candidates? That seems somewhat onerous, wouldn't be better to say you can select up to 14 people who may ore may not get your vote. Making the voting system harder work seems counter intuitive.

STV is prone to "Donkey" voting from players that are not in "alliances that help them fill the preferences" and/or don't know all the candidates. Has the CCP think about this small issue? Will CCP show the candidates in a random order in the ballot?

Results are announced at Fanfest, which is a month later this year.

Will there be increased efforts to raise awareness of the CSM season this time, which in many people's eyes would be a far more important step for the CSM, and do a far better job of making the process legitimate, than fiddling with the voting system? Asking us to vote twice and to make 14 choices on the second occasion isn't exactly going to appeal to the apathetic Joe Random Eve Player.

As far as I know, you'll be able to select any number from 1 to 14. You may be amused to learn that my position was that you should be able to rank all the candidates if you so desired.

This is a system open to neither abuse nor gaming, with absolutely no chance of someone (you know who you are) trying to use it as a "keep the Goons out of the CSM" tool.

I look forward to seeing how our team games the system this time around.

This all looks simpler and more transparent than the previous system.

You will be able to select one, or two, or three, or four... up to 14.

This was my primary concern throughout our internal discussions, the fact that so much time was spent hashing out complicated algorithms for election results calculation and so little time spent discussing how we can promote the election and increase voter turnout. My fear (and we'll see if its founded or not) is that a more complicated voting system will decrease turnout, rather than the other way around.

While I'm glad that enough time was spent that a different, -supposedly- less-exploitable method was selected, we'll see if there's follow-through on the promotional/educational front. We've been told that there will be some more material released closer to the election start, but I'll believe it when I see it. We've asked previously for more promotional support for things like Town Halls and such and received less-than-satisfactory (or just plain tardy) response, so I can only go off of CCP's track record where this is concerned. I would have liked to have seen this been the big push this year instead of voting reform itself, personally.

This all looks simpler and more transparent than the previous system.

Many thanks to the Community team, and especially CCP Veritas for his extra hours put into this, taking the time to make sure this is done in the best way possible. I'm looking forward to this election and seeing how this all plays out!

I wrote a blog post about what this means for wormhole candidates. It is critical to make sure that all candidates ask their supporters to list *all* wormhole candidates at the top of their ballots.

This was my primary concern throughout our internal discussions, the fact that so much time was spent hashing out complicated algorithms for election results calculation and so little time spent discussing how we can promote the election and increase voter turnout. My fear (and we'll see if its founded or not) is that a more complicated voting system will decrease turnout, rather than the other way around.

While I'm glad that enough time was spent that a different, -supposedly- less-exploitable method was selected, we'll see if there's follow-through on the promotional/educational front. We've been told that there will be some more material released closer to the election start, but I'll believe it when I see it. We've asked previously for more promotional support for things like Town Halls and such and received less-than-satisfactory (or just plain tardy) response, so I can only go off of CCP's track record where this is concerned. I would have liked to have seen this been the big push this year instead of voting reform itself, personally.

Encouraging voting for the sake of voting is about as dumb as arranging a system that will land the entire CSM into the lap of one group.

going up to 5 would be better.

People are free to only put in 5 if they wish. Heck, they can only vote for one if that's all the preference they have. The only downside is that they might disenfranchise themselves if noone in their small set of candidates end up having enough support.

Well done CCP for handing control of the CSM to the CFC and the HBC.

Don't say you weren't warned long ago.

Since I will directly benefit this time around I'm not even going to make too much of a fuss about it right now, but the outcome of this change is so obvious that I can't believe it isn't intended. That alone gives me much amusement.

How much extra downtime would be needed to make TQ do it?

Silly Two Step, this change is meant to prevent voting blocs from gaining more influence!

On the flip side, we can then bring in some other people than the top seven instead of being locked in to that predetermined selection.

Granted we know this will generate discussions about the selection, but the flavor of it will be different from the discussions on the matter in the past.

That would be one way to describe it. If by simpler you mean much more complex, and if by transparent, you mean transparently obvious that this will hand the result completely to large voting blocs.

People are free to only put in 5 if they wish. Heck, they can only vote for one if that's all the preference they have. The only downside is that they might disenfranchise themselves if noone in their small set of candidates end up having enough support.

Why wait? 30 seconds thought reveals that you won't even have to try very hard. Just nominate the 7 candidates you like the most and tell your guys to vote for them in any order that pleases them. Bingo: CFC CSM achieved.

I do have an honest question, though. The blog mentions that the top twenty eight candidates from the pre-election process get up for election, and what happens if there are fewer than twenty eight. But what happens if there are more - straight top vote getters are in, I assume?

What Two step is coordinating is identically equivalent to having a primary, except it takes less coordination and is done during the election instead of prior. It has the added benefit of spare "wormhole party" support (as in, leftover votes that aren't enough to elect a "wormhole" candidate) potentially transferring to secondary preferences. The "wormhole party" doesn't magically gain more votes because of the procedural difference - if they account for 2/14 of the vote they'll get 2 seats, if they account for 1/14 they'll get 1 seat, ect.

[quote=Malcanis] But what happens if there are more - straight top vote getters are in, I assume?

Hmm. So an election that struggles to get people to even vote for one candidate is now going to ask us for our top 14? Seems like it's going to be too much form filling for a lot of people.

What Two step is coordinating is identically equivalent to having a primary, except it takes less coordination and is done during the election instead of prior. It has the added benefit of spare "wormhole party" support (as in, leftover votes that aren't enough to elect a "wormhole" candidate) potentially transferring to secondary preferences. The "wormhole party" doesn't magically gain more votes because of the procedural difference - if they account for 2/14 of the vote they'll get 2 seats, if they account for 1/14 they'll get 1 seat, ect.

Pretty much.

For some reason this blog made me think of great solutions in engineering.

Well those groups which are directed by people who can co-ordinate large numbers to vote for the same list are going to get, effectively, 14 votes per voter. Those unco-ordinated demographics who are interested in maybe one or at most 2-3 candidates and will either not vote for anyone else or who will scatter their votes pretty much at random are going to get effectively 1-3 votes per voter.

Can anyone predict how this will end up?

Anyone?

Bueller?

I do have an honest question, though. The blog mentions that the top twenty eight candidates from the pre-election process get up for election, and what happens if there are fewer than twenty eight. But what happens if there are more - straight top vote getters are in, I assume?

I have to ask a broad swath of EVE players to vote for me...
...and then ten or so days later, I have to ask them to do it a second time using a different system?
...and this is your plan to make the CSM more open to non-bloc candidates?

You are misreading the blog. Nowhere does it state that there is a maximum number of candidates. If you get 200 votes, you get on the ballot. However, if less than 28 candidates meet the threshold, it will be lowered so that at least 28 candidates make the final ballot.

To quote the blog: "These votes will then be tallied and candidates that receives 200 or more votes will qualify to be added to the final election ballot. If less than 28 candidates meet this threshold, the next highest ranking candidates will be added until 28 candidates make the final ballot."

On topic: Could candidates post their own voting list suggestions? You know, to get together a strategy to ensure like minded people get elected or unwanted candidates get forced out in favor of less crappy bloc candidates.

People are free to only put in 5 if they wish. Heck, they can only vote for one if that's all the preference they have. The only downside is that they might disenfranchise themselves if noone in their small set of candidates end up having enough support.

I am literally typing up 'GoonSwarm: Crushing Pubbies Via Voting and You' right now

Okay I'm really not because I'm not some midlevel bureaucrat but someone is!

Okay I'm really not because I'm not some midlevel bureaucrat but someone is!

how did it get in your post if you did not type it :ohdear:

Errr... OK. Granted I've only read this once and it's pretty early in the morning, but let me make sure I understand this.

I have to ask a broad swath of EVE players to vote for me...
...and then ten or so days later, I have to ask them to do it a second time using a different system?
...and this is your plan to make the CSM more open to non-bloc candidates?

My inner Garth is screaming that this is truly an election system designed by CCP...

EDIT: And let me be clear: I have no objection to the STV. But isn't the whole point to the STV to eliminate the need for primary elections?

Well those groups which are directed by people who can co-ordinate large numbers to vote for the same list are going to get, effectively, 14 votes per voter. Those unco-ordinated demographics who are interested in maybe one or at most 2-3 candidates and will either not vote for anyone else or who will scatter their votes pretty much at random are going to get effectively 1-3 votes per voter.

Ring-a-ding-ding!

I believe you are confusing STV with condorcet methods, which do suffer from the weakness you are concerned about when used in a multiple-seat context. STV does not.

All ballots have the same amount of voting power in a STV system. A ballot can lose some fraction of its voting power when it becomes exhausted -- there are no more candidates on it to whom it can transfer voting power. So at the margins, a ballot with 14 candidates on it will have slightly more voting power than one with 5, but it is not a huge difference, and it is much less than the difference in voting power in FPTP elections when comparing highly organized, high-information groups with an unorganized electorate.


In the last election, over 24% of the electorate disenfrancised themselves by voting for a losing candidate. Under STV, this percentage is mathematically certain to be lower.

I would encourage those who are interested in investigating the robustness of good STV methods (my preference is Meek-STV) to go out and get some election software and start running simulations (OpenSTV costs $5). You may be surprised at the results.

For example, re-running the previous election under Meek-STV, in a situation where 6 "nullsec candidates" supported each other in strict order, and everyone else went it alone (all of their ballots were votes for them alone, so if they didn't get elected, their ballots were discarded), resulted in a single change to the final results -- 14th position changed.

It's not clearly stated how the actual voting works - the dev blog says you rank candidates 1 through 14 but doesn't mention at any point that you don't HAVE to choose 14, just that 14 is the max amount of candidates you can rank. Only the White Paper clearly spells that out (and thanks to the 6+ pages of filler that the White Paper starts with, you can be sure almost nobody is going to really read it). If one simply read the dev blog, they'd leave with the assumption that you must choose 14 candidates.

The devblog states they haven't even chosen the form of STV they'll use. And that they'll tell us all later, so that we can review.

That must be a joke ... they'll have already coded the system. And there'll be little time to change it even if there were an uproar.

After re-reading the devblog, I am reasonably positive that the new system is way too complex and will disengage potential voters trying to understand what are they doing or why.

Hell, i'm not even sure to vote albeit I did the two last elections as i'm still not sure on what are supposed to be the potential consequences of ranking the candidates. Does it matter hwo I rank them? It's better to vote just one guy or should I vote two or more just in case? What happens with all the votes I don't give (i.e, I have 14 votes, what if I only give 4? Why should i be less represented that someone who uses all his 14 votes?)

This basically sums up my questions. Should I just vote for me and get 14x the benefit of spreading my vote, or should I vote for 14 people and hope they reciprocate?

In the last election, over 24% of the electorate disenfrancised themselves by voting for a losing candidate.

CCP has now instituted a system (that we are quite pleased with BTW) that lets them try to enforce even more unpaid consulting to play survivor in space. The enemy here is clearly CCP and their intentions with regards to the CSM. So the only obvious response is to make that decision very painful.

The devblog states they haven't even chosen the form of STV they'll use. And that they'll tell us all later, so that we can review.

That must be a joke ... they'll have already coded the system. And there'll be little time to change it even if there were an uproar.

I believe, as these two comments give plenty of evidence, that a lot of people will erroneous think they are giving a separate vote for each candidate they choose, while in reality they only have one vote (per account). The vote will go for their top-most choice, in case said candidate receive enough votes to be elected. Failing that, the vote is transferred to their second choice, if that candidate can be elected, so on, so on. But in no moment one vote is added to more than one candidate.

I'm pleased you are happy with the system. Goons in particular make excellent CSMs, they tend to do more than their fair share of the work. Believe it or not, once CCP made the decision to change the balloting system to allow for a multiple-choice ballot, I advocated very strongly for a system that would deliver as close an approximation to proportional representation as possible. I also pushed very hard for the 2+5 system, to ensure that hard-working Goons who did not get elected to the top 7 would get a well-deserved trip to Iceland. So by all means, put up your best candidates, because I for one want as capable a CSM as possible.

If you manage to figure out a way to game the election in order to get more than your fair proportional share of the seats, then congratulations to you. Just be aware that manipulating STV systems is, IIRC, a NP-complete problem. But if you do manage a successful attack, there will be numerous academic papers that will result, and the real world will be better off for your efforts.

I am sure you will be able to exploit some edge effects to optimize your results. But I am also sure that under the older system, you could have optimized them even more.


The choice of ballot counting system is independant of the choice of going to a ranked preference ballot. CCP has chosen to go with a ranked preference system, STV. But how the votes transfer as candidates get elected or eliminated varies depending on the system, and there are tradeoffs. Some systems are simple and easy to understand (they can be done by hand). Others are more complex and arguably more accurate, but have to be done on a computer and are a bit harder to follow.

The exact method is yet to be decided, and I am sure CCP would like input from the community on this. So I encourage people to do the research and express their preference and the reasons why.

I believe, as these two comments give plenty of evidence, that a lot of people will erroneous think they are giving a separate vote for each candidate they choose, while in reality they only have one vote (per account). The vote will go for their top-most choice, in case said candidate receive enough votes to be elected. Failing that, the vote is transferred to their second choice, if that candidate can be elected, so on, so on. But in no moment one vote is added to more than one candidate.

Oh so its that system. Well in that case: bloc CSM it is then. Instead of "wasting" their votes on spending more than enough to get their guy elected, they can efficiently make sure that as many of their guys get elected as they have votes to achieve. Basically it will automatically perfectly co-ordinate bloc voting.

Hilarious.

Excuse me? You say that a candidate must have a chance to be elected in order to get my vote? But how do you determine he haves a chance to be elected without counting the votes first?

I guess that something went awfully wrong between our sides of the screen...

Maybe you mean that if my first choice doens't ranks 14th or higher, then the vote goes to the second choice? Is that?

But then, what's the point in voting my first candidate first unless i think he can make it to the top 14? And how it turned that my vote is a tradeable commodity someone can arbitrarily kick up and down depending on how wisely I chose my first choice(s)?

"Y'know, you were wrong to choose Abe, Barb, Chuck, Donna, Ed, Frank and Gina, but we can proudly inform you that your vote ended up landing Howie into the 14th rank! ".

I seriously wonder what kind of briliant idiot figured that would be satisfactory...

Maybe you mean that if my first choice doens't ranks 14th or higher, then the vote goes to the second choice? Is that?

"Y'know, you were wrong to choose Abe, Barb, Chuck, Donna, Ed, Frank and Gina, but we can proudly inform you that your vote ended up landing Howie into the 14th rank! ".

This system is literally the worst one you could have picked if you didn't want the CSM to be dominated by the two major coalitions.

I could write practically an essay on why, however since my previous ones were ignored I'm going to spend my time helping the HBC and CFC achieve at least half the CSM elected through them instead.

Hello Coalition and EVE University CSM.

I think we should have a pool as to how many CFC, HBC, and EVE University candidates get into the CSM this time around.

There are 14 possible positions.

I am going to guess 4 CFC, 4 HBC, and 2 EVE University.

I wonder who will get the four remaining spots.

Trebor seems to be clear on the specifics. Perhaps he'd like to explain in more detail exactly how it will avoid the weakness I mentioned. I'm prepared to take his word for it that more choices won't outweight less, but I'd like more detail.

Oh so its that system. Well in that case: bloc CSM it is then. Instead of "wasting" their votes on spending more than enough to get their guy elected, they can efficiently make sure that as many of their guys get elected as they have votes to achieve. Basically it will automatically perfectly co-ordinate bloc voting.

Hilarious.

Why, we would never game a system in such a way! Clearly it is impossible from such a well defined and tested election system!

In the real world, I would completely agree gaming it would be an exercise unfeasible. Lucky for us, EVE allows things the real world does not.

I posted these links in my message on the first page of the thread, but here they are again:

* Wikipedia: Single Transferable Vote
* Wikipedia: Counting single transferable votes
* OpenSTV (Software)

The second link has a good explanation of how a simple STV system works. There are many variants that have slight differences in the way vote transfers are done when a candidate reaches quota or is eliminated.

In simple STV systems, each ballot ends up helping elect one candidate (unless it gets discarded because none of the candidates on it are electable). In the more complex systems, ballots can help multiple candidates fractionally in some circumstances, but no ballot counts for more than 1 unit of voting power.

In general, the more complex systems, like Meek-STV, produce a slightly better approximation of proportional representation. But the cost is that following the narrative of the election is a little more difficult.

Arguably the most accurate STV system is Schulze-STV, but it is computationally infeasible for elections with the number of candidates and seats typical of CSM elections.

No voting system is perfect (this is called Arrow's Theorem), and all voting systems have various criteria that they meet or fail to meet (here's a table of them for single-winner elections; the single-winner version of STV is called IRV or AV). However, STV meets several very important criteria, and in particular:

"Later no harm/no help": Adding another candidate to your ballot will not alter the chance one of your higher ranked preferences will get elected. So you should feel free to vote for all the candidates you like, and rank them in the order you prefer. And you should in particular remember to put Trebor Daehdoow in there somewhere!

"Independence of Clone Alternatives": This means that if there are multiple very similar candidates, as long as you add them all to your ballot, they won't destroy each other's chances of getting elected. So in this election, if there are multiple "wormhole candidates", but only enough total votes to elect one of them, then as long as all the wormhole voters vote for them, one will get elected.

Why, we would never game a system in such a way! Clearly it is impossible from such a well defined and tested election system!

For Christ's sake, we're talking about a silly election in a silly videogame were 86% the potential voters didn't bothered to vote not even before the election system became some manic mathematician's toy!

How in the name of the Lord, are you expecting anyone to understand what they're voting when the "simple" explanation holds not less than TWO threshold calculation methods and EIGHT methods for allocating surpluses (and remember, we're talking about votes here, not merchandise).

CCP Xhagen is totally putting the cart before the horse, adressing election system rather than voter involvement. He can pretty much save all efforts and just ask the blocs who they want to have in the CSM, as nobody else is going to invest not even one tenth of the effort I invested so far trying to understand WTF are they going to do to my vote.

- Problem two: representation on the CSM. CCP Veritas pointed out the error in my thinking, what I want is not actually fair representation of ALL EVE players, but of THOSE WHO VOTE (it follows from there that the more people that vote, the better representation we get). The current voting system is sufficient, but there are many vastly better systems out there.

- Problem three: bloc voting. Not a problem in my eyes. If large numbers of people organize themselves, they will get better results than those who do not.

This basically sums up my questions. Should I just vote for me and get 14x the benefit of spreading my vote, or should I vote for 14 people and hope they reciprocate?

Why are we still operating on the premise that CCP was out to make a CSM that better represents the playerbase? I mean...


Emphasis mine.

Or the premise that CCP is out to quash organized bloc voting.



It's a mystery, man.

https://forums.eveonline.com/default.aspx?g=posts&m=1942633#post1942633

As far as CSM "stuffing" this system doesn't really give the CFC/HBC type blocs more advantage than they previously did. Those groups have well developed exit polling techniques that were near-perfectly managing their voting power. What this does is level the playing field so all blocs, including less homogeneous blocs (such as "WH space" which comprise several coalitions and candidates, without the CFC information infrastructure), can do the same thing.

As CSM6 illustrated with the dec system in Inferno (and the deexploiting of abuses in the pre inferno system), stacking the CSM with tons of one playstyle at the expense of other areas of expertise produces suboptimal results. I think Goonswarm (or at least their political leadership) understands that 2 extremely competent and communications savvy CSM delegates will do just as much if not more than those 2 plus 2 silent partners who dont bring anything else to the table, especially if those 2 placeholders come at the expense of candidates from other constituencies who would be more useful for their agenda.

As far as CSM "stuffing" this system doesn't really give the CFC/HBC type blocs more advantage than they previously did. Those groups have well developed exit polling techniques that were near-perfectly managing their voting power. What this does is level the playing field so all blocs, including less homogeneous blocs (such as "WH space" which comprise several coalitions and candidates, without the CFC information infrastructure), can do the same thing.

As CSM6 illustrated with the dec system in Inferno (and the deexploiting of abuses in the pre inferno system), stacking the CSM with tons of one playstyle at the expense of other areas of expertise produces suboptimal results. I think Goonswarm (or at least their political leadership) understands that 2 extremely competent and communications savvy CSM delegates will do just as much if not more than those 2 plus 2 silent partners who dont bring anything else to the table, especially if those 2 placeholders come at the expense of candidates from other constituencies who would be more useful for their agenda.

10,000 votes out of 30,000 matter a lot more than 10,000 votes out of 60,000. By turning the election system into a sophisticate and complex mathematical trick, CCP is effectively rewarding everyone who can take it seriously enough and haves the manpower to rig the election, but all in all are failing to engage with the players who won't even understand the voting system.

I may be wrong, but i am under the impression that the role of the CSM is to communicate the consumers with the producers so the producers can keep the consumers satisfied and retain them.

Under that premise, the election system should be as simple and straightforward as possible, and should valor quantity of voters above the quality of voters. Getting a lot of voters should be the priority, rather than reward the chosen ones who as much as understand what's going on with their votes -or are plain instructed what to do with them.

As far as CSM "stuffing" this system doesn't really give the CFC/HBC type blocs more advantage than they previously did. Those groups have well developed exit polling techniques that were near-perfectly managing their voting power. What this does is level the playing field so all blocs, including less homogeneous blocs (such as "WH space" which comprise several coalitions and candidates, without the CFC information infrastructure), can do the same thing.

As CSM6 illustrated with the dec system in Inferno (and the deexploiting of abuses in the pre inferno system), stacking the CSM with tons of one playstyle at the expense of other areas of expertise produces suboptimal results. I think Goonswarm (or at least their political leadership) understands that 2 extremely competent and communications savvy CSM delegates will do just as much if not more than those 2 plus 2 silent partners who dont bring anything else to the table, especially if those 2 placeholders come at the expense of candidates from other constituencies who would be more useful for their agenda.

As far as CSM "stuffing" this system doesn't really give the CFC/HBC type blocs more advantage than they previously did. Those groups have well developed exit polling techniques that were near-perfectly managing their voting power.

As far as CSM "stuffing" this system doesn't really give the CFC/HBC type blocs more advantage than they previously did. Those groups have well developed exit polling techniques that were near-perfectly managing their voting power. What this does is level the playing field so all blocs, including less homogeneous blocs (such as "WH space" which comprise several coalitions and candidates, without the CFC information infrastructure), can do the same thing.

As CSM6 illustrated with the dec system in Inferno (and the deexploiting of abuses in the pre inferno system), stacking the CSM with tons of one playstyle at the expense of other areas of expertise produces suboptimal results. I think Goonswarm (or at least their political leadership) understands that 2 extremely competent and communications savvy CSM delegates will do just as much if not more than those 2 plus 2 silent partners who dont bring anything else to the table, especially if those 2 placeholders come at the expense of candidates from other constituencies who would be more useful for their agenda.

We fully understand you do not need more than one REALLY good rep. That isn't at issue here. We have shown that we did not seek to stack the CSM previously. However, the new system means you HAVE to stack the CSM if you want the same outcome.

As far as CSM "stuffing" this system doesn't really give the CFC/HBC type blocs more advantage than they previously did. Those groups have well developed exit polling techniques that were near-perfectly managing their voting power. What this does is level the playing field so all blocs, including less homogeneous blocs (such as "WH space" which comprise several coalitions and candidates, without the CFC information infrastructure), can do the same thing.

As CSM6 illustrated with the dec system in Inferno (and the deexploiting of abuses in the pre inferno system), stacking the CSM with tons of one playstyle at the expense of other areas of expertise produces suboptimal results. I think Goonswarm (or at least their political leadership) understands that 2 extremely competent and communications savvy CSM delegates will do just as much if not more than those 2 plus 2 silent partners who dont bring anything else to the table, especially if those 2 placeholders come at the expense of candidates from other constituencies who would be more useful for their agenda.

If you were going to be happy with a controlling vote (i.e.: 2 or 3 seats, meaning you get to dictate who goes to Iceland and who doesn't), would you follow any other policy than "Vote 1 for A, 2 for B, 3 for C"?

Any plans on how you will be handling players abilities to make throw away 21 day accounts to cast additional posts or players abusing the buddy system? It wouldn't be very difficult for large groups of players who want specific candidates to win to manipulate the voting counts this way.

Any active EVE Online account that is at least a full thirty (30) days in age is eligible to vote.

Seriously this. If you think this is going to get you a more varied CSM, I don't know what to say. I do not see this election ending well on that front. Election system monkeying alone is not going to do it for you, so I hope whatever your plans are for reaching out and getting more voter participation are good.

I do have an honest question, though. The blog mentions that the top twenty eight candidates from the pre-election process get up for election, and what happens if there are fewer than twenty eight. But what happens if there are more - straight top vote getters are in, I assume?

When someone calls "CFC and HBC own the CSM now" and the CFC candidate says "Yep" you should expect that this is truth.

All you have to do is count the votes from last year.

Mynnna should get far more than enough votes to get the top spot. After that, Mynnna's overvotes will all roll downhill, ensuring that several other sov null candidates are elected. HBC is even bigger than CFC, if perhaps less disciplined. So the two biggest blocs of coordinated voters will all be voting for the sov null candidates. At an unscientific guess, expect the top 4, if not 6 or more, candidates to be sov null candidates.

Y'know, I know that the white paper is filled with fluff, but...read much?

The summit attendees will use a new 2+5 system, with CCP and the CSM working together to pick the 5 hardest working and most feature relevant CSMs being flown to each summit and the final 2 attendees will be the "most preferred" candidates, chosen by reentering the election results into an STV election to select the top 2 candidates

Y'know, I know that the white paper is filled with fluff, but...read much?

In the worst case, you can expect that the cost is related to C(c,s) where C is the choose function: c!/(s! (c-s)!). c is the number of candidates, and s is the number of seats.

c(40,14), the case for the last election, means you have to compare 23,206,929,840 different possible quorums!

Oh so its that system. Well in that case: bloc CSM it is then. Instead of "wasting" their votes on spending more than enough to get their guy elected, they can efficiently make sure that as many of their guys get elected as they have votes to achieve. Basically it will automatically perfectly co-ordinate bloc voting.

Hilarious.

So much for educating the Populous, it was hard enough to get the majority of the game to vote for one person let alone 14.

And as to "The summit attendees will use a new 2+5 system, with CCP and the CSM working together to pick the 5 hardest working and most feature relevant CSMs being flown to each summit and the final 2 attendees will be the "most preferred" candidates"

CCP have shown us that the CSM is NOT the voice of the players but just a tool for CCP to use or ignore at its whim.

I for one will not be trying to get people to vote for this dreg.

So much for educating the Populous, it was hard enough to get the majority of the game to vote for one person let alone 14.

And as to "The summit attendees will use a new 2+5 system, with CCP and the CSM working together to pick the 5 hardest working and most feature relevant CSMs being flown to each summit and the final 2 attendees will be the "most preferred" candidates"

CCP have shown us that the CSM is NOT the voice of the players but just a tool for CCP to use or ignore at its whim.

I for one will not be trying to get people to vote for this dreg.

Well, like I posted further up the page, Xhagen actually isn't terribly concerned with bloc voting, nor accurately representing the playerbase - only accurately representing those who actually vote. So in that, mission success.

The onus on CCP is now, as before, to increase the size of the voting playerbase. If they successfully do that, I guess we'll see if they're also successful in the quest to more accurately represent those voters.

But then, if we got the IT guy to find us a solution to a minor issue, and the bigger issue is left unattended or actually is harmed, who's fault is this?

Granted we know this will generate discussions about the selection, but the flavor of it will be different from the discussions on the matter in the past.

The voting rule in Pre-election is different from final ballot, right ?

So much for educating the Populous, it was hard enough to get the majority of the game to vote for one person let alone 14.

And as to "The summit attendees will use a new 2+5 system, with CCP and the CSM working together to pick the 5 hardest working and most feature relevant CSMs being flown to each summit and the final 2 attendees will be the "most preferred" candidates"

CCP have shown us that the CSM is NOT the voice of the players but just a tool for CCP to use or ignore at its whim.

I for one will not be trying to get people to vote for this dreg.

Except in the IT guy's own thread about voting reform, the dominant opinion was that low voter turnout was the real issue that needed fixing, and that the voting system itself wasn't the issue (and could actually be detrimental). In that case, it's not "our" fault at all that the IT guy only wanted a solution that could be coded into a website.

Amused that the descendents of the criminal classes of Britain (Australians) can understand and use transferable votes yet EVE players and the current population of Britain can't see the benefits.
Coming and going Malcanis.
I especially like the way you run your mouth off about the system before trying to understand it....exhibiting great CSM potential right there.
Don't worry you are not the only prospective candidate in that territory.

Is there still no change to the policy of revealing candidate real names publically on the Internet? I wonder how many potentially great representatives we're missing out on because they don't want to be e-stalked for blowing up someone's Internet spaceship.

The fact that the mittani has had real life death threats as a result of his name being public makes a pretty clear case for this.

The fact that the mittani has had real life death threats as a result of his name being public makes a pretty clear case for this.

The fact that the mittani has had real life death threats as a result of his name being public makes a pretty clear case for this.

Except in the IT guy's own thread about voting reform, the dominant opinion was that low voter turnout was the real issue that needed fixing, and that the voting system itself wasn't the issue (and could actually be detrimental). In that case, it's not "our" fault at all that the IT guy only wanted a solution that could be coded into a website.

accountability is king.

CCP, has the topic of compulsory voting for the CSM been discussed internally? It may be something to discuss since we are changing the voting system.

CCP, has the topic of compulsory voting for the CSM been discussed internally? It may be something to discuss since we are changing the voting system.

I certainly think they should change their approach to make it seem more relevant. As it stands, a LOT of people don't vote because they feel like it doesn't matter.

accountability is king.

It doesn't matter, and it shouldn't matter. The CSM has no power, and that's a great thing.

As soon as any power accrues to the CSM, you're going to see all of the corruption we get in real life politics. No thanks.

One system that isn't feasible is Schulze-STV. While it is a very good counting method, its computational complexity explodes as the number of candidates and seats goes up. A back-of-the-envelope calculation indicates that a Schulze-STV election with 14 seats and 28 candidates would take over 9 years to compute on a decent PC.

And "you're" grammer is sub year six.

Schulze runs in polynomial time, not exponential time.

The Schulze algorithm is a widest path problem. It has a running time of N^3, where N is the number of candidates. It's quite an efficient algorithm for all candidate sizes.

You're way off base here, Trebor. I realize though that your programming experience is circa-1981.

Schulze-STV is different than regular Schulze. They're similar in that they're both path-based, but Schulze-STV is based on all possible group outcomes, making the graph size grow combinatorially.

Schulze runs in polynomial time, not exponential time.

The Schulze algorithm is a widest path problem. It has a running time of N^3, where N is the number of candidates. It's quite an efficient algorithm for all candidate sizes.

You're way off base here, Trebor.

I realize though that your programming experience is circa-1981.

I'm sorry, but you are incorrect. You are talking about the Schulze Condorcet Algorithm for a single seat election, which is not suitable for a multiple seat election.

For that, you need Schulze-STV. Schulze-STV exhaustively compares all possible quorums to find the best one. As such, it explodes factorially. If you have n candidates and m seats, you need to evaluate n choose m possibilities.

A Schulze-STV election selecting 14 candidates out of 20 requires evaluating 38760 possible results. A test run of this took ~8 minutes or so to execute.

Running last year's election, of 14 candidates out of 40, would require evaluating 23206929840 positions, which would take 598,734 times as long. Even if we can get the code running 100x faster, that's over a month.

tl;dr: I'm sure this will come as a shock to many, and to you in particular, but you don't know what you're talking about.


And clearly I knew more about algorithmic complexity then than you do now.

And clearly I knew more about algorithmic complexity then than you do now.

I don't like the 'if the guy you picked had enough votes already, you get to vote again' idea. That's obviously catering to the big alliance blocs so that their votes won't be wasted.

I DO like the 'if the guy you picked didn't have enough votes to win, you get to vote again' idea.

Selectively transferring undervotes but not overvotes was one of the principles at the core of Trebor's original proposal, actually. The response was adamant assertion by many that this selective transferability constitutes disenfranchisement of larger blocs, a notion that CCP Xhagen appears to agree with. So it's been discussed,but for now, a transferable vote is a transferable vote, and all are eligible. For better or worse, I'm looking forward to seeing how it all plays out.





Oh! ....and inb4 latest round of "HANS IS DEFENDING TREBOR'S PROPOSAL" nonsense.

I'm sorry, but you are incorrect. You are talking about the Schulze Condorcet Algorithm for a single seat election, which is not suitable for a multiple seat election.

For that, you need Schulze-STV. Schulze-STV exhaustively compares all possible quorums to find the best one. As such, it explodes factorially. If you have n candidates and m seats, you need to evaluate n choose m possibilities.

A Schulze-STV election selecting 14 candidates out of 20 requires evaluating 38760 possible results. A test run of this took ~8 minutes or so to execute.

Running last year's election, of 14 candidates out of 40, would require evaluating 23206929840 positions, which would take 598,734 times as long. Even if we can get the code running 100x faster, that's over a month.

tl;dr: I'm sure this will come as a shock to many, and to you in particular, but you don't know what you're talking about.


And clearly I knew more about algorithmic complexity then than you do now.

Selectively transferring undervotes but not overvotes was one of the principles at the core of Trebor's original proposal, actually. The response was adamant assertion by many that this selective transferability constitutes disenfranchisement of larger blocs, a notion that CCP Xhagen appears to agree with. So it's been discussed,but for now, a transferable vote is a transferable vote, and all are eligible. For better or worse, I'm looking forward to seeing how it all plays out.





Oh! ....and inb4 latest round of "HANS IS DEFENDING TREBOR'S PROPOSAL" nonsense.

As for people crying a river over Alex Gianturco, please stop. Nobody should be harassed irl, but nobody should be running around telling people to 'kill yourself irl' and encouraging people to harass a mentally ill person in an effort to get him to commit suicide, either.

Alex Gianturco behaved terribly, apologized, and was removed from CSM. Drop it already.

Y'know, I know that the white paper is filled with fluff, but...read much?

Beyond that the buddy system could be easily abused for this. I'd venture a guess CCP would dish out bans but a buddy account is 'free' for the first 51 days. If it took a 52 day old account then sure, problem solved. As is i think the concern has some merit but won't matter in reality.

Beyond that the buddy system could be easily abused for this. I'd venture a guess CCP would dish out bans but a buddy account is 'free' for the first 51 days. If it took a 52 day old account then sure, problem solved. As is i think the concern has some merit but won't matter in reality.

Is there anything to prevent the largest blocs/alliances from running and ranking identical 14 candidate ballots?

Abuse of the buddy system is considered an exploit. I have no doubt that CCP Sreegs cannot wait to bludgeon the intromissive organs of anyone who tries to mess with the election using the buddy system.

After reading Xhagen's posts, I quite understand where he stands.

CCP must accurately listen to those who put an effort and vote, that's right.

That not right: CCP better figure out away to listen to those that are not voting whom compromise near three quarters thier customer base.

No, but under STV it won't help them. A vote has one unit of voting power. If a group of voters has 4/14 of the voting power, they'll get about 4 seats, whether they run 4-5 candidates or 14.

I must've missed something. In the old system, a group with 4/14 of the votes got one seat because they had to ensure a win. In the STV system, they'll get 4 guys on the CSM. How does that not "help them?"

Are we saying that a group with 4/14 of the votes cast deserves 4 seats? As a tool, a diverse CSM would be most useful.

It just seems to me that this system will work best relative to the number of votes cast. More votes overall could generate a more diverse panel. Maybe. But if that's the case, then working on voter turnout should have preceded the switch.

Unless voting is going to be compulsory, with a popup in the middle of your client that won't go away until you vote, there really is no guarantee folks are going to care more about this election than elections past.

And if the result is a stacked CSM, well, everyone loses.

YK