Welcome to Jester's Trek.
I'm your host, Jester. I've been an EVE Online player for about six years. One of my four mains is Ripard Teg, pictured at left. Sadly, I've succumbed to "bittervet" disease, but I'm wandering the New Eden landscape (and from time to time, the MMO landscape) in search of a cure.
You can follow along, if you want...

Sunday, May 5, 2013

Unrealistic

I had hoped to write this post after we received the full official CSM8 election results, but it looks like those are going to be delayed somewhat.  So I thought I would write a bit on how my election predictions came out in light of the first set of election results.

Needless to say, I'm really pleased!  Of my 14 predictions, 11 of them are on CSM8, including myself.  And I have to say that I'm really pleased at the group that we've ended up with!

First, what I got right.  And that's a pretty good list.  All seven of my upper tier candidates -- mynnna, Sort Dragon, Mangala Solaris, Trebor Daehdoow, Malcanis, myself, and James Arget -- are on CSM8 and clearly put up good showings.  I chose James over Cipreh as the candidate that I figured the hard core wormhole voters would go for, and I got that right.  Malcanis, Trebor, and myself cross-endorsed each other, I figured that we wouldn't have much trouble getting on.  And of course picking Mangala, mynnna and Sort were no-brainers.

I also did pretty well in my bottom tier.  I figured either progodlegend or Ali Aras would get on.  Thanks to TEST clearly dropping the ball with their voters, both of them got in, replacing the third HBC candidate.  Kesper North and Sala Cameron also didn't have much trouble getting on.  I got Nathan Jameson wrong; I figured that the w-space vote would be strong enough to put two candidates on CSM8 and that one of them would be a bit more generalist.  I thought that would be Nathan and it was Chitsa Jason instead.  Likewise, I figured the Russian voter bloc would be strong enough to ensure one candidate.  I guessed Greene Lee would make it and Korvin made it instead.  However, that one was always going to be too close to call.

The place where I (and honestly, everyone else) really dropped the ball: Mike Azariah!  Now don't get me wrong; I'm delighted to see Mike finally make the CSM.  But I think even Mike is a little bit bewildered about how it happened.  Of all the questions I'd like to know the answer to when we get the full results, "How did Mike pass so many other candidates?" is gonna be right on top.

In the middle: I guessed that voting numbers would be equal or lower than last year.  I was a little concerned that they'd be significantly lower than last year but didn't predict they'd be so much lower.  That's a real pity, but predictable in retrospect: with a quieter CSM7, there weren't as many people motivated to vote for CSM8.

And finally... what the hell happened with those permanent seats?

Picking mynnna to take one of the two permanent seats was an easy prediction and I was pretty sure that someone would push Sort Dragon out of the second slot.  And after I had a chance to talk to Trebor further at Fanfest about how STV works, the fact that it wouldn't be Sort in that second slot was kind of obvious.  When you break the voting down to only two seats, in order for the blocs to control both of them, they would have had to have controlled up to three-quarters of the votes.  The fact that they clearly did not do so meant that there was going to be one bloc seat in those top two and one non-bloc seat.  It would just come down to whomever of the non-bloc candidates had the most #1 votes.

I never dreamed it would be me.  I figured it would be Malcanis, or Trebor, or Mangala.  So thank you to all of you for that.  However the numbers come out, you all clearly turned in an unrealistic result on my behalf.  ;-)

As for what happened with the CFC and HBC third seats?  Trebor's got a really good blog post almost finished that will detail what happened there, but the short version?  The CFC and HBC didn't realize that an in STV system, if their weakest candidates didn't have at least a few #1 votes, they'd get eliminated in the early rounds.  And that looks like exactly what happened: they were pushed right off the ballot by mid-range candidates that didn't have the votes they needed to get elected but had enough #1 votes to themselves to keep themselves from being pushed out in the early rounds.  Since Banlish and Kaleb had only very few #1 votes, I think we're going to find that they were eliminated early and didn't survive long enough for the CFC and HBC trickle-down vote to reach them.

It's not a mistake the blocs will make next year.

So all in all?  I'm really pleased with how my predictions came out and I'm really pleased with CSM8.  We've got a really good spread of representatives and in our early discussions it looks like the large bulk of us are going to be active participants in the process this year.  In a typical CSM, you can count on five active members.  In the early going, it looks like CSM8 is going to at least double that and I couldn't be more pleased about it.

Time to get to work!

15 comments:

  1. The sooner CCP can publish some figures about this, the sooner the tinfoil can be packed again.

    Another question might be, why did players votes for Mike. Personally, I had never heard of the guy before this. Never read his blog, and other than a twice glance read, have not been since the election closed. But when certain people I loathe start throwing mud, now I am interested. (general rule of "if you don't have something nice to say...."). I regard myself as a casual player. Mike came over well in his Crossing Zebra interview. He held up very well in a multi-candidate interview against rest being null-sec players. I also voted for Unforgiven Storm because I believed that Industry representation on CSM needs an experienced voice. And was disappointed at the lack of support from his alliance. I hope the guy will run again.

    MT

    ReplyDelete
    Replies
    1. It is already too late for those ballots to be released, without being accompanied by tinfoil. CCP has had plenty of time during the past week to tweak the ballot data to match any results - actual, announced, or otherwise.

      CCP should have released the ballots, as they came in during the election. While this would not prevent "cheating" on CCP's part (they could always "stuff the box"), it would have been much less suspicious than holding on to the ballots for more than a week after the election was over.

      Delete
  2. It just goes to show that EVE is played by people with higher than average IQ scores. Clearly, the "intelligencia" bloc is a force not to be underestimated nor is it likely to fall for obvious meta gaming propoganda from a second rate news site.

    ReplyDelete
  3. I'm curious if the Incursion chat channels vote push was significant or not, but I doubt when the results do come out that'll be measurable. I do know in the Armour channels both our 2 endorsements made it ( you & mike )

    We were happy with the results getting at least 2 CSMs that that've done incursions where the previous 2 CSMs to my knowledge had near 0. Maybe last years Escalation nerf got a lil fire under our behinds to vote :)

    ReplyDelete
    Replies
    1. Actually my Incursion channel(TDF) in the last weeks supported/endorsed a third candidate at the bequest of a few russian Incursioners (Zwo zataki had put him forth ) but I forget if it was Korvin or another russian TBH.

      Delete
  4. Even if the blocs split 1st prefs (say) 3-2-1, they could still easily end up with weaker candidates eliminated.

    Let's say a bloc has raw votes worth just under 2 quanta. Under a simple 1-2-3 voting pattern, A is elected and redistributed, B is almost elected, and C has no votes and is eliminated early.

    Now say they spread the first prefs around. A almost has a quantum (gets half the original votes), B has 2/3 of a quantum, and C has 1/3 of a quantum. In order for A to be elected, either C needs to be eliminated or there needs to be a significant trickle of votes from other early candidates (enough to elect A and spill enough into B and C to carry them over the line).

    We got 49702 voters, which means a quantum is 3550 votes. Had (say) HBC been able to generate 12000 voters (as suggested by the above link), they should have put three candidates over the line to start with, and still had ~1400 votes to spread around. The obvious conclusion is that they didn't get as many votes as expected, and the cross-linking of the minor parties was enough to push those ahead.

    But even if they'd spread their #1 votes, I'm not sure it would have helped that much. Those extra votes have to flow in from *somewhere*, else all they have done is delay when their C candidate is eliminated.

    Actually, I suspect your vote has a big part to play. You got the second highest #1 vote total, and many who voted for you #1 would have followed something like your endorsements. Thus, there's a major flow from you to the "minor candidates", which helps push some of them above the spill from the majors.

    One could conclude that the CSM elections have always been "bottom-heavy" (lots of small-moderate votes spread among lots of candidates), and STV is finally capturing that vote.

    ReplyDelete
    Replies
    1. "Actually, I suspect your vote has a big part to play. You got the second highest #1 vote total,"

      Riptard did get #2 not #1? I didn't realize the numbers were released or are you making an unfounded assumption?

      Delete
    2. The numbers have been released today, looks like Myanna was #1 STV-2 result in CSM8 given by CCP Dolan: Mynnna was elected with 10,348 enfranchised votes and Ripard Teg was elected with 10,301 enfranchised votes. Together that is 20,649 enfranchised votes to elect the STV-2 winners.

      Delete
    3. Apparently it was a completely unfounded assumption, as they used an STV-2 vote for the top two place-getters rather than simple 1st past the post.

      An interesting quirk is that Mangala Solaris and Ripard were almost tied for the 2nd spot (after preference redistribution, of course). This implies that they were the only 2 candidates remaining at that point (Myrnna already having been elected), and thus also that non-bloc candidates received a significant percentage of the primary vote (leftover from Myrnna wasn't sufficient to carry a 2nd bloc candidate into 2nd place).

      Delete
  5. Just don't forget the occasional guide, and you'll have my vote next year. Remeber to enjoy the game and not become a bittervet and you'll get my #1 again. Congrats and best of luck in politics

    ReplyDelete
  6. Those CFC and HBC boys are pretty stupid, aren't they?

    ReplyDelete
  7. This comment has been removed by the author.

    ReplyDelete
  8. There are several misconceptions about STV in your blog and the comments.

    A) To guarantee the top two spots, the HBC/CFC block would have to had 2/3 of the votes, not 3/4.

    B) I will not be surprised to learn that you and mynnna earned the most first place votes, but being ranked anywhere on the most ballots is actually more important. Of course, you need to have enough first place votes to survive the first few rounds. I think it's more likely that you were actually ranked fourth or fifth in terms of number of first place votes.

    C) In Wright STV, overflow votes get reallocated before eliminations, so the bloc votes get allocated down the line immediately. Furthermore, only one candidate is eliminated per round. So the only way that the situation Trebor suggests is applicable is if the HBC/CFC didn't really have enough votes to support a fifth candidate, but Kaleb or Banlish show up on ballots supporting weaker candidates. Since these candidates were primarily relying on coalition support, their failure to get elected suggests that the CFC/HBC simply didn't have the combined votes to support a fifth candidate.

    D) Debir's example about why Trebor is wrong (as I explained in (C)) is pretty good, except the quanta is not 49702/14 + 1 but actually 49702/15 + 1.

    ReplyDelete
  9. Now that the numbers been released it appears that Mike Azariah made it inby the skin of his teeth.

    ReplyDelete
  10. One of the major flaws with the Wright STV system is that the system is overly sensitive to ranking order, due to the methods used for overflow and elimination. This is particularly true when the number of slots on the ballot is greater than 3 (14 is utterly ridiculous, btw).

    Essentially, you can get eliminated *before* votes can trickle down to you. And, "noise" in the system overwhelms accuracy, as you increase the number of iterations, particularly since, in practice, voters are less selective and more arbitrary with the lower ranked slots.

    This is why you rarely will see the Wright STV system used in this particular fashion.

    ReplyDelete

Note: Only a member of this blog may post a comment.