EN24 discord
sov map

CSM 9 election – A review of the vote

May 14, 2014

This is a wall of text for those interested in my analysis of the CSM 9 vote breakdown.

Unlike many Eve players, my first life country uses a STV very similar to the Eve CSM elections for many of our parliamentary elections.  Here are my thoughts.  Any errors are my own. If you have better information please leave a comment.

There is a point I want to make about the single transferable vote system that appears obvious to me but appears strange to others. To be elected, you want first preference votes, but also want preferences from others.  In essence if you get a quota in your own right, you are in. If there are several candidates short of a quota, these candidates need recommendations (i.e. votes) from others.

Disclaimer: I am not entirely unbiased in my Eve politics.  The analysis should still be useful for those that follow such things regardless.

The elephant in the room.

This years total vote dropped from 49,702 for CSM 8 (1) to 31,294 for CSM 9 (2).  That is, for every 5 voters last year, this year we had 3.  Last year, a quota was 3314, this year the quota was 2087.  Regardless of how the vote turns out, this drop is an elephant in the room, and should be talked about.

I expect that those who vote are the more engaged of the Eve player base.  There has been a fall of 37% of those participating in voting.

My question is: Does Eve have a 37% fall in activity amongst our most engaged players?

CCP and CSM, you have an NDA and can talk quietly amongst yourselves if you want, but address this. I don’t need the answer, but CSM 9 please also ask the question about the correlation between voting in CSM elections and whether members are still playing 12 months later.

The new candidates of note

Sugar Kyle is the standout, with 91% of a quota.  A relentless campaign on her blog, and good feelings all around from others.  In my opinion Sugar picked up some of the vote that went to WH candidates last election.

Brave newbies candidates of Matias Otero and Awoxing Pizza-Spymaster McBlueshooter with 86% of a quota between them. The Awoxer with the too long name did assist Matias with preferences.

Wormhole vote comments

Taking Rhava’s list of Wormhole candidates (3) (Asayanami Dei, corbexx, James Arget, Karen Galeo, Proclus Diadochu), there were 1.6 quota’s of votes for them; down from 1.9 on CSM 8.  With suitable preference flows, this could be hoped to translate into 2 candidates.  As only 1 was elected, clearly there were not enough non-WH recommendations for this to occur.  I suspect in part, this is because of some very strong non-block candidates.

CSM8 James Arget remained in the running until there were 18 candidates remaining, finally knocked out at this stage with 887 votes; short what would turn out to be 508 votes to get the final seat.  (More on James below)

Last year, everywhere I went, the wormhole candidates referred to to each other.  This did not happen this year.  Also, last year most groups had the wormhole 5 on their ticket; recommendations that this year went to other candidates.

Asayanami Dei, who was a wormhole candidate was also the last exclusion.  Requiring only another 80 votes,  improving anything in the campaign would have gained those extra votes.

Goonswarm vote comments

While Sion’s vote was down in raw numbers (4,314) from the lead Goonswarm candidate last year (mynnna’s 5,782), it was still a significant increase in relative terms; being over 2 quota in their own right.

If I recall correctly, I believe that last year Goonswarm had still had some of their strategists thinking of first past the post elections and tried to spread some of the vote.  This year they realized that with a single transferable vote system, this is generally unnecessary, and may even be harmful (4).

I assume that they wanted Sion to be a permanent member instead of mynnna.

The top 7 on the Goonswarm ticket apparently was Sion Kumitomo, mynnna, Alia Aras, Xander Phoena, Jayne Fillon, Angry Mustache, Mangala Solaris

Goonswarm candidates (Sion, mynnna, Angry) had 2.6 quotas; electing 2.

Goonswarm + Gentlemans Agreement (Xander) had 3.0 quota; electing 3.

GS + GA + RVB (Mangala Solaris) (5) had 3.4 quota; electing 4.

This last one is curious; In CSM 8, Mangala had a personal vote of 80% of a quota (2681 votes).  This year, with a significant fall in overall vote, he is now in 41% of a quota (857 votes).  This year Mangala gets in on sympathetic preferences, but his personal base has deserted him.  More on Mangala below.

Last year, GS (mynnna + Unforgiven Storm) + GA (Kesper North) + RVB (Mangala Solaris) had 3.1 quota (unless I missed someone).

Rhavas quoted this, but I will too : http://themittani.com/news/gsf-ceo-update-tri-harder

“…  my personal take on this election is that turnout will be at an all-time low. If I’m right, that’s hilarious news for the CFC, because there’s nothing this coalition does quite like marching and voting in lockstep. In a low turnout situation, the impact of organization, unity and force is magnified.”.  There is some justification for this statement, but looking at these primary votes :

  CSM 9 CSM 8 %Fall
Goon 5254 6684 78.61%
Goon + GA 6199 7602 81.54%
Goon+GA+RVB 7056 10283 68.62%

Goonswarm also suffers from the fall in vote that everyone else has, thought to a lesser extent.  The average fall in vote from CSM 8 is 63%.

Goonswarm was not as organised or motivated as it would like to claim.

State of the count with 15 candidates remaining

Each voting system has it’s hard to call spots.  With a multi candidate STV system, it is the last spot.

By the time there are only 15 candidates remaining, the following 12 candidates had received a quota of votes a few in their own right; most on preferences.  They are ‘comfortably elected’, though a candidate would rather be higher up this list than lower down it.

  • Sion Kumitomo
  • corebloodbrothers
  • Sugar Kyle
  • Ali Aras
  • Steve Ronuken
  • progodlegend
  • mynnna
  • Xander Phoena
  • Matias Otero
  • corbexx
  • Mike Azariah
  • DJ FunkyBacon

There are 3 remaining candidates; 1 of which will not be elected

  • Mangala Solaris has 1550
  • Major JSilva has 1474
  • Asayanami Deihas has 1395.  The candidate with the fewest votes is excluded and so all remaining seats are filled.

When I was looking at the GSF + GA + RVB vote above, I was curious as to how Mangala Solaris was elected.  From a personal vote of 857 he picks up 693 preferences from other candidates, never particularly strong preference flows, but significantly better than random.  Mangala and Major are the last two to be elected.

Unlike last year, there are no strong preferences to be gained from the last to be excluded candidates.

The Strong preference flows

In Eve, voters choose their pilots, and while we can provide recommendations, pilots are a prone do not always follow the party line.  Below are the exceptional recommendations where pilots voted in step.  Numbers are taken from the ’15 candidates remaining’ section of the vote:

  • Sion to mynna : 2410 of 2521 votes
  • mynnna to Xander  : 1081 of 1589 votes
  • Corebloodborthers exhausted 864 of 1148 votes

Improved candidates from CSM 8

The candidates that re-ran that next years candidates want to emulate are :

Corebloodbrother.

Last year the combined Provi block (Ali + Corebloodbrother) had 1.1 quota, this year Corebloodbrother received 1.4 quota all by himself; with the surplus largely exhausting.  This was a drop in total ‘Provi block’ but considering Ali still was voted in consuming some of this, it is a significant effort.  CSM 10 Candidates should want to be on his ticket.

Steve Ronuken

Steve picks up significant support not only in quota (made easier by that drop in vote I keep harping on), but also in primary votes as well. Now has 1692 (81% quota) primary votes, up from 1286 (38% quota).  A strong campaign, made stronger by general support from others.

Roughly the same support from CSM 8

Ali Ares, despite changing from Provi to Merc, has in rough terms the same primary support.

progodlegend similarly has roughly the same primary support

mynnna was no longer the Goon primary candidate, but is easily elected on Sion’s preferences.  He however also gets even more support as other candidates are excluded.  I assume there was a decision to put mynnna on the council but not as a permanent member.

Some … other candidates from CSM 8

James Arget.

On again off again, defer your official candidate announcement until the last moment candidate.  In CSM 8, James had a personal vote of 1624 (49% quota).  This year, James had a personal vote of 653 (31% quota).  James, this is not how to run a campaign.

Mangala Solaris

Preferences flowing to Mangala say that those sympathetic to RVB still support him, and off that soft support, Mangala was elected.  A drop of personal votes from 2681 (81% quota) to 857 (41% quota) is truly remarkable and not in a good way.  RVB needs to find their core support again if they want to keep a seat on the CSM.

Last year I observed a strong correlation between primary support and benefit from preference flows.  For me, the real surprise here is that Mangala somehow managed to keep strong preference flows despite his personal vote collapse.

riverini

From Eve News 24. Really?  Coming from last year having 26% of a quota from last year, dropping to 10% of a quota this year is not a good showing.

Other notes

It is good to have most spaces represented this year: High, Low, WH, (lots of) Null, industry, PVP and out of game tools.

I congratulate all successfully elected candidates.  Commiserations to all those who failed to get a seat.  Or given the workload, is it the other way around?

I have slighted some of you by not mentioning you by name, and my apologies.  I just don’t know enough about how you were elected (or not) to add anything meaningful.  This assumes that the words above are meaningful and helpful to those considering future CSM campaigns.

Some footnotes of references

(1) Trebor’s re-run of the votes for CSM8.

(2) community.eveonline.com and from that page: cdn1.eveonline.com I have been using auditlog – 14 seat.txt as the source of voting data.

(3) http://interstellarprivateer.wordpress.com

(4) Please note the qualifications of ‘generally’ and ‘may’.  For the truly tragic election follower, reading up on the election in Braddon in Tasmania/Australia 2014 at kevinbonham.blogspot.com.au is interesting reading.  This was truly a ‘remarkable’ result and has lead to some commentators including myself reconsidering previous advice.  TLDR; Even with single transferable votes it is sometimes better to have a split vote.

(5) Yes I have drunk the kool-aid that says RVB has strong relations with GSF.

My working spreadsheet is available at drive.google.com .  This might or might not be useful to others.

 

Want to read more of Foo’s work? Then check out his Blog

tweetfleet