Greedy Goblin

Monday, March 30, 2015

CSM 10 election analysis

Like last year, I grabbed the voting results to check for cross-supports. The method was simple: if a voter put candidate A to the first position, B to the second, and so on, than I assigned 54.5% vote for A, 24.8 for B, 11.3 for C, each is 1/2.2 of the previous. This was a fitted number, which provided the best approximation of the real STV results. It approximated:
2772 Sion Kumitomo
2098 Sugar Kyle
1616 Manfred Sideous
1545 corbexx
1537 Endie
1376 corebloodbrothers
1364 Chance Ravinne
1347 Steve Ronuken
1221 Sort Dragon
1141 Mike Azariah
1084 Cagali Cagali
1031 Gorga
1002 Thoric Frosthammer
779 Bam Stroker
732 Jayne Fillon
Only Jayne Fillon was not predicted as winner. This is much better than looking at just the first vote (giving 100% to A), which missed two winners, including Endie, who actually passed the quota after Sion and Sugar. I used these for calculating cross-vote supports. For the example case, B has 45% support from A (1/2.2), while A has 100% support from B. So the A-B cross-votes mean how much the voters of A supported B too.

However we must consider low-vote candidates, since CCP in its infinite wisdom allowed every bored troll to run. Imagine a troll who only appeared on one ballot (his own), placing him first and Sugar second. This means 45% cross-vote. However the cross-vote table is not symmetric, while the value in his row, Sugar's column is indeed 45%, the value in the row of Sugar, his column is 0.05%, since the 100% support from this one vote is divided by the 2K votes of Sugar. So in creation of groups I only cared about the rows of the winners and those losers who were eliminated with more than 1000 votes. This does not mean complete removal of the other candidates, as reasonable cross-support can come from them, just see Tora's "group".

Let's see the table!

Yes, having 75 candidates is bad. Let's remove all the candidates who didn't have 1000 votes at elimination and couldn't even get 6%!!! cross-support from one who had:

Much better! The first group had 6 winners: Chance Ravinne, Mike Azariah, Sugar Kyle, corbexx, Steve Ronuken and Jayne Fillon. This means that the voters considered these people similar. 3 "below 1000" losers got into this group, getting some (but clearly not enough) support from the winners. This explains my surprise that only 3 "good" candidates got in. The supporters of Sugar-Steve-Mike disagreed with me about who else is "good". They valued the silly videos of Chance Ravinne and didn't believe me that anyone who signed the Document of Shame is likely a Goon spy. OK, the evens of Fanfest proved me right pretty clearly.

Let's see the next group, centered on the "over 1000" loser Tora Bushido. His voters supported the below-1000 Sabriz, who also gave him lot of support: before the round where Sabriz was eliminated, Tora had 714 votes, but 838 after. The Tora was the only "above-1000" whose voters supported corebloodbrothers. His voters supported no one else and he received support only from Tora and some tiny ones. He also "won" the "most vasted votes" competition, with 41% (last column). His inability to form a "party" and his signature on the Document of Shame are enough for every Provi voters to think about finding a new representative.

The next group is around winner Cagali Cagali, featuring above-1000 Migui X`hyrrn and below-1000 June Ting. Strong cross-support, sporadic outside support to/from the "good" group. This is how a good political party looks like.

The next group is around above-1000 Lorelei Ierendi. While the cross-supports aren't as strong as in the Cagali party, it's OK. Or would be OK if they'd get a bit more votes. Their combined #1 votes are 1600, which is under the quota. Since they had no connection to other group they couldn't receive help from candidates eliminated before Lorelei, nor help anyone OK being elected after he was eliminated.

Now comes the Nullsec "Elite PvP" group: above-1000 UAxDEATH, winners Gorga, Sort Dragon, Manfred Sideous and below-1000 Bobmon. Very strong cross-support, another well-functioning party.

Then comes the somewhat overlapping CFC party, with winners Manfred Sideous, Thoric Frosthammer, Endie and Sion, above-1000 Bam Stroker and below-1000 Bobmon and Xander Phoena. Strong cross-support all over the place. I don't really see what was the gain for Manfred for the cooperation, probably the Goons promised him a permanent seat, which hilariously backfired, leading to the first time since I can remember that Goons didn't get the top position of the CSM.

The last group is the Xenuria-party, named after the above-1000 mastertroll and the lower tier troll Gorski Car. Too bad that he didn't think about trolling Goons by telling his voters to place an anti-ganking candidate after him.

Finally, let's see the permanent seat support. In this election the quota is so high that no one can pass it till the last round when three candidates enter, this time Sugar, Sion and Manfred. If a ballot had any of them, even if just at the last position, it provided a full vote to that candidate (the first mentioned if more than one was listed). So below I separated the ballots based on nothing but their #1 candidate.
These candidates gave more than 20% support to Sugar for permanent seat. No wonder that she had 9080 votes on the final round.

As you can see, much less candidates supported Manfred than Sugar, he had 6228 votes in the last round, getting #2 position only by spillover votes from Sugar. It's funny that Endie got on the 20%+ list, so 1/5 of the people who put Endie to the first place of their ballot preferred Manfred over Sion for permanent seat.

The Sion supporter list is even smaller. Besides two trolls, only official CFC candidates channeled more than 20% of their votes to Sion for permanent seat, providing him 6275 votes that weren't enough to win.

These candidates failed to explain their voters the importance of permanent seat, not directing them to put at least one of the candidates with chance to the end of their ballot. 16K votes were wasted, having no say in the permanent seat votes. These voters could elect any permanent seat holder they want.


Gorski Car said...

Lower tier troll i am honored

Anonymous said...

There is no importance to the permanent seat anymore. It doesn't come with power any more since Sugar pushed for it to be brought back down to the same level as anyone else. Other than you go to Iceland a couple of extra times it means nothing.

Gevlon said...

The "go to Iceland" is actually "being heard during the summits".

Also, if getting excess votes is pointless, than this guy has no clue about EVE politics:

99smite said...


CSM has no influence over devs prioritizing what and where nerf hammers will hit...

Why did only a few people actually vote? Because the majority does not care. The Raven-levelling carebear wants to keep a low profile because he fears that all the 0.0 whining will lead to highsec income nerfs...

Is 0.0 important? Not at all, highsec residents can play without ever having to worry about alliances, wars anything, even impacts on markets are deniable in the long run...

Burn Jita? Who cares? No need to ever go to Jita when a pilot is only levelling his Raven...

Hulkageddon? Just have a dozen pre-fit retrievers with a DC and a bit of tank and laugh off the losses...

Occupancy based sov? While doing SoE missions in Osmon? Hahahahaha....

Even you, Gevlon, have nothing to do with 0.0. If trading in 0.0 was worth it, you would already have gone there....

Why should players waste their time with CSM???

It might be different, if there was the impression that CSM actually had any influence at all, and that CSM members would actually care for "the community" and not only for their lobby...

Anonymous said...

@Gevlon You can't refute an argument that (paraphrasing) "Since last year the permanent seat has no power." with "The Mittani said three years ago the chairman position had power". Not only for the obvious issue of timing, but also because they talk about different positions.
Now I don't doubt that being at both summits is a good thing, but given the CSM is a focus group, CCP would be crazy not to invite those most relevant to the topics to the appropriate summits. Unless a candidate manages to completely alienate themselves from CCP, they're unlikely not to be included whenever relevant.
Proof of this can be seen in Sion: He said publicly some things that ran very far against CCPs official line at the time, and did so in a very abrasive way; despite this he was published on their website, allowed to stand again and invited to talk at fanfest, not to mention CCP adapting some working practices (e.g. confluence) to deal with the issues he raised. CCP are keen to take the best quality feedback they can, as far as I can see, and any candidate that shows competency in discussion and a deep understanding of an area should be pretty relaxed that they'll be given time whenever that understanding is relevant.
If, of course, you want to chip in $0.02 on every issue you know nothing about, then that permanent seat is what you want, but I can't see any of the top three being anything like that given their histories of posting and playing.

Anonymous said...

"The "go to Iceland" is actually "being heard during the summits"."
Which all CSM members do anyway if they attend by video link.

Anonymous said...

1000 is arbitrary. Use a statistically significant figure as a threshold.

Keep in mind that thresholds can always be gamed.

Gevlon said...

1000 is about "half quota".

Anonymous said...

Even with the analysis I still don't know if it's worth trying again for CSM XI. Still number 20 of 75 wasnt that bad.