There was a massive failure of having significantly less voters this year than the previous: 49702 vs 59109, breaking a long positive trend. While CSM7 wasn't something spectacular, it couldn't cause the loss of voters, it could only cause the loss of votes for the CSM7 members re-running. If I find the president horrible, it's more likely to show up for voting next time.
I saw two reasons for failure: one is the over-complicated voting process. The other is the PR of nullsec blocks which tried to make everyone turn away from voting to increase their power. Ironically they decreased their own voters, probably because a "highsec carebear" finds more fun in figuring out a complicated voting process than a "let's shoot stuff" guy. The null blocks meta-gamed since ever, so even the second cause is CCP fault: creating a system that could be communicated as over-complicated, obscure, simply because it was.
Now, before someone would comment: STV is better than first-pass-the-pole voting. It represents the wishes of the voters better and decreases the losses via votes on losers and overvotes. The PR they finally made up "42299 votes affected CSM8 while only 31801 affected CSM7" is true.
However STV needs serious, committed voters who fill up the ballot with well-thought candidates, and this was something they could surely not expect, considering the low voting tendencies. Even CSM7 votes were 15-25% of the total vote-able accounts. CCP tried to save partial votes when large majority of the votes were lost by never being cast. The same effort to publicize CSM, to give CSM in-game tools to communicate with people, to run opinion surveys, to send the CSM "tickets", questions and town-hall-ish requests would have caused much more votes, therefore much wider support.
But still, there is no reason to not optimize the voting system. However their own results shown that most exhausted voting happened on 1-3 long ballots, while almost 40% of the ballots had this short length. It is obvious (and was obvious) that serious amount of people don't have a long priority list. He might like one or two candidates but not more. Interestingly, the 1-long ballot-"winner" was Greene Lee, who did not get on the CSM8.
CCP should have kept it simple. "Simple" is not the opposite of "optimized", as we can have both. Most IT systems offer "simple" and "expert" modes. Even the sell and buy windows of EVE market have it. The CSM election screen should have been a simple mode one, having only one slot and the candidates alphabetically, allowing the casual voters to cast one vote. Besides the "cast vote" button, should have been the "expert mode" button that leads to the current STV screen.
Of course this would have created lot of 1-long ballots, and those had 52% exhaustion value. However such vote still has 48% power, compared to the zero of the uncast vote. Finally "simple mode" wouldn't have to create 1-long ballots. There are two ways to increase the power of simple votes. One is simple ballot copy: if I simple-vote for Mynnna, I express that I trust Mynnna alone. This case it's rightful to create a 14-long ballot for me: the exact copy of the ballot Mynnna casted. Remember, I gave all my trust to Mynnna with my simple-vote. If I'm not happy with that, I can custom-vote. This system would also save lot of time clicking for bloc-voters: "just simple-vote for X guys!". The other method is not processing 1-long ballots as ballots, but directly subtract their number from the quota of the candidate, allowing the longer ballots to exhaust less. This case a simple-vote is 100% used, unless a candidate is elected alone with simple-votes or not elected at all.
Having optimized systems is good. But having simple front-end when you deal with simple people is more important. I hope CSM9 elections will be better organized.
I saw two reasons for failure: one is the over-complicated voting process. The other is the PR of nullsec blocks which tried to make everyone turn away from voting to increase their power. Ironically they decreased their own voters, probably because a "highsec carebear" finds more fun in figuring out a complicated voting process than a "let's shoot stuff" guy. The null blocks meta-gamed since ever, so even the second cause is CCP fault: creating a system that could be communicated as over-complicated, obscure, simply because it was.
Now, before someone would comment: STV is better than first-pass-the-pole voting. It represents the wishes of the voters better and decreases the losses via votes on losers and overvotes. The PR they finally made up "42299 votes affected CSM8 while only 31801 affected CSM7" is true.
However STV needs serious, committed voters who fill up the ballot with well-thought candidates, and this was something they could surely not expect, considering the low voting tendencies. Even CSM7 votes were 15-25% of the total vote-able accounts. CCP tried to save partial votes when large majority of the votes were lost by never being cast. The same effort to publicize CSM, to give CSM in-game tools to communicate with people, to run opinion surveys, to send the CSM "tickets", questions and town-hall-ish requests would have caused much more votes, therefore much wider support.
But still, there is no reason to not optimize the voting system. However their own results shown that most exhausted voting happened on 1-3 long ballots, while almost 40% of the ballots had this short length. It is obvious (and was obvious) that serious amount of people don't have a long priority list. He might like one or two candidates but not more. Interestingly, the 1-long ballot-"winner" was Greene Lee, who did not get on the CSM8.
CCP should have kept it simple. "Simple" is not the opposite of "optimized", as we can have both. Most IT systems offer "simple" and "expert" modes. Even the sell and buy windows of EVE market have it. The CSM election screen should have been a simple mode one, having only one slot and the candidates alphabetically, allowing the casual voters to cast one vote. Besides the "cast vote" button, should have been the "expert mode" button that leads to the current STV screen.
Of course this would have created lot of 1-long ballots, and those had 52% exhaustion value. However such vote still has 48% power, compared to the zero of the uncast vote. Finally "simple mode" wouldn't have to create 1-long ballots. There are two ways to increase the power of simple votes. One is simple ballot copy: if I simple-vote for Mynnna, I express that I trust Mynnna alone. This case it's rightful to create a 14-long ballot for me: the exact copy of the ballot Mynnna casted. Remember, I gave all my trust to Mynnna with my simple-vote. If I'm not happy with that, I can custom-vote. This system would also save lot of time clicking for bloc-voters: "just simple-vote for X guys!". The other method is not processing 1-long ballots as ballots, but directly subtract their number from the quota of the candidate, allowing the longer ballots to exhaust less. This case a simple-vote is 100% used, unless a candidate is elected alone with simple-votes or not elected at all.
Having optimized systems is good. But having simple front-end when you deal with simple people is more important. I hope CSM9 elections will be better organized.
0 comments:
Post a Comment