Using Loomio to select a candidate to represent a group
Olover has just started off a great conversation in the "Welcome and Introductions" thread. I've pulled it out into it's own thread here:
Olover says:
"I am wondering what you suggest is the best way to use Loomio to select several people that will then represent the group on a committee that is being formed.
Would you simply start a discussion and then those who want to nominate themselves as representatives can start a proposal about that, at which point others in the group vote to agree or disagree with electing a specific candidate? It seems like that wouldn’t be so easy, unless we ask people to only vote for one or two candidates and vote disagree on other candidates.
I’d appreciate any ideas you may have!"
My response:
"Right now in one of my groups there’s a new discussion about choosing a new board member.
One of the existing board members has started the discussion with some context about who is on the board and how it works, with links out to relevant documents. Then immediately after starting the discussion, they launched a proposal: “I would like to have a conversation about joining the board”.
This has given people a space to self-nominate, without presupposing the outcome of any decision. That might be a good way for you to start the conversation in your case @olover
Once you have an idea of who is interested in the role, that will make the decision-making much easier.
You could even keep it all in the one discussion thread - the first proposal is to seek self-nominations, and then later once the discussion has progressed, a second proposal could be used to confirm a specific candidate or group of candidates.
What do you think?"
Olover Mon 9 Feb 2015 11:21PM
Thanks @richarddbartlett !! Your idea is an interesting way to go about it, thanks. I do think it might get more complicated past the self-nomination step, but I think we can cross that bridge when we come to it, since we first have to see how many people will want to be on the committee in the first place. After that, I think we can have an individual proposal started for each candidate that would serve as a repository for votes for that candidate and see where that takes us. Or we could use a Google form to poll the participants on which candidate they would like best, though I guess this doesn't address the issue of the vote-splitting spoiler effect mentioned by @gregorycassel. some thoughts...
Alanna Irving Tue 10 Feb 2015 4:20AM
@gregorycassel we have already started using Cobudget for digital dot voting :) They are working on a "points mode" which is about indication preference, not specifically dollars.
Olover Tue 10 Feb 2015 4:59PM
I am wondering if it's possible to join Cobudget? I sent a request I think... would be nice to check it out.
Quentin Grimaud Wed 11 Feb 2015 7:27PM
We applied Richard's method approximately one year ago to decide who should have a given role (spokesperson, discussion is here https://www.loomio.org/d/w4Lr5DgU/termine-candidats-au-tirage-au-sort-pour-etre-porte-parole ) in our horizontal community candidating for european elections. So the thread was mainly a call for candidates for the given role and to discuss about how we would decide who would have the role, and how many "winners" there will be. We decided to draw them by lot using what has then become our usual verifiable method (which is: the winning order of candidates is obtained by removing iteratively the candidate at the position in the list which is the rest of division of the closing value of a stock exchange at a given future date, by the current number of candidates).
Olover Wed 11 Feb 2015 10:23PM
Hi @quentingrimaud , thanks! I do not quite understand how you picked the spokespersons out of the group of candidates. But I suppose it's just a method to pick a random one (or two or three, etc.) of the people who stepped forward. Am I right? If so, any ideas on how to have the group's preferences decide which candidates would be chosen instead of random selection? Thanks!
Quentin Grimaud Thu 12 Feb 2015 3:27PM
Yes you are right :). If you want to exclusively use loomio, I guess it will be difficult or will require a lot of iterative proposals. If you are open to other tools for this specific election, you can have a look at some tools which implement election among more than 2 options, like https://vote.heliosvoting.org/ or http://www.opavote.org/ . The basic idea is either to make voters rank the candidates by order of preference and then apply a Condorcet-compliant method to decide the winning order, or to use approval voting which makes voters select the set of candidates they accept. And from the resulting winning order, you select the top X candidates, depending on the number of winners the vote should have.
Quentin Grimaud Thu 12 Feb 2015 3:29PM
There is a good course on Coursera about voting methods, you may want to check it out, and the members of the loomio community who were interested in following it have created a dedicated group here: https://www.loomio.org/g/Z2iQCjaz/loomio-community-making-better-group-decisions-course-study-group
Greg Cassel · Mon 9 Feb 2015 10:07PM
This is an important subject. I love Loomio, partly because of its focus on binary yes/no proposals. I think that strong focus makes it a less than ideal tool for selection from three or more options.
I broadly advocate preferential voting for selections from more than two options. Ranked voting such as IRV or STV is the most recognizable form of preferential voting. Such techniques allow each voter to indicate their true preferences, without creating the vote-splitting spoiler effect which we often see in choices between three or more candidates.
@richarddbartlett , I agree strongly with your ideas regarding self-nomination process.
I think ultimately that Loomio is best used with other organizational tools, including but not limited to Cobudget. Cobudget is a fantastic way to articulate the proportional preferences of a group. In fact, it could be adapted for voting systems in voluntary groups of good-faith participants. (I would not recommend it for general political use, in its current form, because of the risk of orchestrated last minute 'vote switching' games by competitive factions.)