Here is my draft proposal for moving forward on the crowdmatching mechanism decision:
Aim for alignment, falling back to consent, falling back to majority vote
I would like to aim for alignment (we all favor a decision). If we do a structured discussion process and still fail to find alignment, we will aim to address concerns enough to get full consent (everyone considers a decision acceptable enough even if not everyone’s preferred option). Only if we find ourselves stuck (where some people oppose a decision in itself, not just think it lesser to something else) will we fall back to a vote. That vote will be a simple majority.
Defining the decision space
We collect in a single document all the parameters of a crowdmatching mechanism that we already have complete alignment on. This shall not be limited to the immediate next version. If we know where we’re heading several versions out, and everyone supports that vision without qualification, we capture that too.
Thus, we are closing discussion (for now at least) to options outside of these parameters.
Code some interactive versions of the simplest models
Hopefully @Salt can get someone (working with us and others as needed) to make a simple visual model in R etc. It just needs to be easy to adjust numbers and see the outputs and compare the different outputs given different inputs.
For example (first draft), we set down specs like:
- Project has a dollar goal (which can be adjusted in order to play with the model)
- 10 patrons to play with
- we can specify how many dollars each patron has as absolute max (how much play money they can access at all)
- make this adjustable (so, we could choose to give all patrons only $1 or to give a single patron $1,000 or give them all $10 etc)
- patrons can pledge up to their available dollars but do not need to pledge it all
Have this visualized and with output numbers so people can simply play with what the effect of each tweak is. A patron pledge of $0 is the same as one of the 10 not being pledged.
We can then play this as a game, like set it in any particular state and ask “what if X is changed?” which could be “you are one patron, the others are fixed, what do you pledge?”
We could try setting some explicit goals for players such as “you win by getting the most $ to the project at the least cost to you personally” (i.e. explicit goal of freeriding). We could have a few players actually play the game or ask people to test it and give us their feedback (both their qualitative replies and we see what they do).
Perhaps we go with even smaller, just 2 players (patrons). Try with 2 real people to see how they play the game. Try a different version with 2 patron players and 1 player being the project (setting the goal target). Try with people knowing how much money other players have or being blind to that.
This would all be one-off game, not monthly-continuing.
Important bit: we don’t need to hit the goal point. We can certainly play with numbers that do hit the goal, but to study the behavior of the game and the players, we should focus attention on the behavior assuming the goal is not hit.
We make sure with this single version of the game that we all understand everything about how it works.
Tweak the game
Then, we try this switching one parameter at a time. We can start with the obvious of switching to a crowd-size goal. Play the game more. Notice the actual changes (e.g. the very same lever of a single patron having their goal increase has a different output in the crowd-goal vs the dollar-goal versions). This will be apparent when we are all playing with the same model.
We can also try making the game iterative.
We can specify game states that highlight particular points. For example, a single player pledges, and then we specify that other patrons drop after that (or reduce their pledges) etc., how do people respond to that?
Gathering understanding
Ideally, I’d like @msiep to at least give leadership guidance on how to frame things (how to ask survey questions etc) to get the most neutral response from friends/family/acquaintances who we ask to play with our model.
We only need enough understanding here that we either (A) all have unqualified alignment and move forward or (B) have some agreement about what further questions we have or where we have different understandings that we need to clear up.
Narrowing exact mechanism options
If we do not get clear alignment just from doing the modeling and game playing, we come up with a clear plan around framing (i.e. the UX design, etc) to see if that makes the difference. And we talk directly with outside projects to get their input (see if projects we want to recruit have a strong preference among the variations we’re considering).
If this process doesn’t get us to alignment, we then move through the fallback process to make some decision and move on anyway.
Why do all this?
Although this may seem like a lot of work, I actually think this is as good a process as any to actively recruit interest, engage with people, reinvigorate interest and motivation etc. We should clarify our mission when we talk to people about the game (i.e. that we’re not just aiming for money as end goal but success for FLO and for community-building etc).
Having the model will let us play with how we want to visualize things for the full working site too. And it will show skeptics that we’re really doing our homework, etc.