As opposed to central planning. For example, if multiple "manufacturing initiatives" are making hair irons, there may be a problem in that the best hair irons are not having sufficient automated manufacturing time allocated to meet their demand, so their users could vote (via approval voting) for the hair irons they prefer to be allocated more manufacturing.
Approval voting is where you can vote for any number of options.
bump because dcss is awesome
Voting is not opposed to central planning.
That's just the interface of it. The good thing about approval voting as it is usually defined is that an approval mark doesn't get weaker by approving other non-winning candidates. The vote is not split up by approving several candidates. It's one approval mark = one point, regardless of how many marks are on the ballot. So, if you suspect that your taste is rather unpopular, you can still approve it along with other alternatives you believe are more likely to win, instead of feeling pressure to give more support to a compromise than your true favorite.
However, approval voting is really for decisions where one alternative has to win, and I don't think that this applies as a rule of thumb for consumer items. If some asymmetric thing like a computer mouse can be produced in both a version for the right-handed majority and one for the left-handed minority, shouldn't it be produced in proportion? Or think of clothes in several sizes.
An approval-style ballot can be counted in a proportional way, that's Thiele Approval, invented by a dead Swede a hundred years ago: The first winner is the standard approval winner. To get the second winner, we reduce the weight of approval marks on all ballots that support the first winner to one half each. To get the third winner, we set the weight of any mark on a ballot which either approve the first or second winner to one half each, and for ballots that approve both the first and second winner we set this weight to one third. And so on.
During the count we elect a candidate, fiddle with the weights of the marks, elect another candidate, fiddle with the weights, elect another candidate, and so on. During the count, the weight of a mark on a ballot is = 1 / (number of candidates already elected that are approved on this ballot + 1)
This is still in the spirit of standard Approval in that giving approval marks to non-winning candidates doesn't make your other approval marks weaker.
why are people so deluded with this parlor trick
???
No, instead of centrally deciding on what is manufactured, independent initiatives design things, and ppl vote to give them more manufacturing capacity.
Not "voting" as in to select winners; each vote would only increase capacity.
It's like a market without money; capital can neither be accumulated nor is needed, since anyone can harness the communal automated manufacturing capacity to make a prototype of something, and if more people like it, it or an improved version can be manufactured more.
Cute Gastronok dude.
Can you give a numerical example how you would interpret some votes?
its on the title screen of the tiles version of the game, or atleast, it used to be
If there is no limit to how many variants there are and the decision is of the type how much do we produce rather than whether to produce this or that or not, fractionally counting the ballots is actually okay.
Well, the idea is that votes would increase the rate at which something can be pumped out rather than translate in a specific amount manufactured.
Approval voting usually means that the aggregate ranking is only based on number of approval marks. This is for finding a compromise that many people can live with, the second-highest ranked is the next-best at being such a compromise according to the ballot-data you have, and so on.
Since your goal from what you said doesn't seem to be to find some overall winner, it makes sense to count ballots fractionally.
Compare two scenarios with the same voters voting, and for every alternative that shows up in both scenarios, a voter who approves it in one scenario also approves it in the other, and likewise with not approving. The only difference between the scenarios is that, in the second scenario, one alternative gets cloned. Getting cloned means there is an increase in the number of alternatives in the second scenario, and the new alternatives are treated by the voters in a way that is identical to how they treat the proposal that got cloned. Any voter either approves or disapproves entirely this set of proposals.
If you want to elect a single winner and count the ballots like in normal approval voting, this is robust in that adding clones like described above has no practical effect on the result (if the winner changes, it changes only to a clone of the first-scenario winner).
But if you want to get resources allocated in proportion to support, counting in the standard approval style comes with a big spam vulnerability. That is, replacing one option by several practically identical options will lead to an increase of what these identical options obtain together. So, to reduce that effect I'd rather count that fractionally or with the Thiele rule.
Surely you have some more detailed ideas than that, OP?
Actually maybe we wouldn't have to vote. If there was some sort of general-purpose programmable manufacturing facilities which auto-supplied themselves, and it could efficiently produce single things, everyone could just submit their design blueprints within personal quotas, without having to replicate the "invisible hand" of money.
How does approval voting work with something that's almost infinitely divisible, such as time.
The problem with most voting systems is that people vote blind. They don't know how other people are going to vote. So let's say there's an issue on the table that's really important to a lot of people, and we determine manufacturing allotment according to vote. Lots of people are going to vote to allot manufacturing to this one thing, even if the ideal allotment is X and going above X is not beneficial. I'll use food as an example. Everyone would likely vote in favor of food, but then food would be overproduced because only so much is needed.
If people are going to vote blind on something, it should be along the lines of which algorithm to use to determine production rather than their vote effectively meaning "produce +1 of this."
Honestly though, if we ditch capitalism we ditch the need for planned obsolescence which means a mind-bendingly large increase in efficiency. Most common goods could simply be traded in or bought used because they could be built to last forever. Production would only need to focus on any specific thing for a short time before moving on to something new. Most production time would realistically be spent prototyping innovations, and once some meaningful change was found there would be a burst of production to make pretty much all of that thing that would ever be needed. Ideally production levels could scale dynamically to fit demand. If the workers control them, the only thing stopping them from shutting down factories and firing them back up at will is the physical limitations of the machines.
But as I said, voting would only affect the priority at which something is to be output, and not the actual amount produced, which would depend on demand, so there would be no overproduction.
What you describe is that people under-estimate how much others support something, they then vote in an exaggerated fashion with the goal of making up for that, and because their mental model of how others voted was off, the result is very different from what they actually wanted.
This indeed applies to some systems. It applies to voting with average ratings and to voting with point budgets that are allocated by simple addition. It doesn't apply to voting with median ratings. (Exception in median voting is when number of votes is even and the two middle values are different, strictly speaking the median is the average of these two then, but we can make an extra-rule for that like always taking the lower of the two values in that case.)
Another voting method that curbs exaggeration incentives is to use point budgets that are allocated to different tasks in a way that your ballot's contribution is weighted based on the square root of the points you allocate to the task. Think about it this way: Two tasks are represented by the X and Y axis of a graph, a point on that graph shows the amount of energy dedicated to both.
Suppose you as a voter can directly allocate points to each axis. If you have a good idea where the energy mix is located before adding your own ballot, it makes sense to vote in an extreme way to move the point to the location you want, instead of honestly voting for that location. But with the square-root point allocation the advantage of extreme voting is pythagorased away.
A problem with the square-root point allocation is that this is vulnerable to proposal spamming. That is, cloning a proposal results in that set obtaining more resources together than the original proposal without the clones. This strategy can be nerfed by using something like Thiele in an earlier step of approving proposals for this point-allocation phase.
What do you mean by approval voting, aside from the ballot format being binary? If the voting doesn't determine any quantities, but priority, what does priority mean? Does it mean lexicographic ordering (plan fulfillment for the production of higher-ranked things has infinitely more importance than for lower-ranked)? I would rather use something a bit more expressive than approval voting then.
The more I think about it, the less sense OP's proposal makes to me. If people can vote to differentiate importance of things that take up the same resources, and that differentiation is fixed in plan prior to setting the actual quantities of the more and less important things, these should be fundamental distinctions between product types like what counts as basic and luxury and not details (like OP's different hair irons).
Why do gooks hate chins?