Michael Norrish wrote: >Life would be a lot simpler if we dispensed with ratios entirely, and >framed AIs in terms of differences.
Interesting. I have a concept for quorum which might interact with that idea. There is a problem with the usual way quorum works, that there are situations where voting AGAINST a proposal actually makes it more likely to pass. (For example, if there have been four votes FOR and there is a quorum of five.) I'd like to avoid such perverse situations. So instead of requiring that a certain proportion of the electorate vote on a proposal for it to pass, there should be a requirement that a certain (smaller) proportion of the electorate vote in favour. This is in addition to the requirement that a certain proportion of those expressing an opinion vote in favour. For example, an ordinary proposal might require that at least 10% of the electorate and over 50% of those expressing an opinion vote in favour. A very significant proposal might then require 20% of the electorate and 75% of those expressing an opinion. Working with differences instead of ratios, my quorum replacement would require a proposal to get at least N more votes in favour than against, and at least M votes in favour absolutely. Both of these numbers could be selected as (modified) proportions of the size of the electorate. Another approach to the whole thing would be to express the voting result as the ratio of votes in favour to the total number of votes cast. The maximum voting index would then be 1 (i.e., 100%). We can translate VIs and Powers: 1 -> 50%, 2 -> 67%, 3 -> 75%, 4 -> 80%, Unanimity -> 100%. This still needs a special case for where there are no votes cast, but we don't need infinite numbers there. -zefram
