Fred Gohlke wrote: > I fear you're several steps ahead of me. I'm still struggling to reason my > way through the human relationships inherent in a selection process like > Practical Democracy. So far, although I've visited the links you provided, > I haven't been able to internalize the material.
No problem, I can demonstrate. See below. > re: "I missed your specification of vote shifting. In which post > was it?" > > Members of triads are under no compulsion to select either of their peers. > They may choose as they see fit. However, they may not re-choose. (Vote shifting is not part of your method, so Votorola would have to be restrained in this regard.) > The more I read your post, the more I think we're talking at cross > purposes. I keep thinking we're critiquing the concept of Practical > Democracy, but you seem to be talking about how that concept can be > integrated with your software. Although I've visited your site, I > haven't been able, so far, to get a good handle on what I've > read. Could we start by discussing the purpose of the software, and > then, later, how the software helps achieve that purpose? The general purpose is decision making in the public sphere. But Votorola is suitable, in particular, to support Practical Democracy (PD). This is what I am suggesting. I can demonstrate. (You may need to switch to a fixed width font, in order to see the diagrams below.) 1 (A) (B) (C) (D) (E) (F) (G) (H) (I) (J) (K) (L) (M) (N) (O) (P) (Q) (R) | / | / | / | / | / | / | / | / | / | / | / | / | / | / | / | / | / | / |/ |/ |/ |/ |/ |/ |/ |/ |/ (S) (T) (Y) (U) (V) (Z) (W) (X) (AA) 2 (S) (T) (U) (V) (W) (X) | / | / | / | --- | --- | --- |/ |/ |/ (Y) (Z) (AA) 3 (Y) (Z) | / | ------------------- |/ (AA) FIG 1. The structure of a small PD decision. 27 electors are formed into 9 triads at the first level. Thence 9 candidates are chosen for the second level. Thence 3 are chosen for the third level (Y, Z, AA). Thence the winner (AA). I ran the above election on one of Votorola's test servers. I scripted 27 simulated voters ^[1]. Each voter registered herself and then cast a vote (or not) according to the pattern of figure 1. The results are shown here: http://t.zelea.com:8080/tester/w/Count/?s=test-ward-19-councillor (This is just a test server for Toronto. You can click around, and even cast votes. But bear in mind, it's only an alpha prototype.) As you can see from the above page, the same decision was reached as in figure 1. AA has won the election by unanimous assent. If you follow her supporting votes backward, they trace out the structure of a tree, as shown in figure 2. (The two figures are essentially the same, except the second is drawn contiguously.) (A) (B) (C) (D) (E) (F) (G) (H) (I) (J) (K) (L) (M) (N) (O) (P) (Q) (R) | / | / | / | / | / | / | / | / | / | / | / | / | / | / | / | / | / | / |/ |/ |/ |/ |/ |/ |/ |/ |/ (S) (T) / (U) (V) / (W) (X) / | / / | / / | / / | ----------- | ----------- | ----------- |/ |/ |/ (Y) (Z) / | / / | ------------------------------------------- |/ (AA) FIG 2. The decision as expressed in a delegate cascade. Here is a narrower view of the results, from the perspective of a single voter. This shows the trace of C's vote, along the path (C, T, Y, AA). http://t.zelea.com:8080/tester/w/RoleV/?v=voter-c%40zelea.com&s=test-ward-19-councillor This page also enables C to shift her vote. So this is where Votorola would need to be restrained, for purposes of PD. Otherwise C might do something wild and crazy, like this: (C) / / / (A) (B) (D) (E) (F) | / / | / | / / | / |/ / |/ (S) (T) / | / / | ----------- |/ (Y) | | ----------------- |/ (AA) FIG 3. C shifts her vote from T to B. Votorola is perfectly happy with that, and so am I. But you might not be. Such shifts might occur without regard to the current stage of the PD decision process, messing it up; or they might transform it into an indefinite process, that just keeps running, on and on. Worse still, the voters might ignore the triad boundaries, and thus create groups of non-standard size. C's vote shift in figure 3, for instance, has created two pairs (C, B) and (D, T) in place of the standard monad (B) and triad (C, D, T). So although Votorola can implement PD (as I demonstrate here), it doesn't have the structure to enforce it. The missing structure would have to be added, either internally to the code, or externally to the environment. (Code modification seems the easier choice, as it would involve little more than discounting any vote that breaks a rule. So C's vote would become the equivalent of an abstension. She would be informed of this on casting it, and might then correct it.) Anyhow, there it is. Votorola also has electoral districts and voter authentication, and it comes with a growing collection of user interfaces. Soon it will have an interface to drafting media for the purpose of legislative voting. And it's free and open. (If you come across any other software that can do this stuff, please let me know. I keep track of these things.) [1] The voter simulation script is: http://zelea.com/system/host/t/home/tester/system/bin/vox-voters-abc -- Michael Allan Toronto, 647-436-4521 http://zelea.com/ ---- Election-Methods mailing list - see http://electorama.com/em for list info