Hi Rob, On Thu, Aug 15, 2024 at 10:01:12PM -0700, Rob Lanphier wrote: > Hi folks, > My main reason for writing this mailing list is to ask for help on a > couple of things: > 1. I've written and published the results of many elections as > tabulated by my relatively new software for displaying > election results, and I want some feedback on it. > 2. There's a text format (called "Aggregated Ballot Information > Format" or "ABIF") that I've helped create, and I'm hoping folks > familiar with the devotee codebase would be willing to look at the > format, and maybe even add ABIF support to devotee.
It seems similar to the format that is used by the tool which the Debian Technical Committee uses. I'm personally not involved with Devotee, but the format seems straightforward and I can't see any reason from the sidelines why at least an output file in that format couldn't be done... [...] > A particularly interesting Debian Project Leader result from a few > years ago is this one: > [3]https://abif.electorama.com/id/DPL2003 > Note that awt tabulates the DPL2003 results not only using Condorcet, > but it also tabulates using "instant-runoff voting" (also known as > "IRV" or "RCV" as it's called in the United States). What's > interesting about the DPL2003 election is that Condorcet methods and > IRV result in different results. Pretty much all of the other > elections that the Debian community has had have resulted in the same > Condorcet vs IRV winner. Election nerds (like me) find DPL2003 > fascinating. That is fascinating. When I looked at the Condorcet results, my initial assumption (before I scrolled down) was that Bdale Garbee (who defeated everyone except the eventual winner, Martin Michlmayr) would be the IRV winner. It turns out it's actually Branden Robinson who would win using IRV, and he only defeated Moshe Zadka in Condorcet. In IRV, NOTA is eliminated in the first round, Moshe Zadka in the second, and then with the removal of those ballots, somehow Branden defeats both Martin Michlmayr and Bdale Garbee... > Circling back to my main requests: > 1. Is awt actually giving the correct result? Do y'all see any bugs? I haven't looked at the code, so I don't know. -- w@uter.{be,co.za} wouter@{grep.be,fosdem.org,debian.org} I will have a Tin-Actinium-Potassium mixture, thanks.
