- Election-software - hio.mailmanlists.us

Updates to abif.electorama.com
by Rob Lanphier 19 Aug '25

19 Aug '25

Hi folks I pushed out many updates to https://abif.electorama.com late last night, and it's becoming a lot of fun to use. Some big changes: * Tabbed interface to switch between election methods on any given election * I've added approval voting * it is MUCH easier to just browse tags to find various elections now. I haven't added many new elections, but you may not have noticed many of the elections that were already there before, since the tag navigation was so bad. I did add one notable election: the NYC Democratic Party mayoral race of 2025: https://abif.electorama.com/id/NYC2025-Mayor-Primary-Dem Interestingly, by my count, it would seem that Andrew Cuomo lost in pairwise comparisons to THREE candidates (not just Zohran Mamdani). The November general election is going to be interesting.... The latest version of abif.electorama.com supports approval voting. An interesting bit: it converts ranked ballots to approval ballots. The algorithm uses a funky way of approximating reasonable-ish voter behavior extrapolated from their rankings, which I call the "favorite_viable_half" algorithm: find the candidate with the most first preferences and then determine the minimum number of figurative seats that would need to be open in order for the candidate to exceed the Hare quota with the given first-prefs (an estimate of viable candidates). Assume then that all candidates that the voter ranks above their respective quota of viable candidates are approved, and any ranked below their least favorite in their quota was not approved. Their quota: half of the viable candidates, rounded up. Of course, if they don't rank a candidate, then that candidate isn't approved. Since there isn't a jurisdiction (yet, that I know of) that allows voters to draw an approval line, the algorithm above is my best guess at where they would draw the line between approved and not-approved candidates. In the examples that I've looked at, it intuitively seems rational. For example, let's look at the 2024 SF mayoral race: https://abif.electorama.com/id/sf2024-mayor/approval#approval By the favorite_viable_half algorithm, Daniel Lurie (the winner of the RCV/IRV election) gets 50.1% approval. London Breed (the runner-up incumbent) gets 42.2% approval. That roughly matches what their pre-election approval ratings were, based on my cursory research. Breed had 55% disapproval and 40% approval before the election. It wouldn't be surprising if she netted the 2% difference in the voting booth because voters didn't approve of any of the dozen other candidates running against her. That 42.2% approval is an informed approximation; there were more voters that ranked Breed, but only 42.2% ranked her as one of their two-two favorite viable candidates (of the approximately 4 viable candidates in the election). The media here in SF had a tough time deciding if 4 or 5 of the 13 candidates were viable. The first publicized mayoral debate had 5 candidates, and the remaining debates had 4 candidates. Speaking as a voter who was asked to rank 13 candidates, I say to all y'all who are jealous: be careful what you wish for. The update to abif.electorama.com is all part of the upcoming 0.33 release of abiftool and awt. abiftool is the command-line tool for viewing/analyzing elections (and converting between election-result formats), and then awt is the web interface for abiftool that powers abif.electorama.com. It's all free/open source software, so feel free to download and try it out yourself. Patches/contributors welcome! Also, if you find any bugs (there are many), please file a report in the issue tracker: https://github.com/electorama/awt/issues Don't be bashful about filing feature requests as well. I've got a lot of bugfixing to do before the 0.33.0 release, but after that, I'll be looking for compelling features to implement. Thanks Rob

1 0

abiftool/awt 0.32.0: now with recent SF elections and caching!
by Rob Lanphier 22 Jul '25

22 Jul '25

Hi folks! I'm proud to announce abiftool v0.32.0 and awt v0.32.0, which are: - abiftool <https://electorama.com/abiftool> — a command-line tool for working with election files, converting from other formats into ABIF <https://electowiki.org/wiki/ABIF> (and vice versa), and tallying elections using many systems (Copeland/Condorcet - awt <https://electorama.com/awt> — the web front-end for abiftool. A live instance of awt can be found at abif.electorama.com. New! SF 2024 elections, FPTP support, caching, improved testing, and bug fixes Both awt and abiftool use the abiflib library found in the abiftool repository. Some important advancements in abiflib since the 0.2.0 <https://www.reddit.com/r/EndFPTP/comments/1ihxmw8/over_400_elections_now_at…> release of abiftool back in February 2025: * Handling of San Francisco's latest JSON-based CVRs (see the "sf2024" tag <https://abif.electorama.com/tag/sf2024>). * Initial implementation of First-Past-The-Post (FPTP) tabulation. * Many micro-features hidden in command-line options. * Substantial refactoring of abiflib. * General code cleanup and major improvements throughout. * Much more robust pytests and more of them. Additonally, here's a big improvement in awt * Caching! All prior versions computed the results of the static abif under abif.electorama.com/id/* <https://abif.electorama.com/id> dynamically, which was silly. As a result, the website was often dog slow. Now it is only cat slow. If you tried this out in February and had a poor experience because the many bugs and server timeouts in 0.2.0, come back and try it again. This version is *much* more reliable (he says, automatically jinxing awt to have a serious bug that is obvious to everyone but himself). Bugs, feature requests, other ways to get involved Speaking of bug discovery, bug reporting, and feature requests, there's no end of work to do on abiftool and awt, and your help would be greatly appreciated (even if you aren't a software developer). Please visit abif.electorama.com, and if you find any bugs or think of any features you like to see, file a ticket at github.com/electorama/awt/issues. If you find a problem, don't assume that anyone else knows about it -- getting something in the issue tracker is the best way to ensure someone deals with it. If you're a software developer and want to help, reach out to me; I'm easy to get in touch with (see robla.net and electorama.com) You can help shape the future direction of abiftool and awt: - There's a lot of user interface work that needs to happen with awt. I'm more of a backend developer and text-file twiddler (hence the appeal of ABIF to me), but I recognize many folks have less of tolerance for "ugly" user interfaces than I do. If you fancy yourself a front-end developer looking for a project to make an impact with, talk to me. - Many more formats can be added to and there are many formats that seem worth adding (both for conversion *to* and *from* ABIF). - Keep tweaking at it. There's lot's of smaller features that would be nifty to have (like better display of ties in IRV <https://github.com/electorama/awt/issues/7>). - Merging with other websites/codebases. This is by no means the first public election software (see also stablevoting.org, preflib.org, Votelib <https://github.com/simberaj/votelib>, rob-legrand.github.io <https://rob-legrand.github.io/ranked-ballot-voting-calculator/> and ranked.vote for just a few). There's already some integration/interoperability with some tools [e.g. with the ABIF reader in the pref_voting library <https://github.com/voting-tools/pref_voting/blob/main/pref_voting/io/reader…> or the experimental ABIF branch in Votelib <https://github.com/simberaj/votelib/tree/abif>), but the combinatorial aspects of interoperability between tools mean there's practically infinite work to do. There's a lot of different ways software-developer activitsts can spend their time. What's the *best* way? For that matter, what's a *good enough* way? Join the conversation! To discuss awt and abiftool (and election software more , a good mailing list to join is the election-software mailing list: - https://electorama.com/es ...or if you prefer chat, you can join the Electorama Discord server. - https://electorama.com/discord ...or just get the code and start hacking! Get the code! To get the source code, visit the GitHub projects for awt and abif below: - https://github.com/electorama/awt - https://github.com/electorama/abiftool ...or just visit abif.electorama.com. Rob

1 0

abiftool and awt version 0.31.0: SF 2024 elections, FPTP support, improved testing, and bug fixes
by Rob Lanphier 11 Jul '25

11 Jul '25

Hi folks, Y'all are the first to hear this. Or rather, you're the first to get this email. I've finally released abiftool v0.31.0 and awt v0.31.0. As a refresher: - abiftool <https://electorama.com/abiftool>-- this is a command-line tool for working with election files, converting from other formats into ABIF <https://electowiki.org/wiki/ABIF>, and tallying elections. - awt <https://electorama.com/awt>-- this is the web application version ofabiftool. A live instance ofawt can be found atabif.electorama.com Both awt and abiftool use the abiflib library found in the abiftool repository. Here are some important advancements in abiflib since the 0.2.0 release of abiftool back in February 2025: - Handling of San Francisco's latest JSON-based CVRs. - Initial implementation of First-Past-The-Post (FPTP) tabulation. - Substantial refactoring ofabiflib. - Much more robust pytests and more of them. - Many micro-features hidden in command-line options. - General code cleanup and major improvements throughout. There's a lot of features I would love to add toabiftoolandawt.I'd love your help (especially if you're a Python dev). There's a lot of user interface work that needs to happen withawt, and there are many formats I'd love to add. Even if you aren't a software developer, I'd love for you to visit abif.electorama.com, and if you find any bugs here: https://github.com/electorama/awt/issues To get the source code, visit the GitHub projects forÂ awtÂ andÂ abifÂ below: - https://github.com/electorama/awt - https://github.com/electorama/abiftool Rob

1 0

Release v0.2.0 of abiftool and awt
by Rob Lanphier 04 Feb '25

04 Feb '25

Hi folks, Today, I'm releasing v0.2.0 of "ABIF web tool" (awt) and abiftool. To refresh your memory, ABIF is "Aggregated Ballot Information Format", and is a format for expressing the aggregated results of an election.[1] The release notes for the respective releases can be found here: https://github.com/electorama/abiftool/releases/tag/v0.2.0 https://github.com/electorama/awt/releases/tag/v0.2.0 In short, the marquee feature of this release is the incorporation of compatibility with Brian Olson's "RCV Election Data" from https://bolson.org/voting/votedata/ , along with the display of this data here: https://abif.electorama.com/tag/from-bolson Brian's data significantly augments the modest collection of election data that I was providing before today. While I had slowly been adding other formats to abiftool, this is the first bulk conversion I've performed from another collection. I designed the user interface of the ABIF web tool such that it wouldn't COMPLETELY fall apart if I increased the number of elections, but now that there are over 400 elections shown, we're now testing the limits of the current interface. Please let me know your feedback, either here on this mailing list, or send a personal email to me at roblan(a)gmail.com . Rob [1] https://electowiki.org/wiki/ABIF

1 0

Facebook alternatives?
by Rob Lanphier 19 Jan '25

19 Jan '25

Hi folks, This is a bit off topic for the election-software list, but only a bit. At some point, someone needs to create a voting mechanism that implements many ranked and rated systems, but uses the API of a mainstream social network to filter out duplicate votes by allowing only one vote per account (perhaps limiting things further for social networks that don't mind it if someone sets up dozens or even hundreds of accounts on their system. I think the multi-account per user could be solved by having a group that one's has to be part of, or something along those lines. Still, which API/protocol should we use?Should we focus our efforts on one of the big mainstream corporate networks like Twitter's or Facebook's? Should we focus our efforts on the nascent AT:Protocol <https://en.wikipedia.org/wiki/AT_Protocol>? How about the relatively mature ActivityPub <https://en.wikipedia.org/wiki/ActivityPub>? Since Elon Musk and Mark Zuckerburg are going to be very special guests at Trump's second inauguration, there are many people abandoning Facebook. So, I'd like to share with y'all a wiki page that I'm building out (and would love some help with; just sign up for an account!). The page: https://myndmess.org/wiki/Facebook#Alternatives What are y'all using to communicate with your respective families? What do you wish they'd migrate to? Rob

1 0

Turning distances into coordinates
by Kristofer Munsterhjelm 09 Oct '24

09 Oct '24

(Oops, I had been meaning to write this a long time ago but didn't get around to do it.) Some time ago I suggested that one may use Vivaldi coordinates (https://en.wikipedia.org/wiki/Vivaldi_coordinates) with Kendall tau distance to sort voters into different "camps" based only on the voters' rankings. But the principal value of Vivaldi is that it's distributed, and there's no reason to use it if you're not going to do something distributed. There's a whole class of algorithms that try to solve this problem, which is called nonlinear dimensionality reduction. One example would be multidimensional scaling, https://en.wikipedia.org/wiki/Multidimensional_scaling; the Wikipedia page even has an example graph with voting patterns in the House of Representatives, showing that it can be used for voting analysis. Other examples include isomap and t-SNE. Either of these would probably be better than using Vivaldi :-) The Wikipedia page on nonlinear dimensionality reduction lists a number of other approaches: https://en.wikipedia.org/wiki/Nonlinear_dimensionality_reduction#Important_… and https://en.wikipedia.org/wiki/Nonlinear_dimensionality_reduction#Methods_ba…. -km

1 0

Algorithm for electoral regions from rankings?
by Rob Lanphier 31 Aug '24

31 Aug '24

Hi folks, How might one algorithmically go about generating a 2D placement of affinity groups when one only has ranking information? I've been considering writing something that does this for quite some time, ever since I manually generated this chart of the 2009 mayoral election in Burlington: https://electowiki.org/wiki/File:Burlington2009election-infosheet.png More recently, a discussion broke out on the Center for Election Science's Discord server on this topic. In particular, we started discussing what it would take to generate Yee diagrams from election results. There weren't answers there; there were just more questions. I have some ideas for how to do it (by a vaguely Monte-Carlo-ish evaluation, iterating over placement of candidates and voter affinity groups), but I haven't yet figured out a reasonably simple algorithm. I'm going to set this task aside for now, but if others have ideas, I'd love to see them. Rob

2 1

[Election-software] abiftool now supports SF elections (sorta)
by Rob Lanphier 07 Aug '24

07 Aug '24

Hi folks, I'm planning to start promoting my work on "abiftool" and "awt" some more. In case y'all aren't familiar with either of these, here's the URLs: awt (ABIF web tool): https://awt.electorama.com abiftool: https://github.com/electorama/abiftool Both of these use the same "abiflib" library to parse a variety of formats, transform those formats to ABIF (or "jabmod", which is the internal format that can be used as a bit of an API to abiflib), and then it spits out election results in a variety of ways. abiftool injustice many formats, and I'm happy to add more. It doesn't spit out many formats yet, but I'm happy to add other formats if the developers of those formats work with me to make it happen. If y'all aren't familiar with ABIF, please take a look at the ABIF page on electowiki: https://electowiki.org/wiki/ABIF I would love to get some collaborators on awt and abiftool for pretty much all of the features that are missing from either user interface for either of the tools that I started for ABIF handling. If you want to create an independent implementation (especially in another language besides Python), I'm really interested in that. I'm generally interested in collaborating with folks on software for processing historic elections and theoretical election examples that folks share online. Rob

1 0

[Election-software] Open-ish subscription as of now-ish; "no confidence" votes
by Rob Lanphier 05 Aug '24

05 Aug '24

Hi folks, I've just let a few folks join earlier this afternoon. I had accidentally had the settings too restrictive, but hopefully, it's a bit easier to subscribe now, and my personal button mashing isn't a hurdle to subscription. I'm really happy with how the membership of this mailing list is shaping up, since I know most of you (at least, I know your online presence), and I'm reasonably sure that most of y'all have both programming chops and voting science chops, which is the exact set of folks that I was hoping to see. Please: invite your friends! Also, I'm hoping y'all can all give some advice on moderation policy here. Over at the election-methods list, there were a couple of folks that became very unpopular over the past few months. I won't go into details, but at least one of them left in a huff because I sent them a private email that they found disagreeable (or rather, they want to unsubscribe, but also seem to want white-glove service on a self-service mailing list). Then the other problematic member complained on-list about my private email. It's a little emotionally taxing on me to have to send behavioral emails to members (letting them know just how unpopular they are), and thus, I like mathematical rules for this type of thing. I like the idea of having "no confidence" votes on the membership of the mailing list. I sent (what I thought were) pretty clear instructions for voting to the election-methods mailing list on our first trial "no confidence" vote, which resulted in a chaotic mess of feedback. It was no problem; I got the gist, and learned about some of the more subtle interactions on the list (and I received apologies from people who owed me no apology). I've included my initial email to the EM-list regarding my idea for "no confidence" votes. Despite my complaining about the lack of structure from the feedback, I found the results helpful. To preemptively answer a question you might have: no, I'm not publishing the ballot data. :-) 'Tis a mess on SO MANY levels. My question: assuming this list doesn't lurk in obscurity forever, on the undesirable but probably inevitable occasion that we have a problematic member of the list (or two, or three, or five people who don't play nice with one another), what should we do about it? Rob ---------- Forwarded message --------- From: Rob Lanphier <roblan(a)gmail.com> Date: Sat, Jul 20, 2024 at 11:57 PM Subject: Should we allow "voting off" this mailing list? To: <election-methods(a)lists.electorama.com> Hi folks, I've received a couple of complaints about the tenor of this mailing list. My cursory look through the archives suggests that the problems are with a couple of people. I'm trying to sort it out with the people about whom the complaints are about, but something occurred to me. There's been lots of "votes" on this mailing list, which I've never been a big fan of, because the group of people who are active on this list are self-selected. However, there's one thing y'all are the exact right people to vote on, which is "who should be on this mailing list?". It may be good to have times where y'all get to vote a person or two off of the mailing list "Survivor"-style, though I want to make sure that this list is a safe place for smart people with unconventional communication styles to express unpopular opinions every now and then. I'm announcing today is the first day of voting in this election. Please send me a private ballot (to me only, not to the list) in the following example format: 1. A - This person ("A") is my favorite person posting on the list. (Please only vote for one or two "favorites", though I won't mind if you gush about three or four people in this first round of voting) 2. CCCCC - These people ("C", "C", "C", "C", and "C") are usually pretty reasonable, and though I disagree with them sometimes (or perhaps almost every message they send), their messages are generally thoughtful and well-expressed. It would make me sad if they decided to unsubscribe to this mailing list, and are driven off the list by the more obnoxious/prolific posters. I'm sometimes thinking of these people specifically when I write emails to the list, because I want to reach these people. 3. FFF - These people ("F", "F", and "F") are almost always sending unacceptable messages to this list, and they do it A LOT. It's not just annoying, it's toxic. I've been tempted to unsubscribe mainly because of THEIR messages (and some of you may say "I have done it a time or two in the past"). I really want concise email ballots, though I'm allowing for a little prose to explain why you put someone at the "A" tier or why you're relegating anyone to the "FFF" tier. Try to have at least one person in the "A" tier for every person you put in the "FFF" tier. The "CCCCC" tier should really just be a list. This is an experiment. I'm going to be traveling soon, so I'm hoping to have the first set of ballots by end-of-day Monday (and note: my end-of-day is probably later than your end-of-day, especially if you're not in the United States). I don't mind if you email me personally and say "I'd like a little more time to think about this". I probably won't try figuring out how to aggregate everyone's feedback about other list members until August, but those of you who vote early will have my gratitude. If you want to propose other ballot formats than what I suggest above, that's okay, but I think I'm going to insist on using my three-tier "A", "CCCCC", and "FFF" tiers, and using those names. I'm going to be hand aggregating ballots, so the more novel you are with your ballots, the less likely I'm going to be to consider your ballot. I'm not likely to act soon on the first set of ballots unless there are a lot of ballots that put the same person(s) in the "FFF" category. I'm not planning on holding these votes at any particular interval, but maybe we should devise a set of rules for calling a "vote of no confidence" in the current membership of the mailing list. It may be that only people who post 5 or more times per month are eligible to be voted off the island, so to speak. I haven't decided, but your comments are welcome! Rob p.s. I've started work setting up an alternative mailing list to discuss voting software. I'm more likely to read all of the messages to this alternative mailing list, so I'm less likely to be patient with people that ruin the reading experience for the lurkers on the list.

2 2

[Election-software] Welcome to the election-software mailing list!
by Rob Lanphier 23 Jul '24

23 Jul '24

Hi folks, If you are reading this email message, you're probably reading the archive for the list. As of this writing, I'm the only member. I'm planning on announcing this mailing list in other forums eventually. This mailing list is going to be specifically about software related to elections. This can be everything from the user interface presented to voters through the counting of the ballots all of the way through to the random mechanism used to break ties. Debating the merits of various election methods should be left to the election-methods mailing list ( https://electorama.com/em ) or various other forums where the merits of IRV vs STAR vs Approval vs Condorcet methods are discussed. Code of conduct (for now) is going to be the same as for the election-methods mailing list. I'll describe all of this later. Anyway, I look forward to discussing election software with you! Rob

1 0