Re: [freenet-dev] Democratic process
Arne Babenhauserheide writes: > Ian Clarke writes: > >> On Wed, Sep 28, 2016 2:56 PM, Arne Babenhauserheide arne_...@web.de >> wrote:This, written in a thread where I show a cleaner method of evaluation >> >> along with an implementation of a way to see which reasoning can >> >> actually be taken from the poll as it was conducted. >> >> >> What is your specific proposal, because apparently I missed it. All I saw >> were >> you mentioning a variety of alternatives that are not suitable because they >> do >> not retain the ordinality of people's estimates. > > My specific proposal is to use several different methods of ordering the > tasks by the votes given (value divided by cost estimate) and taking a > subset which is highly ranked in all the different methods of evaluation > (except for the ill-defined one which divides the mean by the spread of > the votes). This is the set for which it is possible to give the robust > answer that it is preferred by the people casting the votes. > > When I do that with the top 10, I get 6 which are ranked in all the > methods. Even just using mean and median should at least make this a bit > more resilient. Clearer, with the preliminary evaluation, out of the top 10 (by mean value), 6 entries are robust in that position and 4 of the 6 are non-controversial. Taken the other way around: * For 4 entries there is *non-controversial* agreement that they deliver the largest value per cost. Their position in the top 10 is robust. * For 2 further entries, there is a *majority* agreement that they deliver the largest value per cost, but the values differ widely. Their position is robust, but there is no consensus. > I do not find the word ordinality in my dictionary. What does it mean exactly? Best wishes, Arne -- Unpolitisch sein heißt politisch sein ohne es zu merken signature.asc Description: PGP signature ___ Devl mailing list Devl@freenetproject.org https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
Re: [freenet-dev] Democratic process
Ian Clarke writes: > On Wed, Sep 28, 2016 2:56 PM, Arne Babenhauserheide arne_...@web.de > wrote:This, written in a thread where I show a cleaner method of evaluation > > along with an implementation of a way to see which reasoning can > > actually be taken from the poll as it was conducted. > > > What is your specific proposal, because apparently I missed it. All I saw were > you mentioning a variety of alternatives that are not suitable because they do > not retain the ordinality of people's estimates. My specific proposal is to use several different methods of ordering the tasks by the votes given (value divided by cost estimate) and taking a subset which is highly ranked in all the different methods of evaluation (except for the ill-defined one which divides the mean by the spread of the votes). This is the set for which it is possible to give the robust answer that it is preferred by the people casting the votes. When I do that with the top 10, I get 6 which are ranked in all the methods. Even just using mean and median should at least make this a bit more resilient. I do not find the word ordinality in my dictionary. What does it mean exactly? Best wishes, Arne -- Unpolitisch sein heißt politisch sein ohne es zu merken signature.asc Description: PGP signature ___ Devl mailing list Devl@freenetproject.org https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
Re: [freenet-dev] Democratic process
On Wed, Sep 28, 2016 2:56 PM, Arne Babenhauserheide arne_...@web.de wrote:This, written in a thread where I show a cleaner method of evaluation along with an implementation of a way to see which reasoning can actually be taken from the poll as it was conducted. What is your specific proposal, because apparently I missed it. All I saw were you mentioning a variety of alternatives that are not suitable because they do not retain the ordinality of people's estimates. Ian. Ian Clarke Founder, The Freenet Project Email: i...@freenetproject.org ___ Devl mailing list Devl@freenetproject.org https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
Re: [freenet-dev] Democratic process
Ian Clarke writes: > On Wed, Sep 28, 2016 10:29 AM, Arne Babenhauserheide arne_...@web.de > wrote: > > I’m trying to make sure that the process is ironclad. > > … > > scale conflicts about it and there’s a split in the userbase. Therefore > > … > > But the decision we’re talking about is controversial. There were large > > > > You used the process without checking what it can actually say? Then you > > were sloppy and people let you get away with it. For uncontroversial > > decisions, no problem with that. > > > I was not sloppy, the problem of strategic voting that you're so concerned > about > was discussed in my original proposal, as a reminder: > A few important things here to ensure we get the true benefit of “wisdom of > the > crowd”: * Don’t be strategic (eg. don’t over-allocate to one area because you >assume others might under-allocate to it) > * Don’t collude > * Don’t consider difficulty in your value allocation > * If you aren’t sure, leave stuff at an even allocation, don’t assign 0 >value Setting rules you cannot enforce is sloppy. What you did might work very well in an environment where you know all people where they all vote more or less honestly due to peer pressure. In an environment with anonymous contributors with people who have very different requirements (some of which we do not know - which is part of the point of asking them), that’s different. Here a method which is very prone to strategic voting is problematic, and the least to do about it is to quantify the size of the problem. Which I did. I did not merely ask you to do it but went for it and did it myself. > All you seem capable of doing is criticizing from the sidelines This, written in a thread where I show a cleaner method of evaluation along with an implementation of a way to see which reasoning can actually be taken from the poll as it was conducted. > Whether it's refusing to use Google Docs because I was mean to you > off-list on an unrelated matter You attacked me for saying that Google Docs is unsuited. It is unsuited. A significant fraction of our votes came via FMS (the Freenet Message System, the main forum system in Freenet) - from anonymous people to whom Google Docs was a no-go. When you attacked me I decided that I would stop trying to discuss that with you and instead just not take part in a process which uses methods which do not work at all for a significant fraction of the the people we ask. > , or now whether you're complaining about a problem that was discussed in > my original proposal as if it's some kind of oversight. Assuming that an arbitrary group of anonymous people will follow the rules without any way to check that (not even a sanity check because we do not know their requirements) is a pretty big oversight. However, from your reactions I have the impression that you do not see it as constructive if someone points out a flaw in something you do, not even if it comes with a way to fix that flaw. Where I come from, pointing out flaws and ways to fix them is called constructive criticism. It is one of the best ways to improve something. Therefore, and to wrap this up, I respectfully disagree with your argumentation. Best wishes, Arne -- Unpolitisch sein heißt politisch sein ohne es zu merken signature.asc Description: PGP signature ___ Devl mailing list Devl@freenetproject.org https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
Re: [freenet-dev] Democratic process
On Wed, Sep 28, 2016 10:29 AM, Arne Babenhauserheide arne_...@web.de wrote:I’m trying to make sure that the process is ironclad. scale conflicts about it and there’s a split in the userbase. Therefore But the decision we’re talking about is controversial. There were large You used the process without checking what it can actually say? Then you were sloppy and people let you get away with it. For uncontroversial decisions, no problem with that. I was not sloppy, the problem of strategic voting that you're so concerned about was discussed in my original proposal, as a reminder: A few important things here to ensure we get the true benefit of “wisdom of the crowd”: * Don’t be strategic (eg. don’t over-allocate to one area because you assume others might under-allocate to it) * Don’t collude * Don’t consider difficulty in your value allocation * If you aren’t sure, leave stuff at an even allocation, don’t assign 0 value All you seem capable of doing is criticizing from the sidelines, demoralizing people like Xor that are actually trying to move things forward. Whether it's refusing to use Google Docs because I was mean to you off-list on an unrelated matter, or now whether you're complaining about a problem that was discussed in my original proposal as if it's some kind of oversight. Either find a way to be constructive, or stay out of it. Ian. Ian Clarke Founder, The Freenet Project Email: i...@freenetproject.org ___ Devl mailing list Devl@freenetproject.org https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
Re: [freenet-dev] Democratic process
Ian Clarke writes: > On Wed, Sep 28, 2016 2:32 AM, Arne Babenhauserheide arne_...@web.de > > This is very much necessary. Without checking what can actually be found ^ very relevant line you removed > > and what cannot, using the poll to support decisions is a mere fraud. > > Now that you've accused me of fraud Notice that I did not. I told you that if you were to use the results without checking what they can actually say, that would be fraud. Like using an uncalibrated meter to sell wares. You used the process without checking what it can actually say? Then you were sloppy and people let you get away with it. For uncontroversial decisions, no problem with that. But the decision we’re talking about is controversial. There were large scale conflicts about it and there’s a split in the userbase. Therefore I’m trying to make sure that the process is ironclad. Best wishes, Arne -- Unpolitisch sein heißt politisch sein ohne es zu merken signature.asc Description: PGP signature ___ Devl mailing list Devl@freenetproject.org https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
Re: [freenet-dev] Democratic process
On Wed, Sep 28, 2016 2:32 AM, Arne Babenhauserheide arne_...@web.de wrote: Ian Clarke writes: > Since you admit that the alternatives you are proposing aren't better than using > mean, which is the obvious choice, The mean is prone to strategic voting, except for single yes/no decisions. Just assume I had given 700 points to a single task. No one voting honestly could have pushed this point out of the first position. We should ignore anyone that is voting strategically, since they clearly violate the instructions. However when I've used this process in the past such "extreme" allocations tend to be averaged out quickly. These are supposed to be ordinal estimates, anything other than using mean to combine them will make the result no-longer ordinal which will invalidate their use as a "value" estimate. If you had asked instead of questioning my motives I would have explained that. Do you want Freenet to be known as the project which fails at even basic and what cannot, using the poll to support decisions is a mere fraud. Now that you've accused me of fraud you've just confirmed to everyone reading that your motives here are malicious. You've been doing everything you can to poison this process almost from the beginning. Fortunately we're all free to ignore you and that is what I recommend everyone does. I suggest that Xor proceed with the plan that was laid out months ago. We'll deal with issues with malicious voting if they arise. Ian. Ian Clarke Founder, The Freenet Project Email: i...@freenetproject.org ___ Devl mailing list Devl@freenetproject.org https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
Re: [freenet-dev] Democratic process
Ian Clarke writes: > Since you admit that the alternatives you are proposing aren't better than > using > mean, which is the obvious choice, The mean is prone to strategic voting, except for single yes/no decisions. Just assume I had given 700 points to a single task. No one voting honestly could have pushed this point out of the first position. Imagine a single Frost user giving 1000 points to "implement a Frost-like system" because he/she deems all the rest irrelevant. That would not actually need to be dishonest (so we could not reject the vote), but it would decide the outcome if all other people at most doubled or removed values. Condorcet voting gets rid of that option, but has the disadvantage that the values compared are not exactly estimated value/cost. The alternatives are not better, but DIFFERENT: They make different tradeoffs. > I really don't understand what your purpose > is here except to unnecessarily complicate the process and/or provoke > pointless > debate. I described my purpose clearly: Showing what can be found from the results and *what cannot*. You officially and publicly promised a democratic process to everyone visiting the webside, and this process currently is not. However it can be salvaged to some degree so that not all the effort which went into it is wasted. This is very much necessary. Without checking what can actually be found and what cannot, using the poll to support decisions is a mere fraud. Do you want Freenet to be known as the project which fails at even basic democratic process? And that’s the deeper reason why I got involved: Anyone who does simplistic pseudo-democracy and then uses it as force amplifier for pushing decisions (regardless of whether they are good or bad) tramples on actual democratic decision-making. I believe in democracy and the farce we did wasn’t democracy. By showing how uncertain the results are (see the differences in ranking between the different methods) I can at least turn it into somewhat more of a democratic process. It’s not the idea behind the process which is problematic (what was promised on the website sounded good), but the implementation - from unclear descriptions of the process over not defining the scope beforehand (what does it actually decide? Spending the 25k, a part of the 25k or creating a fourth iteration of the Freenet roadmap?), not defining the groups eligible for voting (and how their votes contribute to the decision) and not defining how exactly the poll would be evaluated. To make the process robust, a program like the one I wrote would have had to be available *before* the poll (or at least the chosen way to evaluate). But for that someone would have had to actually read up on voting methods before describing the process. A good voting method is robust against most trickery (strategic voting) when the evaluation method is known beforehand. This is not true at all for the mean (except for a single yes/no vote). It is slightly better for the median. And it is best for Condorcet. By screening the results for tasks ranked badly by Condorcet, we can reduce the problems of using the mean. Best wishes, Arne -- Unpolitisch sein heißt politisch sein ohne es zu merken signature.asc Description: PGP signature ___ Devl mailing list Devl@freenetproject.org https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
