User talk:Theorist2

Welcome

Welcome!

Hello, Theorist2, and welcome to Wikipedia! Thank you for your contributions. I hope you like the place and decide to stay. Here are some pages that you might find helpful:

• The five pillars of Wikipedia
• Tutorial
• How to edit a page and How to develop articles
• How to create your first article (using the Article Wizard if you wish)
• Manual of Style

I hope you enjoy editing here and being a Wikipedian! Please sign your messages on discussion pages using four tildes (~~~~); this will automatically insert your username and the date. If you need help, check out Wikipedia:Questions, ask me on my talk page, or ask your question on this page and then place {{helpme}} before the question. Again, welcome! Aboutmovies (talk) 15:57, 25 September 2010 (UTC)

Range voting, et al.

btw... whereas you notice that scores have more information than ranks, you might be interested in Warren's computer-simulated voting research. He takes psuedo-random score votes and translates them into all the other popular voting mechanisms; comparing the "avoidable social utility loss" introduced per-voting-system, and comparing with various strategies per-voting-method. Google "Bayesian Regret", or see the result graph if you're interested. It is on the same site as the other range-voting stuff from this professor. --Osndok (talk) 16:39, 20 October 2010 (UTC)

Thanks for the info. I want to be as unbiased as possible. So I might someday investigate Range voting seriously.

The day may soon come when professionals like political scientists and social choice theorists pay more serious attention to Range voting. This is because professionals such as Michel Balinski and Rida Laraki are also beginning to propose non-preference based methods in academic journals [1].

I do not know if he is intentional or ignorant, but if Warren Smith becomes more academically sincere, then professionals will begin to listen to him. For the moment, as soon as professionals read the statements like "Range voting satisfies all three criteria, accomplishing the "impossible"! Huh?", they would simply stop reading what he writes, believing it is nonsense. Isn't that a sad situation? I hope my revision of Arrow's theorem, which treats Rv fairly, will help change this sad situation.

By the way, we do make a lot of decisions (like choosing someone to be sent abroad) by Range voting at my institution. My own experience of Range voting has not been very positive. Often the result is determined by votes of only a few voters who vote strategically (using the highest and lowest scores). It is just personal, but that's one reason that I cannot be very optimistic about Rv.--Theorist2 (talk) 14:29, 23 October 2010 (UTC)

I'm sorry to hear that you've had such a disfavorable experience with range voting. It is hard for me to understand an objection to using the highest & lowest scores (isn't that just normalized-range?), particularly as every voter has precisely the same theoretical effect. Nevertheless, that is a major complaint with range voting... that in practice it might reduce to approval voting. --Osndok (talk) 21:03, 25 October 2010 (UTC)

The problem is that it does not reduce to approval voting, since some do not vote strategically. Maybe they believe giving a "correct score" to each candidate is important, even when the final result is to choose the candidate or not. I guess Michel Balinski and Rida Laraki's method above is based on that view. In their method, the scores have meaning, and strategic effects are reduced by taking the median of the scores instead of the mean.--Theorist2 (talk) 22:08, 25 October 2010 (UTC)

Yeah, I read about that first on this page of Warren's. To me it seems like the only thing it does is re-introduce the dictatorship problem. According to the same page, median-range actually has *more* of a tactical voting pathology (because one can purposely try to get your ballot in the middle?). Averaging is much computationally easier anyway (can be counted per-district). --Osndok (talk) 18:06, 26 October 2010 (UTC)

btw... I saw the contested footnote/example you gave on the fairvote website against range, but I still don't understand it: if the majority of voters express nearly-zero opinion, why shouldn't the 1% that does count? The issue could just as easily be:

1. 49% voters (1,0,0)
2. 49% voters (0,1,0)
3. 2% voters (0,0,99)

Why shouldn't the third candidate win? they say 98% "prefer a different candidate" (rank logic), it seems to me that 98% of voters have no preference. In your opinion, who do you think should win the above election? --Osndok (talk) 16:53, 27 October 2010 (UTC)

Hello, again, Osndok! Maybe you should write more about yourself on your user page. I'm not sure what you mean by "the contested footnote/example you gave on the fairvote website against range". You mean my footnote about range voting violating Arrow's IIA in the article for Arrow's impossibility theorem? Anyway, let me answer your question. It serves as a practice for writing English.

A rule is something that a society should choose (agree upon) for its own decisions, before a particular situation arises. There must be a reason to choose a particular rule (e.g., market or voting) for particular purpose (e.g., individual or public), as you can see from the fact that most decisions are not made by voting. If the rule is based on rankings, then the third candidate should not win. (If a voter is really indifferent, then she would refrain from voting. As long as she votes, the vote should be counted. That's the choice of the society and the way the rule is supposed to work.) If the rule that the society has chosen is range voting, then I agree that the third candidate should win. (But if how much one cares is really important, maybe market is better than voting.)

Range voting is good for the situation described by your example, but that does not justify the society should adopt range voting. This is because the same rule will be used for many other situations. If this society's always just like above (for all issues, most people are indifferent), then it is probably reasonable for the society to choose range voting. But there do exist some issues that many people care. In your example, suppose 1.9 precent of the first group changes to (30,0,0). Maybe the third candidate should still win according to the argument for range voting. But what if the group of 1.9 percent strategically vote (99,0,0)? I think an argument justifying a rule (which will applied to many situations) by citing a particular situation is weak. That kind of argument is okay for showing that a rule satisfies or violates a certain condition. But it is weak for justifying a rule. --Theorist2 (talk) 22:36, 27 October 2010 (UTC)

Again you said "strategically vote (99,0,0)", if they sincerely desire the first candidate over the other two. How is that strategic? As individuals they can place as much support for any candidate as anyone else. Strategy would be if that group really did support the third candidate [e.g. (99,0,80)]. But thinking that the third candidate would be likely to win, gave him a zero score to favor his desired candidate.

The short story is this: 0%-strategy unarguably good, and 100%-pure-strategic-range-voting (maximum dishonesty) reduces to honest-plurality voting, because there "is never a reason to give your favorite choice less-than-maximum support." Therefore the maximum effect any strategic subgroup has on an election is proportional to it's size (and therefore fair).

It actually would not even require (30,0,0). In this example, if group A even voted as high as (5,0,0) [out of 99 possible] they would beat group C! I am still trying to understand the argument that range voting somehow favors the minority... It seems to me that independently weighing each candidate is not only fair, but always best. That is why I say it is good for society. --Osndok (talk) 17:43, 29 October 2010 (UTC)

If an individual whose values are (30,0,0) votes (31,0,0), that's strategic, provided that you take the viewpoint of cardinal utility. (Even from a practical viewpoint, if a judge in a contest gives a 30 point performer 99 points, then other judges would not call him "sincere.") Your logic is based on a strange (or nice?) blend of cardinal and ordinal utility---I can follow your logic and I would not say it's wrong, but I'm not really ready to accept it. In any case, that's not the point. The point is that an argument based on an example is weak unless you are trying to show that the rule violates a certain condition (i.e., give a counter-example). A mathematically better argument would be (just like social choice theorists do) to write out conditions (axioms) and prove the rule satisfies them. Adios!--Theorist2 (talk) 04:50, 30 October 2010 (UTC)