Slashdot Mirror


User: PurpleBob

PurpleBob's activity in the archive.

Stories
0
Comments
1,030
First seen
Last seen
Profile
(view on slashdot.org)

Comments · 1,030

  1. Re:Its an easy system to stop on Climbing up the Search Ladder · · Score: 1

    So then SEOs make link farms pointing to all the competition's pages instead of their own. What problem did you solve again?

  2. Sampling bias on Top 100 Toys From The '70s or Thereabouts · · Score: 1

    I imagine the toys from today that happen to still be usable 25 years from now will generally be pretty rugged, too.

  3. Re:French Revolutionary Calendar on New Calendar Proposal · · Score: 1

    Whether it's Biblical reasons or not, the 7-day week cycle has never been broken in recorded history; it's even been maintained through calendar changes. By now the 7-day week is ingrained in culture everywhere in the world. So a calendar without a 7-day week will never catch on.

  4. Re:Condorcet and IRV on Networks Ignore 3rd Party Candidates · · Score: 1

    Please read my other post. You don't give points to the candidates; the method you are describing is Borda. And you don't rank individual pairings; that's just confusing and a waste of time.

  5. Re:Arrow's impossibility theorem on Networks Ignore 3rd Party Candidates · · Score: 1

    Okay, you're wrong. Hopefully my other posts explain why.

    And now I've noticed something. The method you've described that you thought was Condorcet (where candidates get a "point" for each candidate they're ranked higher than) happens to be equivalent to the Borda Count, a method that's even more flawed than plurality.

    Your complaints about flaws that you saw in Condorcet were in fact quite accurate complaints about Borda.

  6. Re:Two replies on Networks Ignore 3rd Party Candidates · · Score: 1

    It sounds like you've had Condorcet described to you very badly. Here's how the method works, overall:

    1. Voters each give a ranking of the candidates.
    2. When counting the election, the candidates are considered in pairs. Candidate A is considered to beat Candidate B if A is ranked higher than B on more ballots than B is ranked higher than A.
    3. The vast majority of the time, one candidate beats everyone else pairwise. This candidate is the winner.
    4. In approximately 5% of races with three or more viable parties (the 5% statistic comes from some simulations I've done), instead of a single winner, you get a cycle of people who all beat each other; for example, a majority votes A over B, a majority votes B over C, and a majority votes C over A. This is called a "circular tie". At that point, you need to use another voting method to resolve the cycle.

    Some methods (Winning Votes, Schwartz, Copeland, etc.) are specifically designed to deal with the situation of a circular tie. These are called "Condorcet completion methods". You can also complete Condorcet in the case of a circular tie by using an unrelated method like Plurality, Borda, or IRV.

  7. Re:Arrow's impossibility theorem on Networks Ignore 3rd Party Candidates · · Score: 1

    I'll get to your other questions now.

    On write-ins: hopefully, along with the list of candidates, your ballot contains a few write-in blanks that you can fill in with a name and rank like the other candidates. This probably makes more sense now that you know that you vote with a ranking.

    Your "spoiler" situation with Nader and Gore seems to indicate more confusion about how Condorcet works. I don't know what you mean by a "point"; except in some sub-methods to break circular ties, Condorcet has no points, just a list of pairwise preferences that are held by a majority of voters.

  8. Re:IRV is BROKEN on Networks Ignore 3rd Party Candidates · · Score: 1

    Anyone you don't rank is tied for last place on your ballot.

    You can also mark ties for positions other than last place, which you can't do in IRV.

  9. Re:Arrow's impossibility theorem on Networks Ignore 3rd Party Candidates · · Score: 1

    You don't mark each individual pairwise comparison; that would accomplish little and take forever. You rank the candidates, just like in IRV, except you can rank any candidates you want as tied. Your pairwise preferences are determined from who is higher on your ballot.

    In your case, you could rank the candidates like this:

    (1) Gore and Nader, tied
    (2) everyone else, tied

    Or you could just rank:

    (1) Gore and Nader, tied

    and the fact that everyone else is tied for last would be implied by you not ranking them.

  10. Re:Two replies on Networks Ignore 3rd Party Candidates · · Score: 1

    And I forgot to respond to this question:

    What happens when Democrats only choose their candidate, leaving all other pairings blank?

    Then the Democrats don't get a say in the rankings of other candidates, and they still need to get 50%+1 against every other candidate to win. What did you expect would happen?

    Also, calling the votes "pairings" is commonly used in anti-Condorcet FUD to imply that the ballot is some confusing grid of pairs of candidates. No. It's a ranked ballot, just like in IRV.

  11. Re:Two replies on Networks Ignore 3rd Party Candidates · · Score: 2, Informative

    Condorcet is the one that is flawed. You don't get to vote for the candidate you truly want. What happens when Democrats only choose their candidate, leaving all other pairings blank?

    If you're going to spread FUD about a voting system, back it up. What prevents you from voting for the candidate you want in Condorcet?

    Now to fill in some information.

    General Wikipedia link about voting systems of all kinds: Voting systems

    Useful but contradictory conditions for a single-winner voting system are described at Arrow's impossibility theorem.

    "Monotonicity" is the property that, all things being equal, ranking a candidate higher should never decrease their chances of winning the election. IRV fails monotonicity in cases where there are three viable parties.

  12. Re:Copyrights and money on 'That's All Right' Soon To Enter UK Public Domain · · Score: 1

    The first term of copyright (in your system, the first decade) should be free. Most people don't know that a work they're creating is going to be a hit, so nobody would pay for copyrights until it was too late. Then the _only_ copyrights would belong to corporations and their "hit machines".

  13. Re:Why Fight? on Language Tempest At Orkut · · Score: 1

    Um... Lojban isn't supposed to be a universal language, it's supposed to be a logical language. I'm a Lojbanist, and I think that going around shouting "Everyone should speak Lojban!" (u'i ei ro prenu cu tavla fo la lojban) just damages its reputation, because that's not even one of its goals.

  14. Re:Real news on The Traveling Salesman Problem Meets Starbucks · · Score: 2, Informative

    In English, you can end sentences with prepositions. The rule saying you can't was made up by scholars in previous centuries who thought English was derived from Latin (it isn't at all), and also thought that any deviation from Latin grammar was an error (and even if English came from Latin, why the hell would that be true?) Linguistics was not well understood then, so scholars thought that it ought to be possible to translate word-for-word from Latin to English.

    Specifically, in English, you ask many questions by moving the object (often "what") to the front of the sentence, and leaving nothing in its place. (Examples: "You are doing what?" goes to "What are you doing?", and "You are talking about what?" goes to "What are you talking about?")

    This can indeed leave a preposition at the end of the sentence. Moving the preposition as well is also allowed, but is much less common. (Any English speaker would find "About what are you talking?" to sound stilted.) The preposition at the end of the sentence still has a referent, it's just in another place, and this kind of movement occurs in many languages, as any linguistics course will tell you.

    The only reason you couldn't do this in Latin is because the preposition was PART OF THE WORD, so of course you had to move it with the word. It's a lot like split infinitives, another classic grammatical non-error: Caesar couldn't have split an infinitive if he had tried.

    And so your post had no point except to be elitist about a point whose only basis is in discredited 19th-century linguistics. Don't you feel special?

  15. Re:Socialist? on Slashback: Civilians, Rubyx, Restrictions · · Score: 1

    Clue: the next two words were "compared to". He was making an exaggerated comparison, for effect, and that was in fact the entire point of him mentioning The Jungle at all.

    Were you so eager to rant about socialism that you stopped reading right there, in the middle of the sentence?

  16. Re:Starts with 3GLs. on The History of Programming Languages · · Score: 1

    Well, yes, I know Prolog is spiffy like that. And that's part of it - Prolog is just completely different than any other programming language, and it fills its niche (logic programming) extremely well, so there has never been a reason to branch a new language off of it.

  17. Re:Starts with 3GLs. on The History of Programming Languages · · Score: 1

    Another interesting one is Prolog, which has no interactions with any other language on the chart, just a straight line from 1970 to the present.

  18. Re:It's an "intranet" on Advice On A New-School Old-School BBS · · Score: 1

    Hey, if you look in the source to the parent post, you'll find an HTML comment that says "gullible".

  19. Re:And #11 is a tie between.. on Ten Technologies That Refuse to Die · · Score: 1

    And they get completely bogus responses from it. If your client fakes ident responses, you get on; otherwise, they make you wait a minute to connect because you're not l33t enough.

  20. Re:Just ordered mine on FreeSpace 2 Gets Reissue As Limited Edition · · Score: 1

    Like, apparently, getting you to pay again for a game that they did no extra work on.

    That's not stupid, that's unfortunately smart. I'm sure they care about your money more than the 1/10 feedback.

  21. Re:my reasons....... on Who Needs Case-Sensitivity in Java? · · Score: 1

    Case-sensitive non-homonyms: polish and Polish.

  22. Re:That reminds me on Skeptical Environmentalist Saga Continues · · Score: 1

    Three replies. That was a bit unnecessary.

    An ad hominem attack is to make a personal attack on someone in order to make their argument seem less valid; "you're too stupid to be right." An example said by you was "He thinks he can teach me philosophy!"

    So, that page is saying that we have different definitions of faith. It says that faith has to be theological. This, I assume, is mostly a convenience for the page, which wants to show how non-rational theological faith is, so they make clear that they don't mean anything else when they say "faith".

    But a belief that the scientific method works is the same type of belief, even if you don't call it the same thing.

    Thank you for taking guesses at what I believe. I've already said it: I do believe the scientific method works, and I don't believe in God. I only accept "absolute proofs" as being incontrovertibly true, but I accept evidence as making something likely to be true, just as you do.

    You just have to realize that it's possible to be even more skeptical than your position. Hume came across the fact that induction has no logical basis, said what amounts to "Hmm, that's funny", and then continued to believe it anyway. Bertrand Russell built on Hume and claimed not to believe in induction at all. Of course, with this kind of skepticism, you basically can't be anything but a philosopher.

    It's also possible for people to believe in both God and the scientific method, and to believe that God trumps the scientific method. You can tell these people that they're wrong, but no amount of appeal to evidence will convince them, just as no appeal to faith in God will convince you.

    Of course, there are also people who hold logical contradictions like transsubstantiation in their head, and they are what the skepdic page are attacking most. Indeed, such people are illogical (at least I've never heard a logical explanation of transsubstantiation) and cannot function in a logical argument.

  23. Re:That reminds me on Skeptical Environmentalist Saga Continues · · Score: 2, Insightful

    We "assume" our observations are true because, quite simply, they are evidence.

    Thank you, Captain Tautology.

    Okay. You've pointed me to a joke page on Angelfire. That page, incidentally, is pretending to refute mathematical induction, which is (counterintuitively) totally deductive, not logical induction.

    In response, I point you to Bertrand Russell's "On Induction".

    As long as we've devolved into ad hominem attacks: your attitude is the same as fundamentalists. "I'm right because I'M RIGHT DAMMIT."

  24. Re:That reminds me on Skeptical Environmentalist Saga Continues · · Score: 2, Insightful

    To reply to both:

    A = A is an axiom of deduction. You can also talk about whether deduction is justifiable, but that's off-topic, because I'm talking about induction.

    Induction is not logical at its base. The basic argument of induction goes something like this:

    This is a raven.
    This is black.
    Therefore, all ravens are black.

    An argument that is fallacious on its own, but when repeated many times (with many different ravens) increases in its probability of being valid due to induction.

    You use the word "faith" to mean "all unjustifiable assumptions other than mine", and then you go off on an ad hominem "He thinks he can teach me philosophy" based on the differing definition. Clever.

    We assume that our observations are true because we believe in the scientific method, and we believe there is nothing better to go on.

    Other people believe there is nothing better to go on than God.

    I personally find the scientific method to be useful, and God not to be useful, so I put my faith in the first and not the second.

  25. Re:That reminds me on Skeptical Environmentalist Saga Continues · · Score: 2, Informative

    Y2K is junk science, you say?

    Programmers worked their asses off to fix Y2K bugs before the rollover occurred. When it happened, nothing went wrong because everyone generally did their job right, and then suddenly Y2K is denounced as a hoax.