Den Beste Digest, #5

“CDMA is untraceable, and I can reinvent ad-hoc networking in my spare time” 3002-word version

Stick to the RF stuff, Steven. You obviously know something about that, although even I can tell that your claims about CDMA not giving away location are false. Your server-centric not-really-very-secure ideas about authentication and so on are merely amusing to those who have actually studied those things.

Random Gripe

Why, oh why, in this day and age, are so many email programs so stupid about multiple accounts? So many assume you only want one outgoing server even if you have multiple incoming; quite a few will only allow you to have one outgoing. Don’t the people who write these things realize that, in the real world, incoming and outgoing servers generally come in pairs, and the only smart thing to do when replying to a message received on an account’s incoming server is to send the reply through the same account’s outgoing server? Sheesh. It’s not rocket science.

Making Research Real

A user at America’s Debate provided a link to a fascinating application of social-network theory (the whole “six degrees of separation thing, for those who don’t know) to Mapping Terrorist Networks. I’ve only skimmed it so far, but it looks like it’ll be good reading later.

Geekdom Reloaded

To anyone who has already seen Matrix 2: you’re a dweeb. Lord of the Rings I could understand (though even I didn’t try to attend the very first show), but…the Matrix? Pleeeeease.

Lessons from Texas

Recent events in Texas, where Democratic legislators have left the state to block a redistricting plan that would lose them 5-7 seats, have highlighted another problem with American election systems. An innate characteristic of “first past the post” voting, as compared to any proportional-representation system, is that a party whose support is split across districts cannot get any candidates elected at all. This makes the system very sensitive to how district boundaries are drawn, leading inevitably to high-stakes battles like we’re seeing in Texas. With proportional representation, it wouldn’t be possible to swindle a party out of more than half a seat by redistricting; every party would be fairly represented no matter how the lines were drawn. The result would be less time spent trying to manipulate the system, and more time spent doing the actual work of legislating.

This is also a situation where approval voting or instant-runoff voting don’t help at all. Those are useful approaches for single-seat elections but, as I said before, multi-seat elections are often more interesting and/or important.

This American Life

Cindy and I went to see a live performance of This American Life at the Berklee Theater on Saturday. The guy from Found Magazine was hilarious, but the biggest crack-up line for me was Sarah Vowell’s “what would Jesus do” (answer: die) in the context of discussing the checkered past of the Battle Hymn of the Republic. The accompanying band actually had a theremin player, which was neat to watch. He also played the musical saw, which sounded surprisingly similar. All in all, it was a very good time. If you’ve never listened to the show, I highly recommend trying it once or twice. It’s like nothing else on radio; it’s like an in-depth “slice of life” and/or storytelling kind of thing, with mood music and plenty of humor.

Fingerprinting Text

I’m not entirely sure what got me going, but I’ve been thinking a lot lately about the challenge of distinguishing text written by one person from text written by someone else. It seems to be a much-studied research problem, with a lot of the research oriented toward detecting plagiarism, but there seems to be a surprising lack of available tools. I think a lot of people think of this as a competition, with reputations and quite possibly money on the line, and nobody wants to reveal their “secret sauce”.

In the classic “scratch an itch” tradition, I decided to whip up a text-fingerprinting program of my own. The approach roughly falls into the category of “stylometrics” but it’s pretty naive so far. I’m not going for the whole natural-language-processing thing. So far the program has absolutely no knowledge of words beyond that they are sequences of alphabetic characters. All it does is look at things like the number of letters per word and words per sentence, occurrence rates of certain kinds of punctuation, etc. to generate a “fingerprint” (currently only seven numbers). As brain-dead as the program is, though, the fingerprints it creates already seem pretty accurate as far as distinguishing my own text from other people’s – based on a sample of only a few kilobytes. I have lots of ideas about some other “low-hanging fruit” that I can pick, including some very rudimentary aspects of word usage and things like emphasis or quoting styles, and I think I can make it even better. Maybe it won’t be good for a doctoral thesis, but it might actually be useful instead.

No, I’m not releasing code yet. :-P I plan to, but I want to implement some of these other ideas, and make the interface a little nicer, clean up and comment the code. If anyone’s interested, ping me in a couple of weeks.

Survival of the Meanest

Volleyball has an unusual scoring system. You must already have the serve to score; if you don’t have the serve, you must win a rally to get it. Another unusual thing about volleyball is that the players on a team rotate into different positions whenever that team gets the serve. These two rules combine to create a distinctive scoring pattern. Sometimes there are extended “side out wars” where the two teams trade the serve back and forth without either actually scoring any points. Eventually one team’s strongest server rotates into the service position, or the receiving team has all of their strongest front-row players in the back row or vice versa, and the serving team runs off several points in a row. Smart coaches and captains even plan for this, and arrange their players to take advantage of these effects.

Given these dynamics, what do you think would happen if the members of a volleyball team decided they weren’t going to “run up the score” when they’re already ahead? Maybe they’d switch to serving underhand, or refuse to spike, if they’re up a point. I’ll tell you what would happen: they’d get creamed. If they were up by one point they’d promptly lose the serve, and their opponents would be one step closer to staging a run of their own. If the other team ever got ahead, though, they’d have no qualms about extending their lead and it would be hard for the “keep things equal” team to catch up. The net result (heh) is that such a policy would be tantamount to suicide.

Oddly, though, this behavior really exists – not in volleyball, but in online political debate. Many liberals are pacifists. Many are taught to respect others’ opinions. Quite a few are simply wimps. No matter how you slice it, though, these attitudes put liberals at a distinct disadvantage when debating conservatives and libertarians (collectively, “anti-liberals”) who don’t believe in “playing nice” at all. Liberals rein themselves in, and stop to listen; anti-liberals don’t reciprocate. Put ten liberals and ten anti-liberals in an online forum for a while and you’ll see a predictable little scenario unfold. The liberals will limit the tactics they use, while the anti-liberals will not. One by one, the liberals will get tired of enduring the same old personal attacks and rebutting the same old fallacies, time after time after time. Rather than respond in kind, most of them will quit. Before long you have three liberals and nine anti-liberals left (there’s always some attrition on both sides). In addition to everything else, one of those liberals gets tired of being outnumbered three to one in every conversation, and quits too. That’s all bad enough, but now you add ten more to each side. The new liberals start out being outnumbered nineteen to twelve, with many veteran opponents and few veteran allies, so the situation’s even more lopsided than the first time.

OK, I’m exaggerating a little, but the effect I describe exists even if there’s only a tiny difference in “typical attitudes” between groups, and I’ve seen this scenario play out many times – on M-Net, on Usenet, on any number of web forums. Over time, every forum seems to become more conservative, and more mean-spirited. People often seem surprised to encounter someone like me who’s liberal or moderate but also willing to bite back, because we’re relatively rare. None of this occurs because liberals’ beliefs are less valid, or because their arguments are weaker. It happens only because liberals handicap themselves. They’re like the player in the Prisoners’ Dilemma who always cooperates, and always get screwed. In an ecosystem where cooperators predominate such a strategy can actually work, but the net is not such an ecosystem. Defectors already predominate, and cooperators don’t stand a chance. Maybe that’s something to think about the next time you join your voice to the conservative/libertarian clamor in an attempt to “convert” that one last liberal who’s still holding out in an online forum. What honor is there in “winning” because your opponent gets disgusted and quits?