Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The Price of Anarchy

en.wikipedia.org

Post reply on HN
Print view View original

61–70 of 140 posts

Re: The Price of Anarchy

#61

Post by snrplfth »

monocasa wrote: post #55
snrplfth wrote: post #45

Earlier quoted context omitted.

But if people can freely organize themselves into collectives or communes, then some of those collectives will be more successful than others, and without mandatory redistribution, you'll have inequality again. Also, if people can leave collectives, then inevitably there are going to be collectives with few or only one single member left, and you're back to individual ownership.

If that 'collective of one's legitimately only needs one worker, then that isn't a problem. If someone else joins him later then he has to share the MoP. I fail to see the logical inconsistency.

But are collectives allowed to exclude people? If not, how will they select appropriate workers? If you can just show up at a collective and say "I want to work here, and you can't exclude me or kick me out" then how will they be able to function? People will just leave collectives with poor returns and make collectives with good returns take them on.

Re: The Price of Anarchy

#62

Post by mcguire »

JumpCrisscross wrote: post #22
TeMPOraL wrote: post #20

Earlier quoted context omitted.

Well, what you quote is true, but I think a lot of discussion stems from focusing on the wrong question, namely "does selfishness lead to better outcomes" as opposed to when it does, and when it does not. Too many people seem to approach economy as an ideology rather than a complex system in which a given approach produces good results here and bad results there, and everything should be subject to optimization and c…

> Too many people seem to approach economy as an ideology rather than a complex system Few of these people have any decision-making authority. This mass ignorance of economics does occasionally bubble up on the policy side. But it is generally, at least in the United States, contained by more knowledgeable (and influential) parties.

...like Alan Greenspan?

Re: The Price of Anarchy

#63

Post by danharaj »

di4na wrote: post #51
throwaway729 wrote: post #48

Earlier quoted context omitted.

> ...as they would like it to become... Friendly society... DRO... I think this has always been my primary gripe with ancap -- it's Utopian. And all Utopian thinking has the same problem -- it's either irrational, or very rational but starting from very strong axioms. Ancap writers, to their credit, tend to care a lot about crafting strong, rational arguments. When I think about ancap (or any political theory that se…

Well welcome to nearly all political theory. There is a reason we are still stuck with some fairly well known different systems for all these years. Human bahaviour is the same so far. And saying that we need to modify human behaviour to make it "better" is a complex road too... The thing is, we tryied to did it with the widespread education of since the 50s, with a more or less loose objective and set of policies, i…

> Well welcome to nearly all political theory. There is a reason we are still stuck with some fairly well known different systems for all these years.

There are plenty of anti-utopian and materialist political theories. The real reason why we've seen only a few different systems is that they are the ones that most effectively wield power and violence to maintain their class relations. That's the position of Marxists. Most strains of anarchism are also anti-utopian.

I think it is naive to say that the reason we haven't progressed is because we don't have the right set of ideas. Idealism itself is an obstacle. Understanding the material forces that create and propagate our world is essential. Most mainstream political ideologies pretend the world is driven by ideas when it's the other way around.

Re: The Price of Anarchy

#64

Post by snrplfth »

RodericDay wrote: post #59
snrplfth wrote: post #37

Earlier quoted context omitted.

But the comment he was responding to said "humans in more permissive societies habitually live in relative harmony even more than those in less permissive ones." So the fact that, upon trivial examination, a society is revealed to not actually be that permissive, is not evidence against the assertion that permissive societies are harmonious. I mean, I don't think RodericDay was saying that those two examples were imp…

Parent post said: > One might suspect that all our beneficial habits would break down if there weren't an almighty parental figure lying in wait to punish us I interpreted this in the vein of a libertarian "we don't need a nanny state". And so I assumed that the person was thinking of the USA as a "permissive society in harmony" and of something like North Korea as an "less permissive society in disarray", as opposed…

Okay, then I agree with that. And I would add that its non-permissiveness is a big contributor to its disharmonies.

Re: The Price of Anarchy

#65

Post by mcguire »

jessaustin wrote: post #33
clydethefrog wrote: post #4

Already in the first line it claims that anarchy is a system wherein >a system degrades due to selfish behavior of its agents To my point of view, most game theory is just capitalistic ideology disguised in mathematical models.

There is a bit of ideology here, in that there's no particular reason that this tiny little observation about network routing had to be called "The Price of Anarchy ". Just the same, it's a mistake to think this topic itself has anything to do with politics or economics. A mistake made intentionally, perhaps, by those with particular political motives.

Network routing is an example. Just because Einstein's original thought experiment involved a train doesn't make special relativity railroad engineering.

Re: The Price of Anarchy

#66

Post by danharaj »

bo1024 wrote: post #38

I just want to mention that many commenters here are reading much more into the name politically than do the game theorists/computer scientists who actually do research on it. The research is focused on the game theory and not meant to make any sort of political points or value judgement. The main question asked is how to design systems so that even in a game-theoretic equilibrium, the total "welfare" is close to the…

Having no intention to be political just means they uncritically reproduce aspects of the ideology they passively go along with. Associating anarchy with Bad Things was a deliberate government program to suppress its movements(quite popular at the turn of the previous century). Now people do it reflexively and that is very much political. The political is not merely intent.

Re: The Price of Anarchy

#67

Post by sebastos »

bo1024 wrote: post #38

I just want to mention that many commenters here are reading much more into the name politically than do the game theorists/computer scientists who actually do research on it. The research is focused on the game theory and not meant to make any sort of political points or value judgement. The main question asked is how to design systems so that even in a game-theoretic equilibrium, the total "welfare" is close to the…

Ding ding ding. The Price of Anarchy originally described the surprising result that internet routing seems to work despite the lack of a central planner. [1] The "central planner" is a theoretical maximum for the routing problem - it represents an oracle that automatically knows and applies an optimal routing solution. It's not some kind of government.

It's kind of hilarious seeing all of these people commenting about how this is "utopian" thinking, ancap-blah-blah, and "who says the central planner would be just?" People read what they want to read, I suppose.

If you do want to read something interesting about the price of anarchy, check out this other paper [2], written by one of the original authors, which shows that the PoA = 1 (that is, there is no cost) when there are private "owners" of the edges in a routing network who are allowed to charge a "toll" for traversing that edge. Makes you wonder if a globally-standard-but-locally-owned electronic tolling system could replace gas taxes to fund our roads in a fairer way. This could remove the undue toll burden shouldered by certain unlucky people who have to commute across a heavily tolled bridge. (Seattlites should know what I'm talking about).

[1] http://www.cs.yale.edu/homes/jf/poa.pdf [2] https://people.eecs.berkeley.edu/~christos/papers/selfishrou... [3] http://theory.stanford.edu/~tim/papers/routing.pdf

Re: The Price of Anarchy

#68

Post by snrplfth »

throwaway729 wrote: post #56
snrplfth wrote: post #45

Earlier quoted context omitted.

But if people can freely organize themselves into collectives or communes, then some of those collectives will be more successful than others, and without mandatory redistribution, you'll have inequality again. Also, if people can leave collectives, then inevitably there are going to be collectives with few or only one single member left, and you're back to individual ownership.

A lot of communists posit that capitalism (and perhaps in some cases statism) cause and/or arise from something kind of resembling a mental illness. Marx's "Alienation" and especially Delueze's "Desire" have this flavor. I think this is one reason why there's a whole brand of communism that's closely tied to Freudian/Lacanian style psychology. That psychological aspect usually plays some role in transitionary theorie…

Well, that's a big weakness of these theories, I think. My question is not terribly complex or dependent on unusual or unrealistic situations.

I would also note that sometimes, people operating off a 'psychologist' basis like you've identified, misunderstand the critiques of others as being also 'psychologist'. For example, the Hayekian critique of central economic planning is often characterized as a psychological argument that "people are naturally selfish and greedy". Thus, if you can get them to stop being so selfish, the objection is overcome. However, what the critique is actually about is inherent limits to knowledge and information, which would apply regardless of psychology.

Re: The Price of Anarchy

#69

Post by Hoasi »

nerdponx wrote: post #42
CaptainSwing wrote: post #9

I think that this is a really interesting concept, but dislike the name. All social anarchists (anarchist-communists, anarchist-syndicalists, even mutualists etc) would see anarchy as a social process rather than a purely individualised "every man for himself" system. It disregards the possibility for non-coercive collective forms of decision making. but, saying this, a concept that discovers the loss in a system whe…

Economists tend to go for whimsical names for their high concepts, cf. "Tragedy of the Commons"

That's when their inner poet does the talking.

Re: The Price of Anarchy

#70

Post by sebastos »

madmaniak wrote: post #16
madmaniak wrote: post #3

"In the 'centralized' solution, a central authority can tell each agent which path to take in order to minimize the average travel time" That's a wishful thinking. Central authority can and usually is selfish and corrupted but the impact of it is much bigger. So we can call this piece of math a science but voting for centralized systems based on that is super naive.

I see my comment is not well received. Who will explain me why we assume that a single unit is selfish, but a big one contained of many selfish units is altruistic?

I'm guessing it's because you're projecting political baggage onto something that is more of just an interesting theoretical result. The Price of Anarchy's original setting was the issue of internet routing, and the "mystery" of how it could work without any central oversight. The answer is that there is a price you pay for having completely decentralized, selfish decision making, but it's bounded. The central oversight bears little relation to a government - it's an oracle that can calculate a theoretical optimum, not some kind of decision-making bureaucracy or something.
Post reply on HN