This seems to actually underestimate the value of voting, in that it assumes that a vote is only significant if it flips the winner of the election. But as Eliezer wrote:
But a vote for a losing candidate is not “thrown away”; it sends a message to mainstream candidates that you vote, but they have to work harder to appeal to your interest group to get your vote. Readers in non-swing states especially should consider what message they’re sending with their vote before voting for any candidate, in any election, that they don’t actually like.
Also, rationalists are supposed to win. If we end up doing a fancy expected utility calculation and then neglect voting, all the while supposedly irrational voters ignore all of that and vote for their favored candidates and get them elected while ours lose… then that’s, well, losing.
it sends a message to mainstream candidates that you vote, but they have to work harder to appeal to your interest group to get your vote.
I recently heard an argument to the contrary: Viewing voter preferences along one dimension for simplicity, if a small percentage on the left breaks away and votes for an extreme-left candidate, the mainstream left candidate may actually move further to the right—since the majority of undecided voters are in the middle, not along the boundary between left and extreme-left.
This may not generalize to a hyperplane separating a particular non-mainstream candidate from other candidates in n-dimensional policy-space, but I don’t know if presidential campaigns are set up to do that level of analysis.
You were right when you described “along one dimension” as being simplistic. There are other options than extreme-left, left, centrist, right, and extreme-right (for instance). Engaging in false dilemma reasoning as an excuse to vote for a mainstream candidate with no interest in sending political messages encouraging reform is not particularly rational.
If irrational voters are the supermajority, and elect their candidate, then you lose, yes. If you waste your time voting for someone rational who can’t win, you lose more.
Rational voters never being large enough of a block to influence the outcome of any election seems quite unlikely, especially so if we don’t require the rationalists’ favored candidates to necessarily win. I don’t know about the US, but at least in Finland, even a candidate who doesn’t get elected but does get a considerable amount of votes will still have more influence within his party (and with the actual elected candidates) than somebody who got close to no votes.
My claim isn’t that this can never be the case but that it’s not the case now, and in general it’s the most important factor in whether a rational voter can win by voting.
? The long game makes voting when you can’t make a decent impact even less rational compared to anything else you could be doing that would give you long term gains. Making money you can invest, taking time to learn a skill or network, getting more information on almost anything, convincing people to follow your beliefs or teaching others about information, donating to x-risk or other charities, working on inventing. Each of these are “long game” activities.
Of course almost no one spends all their time doing this sort of thing, and I don’t care if you take 20 minutes out of one day to go vote because it gives you fuzzies. But don’t pretend it’s a great thing you do.
I won’t pretend it’s a great thing to vote if you promise you’ll stop pretending I pretended any such thing, or that I was talking about anything other than comparisons of voting strategies.
The US suffers from a major problem with institutionalizing false dilemmas in politics. Playing the long game as a voter might well involve actions intended to lead to eventual disillusionment in that regard. Whether your time is better spent, in the long run, doing something other than voting (and learning about your voting options) is a somewhat distinct matter.
In short, you suggested that at this time rational voters cannot win by voting, which I took to mean you meant they could not get a winning result in the election in which they vote right now. My response was meant to convey the idea that there are voting strategies which could lead to a win several elections down the line (as part of a larger strategy). You then replied, for some reason, by suggesting that voting is not as useful in general as inventing something—which may be true without in any way contradicting my point.
It’s ridiculous to condemn me for trying to interpret actual meaning out of your vague one sentence reply and then respond with 2 paragraphs of what you “meant to convey”, none of which was any more obviously implied than what I read into your comment.
To respond to THIS point: So what? Each vote is a distinct event. It can easily make sense that you can influence elections positively in the future without you having that ability in any relevant way today.
I fail to see how not knowing what someone meant somehow compels you to make up elaborate fantasies about what the person meant, or even excuses it.
. . . and of course nobody ever does anything other than actually cast a vote when strategizing for the future. There’s no way anyone could possibly, say, make the voting part of a grander strategy.
. . . and I suppose you probably think that I think voting is a winning strategy in some way, basically because I pointed out some possible strategies that might seem like a good idea to someone, somewhere, as part of an attempt to remind you that the one-vote-right-now tactic may not be the only reason someone casts a vote.
In short, you assume far too much, then blame me. Good job. That’s certainly rational.
But a vote for a losing candidate is not “thrown away”; it sends a message to mainstream candidates that you vote, but they have to work harder to appeal to your interest group to get your vote. Readers in non-swing states especially should consider what message they’re sending with their vote before voting for any candidate, in any election, that they don’t actually like.
But that point can still be subject to the same (invalid, IMHO) argument against voting: your vote alone is not going to change the poll’s percentages by any noticeable extent, hence you could as well not vote and nobody will notice the difference.
I’ll explain why I think this line of argument is invalid in another comment. EDIT: here
Also, rationalists are supposed to win. If we end up doing a fancy expected utility calculation and then neglect voting, all the while supposedly irrational voters ignore all of that and vote for their favored candidates and get them elected while ours lose… then that’s, well, losing.
That’s actually a better point, but it opens a can of worms: ideally, istrumentally rational agents should always win (or maximize their chance of winning, if uncertainty is involved), but does a consistent form of rationality that allows that actually exist?
Consider two pairs of players playing a standard one-shot prisoner’s dilemma, where the players are not allowed to credibly commit or communicate in any way.
In one case the players are both CooperateBots: they always cooperate because they think that God will punish them if they defect, or they feel a sense of tribal loyalty towards each other, or whatever else. These players win.
In the other case, the players are both utility maximizing rational agents. What outcome do they obtain?
By having two agents play the same game against different opposition, you compare two scenarios that may seem similar on the surface but are fundamentally different. Obviously, making sure your opponent cooperates is not part of PD, so you can’t call this winning. And as soon as you delve into the depths of meta-PD, where players can influence other players’ decisions beforehand and/or hand out additional punishment afterwards, like for example in most real life situations, the rational agents will devise methods by which mutual cooperation can be assured much better than by loyalty or altruism or whatever. Anyone moderately rational will cooperate if the PD matrix is “cooperate and get [whatever] or defect and have all your winnings taken away by the player community and given to the other player”, and accordingly win against irrational players, while any non-playing rationalist would support such kind of convention; although, depending on how/why PD games happen in the first place, this may evolve into “cooperate and have all winnings taken away by the player community or defect and additionally get punished in an unpleasant way”.
By the way, the term CooperateBot only really makes sense when talking about iterated PD, where it refers to an agent always cooperating regardless of the results of any previous rounds.
This seems to actually underestimate the value of voting, in that it assumes that a vote is only significant if it flips the winner of the election. But as Eliezer wrote:
Also, rationalists are supposed to win. If we end up doing a fancy expected utility calculation and then neglect voting, all the while supposedly irrational voters ignore all of that and vote for their favored candidates and get them elected while ours lose… then that’s, well, losing.
I recently heard an argument to the contrary: Viewing voter preferences along one dimension for simplicity, if a small percentage on the left breaks away and votes for an extreme-left candidate, the mainstream left candidate may actually move further to the right—since the majority of undecided voters are in the middle, not along the boundary between left and extreme-left.
This may not generalize to a hyperplane separating a particular non-mainstream candidate from other candidates in n-dimensional policy-space, but I don’t know if presidential campaigns are set up to do that level of analysis.
You were right when you described “along one dimension” as being simplistic. There are other options than extreme-left, left, centrist, right, and extreme-right (for instance). Engaging in false dilemma reasoning as an excuse to vote for a mainstream candidate with no interest in sending political messages encouraging reform is not particularly rational.
One might argue that you could send a better message by writing about an issue for 1 hr rather than waiting 1 hr in line at the polls.
Laptops! Do both.
If irrational voters are the supermajority, and elect their candidate, then you lose, yes. If you waste your time voting for someone rational who can’t win, you lose more.
Rational voters never being large enough of a block to influence the outcome of any election seems quite unlikely, especially so if we don’t require the rationalists’ favored candidates to necessarily win. I don’t know about the US, but at least in Finland, even a candidate who doesn’t get elected but does get a considerable amount of votes will still have more influence within his party (and with the actual elected candidates) than somebody who got close to no votes.
My claim isn’t that this can never be the case but that it’s not the case now, and in general it’s the most important factor in whether a rational voter can win by voting.
Don’t forget to take the long game into account.
? The long game makes voting when you can’t make a decent impact even less rational compared to anything else you could be doing that would give you long term gains. Making money you can invest, taking time to learn a skill or network, getting more information on almost anything, convincing people to follow your beliefs or teaching others about information, donating to x-risk or other charities, working on inventing. Each of these are “long game” activities.
Of course almost no one spends all their time doing this sort of thing, and I don’t care if you take 20 minutes out of one day to go vote because it gives you fuzzies. But don’t pretend it’s a great thing you do.
I won’t pretend it’s a great thing to vote if you promise you’ll stop pretending I pretended any such thing, or that I was talking about anything other than comparisons of voting strategies.
The US suffers from a major problem with institutionalizing false dilemmas in politics. Playing the long game as a voter might well involve actions intended to lead to eventual disillusionment in that regard. Whether your time is better spent, in the long run, doing something other than voting (and learning about your voting options) is a somewhat distinct matter.
In short, you suggested that at this time rational voters cannot win by voting, which I took to mean you meant they could not get a winning result in the election in which they vote right now. My response was meant to convey the idea that there are voting strategies which could lead to a win several elections down the line (as part of a larger strategy). You then replied, for some reason, by suggesting that voting is not as useful in general as inventing something—which may be true without in any way contradicting my point.
It’s ridiculous to condemn me for trying to interpret actual meaning out of your vague one sentence reply and then respond with 2 paragraphs of what you “meant to convey”, none of which was any more obviously implied than what I read into your comment.
To respond to THIS point: So what? Each vote is a distinct event. It can easily make sense that you can influence elections positively in the future without you having that ability in any relevant way today.
I fail to see how not knowing what someone meant somehow compels you to make up elaborate fantasies about what the person meant, or even excuses it.
. . . and of course nobody ever does anything other than actually cast a vote when strategizing for the future. There’s no way anyone could possibly, say, make the voting part of a grander strategy.
. . . and I suppose you probably think that I think voting is a winning strategy in some way, basically because I pointed out some possible strategies that might seem like a good idea to someone, somewhere, as part of an attempt to remind you that the one-vote-right-now tactic may not be the only reason someone casts a vote.
In short, you assume far too much, then blame me. Good job. That’s certainly rational.
But that point can still be subject to the same (invalid, IMHO) argument against voting: your vote alone is not going to change the poll’s percentages by any noticeable extent, hence you could as well not vote and nobody will notice the difference.
I’ll explain why I think this line of argument is invalid in another comment. EDIT: here
That’s actually a better point, but it opens a can of worms: ideally, istrumentally rational agents should always win (or maximize their chance of winning, if uncertainty is involved), but does a consistent form of rationality that allows that actually exist?
Consider two pairs of players playing a standard one-shot prisoner’s dilemma, where the players are not allowed to credibly commit or communicate in any way.
In one case the players are both CooperateBots: they always cooperate because they think that God will punish them if they defect, or they feel a sense of tribal loyalty towards each other, or whatever else. These players win.
In the other case, the players are both utility maximizing rational agents. What outcome do they obtain?
By having two agents play the same game against different opposition, you compare two scenarios that may seem similar on the surface but are fundamentally different. Obviously, making sure your opponent cooperates is not part of PD, so you can’t call this winning. And as soon as you delve into the depths of meta-PD, where players can influence other players’ decisions beforehand and/or hand out additional punishment afterwards, like for example in most real life situations, the rational agents will devise methods by which mutual cooperation can be assured much better than by loyalty or altruism or whatever. Anyone moderately rational will cooperate if the PD matrix is “cooperate and get [whatever] or defect and have all your winnings taken away by the player community and given to the other player”, and accordingly win against irrational players, while any non-playing rationalist would support such kind of convention; although, depending on how/why PD games happen in the first place, this may evolve into “cooperate and have all winnings taken away by the player community or defect and additionally get punished in an unpleasant way”.
By the way, the term CooperateBot only really makes sense when talking about iterated PD, where it refers to an agent always cooperating regardless of the results of any previous rounds.