Tuesday, 1 October 2013

Representative Democracy

To win in a representative democracy you have to be the party or coalition that has the largest number of representatives.  There are quite few variants of representative democracy.  Around here we play a variant misleadingly called "First Past the Post," but I will call it "British Rules" since that's shorter and less informative.

So, how do you win this game?  The game is played in two phases: the election phase and the representation phase.  If you are playing the Three-Syllable-I expansion that allows 12+ players then you can't realistically get the majority of representatives during the election phase so you need to do a lot of playing in the representation phase.  Under British Rules you are playing with two to five players, and you can try to win outright during the election phase.  In fact, under British Rules people barely even bother playing the representation phase, usually just skipping to the round if no one wins in the election phase.  The majority British Rules players play with only two player so someone has to win during the election phase.

Obviously, most people play games just to have fun.  Maybe in your game of Puerto Rico you want to build a University so that your people can be educated.  Similarly, in a game of representative democracy you might wan to just come up with the best policies you can and see if they earn you more representatives or not.  But realistically, when a game is being played at a high level - like national championships - it's not worth considering this method of play.

There are two different tactics during the election phase.  One is the suasion tactic where you try to be more appealing to the electorate than your opponent.  The other is the vote manipulation tactic where you try to set up a situation where it is easier to vote for you and more difficult to vote for your opponents.  Most pro players use a combination of both, but for various reasons vote manipulation is much more effective.

First of all, even though the game has been around for a long time no one is even close to solving the suasion portion of the game.  In fact, even professional players are downright terrible at it.  This is partly because professional representative democracy players are largely the kind of results-oriented idiots who would pat themselves on the back for winning a coin-flip competition, but it is also harder to practice than you would think.  With British Rules national championships occurring in only four countries and only about every four years in each, there just isn't a lot of time to get good at it.  You can practice locally but really the game is so different on the national level its almost a disadvantage to walk in thinking you know what you are doing.  Plus, the game has so much randomness in it you'll never know exactly why you won or lost in the past.

Despite being actually stupid, results-based thinking does lend itself to an evolutionary process, so those who, for whatever reason, tend to employ more successful strategies are the ones who will tend to creep their way up the ranks and be sent up by the national teams.  It's a lot worse than a good analysis, but in a game where a good analysis might be impossible, using evolution to find the best players might be the way to go.

Through this process, the national teams have put forward players who play essentially at random as far as suasion is concerned and go all in on vote manipulation.

What are the pros doing in vote manipulation right now?  We'll take a look at top strategies.  Notice that many of these strategies actually sacrifice a lot in terms of suasion in order to make gains in vote manipulation.

1. Abandon areas where you are significantly behind or significantly ahead and focus entirely on areas where you are slightly behind or slightly ahead.  Since each area produces only one representative, you want to be one elector ahead of your opponents in half the areas, not to blow them out in a small number of areas or to be a close second.  This strategy is basically used by all serious players of the game.

2. Focus your suasion on a very narrow part of the electorate and then try to alienate absolutely everyone from the political process as a whole.  The goal is to build up a group of fanatical electors who will vote for you and then try to get as few other people as possible to vote.  It's better to win two to one than it is to lose 9999 to 10,000.

3. Within each area there will be sub areas that favour you and sub areas that favour your opponent.  Use agents of the state to restrict freedom of movement during the election in sub areas that favour your opponent.  For example, you could have police conduct more traffic stops there.  This strategy seems very weak because it has a minimal effect on the election and it seems like it would cost you a lot in terms of suasion, but in practice the loss of suasion seems to be next to nothing as long as the targeted group is generally discriminated against anyway.

4. Like 3, but instead of preventing people from getting to vote, you use trickery to make them think the election is on a different day or that they are voting in a different place.  Unlike 3, unfortunately, this seems much more likely to lose you suasion in the next round as your involvement will be far clearer.  Of course, this only matters if you care about the next round and not just about this one.

5. Gerrymandering is the strategy where, in the representation phase, you remix the sub areas into new areas so that they are more likely to either have you win by a bit or lose by a lot.  Pros at the state level in the US are really big on this tactic.  Personally I worry that it is a very bad long-term tactic, but if you are keeping your eyes on the next prize, it is virtually unbeatable.

6. You could pass laws to suppress the voting rights of those who aren't going to vote for you.  Like 3, as long as you target people who are generally discriminated against already, you may be able to do this without much loss.  What makes this work so well is that the people who are going to vote you for are mostly going to be discriminating against people who aren't going to vote for you, so you don't lose votes from the target group because they weren't voting for you anyway.

7. You could threaten to and actually begin to imprison or kill people who don't vote for you.  This is currently not allowed in British Rules.

Generally you need to win the previous election phase to have much success with 3, 5, 6, and if you need to win the current election to avoid the consequences of 4 and 7.  Because of this you'll see a lot of pros focusing more on 1 and 2.  It's not that they won't employ the others, it's just that whoever won last round already has an advantage in the coming round and it's better not to focus on winning more if you want to up your overall record.

What we haven't seen recently is anyone really go all in on 6 and house-rule in 7.  I think the general feeling of most people who play British Rules is that if you go that way then people will stop playing with you altogether, but if you really want to win a few rounds without any regard for what happens after that, this is quite obviously the way to play the game.  Since the national teams seem increasingly focused on one round at a time, I think the time is getting right for someone to make a move in that direction.  My expectation is that such a tactic would be played in the US at the state level first, and then if it worked it would move to the national level and then get played in Canada a few years later.  There are already many states that use 6 against very disliked segments of the population, and there are some states where it is being tried more widely.

Hopefully when people do get around to killing voters who don't vote for them and making it illegal to vote if you aren't part of the group that will support the government they will broadcast the national finals on Twitch so we can all watch.

No comments:

Post a Comment