The Darwin Game - Rounds 10 to 20

by lsusr4 min read17th Nov 20208 comments

31

Rationality
Personal Blog

MeasureBot maintains its lead.

Rounds 10-20

Everything so far

Today's Obituary

Bot Team Summary Round
Silly Random Invert Bot 2-3 NPCs Returns 2 or 4 on the first round. Returns 5 - <opponents_last_move> on subsequent rounds. 10
Silly 3 Bot NPCs Always returns 3. 10
CooperateBot [Larks] Chaos Army "For the first 10 turns: return 3. For all subsequent turns: return the greater of 3 and (5 - the maximum value they have ever submitted)" 10
Silly Cement Bot 2 NPCs Returns 2 on the first turn. Otherwise, returns 5 - opponent_first_move. 12
Silly Counter Invert Bot NPCs Starts by randomly playing 2 or 3. Then always returns 5 -opponent_previous_move. 12
Silly Invert Bot 5 NPCs Returns 5 on the first round. Returns 5 - <opponents_last_move> on subsequent rounds. 12
Silly Cement Bot 3 NPCs Returns 3 on the first turn. Otherwise, returns 5 - opponent_first_move. 14
Silly Cement Bot 2-3 NPCs Returns 2 or 3 on the first turn. Otherwise, returns 5 - opponent_first_move. 14
Silly Invert Bot 3 NPCs Returns 3 on the first round. Returns 5 - <opponents_last_move> on subsequent rounds. 15
Silly Invert Bot 4 NPCs Returns 4 on the first round. Returns 5 - <opponents_last_move> on subsequent rounds. 17
Random-start-turn-taking Chaos Army Selects 3 or 2 randomly until symmetry is broken. Then oscillates between 2 and 3. 17

This alternate timeline will conclude on November 20, at 5 pm Pacific Time.

Rationality2
Personal Blog

31

8 comments, sorted by Highlighting new comments since Today at 6:06 AM
New Comment

I notice I was slightly declining for a bit until round 10, where I started shooting up again. I'm not sure if it's because I changed my strategy at that point and scored more or because a bunch of other people changed their strategy at that point and scored less. I think it's more the latter, particularly increasing clone hostility.

Clones have slightly lost ground since last time. Without critical mass, their increasing hostility will hurt them more than the opponents. It looks like we're heading for a repeat of history, with Measure as the Zvi to my David. Because MeasureBot always starts 3 in the endgame and I randomize 50/50, I think I slowly lose if its starting population is bigger than mine and it's just us. If there are multiple endgame bots my more cooperative nature could be an advantage.

Actual Zvi's BendBot has gained significant ground after being in the middle of the pack in earlier rounds. Maybe it handles the middle game especially well. LiamGoddard in fourth place is the highest ranked bot we haven't gotten any explanation of.

The "true timeline" with AbstractSpyTreeBot is probably going to be this but more extreme, since ASTB feeds Measure even more.

Larks, excellent name choice for your AttackBot.

Larks, excellent name choice for your AttackBot.

Thanks! I figured it was in the spirit of a DefectBot to defect linguistically as well, and there was a tiny chance someone might be doing naive string-matching. 

LiamGoddard is an EquityBot. It plays 3232 on the first four rounds and then determines the sequence for the rest of the game based on the opponent's sequence on the first four rounds- if they played 2323, then continue playing 32323232, if they played 3232, then play 232323, if they played 3333 then play a pattern of 3s and 2s that makes sure they don't outperform cooperation while maximizing my score, if they played something random then just try to keep cooperating. No matter what they played, my selected pattern will continue for the rest of the game.

It is really simple, but I don't know how to code myself so I wanted to be sure that it was specified carefully. I also didn't realize at the time that simulators would be allowed. Nevertheless, it's reached fourth place, which is better than I had expected. Long live the Dark Lord Liam!

My entry BendBot is also essentially a deterministic EquityBot with a different opening sequence I thought was more likely to find cooperation faster. Also, I think I'm somewhat harsher against attacks, which makes me think my opening is indeed slightly more efficient. Assuming both were coded properly, 50% of the time when we meet we get perfect cooperation, 50% of the time we get a perfect 250-250, 50% of the time we get 247-247 because we each start 23, so eventually whoever is bigger will very very slowly grind out the other if we went heads up, and we effectively are acting as a block vs. the world at this point and combine to similar size to MeasureBot.

If OscillatingTwoThreeBot (sixth place) is exactly what it says on the tin and always plays 23232323..., you get perfect cooperation with it 100% of the time. Could be a nice minor advantage.

Indeed, OscillatingTwoThreeBot does behave like that. Thanks for the cooperation LiamGoddard!

This alternate timeline will conclude on November 14, at 5 pm Pacific Time.

It's obvious there's some time travel going on, but I'm still confused.