riddler-castles/castle-solutions: 600
Data license: CC Attribution 4.0 License · Data source: fivethirtyeight/data on GitHub · About: simonw/fivethirtyeight-datasette
rowid | Castle 1 | Castle 2 | Castle 3 | Castle 4 | Castle 5 | Castle 6 | Castle 7 | Castle 8 | Castle 9 | Castle 10 | Why did you choose your troop deployment? |
---|---|---|---|---|---|---|---|---|---|---|---|
600 | 1 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | For this, I didn't try to get in the minds of my opponents. I developed a random deployment generator - or tried to. Then pitted each deployment against every other deployment. I first selected a random number from 1 to 10. This would be the first castle I would deploy troops to. I selected a random number from 0 to 100 to determine the number of troops. Then I selected another number from 1 to 10. If it was the same as the first draw, I drew again. Then selected the number of troops to send between 0 and 100 less the number at the first castle. Repeat this until all castles have been assigned troops. This produced a lot of identical deployments, which made me question the randomness of this. After further thought, I should chosen the second castle with a random number from 1 to 9. But because I'm up against a deadline, I don't have the time to develop truly random deployments. I can still glean some insights from my faux random deployments. I got about 10,000 unique deployments and ran them against each other - round robin style. I added in a few of my own (e.g. even distribution, weighted high, weighted low, etc.) I gave 1 point for a win, 0 for a loss and 0.5 for a tie. The winning deployment earned 9,663.5 points and was one of my chosen strategies. This beat out the next best strategy (a random one) by almost 40 points. Honestly, I'm disappointed in this answer. I was hoping that one of the random ones would prove better. |