Monthly Archives: January 2016

Keep Talking and Nobody Explodes

This week’s Wednesday meeting saw the Queens’ CompScis participate in a somewhat different activity: bomb defusal.

In Keep Talking and Nobody Explodes, a player is locked in a room with a bomb that will imminently go off, and has five minutes (or less) to defuse the bomb by disarming several complicated modules. However, they have absolutely no idea how to defuse the bomb. Thankfully, the other players are equipped with a bomb defusal manual — but they can’t see the bomb, so everyone involved has to communicate quickly and clearly to stand a chance of survival.


Having arranged ourselves into threes and decided on team names, we got to work. The game proved to be quite interesting, and certainly tested our ability to convey information accurately and under pressure. We deciphered confusing symbols, disambiguated between lots of homophones, and tried to tackle a light displaying Morse Code, although around half of us found it impossible to coherently read out the rapid sequence of dots and dashes.

Overall, we enjoyed improving our communications skills in a fast-paced and competitive game environment. Congratulations to Dhruv, Henry, and Tamara, who managed to defuse 9 bombs by the end of the session! Also thanks to the developers of Keep Talking and Nobody Explodes for providing us with copies of the game.

Game Theory – NIM and Sprague-Grundy

Last Wednesday we were given a talk by Radu, one of the third year students at Queens’. His talk introduced us to some interesting and cool concepts in game theory – a topic that relates well to Computer Science, such as in problems where the optimal play must be taken.

Radu’s talk focused on NIM-style games. The aim of the game is to be the last player to be able to play a move. The image below shows an example – three piles of coins exist, and a each turn a play is allowed to take as many coins as they like, from a single pile (they must take at least one coin). IE, the player could take all the coins from pile one, and then it would be the opposing players turn.

Screenshot from 2016-01-26 12-09-24

Through the clever use of XORing the numbers, Radu went on to show that calculating and making the optimal play is very easy. Once a player has forced themselves into a winning state (where the XOR-Sum of all numbers is 0), as long as the player maintains the optimal play, they cannot lose.

Screenshot from 2016-01-26 12-16-34

Moving on from the NIM game, Radu went on to show (Via the Sprague-Grundy theorem) that many games can be reduced into a NIM game, and thus an optimal solution can be found. One such example is represented in the image above, the “Poison Chocolate Problem”, where a player can take any row(s) or column(s) in a single turn, but must not be the player to eat the poisoned block.