Yesterday, in one of my brief respites from feeling terrible, I came across a mathematical analysis of Candy Land. He does an analytic analysis which assumes independent draws (i.e. you draw a card and then put it back in and reshuffle the deck) and a Monte Carlo analysis that’s much more realistic. According to Lou Scheffer, a two-person game has a mode of 31 turns.
But wait! There are other analyses! This one is an analytical solution using Maple, while this one uses MATLAB to compare Cootie, Candy Land, and Chutes and Ladders to decide which is the least painful to play. I’ve said that Candy Land is a deeply stupid and annoying game, and the last link backs me up on that — according to his analysis, of the three games, Candy Land is likely to take the longest time to play.
As Barry Wise, author of the final article, says, “This article demonstrates how Monte Carlo simulation can be used to solve a real-world, every day problem: Of these three games, which one will provide entertainment for my four-year-old yet let me retain my sanity?” And people say you can’t do anything useful with math!