This is a continuation of my previous post about this game. Because it is impartial and the rules are simple, one can write a computer program that computes for a given position of dirt pieces the size of the equivalent Nim heap (which is also called its *Grundy number* or its *nimber*). Because this is computationally prohibitive, one does this for simple shapes, discovers patterns, and proves these. We did this for rectangles the last time. To warm up, we do it for L-shapes today. Here is. (5,3)-L:

The nimber of this position is 1, which means in particular that there is a winning move for the first player. You can for instance do vertical swipe in the second column, leaving the second player with two disconnected row/column of three dirt pieces each. From then on, we play symmetrically and win.

For the general (alb)-L, the nimbers are as follows:

a\b | 2 | 3 | 4 | 5 | 6 |

2 | 0 | 4 | 4 | 0 | 3 |

3 | 4 | 1 | 0 | 1 | 0 |

4 | 3 | 0 | 3 | 0 | 3 |

5 | 0 | 1 | 0 | 1 | 0 |

6 | 3 | 0 | 3 | 0 | 3 |

In other words, the nimber of an L with legs at least 3 dirt pieces long behaves quite simple.

This is good, because if a player can easily memorize the nimbers of simple positions, and these nimbers are small, then the player can usually easily win against players who lack this knowledge.

But maybe all this talk about nimbers is just vain traditional mathematics, and there is an alternative way to understand this game and win easily, without any theory, just by being *smart and tough*?

Let’s look at at another type of *Spring Cleaning* positions which I call *zigzags*. Below are the zigzags Z(1) through Z(7),

and here is the table of the first few nimbers:

n | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 |

nimber | 0 | 0 | 1 | 2 | 0 | 1 | 2 | 3 | 1 | 2 | 3 | 4 | 0 | 3 | 4 |

Any pattern emerging? No? There is a solution: Whenever you have a sequence of integers you are clueless about, you had to the Online Encyclopedia of Integer Sequences, the best thing the internet has produced ever, and type it in.

It turns out that the sequence at hand is known as the sequence of nimbers of another impartial game which is called *Couples are Forever*. The rules are simple: The game is is being played with several piles of tokens. Both players take turns splitting any one pile of size at least three tokens into two piles. The game ends when no piles with more than two tokens are left. This impartial game, beginning with a pile of size n, has (with a shift of two) the same nimbers as the zigzags in *Spring Cleaning*.

Moreover, for *Couples are Forever*, the nimbers have been computed into the millions, and no pattern has been found whatsoever. It is not even known whether the nimbers remain bounded.

Above is the graph for the nimbers of *Couples are Forever* (or zigzags in *Spring Cleaning*) for pile sizes up to 25,000. At first (up to 10,000 or so), it looks like the nimbers are growing linearly, but then they appear to even out. Nobody knows. This is one of the famous open problems in combinatorial game theory.

It tells us also that there won’t be a tough&smart solution for this game, or for *Spring Cleaning*, or for reality. Which is a good thing.