And it’s over! Spider GM played well to win 8 games, but monkey went one better with 9 victories. Most of the games were easy, but there were a few exceptions. Game 3 was just horrible: the final round was 833A5A8jk4 which is close to an instant loss (assuming no lucky suited connector such as a 3 landing on a 4 of the same suit). And that was not the only problem. Both human and monkey “agreed” on every game (i.e. both win or both lose) except game 4. Spider GM never found an empty column since the tactics were all wrong. Even so, 64% by the monkey was not the most convincing of victories. The conclusion is that the monkey’s win rate should have some correlation with an expert playing 4-suit solitaire. In other words, the effects of making random moves and playing at the one-suit level pretty much cancel each other out.
Game | SpiderGM | Monkey |
1 | win | 89% |
2 | win | 64% |
3 | lose | 26% |
4 | lose | 64% |
5 | win | 67% |
6 | win | 72% |
7 | win | 79% |
8 | win | 55% |
9 | win | 62% |
10 | loss | 36% |
points | 7 | eight |
Of course 10 trials is not a lot of data, and perhaps I need more games to say for sure. Also, lesser players may find similar results, but the threshold should be e.g. 30% not 50%.
Congrats to Monkey for a well-earned win and commiserations to Spider GM.
BTW If anyone knows how to change the word “eight” in the above spreadsheet into a number 8 without the unwanted percentage sign, please let me know 🙂