0 reactions
tI solved Guess Who 100% and Mark Rober was Wrong about the Optimal Strategy
**Don't use this against your kids :) This is just for fun! But ABSOLUTELY use it against your cousins!**
This all started when I was watching Mark Rober's video on how to win Guess Who, the MYSTERY board game, 96% of the time.
His video is here: https://www.youtube.com/watch?v=FRlbNOno5VA
I noticed that some things did not add up. I spent about 3 weeks investigating. Yes, I went overboard. I know. I checked all of the math, I checked the computer program he used, and I found that his strategy is missing some critical end game logic. It also does not really win 96% of the time.
This got my detective brain firing. I had to know what the real best strategy is. I wrote my own computer program to recursively calculate every possible position the game could be in and what moves lead to what odds of winning. I recorded all of my research in my own video here:Why Mark is Wrong and Best Strategy: https://www.youtube.com/watch?v=tzTQ4JcUzxM
And I wrote a computer program to calculate 100% the odds of winning at any time https://github.com/DetectiveInspectorJoeyG/GuessWhoBestMoveI ried to make it as easy as possible to run but if you just want the answer I describe it in the video. The idea is to combine even splits (but perfect splits are not always necessary) with forced end game guessing.
In a perfect game Player 1 wins about 56% of the time and in the mercy rule variant it rises to 66%.
One last thing. I know that the detective LearnYouSomeMaths also figured out Mark Rober was wrong a couple of years ago, however, I found the LearnYouSomeMaths strategy to be wrong as well. I also talk about how it is slightly less optimal in my video. I did not see their work until I finished my own investigation. Still they did figure out a better strategy than Mark so I wanted to credit them here.
**Don't use this against your kids :) This is just for fun! But ABSOLUTELY use it against your cousins!**
This all started when I was watching Mark Rober's video on how to win Guess Who, the MYSTERY board game, 96% of the time.
His video is here: https://www.youtube.com/watch?v=FRlbNOno5VA
I noticed that some things did not add up. I spent about 3 weeks investigating. Yes, I went overboard. I know. I checked all of the math, I checked the computer program he used, and I found that his strategy is missing some critical end game logic. It also does not really win 96% of the time.
This got my detective brain firing. I had to know what the real best strategy is. I wrote my own computer program to recursively calculate every possible position the game could be in and what moves lead to what odds of winning. I recorded all of my research in my own video here:Why Mark is Wrong and Best Strategy: https://www.youtube.com/watch?v=tzTQ4JcUzxM
And I wrote a computer program to calculate 100% the odds of winning at any time https://github.com/DetectiveInspectorJoeyG/GuessWhoBestMoveI ried to make it as easy as possible to run but if you just want the answer I describe it in the video. The idea is to combine even splits (but perfect splits are not always necessary) with forced end game guessing.
In a perfect game Player 1 wins about 56% of the time and in the mercy rule variant it rises to 66%.
One last thing. I know that the detective LearnYouSomeMaths also figured out Mark Rober was wrong a couple of years ago, however, I found the LearnYouSomeMaths strategy to be wrong as well. I also talk about how it is slightly less optimal in my video. I did not see their work until I finished my own investigation. Still they did figure out a better strategy than Mark so I wanted to credit them here.
| 0 mlx | |
| 63 | |
| 0 |