What works for me is scaling the number of doors to 100 (or more). Now the initial guess is correct only 1% of the time, while switching is 99%. Simulated many

What works for me is scaling the number of doors to 100 (or more). Now the initial guess is correct only 1% of the time, while switching is 99%. Simulated many more variations at https://simonduff.net/monty_hall/ https://simonduff.net/monty_hall/