Monte Hall Hoax

[Hanarchy Montanarchy]

There are no 1/3 odds when you make your second choice either. The incentive to switch in the Monte Hall paradox assumes collapsing 2/3 odds onto one door.

No, you're making the mistake of thinking that the revelation of goats has somehow affected the odds of your initial decision, and it has not. It can only affect the odds of your next decision. If you make no next decision (aka you make no new door picks) then you're holding a door that carries its initial odds. The trickery of revealing goats that the host always knew he could reveal regardless has confused your reasoning. Monte could reveal 1 goat, or (in my example) 98 goats no matter what you pick. The only thing changing the odds of your initial pick is making a new choice.

We agree that the odds change to 1/2 on the second iteration, I believe. The Monte Hall problem states that the odds are different between the two doors. 1/3 for your first choice, 2/3 on the door you didn't choose, so you should switch. I am suggesting that the trick built into the computer models testing the Monty Hall problem is that you could still theoretically choose the revealed goat, so the goat/car ratio remains the same. In reality, Monty is taking away an option, not shuffling the odds of the same n/3 choice.

[Martin Hash]

Won't you guys feel silly once I get famous?

[DBTrek]

We could always meet at Martin's place, and each bring $1000 with us. Play three thousand rounds of Monte Hall at $1 apiece. You two always stick with your initial door picks and I'll always switch.

[Hanarchy Montanarchy]

We could always meet at Martin's place, and each bring $1000 with us. Play three thousand rounds of Monte Hall at $1 apiece. You two always stick with your initial door picks and I'll always switch.

It only works with goats. I can't afford 1000 goats.

[Martin Hash]

We could always meet at Martin's place, and each bring $1000 with us. Play three thousand rounds of Monte Hall at $1 apiece. You two always stick with your initial door picks and I'll always switch.

Dude, that supposed "test;" I went and looked; it was coded by someone who fell for the hoax.

IF THE ODDS ARE 50/50 TODAY, THEY WERE 50/50 A WEEK AGO & IF THE ODDS ARE 50/50 TO SOMEBODY ELSE, YOU HAVE THE SAME & IF THE ODDS OF PICKING THE GOAT ARE 50/50 SO ARE THE ODDS OF PICKING THE CAR

[DBTrek]

We could always meet at Martin's place, and each bring $1000 with us. Play three thousand rounds of Monte Hall at $1 apiece. You two always stick with your initial door picks and I'll always switch.

Dude, that supposed "test;" I went and looked; it was coded by someone who fell for the hoax.

IF THE ODDS ARE 50/50 TODAY, THEY WERE 50/50 A WEEK AGO & IF THE ODDS ARE 50/50 TO SOMEBODY ELSE, YOU HAVE THE SAME & IF THE ODDS OF PICKING THE GOAT ARE 50/50 SO ARE THE ODDS OF PICKING THE CAR.

It doesn't even have to be coded, it can be directly mapped (as it was in the early section of the paper):




/shrug

[Martin Hash]

We could always meet at Martin's place, and each bring $1000 with us. Play three thousand rounds of Monte Hall at $1 apiece. You two always stick with your initial door picks and I'll always switch.

Dude, that supposed "test;" I went and looked; it was coded by someone who fell for the hoax.

IF THE ODDS ARE 50/50 TODAY, THEY WERE 50/50 A WEEK AGO & IF THE ODDS ARE 50/50 TO SOMEBODY ELSE, YOU HAVE THE SAME & IF THE ODDS OF PICKING THE GOAT ARE 50/50 SO ARE THE ODDS OF PICKING THE CAR

It doesn't even have to be coded, it can be directly mapped (as it was in the early section of the paper):




/shrug

That was a simulation.

[DBTrek]

Looks like a simple decision tree that covers all combinations of possible outcomes.
What did they get wrong?

[Martin Hash]

Looks like a simple decision tree that covers all combinations of possible outcomes.
What did they get wrong?

Dude, it's a good illusion. What? You've never seen anybody exploit the foibles of statistic before?

[Hanarchy Montanarchy]

Looks like a simple decision tree that covers all combinations of possible outcomes.
What did they get wrong?

"Monty opens 2 or 3"

This collapses two loss conditions over multiple iterations of the game into 1. And it happens three times. There aren't 9 possibilities with 6 victory conditions for switchers and 3 for stickers. There are 12 possibilities evenly split between switchers and stickers.