• Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.

Okamiden

Member
Oct 25, 2017
351
The original problem, for those unfamiliar:


The variant:

You have 100 closed doors in front of you. Only one door has a grand prize in it. $10,000 and a limited edition version of whatever your favorite game is. The other doors have N-Gages.

The game is in two phases. You pick one door at random. It stays closed. No peeking.

Then, the host opens 98 doors at random, but never opening yours. He doesn't know what's behind the doors he opens. If he opens the prize door, oh no! You lose. Unless you want the N-Gage.

But on this particular game you're in, by sheer luck, he manages to open 98 doors without a single one being the prize door. He picked at random, so that's pretty lucky.

There are now two doors left. You get an ultimate choice. You can now decide to switch doors with the other door left, or keep yours. What do you decide, and why?

Yeah, I was bored.

I expect people to solve this quickly so maybe consider using spoilers
 

jdmc13

Member
Mar 14, 2019
2,885
I switch.

1% chance I picked the correct door the first time which means there is a 99% chance it is in the only other remaining door.
 

fr0st

Member
Oct 28, 2017
4,485
Pretty sure you always switch since the chance of you getting the prize is extremely high
 

Chopchop

Member
Oct 25, 2017
14,171
Isn't that
just 50/50 then? In the original problem, the host knows which door has the prize. If the host doesn't know, then it's just pure luck, right
?
 

RecRoulette

One Winged Slayer
Member
Oct 25, 2017
26,044
It's just a 50/50 at that point since he doesn't know?

Initial thought was "always switch" but in this case it's always ngage.
 

Platy

Member
Oct 25, 2017
27,636
Brazil
i stay on my own door because the retro market is bordering on saturation and overpricing and considering how low is my country's money compared to the dollar, if I sell the n-gage on the ebay i will get a pretty big moeny
 

WedgeX

Member
Oct 27, 2017
13,168
If I had the time to rebuild my Monte hall problem worksheet in excel, I'd tell you.
 

Stalker

The Fallen
Oct 25, 2017
6,726
You switch. The probability you got the correct door out of 1 hundred is extremely low when all the doors have been eliminated (and the game is still going) the higher probability is the door you didn't pick.
 

thefro

Member
Oct 25, 2017
5,996
It's still the same solution as the regular version of the problem.

Yep, and the original problem is silly because it assumes the host always does the same thing or is acting in good faith, which doesn't fit the actual "Let's Make a Deal" show.

It's basically the original version of one of those shitty Facebook memes where they have a problem with confusing instructions.
 

Mr.Awesome

Banned
Nov 4, 2017
3,077
The Monty hall problem is driven off the fact that they are giving you an incorrect door.

I dont see that in the op so it's just 50/50.
 

Border

Banned
Oct 25, 2017
14,859
I will probably go to my grave never really understanding the logic behind this problem.

The idea is that my first choice only had a 1% probability of being correct. But if I switch after all the reveals, that choice would have a 50% probability of being correct. But why doesn't my original choice also have the same 50% probability at that point? Isn't "the decision not to switch" a choice that is independent of my original selection?
 
OP
OP
Okamiden

Okamiden

Member
Oct 25, 2017
351
Yep, and the original problem is silly because it assumes the host always does the same thing or is acting in good faith, which doesn't fit the actual "Let's Make a Deal" show.

It's basically the original version of one of those shitty Facebook memes where they have a problem with confusing instructions.

Well in this case, the host is picking completely at random. So there's no bad faith.
 

Stalker

The Fallen
Oct 25, 2017
6,726
Giving you an incorrect door? I don't understand what you mean.

In the monty hall problem you a given a door by the host that doesn't have the car.

3 doors

you pick one and the host gives you one as proof it doesn't have a car. The host knows so he didn't pick the car one. You had a lucky chance out of 3 doors but now your being told it's not 1 of the doors your odds increase if you switch
 

Burly

Member
Oct 25, 2017
5,070
In this situation I don't think it matters. The original problem always had the addendum that the host would never open the grand prize door, it would always be "other" doors. If both picks are truly random, then I don't think it matters.
 

Stinkles

Banned
Oct 25, 2017
20,459
Trick question -- you sidetalk to open all the doors simultaneously.

True Story I once had an n-gage and every available game.
 

Fugu

Member
Oct 26, 2017
2,730
The odds here are 50/50 so it doesn't matter if you switch or not. The host, by pure fluke, opened nothing but ngage doors.

What you did was took the "easy Monty Hall problem explanation" and changed a critical element to make the original reasoning behind switching no longer true. In that problem there's a billion minus one doors with goats behind them and one with a car, so the host offers to open all doors but two for you, with the key difference being that the host isn't allowed to open the prize door. The reason switching is advantageous is because it leverages the host's knowledge against him. To not switch is to take the one in a billion odds that you nailed the prize right at the outset.
 

Border

Banned
Oct 25, 2017
14,859
Yeah, the original problem is kinda dependent on the fact that they are intentionally revealing all the "Losing" doors and only leaving you with your original choice + 1 alternative.

In this variant, it's merely luck or happenstance that the winning door was not revealed. It's unlikely that you would be able to pick the Winner out of 100 choices, but it's also highly unlikely that they would be able to reveal 98 doors and never hit the winner.
 

ABIC

Banned
Nov 19, 2017
1,170
I will probably go to my grave never really understanding the logic behind this problem.

The idea is that my first choice only had a 1% probability of being correct. But if I switch after all the reveals, that choice would have a 50% probability of being correct. But why doesn't my original choice also have the same 50% probability at that point? Isn't "the decision not to switch" a choice that is independent of my original selection?

Yeah, I think you're mostly there.

Your first choice has a 1% probability of being correct also means there's a 99% probability the correct door is some other door.

One other way to think about it is to run this simulation hundreds of times -- imagine if Bob and Paul had to play this game 500 times, for example.

You see Bob and Paul pick a door, and then 98 other doors disappear leaving one.
Bob never switches.
Paul always switches.
They do this again and again and again and again for 500x.

It becomes more logical to see that Paul will win more often than Bob. In fact, if you ran this simulation through code, that's exactly what would happen.
 

HommePomme

Member
Oct 30, 2017
1,052
Yeah, the original problem is kinda dependent on the fact that they are intentionally revealing all the "Losing" doors and only leaving you with your original choice + 1 alternative.

In this variant, it's merely luck or happenstance that the winning door was not revealed. It's unlikely that you would be able to pick the Winner out of 100 choices, but it's also highly unlikely that they would be able to reveal 98 doors and never hit the winner.

Yeah I think it doesn't matter that it''s "lucky" that they opened 98 doors and never hit the winner, you need to switch to take advantage of that luck
 

ItIsOkBro

Happy New Year!!
The Fallen
Oct 25, 2017
9,476
think of it from the perspective of the host. there are two doors left and he's just yolo opening doors. let's say he could open your door or the other door. he wouldn't have different odds for opening your door. it's 50/50 for him and 50/50 for you if to switch.
 

Parthenios

The Fallen
Oct 28, 2017
13,602
I still believe the correct answer is to switch.

When you picked the first time, your odds of choosing correctly were 1/100, and those odds never change regardless of how many doors are revealed. If you switch, your odds of having picked correctly change to 1/2.

In the classinc Monty Hall, switch actually is correct 2/3 of the time, not 1/2, so maybe that specific part is changed based on the host knowing or not knowing, but the correct choice is switch regardless.
 

Phatmanny

Member
Nov 14, 2017
209
The fact that you said he chose at random means they both have an equal chance of containing the prize 1/100

it's because in the original problem he deliberately opens a losing door is what modifies the odds.
 

Phatmanny

Member
Nov 14, 2017
209
No new information is added by opening doors at random. While if he chose 98 "losing" doors deliberately then it modifies the odds
 

Fugu

Member
Oct 26, 2017
2,730
Yeah, I think you're mostly there.

Your first choice has a 1% probability of being correct also means there's a 99% probability the correct door is some other door.

One other way to think about it is to run this simulation hundreds of times -- imagine if Bob and Paul had to play this game 500 times, for example.

You see Bob and Paul pick a door, and then 98 other doors disappear leaving one.
Bob never switches.
Paul always switches.
They do this again and again and again and again for 500x.

It becomes more logical to see that Paul will win more often than Bob. In fact, if you ran this simulation through code, that's exactly what would happen.
This logic is applicable to the real Monty Hall problem, but not the question the OP is posing.
 

LakeEarth

Member
Oct 27, 2017
8,174
Ontario
I will probably go to my grave never really understanding the logic behind this problem.

The idea is that my first choice only had a 1% probability of being correct. But if I switch after all the reveals, that choice would have a 50% probability of being correct. But why doesn't my original choice also have the same 50% probability at that point? Isn't "the decision not to switch" a choice that is independent of my original selection?
The OP example is messed up because the host randomly opened the doors, and just so happened to not open the one with the prize. In the real version of the Monty Hall problem, the host knows what door has the prize and purposely never opens it when opening all other doors.

So with the real Monty Hall problem, it's not 50/50 because it isn't REALLY a choice about switching from your door to the other unopened door. It's a choice between the door you picked, and the other 99 doors. Him opening the 98 doors with no prize before allowing you to switch is masking this fact.

If you still have a problem imagining this, expand it. There are a million doors now. You pick one of them. Monty Hall eliminates 999,998 of the doors he knows don't have the prize in them. Is switching to that lone door he didn't open still a 50/50 shot to win now?
 

ss_lemonade

Member
Oct 27, 2017
6,648
Isn't this just the exact same problem? You still have a significantly higher chance of getting it right if you switch
 

ABIC

Banned
Nov 19, 2017
1,170
This logic is applicable to the real Monty Hall problem, but not the question the OP is posing.

Yeah, you're right, but I was mainly replying to a poster not the OP.

The OP question is more interesting with the host being able to open the prize door.
 

Ellyshia

Member
Oct 27, 2017
451
I went through this problem years ago cause of deal or no deal. Ultimately, because it's random chance since the host also doesn't know, the probability is 50/50. The easiest way to see this is to draw out the tree of possibilities with 3 prizes.

1 - You pick the winner(A) host picks a loser(B) = Don't switch
2 - You pick the winner(A) host picks a loser(C) = Don't switch
3 - You pick a loser (B), host picks a loser(C) = Switch
4 - You pick a loser (B), host picks a winner(A) = Doesn't happen in this reality
5 - You pick a loser (C), host picks a loser(B) = Switch
6 - You pick a loser (C), host picks a winner(A) = Doesn't happen in this reality

So basically there are 6 possible outcomes, but the premise already removes 2 of those outcomes. Of the 4 remaining outcomes, 2 of them you want to switch, 2 of them you don't. So 50% chance. It's a coin flip.
 
Last edited:

Nappuccino

The Fallen
Oct 25, 2017
13,007
The fact that you said he chose at random means they both have an equal chance of containing the prize 1/100

it's because in the original problem he deliberately opens a losing door is what modifies the odds.
Aren't you just describing why people fall for the Monty Hall problem? Intentionally revealed or not, you've still seen them only open losing doors.
 

Sybil

Avenger
Oct 27, 2017
1,642
The real problem is the fact this game show managed bought and kept 99 N-Gages. Nothing good can happen from hoarding all of those. We have to stop them, 1 N-Gage at a time
 

Lumination

Member
Oct 26, 2017
12,464
It doesn't matter because the chance is the same. The original problem leverages the fact that the host knowingly avoids the prize. So you are essentially making a 1% guess (with 100 doors) vs a 99% guess. Since the host doesn't know and is just very unlucky, the chance of either door containing the prize is both the original 1/100, or 50/50 when it's down to the last two.
 

Mezentine

Member
Oct 25, 2017
9,969
This actually makes the Monty Hall Problem much easier to understand, oddly enough. The math of why it works isn't quite as counterintuitive, and as someone else says this makes it clearer that what's happening when you switch is you're taking advantage of the other person's "luck" (in a neutral sense) by not revealing the prize with their picks.