Marilyn vos Savant and the Monty Hall Problem (2015)

https://priceonomics.com/the-time-everyone-corrected-the-worlds-smartest/

by hecubus • 2 years ago

101 300 2 years ago

dang• 2 years ago

The Time Everyone “Corrected” the World’s Smartest Woman (2015) - https://news.ycombinator.com/item?id=27092258 - May 2021 (303 comments)

The Time Everyone “Corrected” the World’s Smartest Woman (2015) - https://news.ycombinator.com/item?id=17686279 - Aug 2018 (112 comments)

Marilyn vos Savant and the Monty Hall Problem - https://news.ycombinator.com/item?id=9078660 - Feb 2015 (142 comments)

Game Show Problem (1990) - https://news.ycombinator.com/item?id=6386289 - Sept 2013 (31 comments)

swayvil3• 2 years ago

Am I the only person who wonders how "Marilyn Vos Savant" also happens to be the smartest person?

If you didn't notice, "Savant" means smart.

So is it just a coincidence? Karma? The pressure of carrying such a name drove her to smartness?

Jimmc4143• 2 years ago

Perhaps Bill Ackman was right about impact of a person’s name. I doubt it, but it’s an interesting observation.

paulpauper• 2 years ago

like the surname "mason" or "farmer"

ksherlock• 2 years ago

Per Wikipedia:

Marilyn vos Savant was born Marilyn Mach on August 11, 1946, in St. Louis, Missouri, to parents Joseph Mach and Marina vos Savant.

... so only 50% coincidence.

bemusedthrow7524• 2 years ago

Fittingly it looks like informed choice.

kevinmchugh• 2 years ago

The "smartest person in the world" tag is a marketing push. It works with her name. It wouldn't work as well if she was named Marilyn Smith.

WarOnPrivacy3• 2 years ago

The "smartest person in the world" tag is a marketing push.

It reflects some broad understandings of the 1980s and what we valued.

tdhoot• 2 years ago

Nominative Determinisn: https://en.wikipedia.org/wiki/Nominative_determinism

dghughes• 2 years ago

She said this:.

Marva Watkins in Chicago, Illinois, writes: What is your real last name?

Marilyn responds: It's "vos Savant," which was my mother's maiden name and is my own legal name. (My mother's maiden name is just as real as my father's boyhood name.) My maternal grandfather was Giuseppe vos Savant. Coincidentally, he married a woman named Maria Savant, my maternal grandmother. Growing up, I never thought about "Savant" being a word, too. I'm sure that people with the name "Miller" don't think about it, either! By the way, my mother changed her surname to my father's surname when she got married, which was traditional at the time. I use her name, and my brother uses his name.

https://parade.com/418148/marilynvossavant/marilyns-surname/

tromp3• 2 years ago

Imagine that you’re on a television game show and the host presents you with three closed doors. Behind one of them, sits a sparkling, brand-new Lincoln Continental; behind the other two, are smelly old goats. The host implores you to pick a door, and you select door #1. Then, the host, who is well-aware of what’s going on behind the scenes, opens door #3, revealing one of the goats.
“Now,” he says, turning toward you, “do you want to keep door #1, or do you want to switch to door #2?”

There are some implied assumptions. You need to know that the car is randomly placed (or that your door choice was made at random), and that the host always has to show you an empty (ok, goat) door before offering you to switch.

Under those assumptions, the winning odds of the switching strategy equal the odds of your initial pick NOT being the car.

paulcole5• 2 years ago

You need to know that the car is randomly placed

You don't need to know this.

and that the host always has to show you an empty (ok, goat) door before offering you to switch

You know this through the line, "Then, the host, who is well-aware of what’s going on behind the scenes." If you can't understand that a game show host won't spoil the game they're hosting, that's on you and a good opportunity to pick up common sense.

tromp3• 2 years ago

It might be that he only does the reveal-and-switch offer if your original pick was the car.

AnimalMuppet• 2 years ago

Under those assumptions, the winning odds of the switching strategy equal the odds of your initial pick NOT being the car.

Only if there are three doors - or rather, only if the host eliminates all but two doors, one of which is the one you chose.

If there are, say, four doors, and the host eliminates one, your odds of success on switching are not 3/4.

pjdemers• 2 years ago

I got the problem wrong because I didn't know he never opens the door picked nor a door with a prize behind it. That's because I hadn't seen the show since I was about 5 years old, because that's about the time I stopped watching daytime TV.

bemusedthrow7524• 2 years ago

That is also in the problem description. Again, it doesn't matter whether he does or doesn't ever do it (though no game show host would ruin the game).

It is the described problem that matters.

There was so, so much writing like this at the time, people saying "well, if...".

bemusedthrow7524• 2 years ago

This is something I remember from the time.

I could not grasp the probability maths (still shit at it, sorry Ms Von Savant) but I could write a program to experimentally demonstrate it, so I did.

In QBasic. Because that's what we had and we were happy.

nickpsecurity2• 2 years ago

That’s exactly what I did. In FreeBASIC.

The way the problem works started coming together in my head as I coded it. The code’s structure reflected the problem’s structure. I realized I previously misunderstood something about the problem statement. Later, I read someone saying many mathematicians got this problem wrong because of how it’s worded.

So, it may be challenging, not because probability was challenging, but because it was unclearly specified. That’s the cause of many software errors, too.

bemusedthrow7524• 2 years ago

It's not unclearly specified. It's very tight.

People who get it wrong are usually projecting some implied understanding onto it, which is why they get it wrong.

Which is the point of the problem. It's designed to reveal this tendency in analytical thinking leading to unexpected outcomes. You get it wrong, you're gobsmacked, you understand why, you gain some enlightenment from it. It's fun to be wrong in ways you later understand. (This is one of the thrills of programming: that moment during debugging when you shriek with joy: "YES, IT BROKE!")

The fact that it also exposes the way some overly confident people lash out with anger that exposes other biases is the point of the article, I think.

Karellen2• 2 years ago

It was the 100-door variant (described in the article) that was my key to properly understanding the result.

bemusedthrow7524• 2 years ago

Yeah. I think my program code was ten doors or something.

It was years before I understood the actual lesson here -- both the probability lesson and also the way it hints at the fundamental truth that in life, binary choices may have been weighted in ways you don't understand, perhaps by the people asking you to make the choice.

zuminator• 2 years ago

Ha, I did the same thing after first reading about the problem. I understood the probabilities from her explanation, but it felt more indisputable after seeing the results play out after a thousand tries. Guess you could call it a "Monty" Carlo simulation.

jawns• 2 years ago

"You are the goat!" one reader wrote.

Turns out, instead, she's the G.O.A.T.

WarOnPrivacy3• 2 years ago

MvS was a pretty good role model for me. Her influence helped me define positive human attributes and then distinguish them from potential advantages.

My character hinges on that understanding now.

Jimmc4143• 2 years ago

I never understood this one. Not choosing a new door seems like an implied 50/50 choice to keep your original selection.

edit: I know it works out on paper, but if I put myself in the situation, I know that if I decided not to choose again that is a choice and I am re-choosing the same door.

CoastalCoder• 2 years ago

I think that's what trips up most people.

I found working through it to be really rewarding, you might too.

Spoiler alert: consider how Monty's choices are constrained by what he knows, and how the player can capitalize on that.

anonzzzies• 2 years ago

Yes, although I wasn't very good at statistics in university (I found it too vague and non-math-y...), this one I never had much problems with; the fact that Monty is guaranteed to open a door with a goat and you already selected a door moved 1/3 to the door we both didn't select. You have the 1/3rd that you DID select the car, so the other he DIDN't select must have 2/3rd.

xorcist• 2 years ago

It is, strictly as written. The wording, both in the original article and the many rewrites, does not fully explain the problem as intended. The show host not only knows where the prize is, but actively chooses never to reveal the door where the prize is. The reveal is not random. The reveal is a tell where the prize isn't!

Worded more carefully, the answer is more intuitive, and this is probably why there was controversy surrounding it.

The naming is taken from an actual game show, which perhaps confusingly seems to have been a third variant on the same problem, using more steps and trick monetary incentives to make fore more exciting viewing.

grumbel2• 2 years ago

The remaining doors are not random. The one you picked still has a 1/3 chance of winning, that doesn't change with opening another door. However switching doors is essentially the same as picking all the other doors at once, since the door with a goat is not opened, thus the 2/3.

Imagine the thing with 100 doors and imagine the host isn't opening any doors, but just telling you what's behind them and it becomes pretty obvious.

Jimmc4143• 2 years ago

The one you picked still has a 1/3 chance of winning

If there are only two choices remaining, I guess I don't understand the assertion that there is still a 1/3 chance of winning.

Suppose you were new to the contest and were allowed to put money down on the choice at the point that there were two doors left.

Then you repeated the bet every new contestant.

Would one door would be more profitable than the other?

mayd• 2 years ago

Case 1: Suppose there are n doors and Monty does not know what is behind any of the doors. You choose one door at random. Then Monty opens (n-2) of the remaining doors AT RANDOM until there is only one other door left. By chance none of the doors he opened had the car behind it. Then he asks you if you want to switch. Should you switch or not?

Case 2: Suppose there are n doors and Monty knows what is behind every one of the doors. You choose one door at random. Monty deliberately opens (n-2) of the remaining doors from the left to right, skipping the door with the car. Then he asks you if you want to switch. Should you switch or not?

Whether n=3 or n=100, it seems to me that it does not matter whether Monty Hall has complete knowledge or zero knowledge of the location of the car. You are required to make a choice under the condition where there is only one other unopened door and all the other doors did not reveal a car. The player's original choice was correct with probability 1/n and the probability of the complementary event must be (n-1)/n. The player's strategy of switching will result in a win with probability of (n-1)/n.

habinero2• 2 years ago

The difference is MH knows what's behind the doors and deliberately shows you a goat.

It's not 50/50 because the choice wasn't random.

gerad• 2 years ago

The intuitive way for me to understand the Monte Hall problem is to pretended there are 1000 doors. You pick 1. There’s a 1 in 1000 chance you get it right. The host then opens 998 doors that don’t have the prize. Do you keep your original or do you switch? Are the odds 50:50?

Karellen2• 2 years ago

What improvement does your 1000-door variant have over the 100-door variant described in the article?

michaelcampbell3• 2 years ago

Bigger numbers make it easier to convince yourself if skeptical.

The problem stated with 3 doors is hard, with 5 is still not quite there for many. 100, 100, 10_000_000... becomes almost absurd.

tshaddox4• 2 years ago

The most common unintended interpretation seems to be that Monty Hall is only coincidentally opening a door with a goat behind it. I don't quite understand how people arrive at that interpretation, but the 1,000-door variant should make it even more of a stretch to interpret it as Monty Hall coincidentally doing something so extremely improbable.

mithr2• 2 years ago

I think the reason it's so hard to understand is that we try to break the problem down into more digestible problems, and in this case, if you only consider the last stage of the game, it certainly seems like it should be 50:50 odds.

In other words, at the final round, you are presented with two doors. Ignore the fact that you already chose one, and assume you were starting at this point: you enter the game, see two doors, and need to pick the one you think hides a car. So intuitively, your odds of choosing which of two doors has the car are 1/2.

It's hard to really understand why it matters that the previous rounds occurred at all — why it matters how you narrowed the selection to these two doors. I don't think the 1000 door version makes that easier to understand, because the same thing is true there — if you come into a game, see two doors, and are asked to choose between them, it's very hard to understand why it matters whether in a previous round there were 2, 3, 10, or 1000 doors — there are only two now, when you make your choice.

rspeele8• 2 years ago

That's especially tempting since we're used to probability problems often being set up where considering previous state is the sucker's answer. E.g. a fair coin has landed on heads three times in a row, what are the odds of it being tails on the next throw?

polygotdomain• 2 years ago

It's hard to really understand why it matters that the previous rounds occurred at all

It's not hard at all. Humans are not perfectly rational beings. It's not purely about odds, it's about emotions and psychology. In a version where there's no previous selection and it's 50:50, than I pick one and live with it. In the canonical example, the previous selection means that the contestant has already staked their claim, and changing it to the loosing door would have a different emotional response than a simple 50:50 shot with no previous selection. There's a reason why Roulette shows you the last X spins and whether they're odd/even, red/black... because humans make the assumption that the last disconnected data point somehow impacts the current one.

You're also glossing over the fact that there was a 1:3 chance you picked the right door to begin with, and 2:3 chance the "other" door was right. The last round isn't 50:50, as you so claim, because there was prior information

This is the same reason why the 1000 door example helps explain things; because the math is fundamentally the same, yet significantly more imbalanced. We can also think about how we'd feel about our selection in the 1000 door version, which is likely significantly less confidence, and therefore more likelihood of switching. Whether it's 3 doors, or 1000, the math still say switching is optimal, and our psychology and emotions deal with the choice of switching different in each case.

rspeele8• 2 years ago

The explanation that worked for me was:

Suppose after you make your initial choice, instead of opening a door, Monty simply asks if you'd like to switch to BOTH the other two doors, such that you win if the prize is behind EITHER of them.

That switch is intuitively a great deal, giving you 2/3 odds. The only way you can lose is in the 1/3rd case where you already picked a winner. The original scenario is equivalent to this, since by revealing a goat Monty is allowing you to pick "the best of" the two doors.

---

Often people who don't like this problem complain that actually Monty, with his knowledge of the winning door, may be trying to trick you into losing. Maybe he offers this trade conditionally based on whether you've already selected a winner. Of course, this wouldn't support the most common intuitive answer of "it's 50/50", either, making this a bad excuse. If we accept this behavior from Monty -- not stated in the problem -- it becomes a very uninteresting problem based on whether we are dealing with Evil Monty who only offers the trade when it's a bad one (stay wins 100% of the time), Friendly Monty who only offers the trade when it's a good one (switch wins 100% of the time), or somewhere in the middle. It has no analytical answer. So for this to be a solvable problem at all we must assume we are dealing with Fair Monty who offers the trade all the time, or at least, offers it at random times not based on the contestant's initial pick.

mithr2• 2 years ago

I do think this modified scenario makes the solution much clearer. At the same time, I also think that it feels like a significantly-enough different scenario from the original that saying the two scenarios have the same probability then becomes the non-intuitive part. The act of revealing what's behind the second door seems like it should change the probability from 1/3 (one door) vs. 2/3 (one of two doors) to 1/2 (one door of the remaining unopened two) vs. 1/2 (one door of the remaining unopened two, since one was "eliminated").

It's amazing how even seeing the probabilities written out, or running simulations, doesn't really make it easier to truly understand the result.

mcv2• 2 years ago

I think the important part of the problem is the unstated assumptions around it, and I suspect some people may see different assumptions than others.

If Monty always opens a door, and uses his knowledge of which door has the prize to ensure that the door he opens is always empty, then you should switch, because he's providing you with extra information about which door has the prize.

However, if Monty changes it up and sometimes doesn't open a door, or opens the door with the prize, the whole problem changes. If he only opens a door and offers you the chance to switch when you've picked the door with the prize, you should never switch of course. If he opens a random door and might reveal the prize (after which he obviously won't let you switch anymore), switching doesn't change your odds. I think a lot of people see it as one of those two situations.

dekhn8• 2 years ago

a single sentence convinced me: "When Monty opens the door, he reveals information about the state behind the doors".

tromp3• 2 years ago

Your formulation is once again ambiguous. Instead of clearly stating that Monty had to open 998 empty doors, you state that he did that. Which could mean he happened to do that.

Had to vs happened to makes a big difference.

Had he opened them at random and by (extremely small) chance they happened to be empty, the odds are quite different.

mathstuf• 2 years ago

There's a variant where Monty trips and opens one of the doors by accident. Then it is 50/50 again (because he could have tripped into the car door). See also this video where cutting off a question in the middle changes the probability because of the information it provides: https://www.youtube.com/watch?v=bDZieLmya_I

niel• 2 years ago

I'm trying to understand why the odds would be different if Monty opened empty doors by chance versus on purpose.

Does it really change anything?

topaz011• 2 years ago

TFA also presents this intuitive argument.

fortran77• 2 years ago

But this is wrong. The odds become 2/3 in the original problem, not 50:50. Your "intuition" isn't helping you.

topaz011• 2 years ago

I think GP is implying that you would reject 50:50 as absurd in this case (answer that rhetorical question with a "no"), and therefore should also reject it in the original. I agree that was not clear, though.

dynm• 2 years ago

I actually like to back up even further and start with this problem: There are 1000 doors. You pick one. There's a 1 in 1000 chance you get it right. Now, do you want what's behind your door, or what's behind the other 999 doors? https://dynomight.net/2020/09/17/making-the-monty-hall-probl...

mihaic• 2 years ago

I think this is the clearest explanation I've heard of the principle, kudos!

cynix• 2 years ago

Do you really want 998 goats though? It must cost a fortune to transport them home from the studio, finding a space for them to live in, and feed them, etc.

HumblyTossed2• 2 years ago

The way I think about it, in general, is that Monty, by purposely not opening the correct door, is giving you a huge hint and thus changing the game which changes the odds. Think of it similarly to those people who play the same lotto #s all the time because they think eventually it has to hit. Well, no. It's a new game every time. If the lotto commission retired each combination that won each week, then sure. And this is essentially what Monty is doing. He's retiring some doors and changing the odds.

lelanthran4• 2 years ago

And this is essentially what Monty is doing. He's retiring some doors and changing the odds.

That still leads to the wrong answer, because he is retiring one of the losers, leaving a 50:50 chance of getting a prize.

hinkley• 2 years ago

No explanation worked for me. I did a truth table, and said, "...huh."

You don't need to model all three door choices. Just say you pick Door A, not Door 1. The first door you didn't pick is B, and C is the other door. Then map the prize behind each of 3 doors, and whether you stay or change your answer.

Then count the number of successes for change versus stay. Spoilers: It's 3/6 vs 2/6.

shadowgovt2• 2 years ago

My experience was I coded a simulation in Python and got 50/50.

... this revealed a major flaw in my intuition that helped me to grasp it. I had subconsciously been assuming Monty sometimes opens the door revealing the prize instead of always opening a goat door. When I coded that into the simulation, of course it broke (because the player was now playing foolishly, looking at the prize and intentionally choosing the closed door that definitely still had a goat behind it).

lelanthran4• 2 years ago

I drew a decision tree once to show all possible states, with the 1/3 and 2/3 probabilities along each edge, which finally made it click for the coworkers I was trying to convince.

In the tree it becomes blindingly obvious that the probabilities are.

full_monty3• 2 years ago

I never understood why the 1000 door version was more intuitive. Here's how I understand it.

There's 1/3 chance of picking the car. 1/3 of the time you pick the car, switch, and lose. 2/3 of the time you pick a goat, switch, and win. Why? Because 2/3 of the time you picked a goat, Monty shows you the other goat, so if you switch you definitely get the car.

michaelcampbell3• 2 years ago

username checks out =D

I never liked the 10^x other doors version either. My thought was, what say me and a friend play a the same time, and pick the same initial door? If I NEVER change, and he ALWAYS does... I'm still going to win 1/3 of the time, and if Monty always shows the goat, ONE OF US has to win, so if I'm 1/3 by not switching, he HAS to be 2/3 by switching.

michaelcampbell3• 2 years ago

When first confronted with this, I wrote a simulation in GW-Basic on a PC with an orange monochrome monitor and ran it a bunch of times to convince myself switching was the right thing to do, statistically.

I came up with another way to visualize it that made more sense than increasing the number of doors, but it requires that one accept the fact that your chances of winning are 1/3 if you don't switch, and you know Monty always shows a losing door.

andrewla4• 2 years ago

Did the host have to open 998 doors, or could he have chosen to just let you open your selected door? If he's "out to get you" then you get very little information here, because he could have just let you open your door. The only reason he'd open the other doors is to try to get you to switch, which means that it's pretty close to 50-50.

DerekL2• 2 years ago

*Monty Hall

mhb2• 2 years ago

Yes. Just like in TFA.

niklasbuschmann• 2 years ago

Suppose you're on a game show, and you're given the choice of three doors: Behind one door is a car; behind the others, goats. You pick a door, say No. 1, and the host, who knows what's behind the doors, opens another door, say No. 3, which has a goat. He then says to you, "Do you want to pick door No. 2?" Is it to your advantage to switch your choice?

Without additional assumptions this original wording is not clear, it could be that the host always shows a goat, then one should switch, or the host shows something at random and we just look at the cases where he shows a goat, then it doesn't matter.

bemusedthrow7524• 2 years ago

Without additional assumptions this original wording is not clear

The only problem you are being asked to solve is the one described only in the words of the problem.

That's actually what this is all about. This is a learning exercise, not just about probability, but about comprehension of what the problem as described exactly says.

That's what makes it so enormously valuable.

jncfhnb21• 2 years ago

If you’re literally following the words then you cannot solve the problem. There is no information gain.

kergonath2• 2 years ago

it could be that the host always shows a goat

That’s what the sentence says: “and the host, who knows what's behind the doors, opens another door […] which has a goat”.

or the host shows something at random and we just look at the cases where he shows a goat

Nothing says that the host opens a door at random. Quite the contrary, we know that the host has a perfect knowledge of the situation.

jncfhnb21• 2 years ago

Incorrect. The host doesn’t need to choose randomly. His purposeful behavior could also be random. The only way you know the switching to be optimal is if it specifically says he picks a goat door to open.

If it says he opens a door, and it has a goat, then you’ve learned nothing about the remaining two.

Workaccount2• 2 years ago

Total aside here, but this article is a good example of the problem, so I just want to point it out.

People go digging, usually pretty hard, to find insults to use as capital in whatever point they are trying to make. In this case, yes, she received a lot of criticism and ad-hominem attacks for her correct answer. But she wasn't really criticized for being a woman, despite it being repeated multiple times in the article.

Look at the given list of people who wrote her angry letters and what they wrote:

     Ph.D
     Ph.D
     Ph.D
     Ph.D
     Ph.D
     Non-doctorate professor?
     Don Rivers from Sunriver Oregon who made a sexist remark.

Why is a comment from Don Rivers given so much weight? Who the fuck is Don Rivers from the town of 1,400 people in central Oregon? Man or woman, if you receive 10,000 hate letters from the general population, you can surely find whatever criticism you want if you dig hard enough. But why pay attention or give any weight to these absolute nobodies?

glimshe• 2 years ago

Because there would be no article and thus no clicks and no ad money without the comment from Don Rivers from central Oregon.

bemusedthrow7524• 2 years ago

But she wasn't really criticized for being a woman

Nobody is saying she was criticised for being a woman, though. What they are observing is that it was the basis of and the tenor of attacks on her character, intelligence, certainty etc.

It was nasty and it really happened, is all I'm saying. Are we better? I hope so. But I don't know that I think so.

desert_rue• 2 years ago

Exactly. Would the tone of the letters been different if the column was said to be have written by a man? Likely they would be different.

This is what makes discussions about discrimination so hard- it can be very subtle. And it can include situations like this one, where we don’t have a test environment where we can compare responses to the article written by a woman to the one written by a man.

ajross• 2 years ago

Who the fuck is Don Rivers from the town of 1,400 people in central Oregon?

FWIW, Sunriver is a resort community outside of Bend. It's not the middle of nowhere, it's filled with tech industry vacationers and wealthy retirees.

btilly5• 2 years ago

As stated, the answer is indeterminate. To demonstrate, consider these scenarios.

- Savant's Monty: The host knows where the car is, and wants to prolong the game. Switching wins 2/3 of the time.

- Ignorant Monty: The host has no idea where the car is. Revealing the goat was luck and your odds remain 50-50.

- Malicious Monty: The host knows where the car is, and wants you to lose. The fact that the host didn't reveal the car means that you have the right door. You'll have no chance of winning if you switch.

Nowhere in the problem are Monty's knowledge and motivation stated. When explained, most go along with the assumptions that Savant's argument introduces. When the actual Monty was asked, it turned out that he was usually ignorant. And the malicious scenario demonstrates how different the answer can be.

The fact that all three answers are consistent with the statement of the problem means that it is indeterminate.

saalweachter2• 2 years ago

_Let's Make A Deal_ was on the air for twenty (thirty?) years at the time of this controversy.

There was sufficient evidence for Savant's Monty -- the car was revealed 0% of the time, not 33 or 66% of the time.

orblivion2• 2 years ago

I've never seen the show; I'm curious, did he always open one of the doors?

orblivion2• 2 years ago

So long as you describe a Malicious Monty that might not have shown you the door, you could also have a Benevolent Monty who won't show you the door if you've gotten it right.

Though I'm not sure I've ever considered the Malicious Monty, only the other two.

KrazyButTrue2• 2 years ago

Nowhere in the problem are Monty's knowledge and motivation stated.

"and the host, who knows what’s behind the doors, opens another door [..] which has a goat."

The question is clear: The host (1) knows what is behind each door and (2) always shows a goat. It's clearly a determinate problem.

jncfhnb21• 2 years ago

Doesn’t matter. If you don’t have a guarantee that he will always open a goat door, his opening of the goat door doesn’t give you information.

btilly5• 2 years ago

(Rereads the article.)

You're right. The problem is stated multiple times in the article. As is usual, most of the statements do not address knowledge. But Marilyn's own statement did. However she did not address motivation.

Monty himself claims that his motivation varied depending on his mood. The actual game had more complications. And his statement about the real game was, "My only advice is, if you can get me to offer you $5,000 not to open the door, take the money and go home."

You can find the article I got that quote from at https://www.nytimes.com/1991/07/21/us/behind-monty-hall-s-do....

pessimizer• 2 years ago

Ignorant Monty can't exist. Monty is the host of a game show, not a contestant. He's not trying to win a prize, and if he opens a prize door, it's a production failure. Whether he knows or not, somebody told him which door to open, and that door does not have the prize behind it.

Malicious Monty, however, can exist. He only ever opens a second door when you've picked correctly, but when you haven't simply accepts your choice.

The problem was stated correctly and clearly, though, as long as you assume that Monty is not a person, has no motivation at all, and as part of a math problem simply wants to reveal a goat after you've made a choice and has the means to do it. That's Savant's Monty: the Monty that doesn't add significant features to the question.

btilly5• 2 years ago

Ignorant Monty absolutely can exist. The person standing in front of the audience is not necessarily the person who arranged the prize. And if he's not, he knows nothing more than the audience.

I was, however, wrong that Monty was ignorant. By his own statement, he always knew the answer. But whether or not he was on your side depended on his mood. And he applied various forms of pressure to get you to make the right, or wrong, decision.

Therefore the problem was not stated correctly. A correct statement needs to not just state what you witness, but the counterfactuals about possibilities that you didn't witness.

rmacqueen2• 2 years ago

I swear this problem only ever causes confusion because the original wording is ambiguous about what the host's motivations are. Once you realize the host is trying to get you to lose and knows what's behind each door and must open a door, then it's clearer that you should switch.

kergonath2• 2 years ago

The host’s motives don’t matter at all. If the first choice is a goat, then the host does not have any choice about the door to open, so their intent is irrelevant. If the first choice is a goat, then the outcome is the same regardless of the door the host chooses.

The host does not really have any agency here.

The rules are intended to confuse people and trick them into thinking both remaining doors have the same probability. I can see how this would favour the house on average because once people committed and chose a door, they are invested and would tend not to change their mind. But the host’s actions have no effect. And even then, this might be an intuition, but does not provide any understanding as to what exactly happens.

rmacqueen2• 2 years ago

I'm not sure if you would call it 'agency' but the host is following it's own specific door-picking policy - a policy which will never result in a car being revealed. It is precisely this policy and no other which makes the contestant switch the correct move. But the original wording doesn't say that - it just says the host opens a door and it happens to have a goat behind it.

hardlianotion2• 2 years ago

Judging by the hazing Marilyn vos Savant received for her resolution of the Monty Hall problem, it seems like Twitter's preferred style of commentary isn't all that new.

pohl• 2 years ago

But we already knew that because usenet.

hardlianotion2• 2 years ago

Fair.

bilsbie• 2 years ago

Has anyone else spent hours on understanding this and just accepted they’ll never accept it?

I completely get all the explanations but it just feels too weird.

bemusedthrow7524• 2 years ago

No. I am not a mathematician but I got it once I understood that this is not drawing coloured balls randomly from a bag (which is how all my school probability problems seemed to go).

That is:

- the setup of the system matters.

- The state of the system at the point of the decision to switch matters.

- The choices don’t get re-randomised.

So the probabilities assigned to the original choice (and the remaining alternative) still count.

If the host had closed a curtain over the stage and randomised the remaining doors, then it would be 50:50.

But he didn’t. So you’re still in the probabilities of the original choice.

One of the goats has been removed. The car and one goat remain: you know this for sure.

You are being offered a door knowing that behind it must, necessarily, be the opposite of your original choice, and the probabilities have not been reset.

If you originally picked the goat, that door absolutely has a car behind it. And there's a 2/3 chance you picked the goat originally. So by inference there's a 2/3 chance the door has a car behind it. You should switch.

I didn’t get it until I had written a simulation to see it for myself though!

full_monty3• 2 years ago

I think I have a simple explanation. There's 1/3 chance of picking the car. 1/3 of the time you pick the car, switch and lose. 2/3 of the time you pick a goat, switch and win. Why? Because 2/3 of the time you picked a goat, Monty shows you the other goat, so if you switch you definitely get the car.

bemusedthrow7524• 2 years ago

Yes but in fact Monty always shows you a goat. There's always a goat for him to choose regardless of your choice, he knows where it is, and he's not going show you the car or the goat you already picked.

Showing you the goat is the event that, as you say, guarantees that the other door has the opposite of your original choice behind it. Because nobody closed the curtain to shuffle the choices.

One of the things I think people struggle with -- and I struggle with -- is that probability isn't about hypothesising about a single event that happened and how it might have happened. It's about encapsulating all the possible ways a single specified scenario can play out in a single expression.

Monty shows you a goat this time, but this means Monty always shows you a goat. There's no scenario where he is unable to show you a goat. Just like you always only pick one door. And there's always only two goats and one car.

(Full marks for username choice)

Sohcahtoa82• 2 years ago

Based on other comments, it seems the people that refuse to accept the answer that it's better to switch and not 50/50 are nitpicking how the game is described.

The idea is supposed to be that Monty always reveals a goat. He does not pick a door at random to reveal before offering the choice to switch. If you pick door A, and the car is in door B, he will reveal door C and ask if you want to switch. If you pick A, the car is in C, then he will open door B and ask if you want to switch.

Expanding it to 100 doors with Monty opening 98 of them and then offering you the opportunity to switch should make it very intuitive. Let's say you pick door 7, and Monty then opens every door except your door and say, door 48, all revealing goats, then it should make you think "Wow...very odd that he didn't reveal door 48...there must be a reason he skipped that specific door".

Make it 1,000,000 doors. You choose door 1. Monty opens every door except number 423,901, and all the doors he opened had goats. You're tired and hungry because all you've done for 4 days is watch Monty open doors to goats, but you should certainly be thinking it's odd that he skipped that one specific door. You should probably switch, though maybe take a nap before claiming your car.

NoboruWataya2• 2 years ago

In addition to the "100 door" formulation that helped me a lot, the way I rationalise it is that:

* When you choose your first door, the chance the car is behind that door is 1/3, for obvious reasons.

* When Monty opens the goat door, nothing actually changes that is relevant to your odds. He is always going to open a goat door, whether your first choice is a car or a goat, so it's basically irrelevant. He hasn't move the car, he hasn't given you any information you didn't already know when you made your first choice. So the odds of your first choice being correct can't have changed. They are still 1/3.

* The only possible outcomes are that your first-chosen door has the car, or the other remaining door has the car. The probability if your first-chosen door having the car is (still) 1/3, and the probabilities of all possible outcomes must add up to 1, so the probability of the other door having the car are 1 - 1/3 = 2/3.

tuatoru• 2 years ago

The Monty Hall problem controversy is good evidence that evolution has ruthlessly constrained human intelligence.

HumblyTossed2• 2 years ago

I recall reading a lot of the replies and one always struck me - a (I believe) teenager wrote a program that proved her right.

That person wrote a program and proved her right, while all these stuffy old men got just downright testy with her - one even getting misogynistic with here. It was gross to read about.

But good on the teenager for actually sciencing it.

shadowgovt2• 2 years ago

Arguably the most fascinating thing about the Monty Hall problem is that although this solution is extremely widely known in academic circles, the game is still played today on Let's Make a Deal! in the exact same way: three doors, pick one, a Zonk is revealed behind another, switch or keep?

(Although there is one variable I don't know the answer to: I don't know if they always give you the offer to switch. Sometimes I believe it's a different offer, "Keep the door or take this money out of my hand". The host being able to choose which offer is made shifts the odds a bit, because the host may be choosing the offer based on whether you've picked the right door initially, and may be making that choice in a biased or randomized way. There's hidden data there we don't have as observers).

dekhn8• 2 years ago

There's the Monty Hall Problem and then there's the Second Order Monty Hall Problem Problem.

That is, did the wording as it was originally printed unambiguously define the problem to be solved, or was it ambiguous enough that some people who got it wrong got it wrong for the right reason?

Like many others, I completely missed that when Monty opens a door to show you a goat, he always shows you a goat. I think under close inspection, the problem doesn't actually allow you to interpret otherwise. But, should a problem like this depend on parsing and identifying the ambiguity, or should the problem be defined with explicit details about the door opening?

Given how many otherwise extremely smart people have misunderstood the problem, I am still unsure whether this was an intentional choice in framing the question. I got it wrong and thought Vos Savant was wrong until I read one of the replies in her follow-up article (yes I read this when it first came out!) said they wrote a computer program and reproduced her results. The program implemention worked because they correctly modelled Monty's behavior. I think when people end up writing the program out the ambiguity becomes more obvious.

ghaff2• 2 years ago

I think this is the original "Ask Marilyn":

Suppose you’re on a game show, and you’re given the choice of three doors. Behind one door is a car, behind the others, goats. You pick a door, say #1, and the host, who knows what’s behind the doors, opens another door, say #3, which has a goat. He says to you, "Do you want to pick door #2?" Is it to your advantage to switch your choice of doors? Craig F. Whitaker Columbia, Maryland

Yes; you should switch. The first door has a 1/3 chance of winning, but the second door has a 2/3 chance. Here’s a good way to visualize what happened. Suppose there are a million doors, and you pick door #1. Then the host, who knows what’s behind the doors and will always avoid the one with the prize, opens them all except door #777,777. You’d switch to that door pretty fast, wouldn’t you?

-----

Yes, the question as posed by a reader was ambiguous but Marilyn made a very reasonable interpretation based on what someone at that time would probably tend to know about game shows. And she even gave an explanation (using more doors) that I find tends to work with folks combined with the explicit assumption that the host will always pick goats.

dekhn8• 2 years ago

Yes, I'm literally talking about this sentence: "host, who knows what’s behind the doors, opens another door, say #3, which has a goat."

I parsed that as "50% of the time, monty opens another door and it has a car and you win immediately, and 50% of the time, monty opens another door and it has a goat". In retrospect I think my brain just sort of pictured that and proceeded to assume there was no reason to switch, and it wasn't until I read some answers (not her explanation) that I understood I was wrong (and I went through a lot of anger, like many of the commenters, and thought she was very wrong before that).

If there had been exactly one more sentence saying "monty always opens a door showing a goat", I'm pretty sure I would have recognized that. BTW I was familiar with the game (I liked Jeopardy better, as Let's Make a Deal was a fairly dull show) and I don't think they had a game that was identical to the problem as described.

Scarblac2• 2 years ago

I assume that at the time of writing, most readers would have been familiar with the actual game show and the fact that the host always opens a door with a goat?

dekhn8• 2 years ago

The show uses a different game.

dboreham• 2 years ago

I had never heard of the show when I was introduced to the problem so all my knowledge comes from the problem description.

jcl• 2 years ago

As Monty Hall pointed out in interviews, he was not obligated to offer the opportunity to switch and sometimes did not. So in this case knowledge of the actual show may have contributed to confusion.

https://en.wikipedia.org/wiki/Monty_Hall#Monty_Hall_Problem

(But vos Savant stated in later columns that most of the critical responses she received assumed that the host was obligated to offer the switch, so they were genuinely confused by the paradox.)

kristjansson3• 2 years ago

As someone who misunderstood the problem as you describe, I'm much more invested in this version of the problem :)

The problem statement is just a mess:

You pick a door, say #1, and the host, who knows what’s behind the doors, opens another door, say #3, which has a goat.

even tiny edits like:

You pick a door, say #1, and the host, who knows what’s behind the doors, opens another door which has a goat, say #3.

or even

You pick a door, say #1, and the host, who knows what’s behind the doors, opens another door, say #3, which he knows has a goat.

make it clear that the last phrase is describing an invariant of the problem, not an artifact of the illustrative hypothetical.

svantana• 2 years ago

That's much better, but still not enough. We also need to add that the host will open a door regardless of which door I initially choose.

empyrrhicist2• 2 years ago

Your re-framing is actually not quite right either (or at least, not complete). Key point: if the host picks the car, the game essentially re-sets.

Imagine the extreme scenario with 999 goats and 1 car. You select the first door, the host opens 998 doors at random, leaving your selected door and one other. Two scenarios are now possible:

1. There was a car in the 998 doors that got opened. Tough luck, you lose the game automatically now (game resets).

2. There was not a car in the 998 doors. In this scenario, you are still advised to switch.

Edit - I guess I misremembered a simulation I once did about it.

Swizec• 2 years ago

Also you're picking between Reward and Bigger Reward. Positive EV no matter your choice so eh don't worry about it too much :)

A nice goat fetches up to $1000 to according to a quick google.

pxx• 2 years ago

Nope. Write the simulation for 2.

Or look at this case analysis for the three element case. https://news.ycombinator.com/item?id=24714612

You can get intuition by understanding that the cases where the game is ruined are only the cases where you didn't choose the car.

topaz011• 2 years ago

As I explained in my reply to your comment on mine, this analysis doesn't work out, even in this very-lopsided case. If the game resets when a car is revealed, then games where you first choice happened to be a car will always complete, and games where your first choice was a goat will almost always reset. On average, of 1000 games, your first pick will be a goat 999 times, and of those, 998 will be reset because MH opens the car and 1 will complete and give you the car if you switch, whereas in 1 in 1000 you will pick the car first, in which case the game is certain to complete but you will lose the car if you switch. Therefore, most games will reset, but of the ones that complete, there is no advantage to switching.

TotempaaltJ3• 2 years ago

Vos Savant mentions this, and she's kept track of who did and didn't understand the conditions:

And a very small percentage of readers feel convinced that the furor is resulting from people not realizing that the host is opening a losing door on purpose. (But they haven’t read my mail! The great majority of people understand the conditions perfectly.)

https://web.archive.org/web/20181118225305/http://marilynvos...

topaz011• 2 years ago

Nice find! I have wondered this. It's easy to say after you have come around that you had first misunderstood the scenario and done the math right in that context, even if in truth you (implicitly) understood the scenario but just neglected to account for the constraints.

topaz011• 2 years ago

What astonishes me is not that people interpret the scenario differently, but that they are so quick to conclude that others are doing the math wrong and feel superior or even angry, rather than thinking about whether you might be talking at cross purposes about different scenarios.

So I see the second order problem a bit differently: it's not "was the statement unambiguous", it's about what you do with the (possibility of) ambiguity: do you assume you can't be wrong and therefore others must be, or do you try to figure out whether you were solving different problems?

kristjansson3• 2 years ago

Humans don't backtrack well :)

Verdex• 2 years ago

It's kind of like the xkcd airplane on a treadmill blogpost:

https://blog.xkcd.com/2008/09/09/the-goddamn-airplane-on-the...

There appears to be problems that make it easy for people to assume the worst of others.

And on a side note, I discovered about 6 months ago that mythbusters actually did this experiment. At the end of the day, few things work better than seeing something happen in real time.

juped• 2 years ago

No, there's not.

There's the Monty Hall problem and then there's the "can I get away with falling back on claiming the question was somehow wrong after embarrassing myself by getting the Monty Hall problem incorrect" problem. No one who claimed it was unclear or ambiguous did so before getting it wrong.

It's really tiresome to have to always pretend to assume good faith when it's really clear what's going on.

c2occnw• 2 years ago

That interpretation requires believing that the host, who knows what’s behind the doors, would sometimes open the door revealing the automobile and ask the contestant if they want to switch to that door. It doesn't really make sense, so that should tip off readers that they've misunderstood the show format.

Marilyn even made this explicit in her answer, writing that the host "knows what's behind the doors and will always avoid the one with the prize" [0]. So readers were arguing her answer was wrong even given this understanding of how the show works.

[0] https://www.newspapers.com/article/the-missoulian-the-monty-...

mlyle• 2 years ago

That interpretation requires believing that the host, who knows what’s behind the doors

The host has to always open the door, too. Else-- if your strategy is always to switch when shown a goat, the host can choose to only show you a goat if you have already picked the car (which would cause you to lose 100%). Or various mixed strategies.

This is one of those problems that is hard to grasp at baseline, but a little harder to grasp because the situation isn't fully specified. As a result it tends to create a lot of controversy, like the one about an airplane on a treadmill.

bhk• 2 years ago

the problem doesn't actually allow you to interpret otherwise

How so? The problem is typically presented as a single trial, with nothing that clarifies the intent of the host.

topaz011• 2 years ago

The explanation in the article is fine as far as it goes, but I think it's much more helpful to emphasize one key fact, which is that MH has to show a goat, and can't open your first choice door, even if it's a goat. That means that if you did pick a goat the first time around, he can only have revealed the other goat, in which case switching will necessarily give you the car. That happens with probability 2/3.

This is implicit in the "100 doors, MH opens 98 of them" construction discussed in the article but making it explicit makes the idea quite clear and intuitive, even with just 3 doors.

I'm moderately sympathetic to the idea that this constraint is not immediately obvious from how the scenario is described, and could technically be ambiguous in some versions. I do think most people will agree that it makes sense to infer this constraint once you think about it. Obviously if he opened the door that you initially picked and it was a goat, you would know to switch to one of the other two (and have 1/2 chance of finding the car). (Or, more absurdly, he could just open the door with the car and you could choose it directly). Or you could imagine a game show where MH can choose to open a door or not, in which case the problem is not well-posed, because you would need to know how he makes the choice.

ghaff2• 2 years ago

There are probably a lot of people at this point who have never watched one of these types of game shows. Maybe if they thought about it, they'd realize that MH probably will always show a goat but it wasn't explicit in the initial question that Marilyn answered.

Certainly, that constraint is important to the final solution.

topaz011• 2 years ago

I guess my point is that in the inevitable arguments, whether they are in advice columns or HN comment threads or around dinner tables, sometimes seem to their participants to be about how math works, when they're really about the problem description. Sometimes the disagreement amounts to "I agree that those constraints make sense and I wasn't properly accounting for them" and other times it amounts to "I was actually picturing a scenario where MH might have shown you a car". Either is ok, and there might be room for more discussion after getting there, but it feels like a big waste for either of these to get confused with a fruitless argument where both sides think the other is just doing the math wrong.

andrewla4• 2 years ago

No, he wouldn't always open a door -- sometimes he'd offer a trade; like "you can take $1,000 cash from me now, or have your door".

I wonder if anyone has gone through every episode of "Let's Make a Deal" where this scenario occurs and look at the empirical distribution of switch/not-switch.

empyrrhicist2• 2 years ago

nvm

topaz011• 2 years ago

Actually it does matter. If you lose automatically when a car is shown, that will happen in 1/3 of games: none of the times your first pick was a car (0 * 1/3 = 0), half of the times your first pick was a goat (1/2 * 2/3 = 1/3). The remaining 2/3 are equally split between the times you chose a car first (which are guaranteed to complete) and the times you chose a goat first and MH randomly chose the other goat rather than the car.

(Note: this analysis does depend on the constraint that MH doesn't open your first pick door! Edit: actually the answer also becomes equal odds if he can pick any door, regardless of whether it was your first pick or the car. In that case, 1/3 of games will not complete, and the rest will be 50/50 between the unopened doors.)

graeme• 2 years ago

Monty never shows the car. That was stated in the original reply from Marilyn, and it should be obvious to anyone who's watched a game show. Why on earth would monty open a door, show you the prize (which you now can't get) and then ask you if you want to switch your door?

No one would design a game show like that, so it means the question only has sense if interpreted as "Monty always opens a goat door".

SeanLuke• 2 years ago

The explanation in the article is fine as far as it goes, but I think it's much more helpful to emphasize one key fact, which is that MH has to show a goat, and can't open your first choice door, even if it's a goat.

The second part I agree with, but not the first part; because it sounds like you're saying MH is required by the rules to reveal a goat. But the problem works fine if he always opens a remaining door at random and it happens to be a goat.

I'd say it as: once you had chosen a door, Monte was required to open some other door. And the door he happened to open, for whatever reason, had a goat.

Wouldn't that work?

topaz011• 2 years ago

This has been addressed several times in the comments on this post, and the answer is no, it does matter that he has to select a goat. If he might have selected a car and just got lucky in selecting the goat, then all of the following had equal probability:

1. you picked car first, he picked the first goat

2. you picked car first, he picked the other goat

3. you picked first goat first, he picked the car

4. you picked first goat first, he picked the other goat

5. you picked other goat first, he picked the car

6. you picked other goat first, he picked the first goat

You happen to know that he picked a goat, so you can eliminate 3 and 5, but the remaining four possibilities are still equally likely. Half of them let you win by switching. This is also the "MH stumbles into a door on accident" variant that others have mentioned.

Another way of getting intuition about your version is to realize that, if he might have shown you a car, then the fact that he has opened a goat is evidence against the hypothesis that your first pick was a goat, because half of the times you choose a goat he would open a car, whereas if you chose a car he would have to show you a goat every time. It eliminates some of the ways things could have turned out if you started on a goat without eliminating any of the ways things could have turned out if you started on a car, so the former is relatively less likely in retrospect than it was before MH opened the goat. More obvious in the 100-door case: if your first pick was a goat, then the fact that he hasn't shown you a car after opening 98 doors at random (or equivalently leaving one door closed at random) either means that you were very lucky in the first pick and got the can (in which case it's easy to open 98 doors without a car) or he got very lucky in his random choices. Most games played in this way (98%) would end with MH revealing the car, and the fact that you don't happen to be in one of them only tells you that one or the other of you got very lucky. The probabilities work out that the two options are equally likely.

GMoromisato• 2 years ago

I remember reading Marilyn vos Savant every week as a teenager. At the time I half suspected that she wasn't real--just a character made by a team or something. [This was before the internet, so I couldn't just ask Google.]

I'm happy that she's real and as smart as portrayed. But what really impressed me is how calmly she received all the negative feedback and personal insults. No matter what, she never lashed back. I have a feeling that's authentic too.

tempodox• 2 years ago

I wonder where the “vos” in her name comes from. I don't remember seeing it anywhere else before.

TotempaaltJ3• 2 years ago

I've read through this thread, and for anyone criticising the Monty Hall problem, the article, or Vos Savant herself, I recommend reading the original responses by Vos Savant, available on her website on te Internet Archive: https://web.archive.org/web/20181118225305/http://marilynvos...

They address every single (meta) concern I've read here.

iancmceachern• 2 years ago

I met her several times. I worked for her husband, Dr. Robert Jarvik in Manhattan building artificial hearts. Lots of weird stories about this family...

neverartful• 2 years ago

The following visualization helps to understand the answer:

https://imgur.com/a/hkme8jG

For the 3 different scenarios, switching is the correct choice in 2 instances. Not switching is the correct choice in 1 instance. 2/3 chance that switching will help you to win.

NoboruWataya2• 2 years ago

Wow, I knew about the problem and was vaguely aware that it had generated some controversy amongst statisticians but I had no idea about the insanely arrogant and obnoxious (not to mention wrong) abuse Marilyn had received from the "intellectual elite".

I have little sympathy for the "ambiguous question" defence. Not only is Marilyn's interpretation grammatically valid, it just wouldn't make sense in the context of the game show for the presenter (who knows where the goats are) to ever open a door to reveal the car and give you the option to switch. The game would be reduced to "does Monty feel like gifting you a car or a goat today".

Furthermore, Marilyn's own explanation of her answer makes it very clear that she interpreted the question as meaning that the presenter will only ever open goat doors. And all the attacks seem to focus on her grasp of the math rather than her interpretation of the problem question.

To be clear, I think it's perfectly reasonable for a person to interpret the problem the other way without thinking too much about it, and casually throw out an incorrect answer. But I would expect an academic to devote at least another 30 seconds of thought to the problem before publicly excoriating a woman known for her high intelligence.

tialaramex• 2 years ago

Also regardless of whether she's a woman or she's known for intelligence, a response in which you say "You're wrong" needs an extra second's thought to consider whether, in fact, they are wrong, or maybe just their understanding is different from yours and you need to reconsider with their context.

Some time back in an HN thread about programming languages I read a response in a thread about inference which said that a particular feature doesn't really save on typing. And it seemed ludicrous. Until I realised the author of the comment has seen the word "typing" (which everybody else in the thread is reading as "The thing you do with a keyboard") and they're interpreting it as "The thing the compiler does". And yeah, with that perspective of course inference isn't saving typing, all your variables still have types, we were just happy because we spent less time operating the keyboard to write a program. Their comment was not wrong and we weren't correcting them, we'd just understood things differently!

Edited to add, link: https://news.ycombinator.com/item?id=36593110

city413• 2 years ago

I almost never say "you're wrong", no matter how confident I am. Because I can be, and often am, wrong myself. If there is a disagreement, a miscommunication, etc, why not instead work with the person to find where you two differ and look for common ground? If the other person really is wrong, it's almost always naturally revealed that way.

bazoom423• 2 years ago

it just wouldn't make sense in the context of the game show for the presenter (who knows where the goats are) to ever open a door to reveal the car and give you the option to switch.

None of it makes sense because it is not a real game show, it is a thought experiment.

apetresc• 2 years ago

Apropos the topic of being confidently incorrect, it was indeed a real show: https://en.wikipedia.org/wiki/Let%27s_Make_a_Deal

Hosted by a real Monty Hall: https://en.wikipedia.org/wiki/Monty_Hall

It even had the gag prizes like goats (or, as in the photo from the wiki page, a llama)

XCabbage• 2 years ago

it just wouldn't make sense in the context of the game show for the presenter (who knows where the goats are) to ever open a door to reveal the car and give you the option to switch

No, but it would make perfect sense for him to open a door at random and - if doing so reveals the car - tell you sadly that you've lost as a disappointed klaxon plays. Indeed this is almost the only possible way that the initial door opening could realistically make sense in a game show; if it works in any other way, there's no tension when the first door is opened since there's no risk!

psd1• 2 years ago

there's no tension

There certainly is. It's a high-stakes scenario and the door-opening presents new information. It makes the conclusion feel less certain.

Revealing the car would conclude the game, of course. Unless the contestant were holding out for the better goat.

flomo2• 2 years ago

In reality, Monty Hall often did not behave in the way the problem describes, that might be the source of confusion surrounding this. (e.g. sometimes he would reveal a prize and imply there might be a better prize behind one of the doors. sometimes there were only prizes and no goats, etc.)

lucas_membrane2• 2 years ago

> The game would be reduced to "does Monty feel like gifting you a car or a goat today".

But, in fact, that is the way the real show worked. He was more likely to give you a car when he hadn't given away any cars for a while. And Marilyn's description of the problem included no information about whether or not Monty offered to swap doors rarely, frequently, or always, or about the previously observed outcomes depending on whether or not the contestant took the bait. And she surely did not say that Monty was a robot and had no interest in what happened on his show.

The way the game actually worked is like how poker and life work until proven otherwise. Every high-stakes game has a sucker, and if you do not know who is the sucker, it is you. Read the story in which a NY Times reporter interviewed Monty about the controversy and you will learn how it really worked.

harshreality• 2 years ago

If you're playing the game as stated, and if the host opens a door revealing a goat, you must switch for the best odds.

The offer to switch doors is part of the problem statement. You can't justify your decision based on the possibility in the real game that you might not get the offer to switch (which would mean that when you do it might be an intentional red herring).

Even if there's a 50% chance there are all goats, or a 75% chance, you still switch. That reduces the overall probability of a win, but it doesn't affect the relative probability from switching. The only reason you might not want to switch is if the first choice had been forced, such that there was an above-random chance that your first pick was the car.

digitalscribe2• 2 years ago

Not only is Marilyn's interpretation grammatically valid

I would consider both interpretations equally valid. If anything, in law we have the "rule of the last antecedent" (phrase/clause modifies nearest antecedent noun) so that the hater's interpretation could be more correct if based only on this rule.

This language is ambiguous: "host ... opens another door, say #3, which has a goat."

The two options are Marilyn's interpretation of the door always containing a goat (invariant version), or that one door he opens (say #3) just happens to contain a goat (hater's version) but it could have contained a car.

As suggested by someone below, a much clearer formation could have said "host opens a door with a goat, say door #3" to avoid this particular ambiguity.

There's also the issue of usage of "which" signifying a non-restrictive non-essential clause.

hulium2• 2 years ago

If the the door contains a car, then it doesn't matter whether you switch or not. If it is a goat you get the problem as intended.

You still have the intended problem as a subproblem, so this is just a distraction and not a serious objection.

wslh• 2 years ago

On this topic, please watch what I think is one of the greatest scenes in movies: Kaspar Hauser directed by Warner Herzog [1].

[1] https://youtu.be/C9uqPeIYMik?si=ZjJIH1RhQFTN_ARR

swayvil3• 2 years ago

Hey that riddle was in Labyrinth (with David Bowie as the goblin king!)

HarHarVeryFunny5• 2 years ago

I think the problem statement is clear, and is independent of how the TV show actually operated.

The problem posed is that you have three closed doors, behind one of which is a car, and behind the other two are goats. You get to pick one of the closed doors, and will win whatever is behind it (you want the car).

One of the doors you did not pick is now opened, revealing a goat. You therefore now know the car is either behind the door you picked or behind the other closed door.

In this setup, the question is: if offered the chance to switch your selection from the door you initially picked to the other closed door, should you switch? Does it make a difference?

Mariyln Vos Savant said switch. The peanut gallery (some statisticians who should have known better) said it makes no difference - with two doors left you've got a 50/50 chance of winning the car whether you switch or not.

First off, you can run multiple simulations of the game (with random positions of the goats and car), and prove for yourself that always switching is a better strategy, so it's sad that anyone would insist that it's not, even if that seems logical.

The simplest way to explain why switching is the correct strategy is this:

When you make your initial selection, you have:

a) A 1/3 chance of selecting the car, and

b) A 2/3 chance of selecting one of the two goats

So, 1/3 of the time you selected the car and will therefore lose if you switch, and 2/3 of the time you selected a goat and will therefore win if you switch. Switching will therefore be a winning strategy 2/3 of the time.

petesergeant2• 2 years ago

The simplest way to explain why switching is the correct strategy

In my experience the most _intuitive_ explanation is to simply ramp it up to 100 doors, with Monty opening 98 of them, to make it clear that switching offers you the benefit of all unopened doors.

heleninboodler2• 2 years ago

It's a lot easier to see the 2/3 nature if you imagine that instead of revealing the goat and then asking if you want to switch, Monty said "you've chosen this door that has a 1/3 chance. I'm offering you to switch to choosing both of these two remaining doors, at least one of which definitely has a goat by definition." This is functionally the same offer. It's 2/3 to switch because you're effectively choosing two doors and having Monty automatically grant you the car if it's behind one of them.

listenallyall3• 2 years ago

It absolutely is not "independent of how the show operated." The host knowing in advance where the car is, (thereby ensuring the revealed door is always a zonk) is essential. If the host just picked a random door every day, and sometimes did reveal the car prematurely, then it would in fact be a 50-50 decision to switch.

lamontcg3• 2 years ago

The asymmetry is that if you pick the one with the car first (only 1/3rd chance) then Monty Hall can show you either of the other doors. If you pick one without a car first (2/3rd chance) then Monty Hall must show you the door that doesn't have a car and the remaining door always has the car.

So 2/3rd of the time you have a 100% chance of a car if you switch, 1/3rd of the time you have 0% chance of a car if you switch. So switching is a 66.7% probability of a win and is the 'inverse' of if you picked the car right on the first guess or not.

The choice of which door Monty Hall shows you is constrained and isn't an independently random event since he never shows you the car when he opens a door.

You can also laboriously go through every event and "bifurcate the world" and then add up the probabilities, but you need to be careful about how you do it:

    1.  assign the car to a door (there are three possibilities)
    2.  have the contestant pick a door (there are now nine possibilities)
    3.  handle which door Monty Hall shows (this is tricky because for the three permutations in #2 where the contestant picked the car, then the door Monty Hall picks is unconstrained and so the world splits, but you are slicing a 11.1% chance in half for ~5.6%, while for the six permutations where the contestant didn't pick the car, Monty Hall is constrained so those remain 11.1% chances (at every level the probabilities have to add up to the probabilities before them and you can't violate 100% when you sum across all of them).
    4.  then have the contestant either switch or not switch and cut all your probabilities in half again.
    5.  now color everything according to winning or not and sum it all up.

I call that the "Quantum Many Worlds Interpretation Solution to the Monty Hall Problem" and if you do step #3 correctly you should get vos Savant's answer.

Lendal• 2 years ago

Aha! This is the first explanation that reveals the most common error. The fact that the reveal isn't just a random door. The host must not reveal the car. That changes everything. I believe that's what most people are missing. But if you actually stated that explicitly, you would give away the answer and the problem wouldn't be fun anymore.

The fun is in not explicitly stating every unconscious assumption you have about the problem.

calderknight5• 2 years ago

That's clear. Otherwise the question would just be about speculating on the host's personality.

munificent• 2 years ago

You would be surprised how many people do correctly understand the problem (that the host must open a door and must not reveal the car) but still get the resulting probability wrong.

calderknight5• 2 years ago

It's simple. To begin with it's 1/3 your pick is right. You know that a goat will be revealed. So it's absurd to suppose that the fact that a goat is revealed has increased the probability of your initial guess being correct.

Which door is opened to reveal a goat is totally irrelevant, as either would (under the absurd supposition) increase the probability of the initial guess to 50%. In other words, an increase to 50% would be guaranteed from the very beginning. So a contradiction is reached with the obvious fact that initial prob is 1/3.

elteto2• 2 years ago

Fantastic answer. I went through and built the tree and it just shows everything so neatly. There's just a small mistake, when you say:

    4.  then have the contestant either switch or not switch and cut all your probabilities in half again.

You don't halve probabilities again because the strategy we are evaluating is _to switch_. So we always switch. It is deterministic.

If we switch deterministically we end up with 6 cases (leaves in the tree) with probability 1/9 where we win. 6/9 = 2/3!

If we switch randomly then we do split probabilities again, and end up with 6 cases with probability 1/18. 6/18 = 1/3.

lamontcg3• 2 years ago

I don't really consider that a mistake it lets you compare the switching strategy to the not-switching strategy. YMMV.

bazoom423• 2 years ago

host, who knows what’s behind the doors, opens another door, say #3, which has a goat.

There is quite a lot of ambiguitiy in how this is phrased. So the host knows what is behind the doors, but how does this affect his choice to open another door?

Does he always open another door, or does it depend on what is behind the door the contestant selected?

Is the second door selected by random or will it always be one with a goat?

The implicit rule is that the host: 1) always opens one of the other two doors, regardless of what is behind the selected door. 2) deliberately opens one of the two doors which doest have a car behind it.

dunk010• 2 years ago

The key is that if the contestant chose a goat then Monty is FORCED to reveal the other goat. It’s this which skews the odds. He has (often) added information to the game.

jlarocco• 2 years ago

I think the phrase ", say #3," confuses people. Taking it out removes the ambiguity:

"[The] host, who knows what’s behind the doors, opens another door which [...] has a goat."

So he always shows a goat. If he opened the door with the car then the game would be ruined.

snakeyjake• 2 years ago

All of those PhDs and not a single one of them decided to sit down, take 17-and-a-half minutes to write a program, and experimentally prove it. Did PhDs in the early 90s not have computers? I had a computer and I'm a scrub.

andrewla4• 2 years ago

The reason the scenario is kind of twisted is that in the iterated game of "you pick door, he shows goat, switch or no", vos Savant's answer is clearly correct.

But in the iterated game of "you pick door, and host chooses whether to show a goat or to let you open the door you initially chose" I don't think you (or he) can get better than 50-50. If you chose the correct door initially, then he can try to trick you into switching by opening a goat-door. But if you chose a goat-door, then he can let you open your selected door (or offer some other deal -- $1,000 cash or the door you chose). If he only shows the goat-door when you have selected the car-door, then you should never switch, so his optimal strategy is to show a goat-door when you choose a goat-door 1/3 of the time, so you basically get no information from his actions.

The "always switch" solution imagines this as a one-player game rather than a two-player game, with Monty Hall actively trying to sabotage.

flomo2• 2 years ago

That's correct, the premise of the show was more Monty Hall doing something to challenge the contestant. That's probably why people missed the problem's premise where he always opens a goat door.

grraaaaahhh• 2 years ago

I don't think you (or he) can get better than 50-50.

Doesn't Monty reduce your win chances to 33% just by never opening a goat door?

andrewla4• 2 years ago

That's an excellent point.

dwpdwpdwpdwpdwp• 2 years ago

I taught this problem to a group of high schoolers as part of a statistics curriculum. It was engaging for two reasons:

The topic of the vitriolic responses to Marilyn is not boring.

It’s a great problem for illustrating the importance of simulation in statistics. By breaking off into pairs, half of whom played the switch strategy and half of whom played the stay strategy, we did hundreds of trials and found empirical evidence in favor of Marilyn’s answer. Furthermore, by actually playing the game, whatever any ambiguity in the wording of the problem was quickly resolved.

samatman• 2 years ago

It remains amusing that this became dubbed the Monty Hall Problem, since Let's Make a Deal did not, in fact, work that way.

Here's Monty Hall himself on the subject: https://twitter.com/kareem_carr/status/1749463327464456644

bumbledraven• 2 years ago

University of Toronto stats prof. Jeffrey S. Rosenthal wrote a nice paper called "Monty Hall, Monty Fall, Monty Crawl" (https://web.archive.org/web/20230706235720/https://probabili...) in Math Horizons (Sep 2008). It discusses two variations of the Monty Hall problem that are likely to be answered incorrectly by those who used flawed logic to arrive at the correct answer to the original:

Monty Fall Problem: In this variant, once you have selected one of the three doors, the host slips on a banana peel and accidentally pushes open another door, which just happens not to contain the car. Now what are the probabilities that you will win the car if you stick with your original selection, versus if you switch to the remaining door?
Monty Crawl Problem: As in the original problem, once you have selected one of the three doors, the host then reveals one non-selected door which does not contain the car. However, the host is very tired, and crawls from his position (near Door #1) to the door he is to open. In particular, if he has a choice of doors to open (i.e., if your original selection happened to be correct), then he opens the smallest number available door. (For example, if you selected Door #1 and the car was indeed behind Door #1, then the host would always open Door #2, never Door #3.) What are the probabilities that you will win the car if you stick versus if you switch?

failrate• 2 years ago

The great thing about the apparent paradox is that you can write up a quick script in your favorite language and test it. I did not (quite) jive with this answer until I simulated it myself and golly gee!

deeg• 2 years ago

A while back I wrote up an explanation of the MH problem using code: https://github.com/DeegC/monty_hall_paradox

mhb2• 2 years ago

She should have played the game with the detractors for money.

asdfasvea• 2 years ago

Jeez, this thread should have it's own thread where we discuss it:

"Marilyn vos Savant and the Monty Hall Problem (2015) Problem (2024)"

That thread wouldn't be about the problem, or about the decades ago backlash to the correct answer about the problem, but about the backlash to how everyone in 2024 semantically picked apart the problem.

I lived though the initial backlash and it was annoyingly entertaining how high level mathematicians tried to show how she was wrong. Now everyone is trying to use semantics to show how they are right--or at least not wrong.

I don't know which is worse--pompous incorrectness (then) or pedantic semantics (now).

perrygeo• 2 years ago

Let's build a realistic model of a game show host. What is the causal process at work?

The game show host isn't going to reveal YOUR chosen door - that would ruin the whole game.

Likewise, the host isn't going to reveal the CAR and end the game immediately without a winner - it's bad television.

So the host opens a third class of "irrelevant goat door" in order to offer you the suspenseful proposition that is sure to get millions on the edge of their seat - switch or stick?

In the case of 3 doors there is only 1 irrelevant goat door. In the case of 100 doors, there are 98 irrelevant goat doors, driving the point home further.

We have to assume the host knows which are irrelevant. All irrelevant goats are eliminated. Which means the probability of the car behind your door is 1/n and the probability of car behind the remaining door is always (n-1)/n. Always. Given this assumed behavior of the host.

Knowing the above gives you a clear empirical advantage for modeling and inference. Why would you reject knowledge that may be directly relevant to the question at hand?

YeGoblynQueenne• 2 years ago

FTA:

> Despite her status as the “world’s smartest woman,” vos Savant maintained that attempts to measure intelligence were “useless,” and she rejected IQ tests as unreliable.

The word "maintained" is a link pointing to this blog post, reproducing an interview of vos Savant [1]. In the interview, vos Savant is quoted as saying:

> Not that IQ tests are useless. Far from it.

______________

[1] http://sailom.blogspot.com/2007/04/are-men-smarter-than-wome...

Marilyn vos Savant and the Monty Hall Problem (2015)

Related Stories