The first time someone introduced me to the Monty Hall problem I got so angry. Not because I didn't get it, but because they didn't get it. The way they said it they never stressed the host HAD to show a false door after the contestant selected and before offering the option to switch. I realized the math was enticing to switch, assumed the game host was trying to entice the switch by showing a false door, and opted to not switch. Somehow in the moment neither of us realized the disconnect -- I was arguing over the psychology of the host and they were arguing about the statistical incentive. We were only at odds because I assumed the host was a second rational actor, acting in opposition to the contestant, and they failed to explain this was a scenario with only one rational actor, the contestant.
It's annoying, isn't it? Feel like this is why a lot of people struggle at first - bad framing. I wonder if framing it as 'A friend of yours picks one of the two remaining doors, opens it, and reveals a goat' would make people better understand that the situation has changed. Naturally, the relationship with a game show host tends to be an adversarial one, but in the problem as stated Monty provides consistent and helpful information. Also while I hate to just regurgitate wikipedia, the articles on the problem reference a newspaper interview in which Monty stated that the real show ("Let's Make a Deal") did not, in fact, work like this, and that he would deliberately play with the contestant's expectations. Sometimes his offer would help, and sometimes it would be a trap, and sometimes he would simply make them accept the door they picked - whatever seemed best for the show. "It all depends on [the host's] mood," he said.
I feel like this is misrepresenting what you can do with game theory. Game theory isnt necessarily presctiptive. You dont have to take it as "this is what you should do". Its the question: "If everyone acts to maximize their own value, what happens?" and in a world where many people and corporations act that way, its usefull to understand why something is happening. Things like: how do we design systems so they cant be exploited by greedy people? How to make better voting systems? How to share items that have different value to different people in some way that you could call "fair"? Or "how do you get corporations that only care about their profit to act for some common good (like fixing the climate or some equality notions)? Are all questions that gametheory can be helpfull to answer. I totally agree that you should not live your day to day life "according to gametheory", but that doesnt mean the field is useless or bad. Otherwise great explanation of Monty Hall!
I feel like this runs into the same problem as "economics" because yes whether or not a field is descriptive or prescriptive isn't like, anything inherent to it; but it is shaped by people in that field and how they choose to behave. I think if we're going to be descriptive (as we should) we should notice that a lot of people are using game theory to justify and reinforce the "what happens" and even twist it into "and this is what should happen" rather than just understand it. idk your comment is in no way wrong, nuanced disagreement.
@@FinetalPies yea, there definitly are people that look at game theoretic solution concepts and their takeaway is "everyone should act like this, anything else is wrong" or "acting like this is superior / how things should be", and those are really not people we should listen to. Those people are probably the reason that Jorbs shows such a negative view of the field in this video, and i totally get that, those people suck. I just really wanted to highlight that these are not the only people doing game theory, and that there are more positive (in terms of social wellfare/common good) ways to use game theoretic ideas.
The problem with game theory presented here is how it is often presented as "the rational way to make decisions" without acknowledging that it's only rational within a huge range of cultural assumptions and arbitrary value judgements.
The preoccupation with not whether or not we want the goat pays off in this video, but I always viewed the Sports Car/Goats as symbolic for "Desirable outcome/Undesirable outcome". To me it's pretty clear that those things were picked because that was an easy shorthand from the framers culture which fit context being constructed around the problem. It's irrelevant what's behind the doors to the theory itself, the only thing that matters if that you want to open a particular one...
@@karma5390 by YOU do you mean the audience? the people who are going to have access to the show because it is the 1960s and not the internet. sure not everyone in the possible audience value the car over the goat, but would 95% of them? And also its not required for it to match the values of everyone who has ever lived, thats impossible, just enough of the people there.
@@MrMarson123456 The world has progressed since the 1960s. People use the same Monty Hall problem, but now it's widely been brought to have a larger meaning than it did in its original game show context in the same way that the word "broadcast" has come to harbor a more symbolic meaning, rather than the literal casting of seeds broadly. Yet, we still use the goat and the sports car. The reason jorbs brings up the issue of why the rewards are or are not culturally relevant is because the Monty Hall problem, much like the Prisoner's Dilemma, is presented with the text that it is here, explicitly for the purpose of making it more readable to a certain audience for whom it was made. That makes it easier for certain people to intuit than others. Nobody is claiming that it is required for every hypothetical to match the values of everyone who has ever lived, that's absurd, and I don't know what in the video made you receive that message. This video is asking you to consider this typically theoretical problem and its implications on decision-making in the context of real-world decision-making, which has many more nuances and considerations. Yes, the goat and car are useful symbolic shorthands, but we are trying to think about more than just the symbols.
@@gg829 I mean same with scientific racism, which is the point of comparison here. It wasn't that people invented statistics and decided they will be racist because of it, they invented transatlantic slave trade and the next several centuries of thought got defined by the need for convincing excuses.
@@gg829 I don't think this is a convincing summary of what stw7120 said. Yes, material reality can shape how we think, but I think it is ridiculous to claim that the way people think doesn't shape material reality. Humans, more than any other creature, have the capability to impose their thoughts upon the world; to shape the world in a way befitting of on contingent on the way one thinks.
Trying to hold back the frustration I felt watching this video - my major takeaway is this: Maybe this video is better suited for someone else who doesn't have a healthy separation between ideological tools and self. I do not hold onto ideas like they are permanent augmentations of my self. As such, I can simultaneously value the Monty Hall game theory while also casting it aside when presented with information that is outside the bounds of its applicability. Side note: Who the fuck has ever been presented with a real life monty hall problem? That shit ain't real, it basically only exists as a theoretical mathematical proof.
@@MysticalRefpanel Good question, I worded the beginning poorly - probably because I was processing as I was typing lol. IICR; I was not frustrated so much after I had discovered the appropriate takeaway for me. I felt frustration for the dogmatism and lack of nuance used in the video. The cope I needed was interpreting those tools as tactics to persuade certain listeners and not as heuristics all people should follow.
I feel like this is being unnecessarily harsh on statistics, for example you seem to dislike that statistics is "based on rigid assumptions which are knowably and provably untrue", but simplifying assumptions are important and useful in many cases. When you learn about projectile motion in physics, you are taught to assume that air resistance due to friction is 0 which is provably false but does that make the underlying equations bad? No, they are still very useful even though they're technically not "true". Relating this back to statistics, yes you're correct that we're assuming that the car/goat/whichever outcome you like best has an equally likely chance of being behind any of the three doors and that this assumption may not be correct, but isn't it still useful and cool that we can know what the right answer is IF that assumption were correct? And isn't it useful/cool that if we relaxed our assumption (and figured out the true probabilities), we could still use the framework of statistics to come up with an accurate answer? Idk I guess I'm just saying that from my layman point of view, statistics seems kind of amazing and I feel like you're being overly dismissive of it. Of course, your perspective is probably more informed than mine given that you've studied (and taught I think?) statistics so I acknowledge that I'm probably underlooking some things and missing some perspective. Also fwiw, I'm not a huge fan of the argument that statistics is bad because it was originally used for bad things (eugenics in this case). Statistics can be used for good or bad things and what it's used for doesn't make it inherently good or bad, it's just a tool. It's our job as humans to try to use it for good things. Anyway, thanks for the video, it was interesting and thought provoking at the very least even if I don't fully agree with everything presented.
I agree statistics is fascinating, and some results indubitably useful, even potentially wholesome, and shouldn't be totally discounted even if its original purpose was malicious. But the argument here wasn't that it's "bad," but that we shouldn't worship it the way it often is. It also helps put in perspective why these concepts might be so hard to grasp intuitively. Not because people are dumb, but because our brains and social concept developed with different sets of assumptions. It's politically and philosophically interesting to ask why those assumptions are so different.
Many statistical models leave out some aspect of reality with no clear statement or explanation of why, and there aren't always explicit assumptions and oversights presented with those models. Those kinds of details are often important when deciding what lessons are transferable. From my personal experience, strategy gamers who use statistics for their own arguments tend to frame arguments as "here are numbers that you can't argue against because all of this math is correct", without recognizing that their fundamental assumptions don't align with the real life situation. I agree that there's still a lot of value from statistics even from imperfect models, but I think that the average person shouldn't blindly use statistics with absolute confidence
@@ZaItan1 I definitely agree it's good to question assumptions and to think critically about our sources of knowledge. But I will also note that the last slide is heavily implying that trying to use statistics to solve problems is no better than praying to Athena which to me seems like the argument is indeed "statistics is bad". Although, maybe Jorbs is just saying something like "statistics is bad when you don't understand how to do statistics" which does sound a lot more defensible to me and I would probably agree with that. But then he's also saying stuff like "...[statistics] doesn't mean anything if you get it right" (third-to-last slide) which kinda sounds like a condemnation of the whole field.
I think it's like the weather forecast. It's unreliable; in earlier years because we didn't know how to calculate it, nowadays because we lack the data to make exact predictions. Just because the model is used without all the data filled in, that doesn't mean the model is wrong or useless
@reidmcinroy8629 if the point of the video was to demonstrate that statistcs sometimes ignores reality with no clear statement of why, then the monty hall problem is a horrible choice for an example. That's my main ossue with the video. Every assumption made in the monty hall problem is clearly stated basically every time it's posed. Yes, you do want the car. Yes, Monty does have to open a door. Yes, the door which contains the car is completely random. The formal definition of this problem isnt vague at all. There are plenty of times stastics leads to incorrect assumptions because they make unspoken assumptions. That is a good lesson to teach. Unfortunately, the example chosen just ... isnt one of them and it makes the message muddled.
Never knew that there were actually many written versions of the Iliad spread across multiple cultures. Probably the most interesting part of the video to me.
This could use a clarifying example of a problem that can be solved with game theory and with the ancient technology, yielding alternative solutions. As it is dual-determinism doesn't sound like a decision-making process. What uncertainty is Achilles trying to get over when he decides to work on being a warrior while favored by his deity? Would Achilles' approach to the Monty Hall problem surprise us with new insights the probability calculations couldn't get at? Associating game theory with eugenics is suggesting something you don't say explicitly. What is it? It's not as if humanity's capacity for racism and reverse-justification reasoning for racism lay dormant until the advent of game theory. Anti-jorbs could harshly criticize dual-determinism by highlighting its role in some historic atrocity, while gesturing towards comparatively innocuous uses of game theory. That wouldn't illuminate much either. Hence the need for a like-to-like comparison of solutions to, well, compare the two decision-making methods
one thing about dual determinism is that it collapses your option space a lot. achilles chooses between becoming a famous warrior (because he is a strong warrior and favored by zeus) or becoming a farmer on an island (because he is good at farming and his mother is a sea goddess). he has clarity that he can achieve either, but few thoughts that he could do something else. a large benefit of dual determinism as a decision-making technology is that it eliminates many options immediately, so you're not seeing the payoff in the same part of the decision-making process as you do with game theory. he ultimately makes his choice between the two options emotionally, which is absolutely another decision-making technology we have, and one a lot of us could be more honest about using regularly xD. in all cases though, these are just decision-making technologies. none of them guarantee a better outcome than others in all situations, and a "best" life includes utilizing many different ones as needed. i'm not speaking to the idea that there are other types of decision-making technology which we should all adopt over game theory, but rather against the idea that game theory is a sensible way for everyone to make lots of decisions. it's a thought which nags at me whenever game theory enters public discourse, like with elections, collective action problems, etc.; i often want people to realize that not everyone makes decisions based on game theory, that they themselves don't have to, and that doing so doesn't necessitate better outcomes than using other decision-making technologies.
small addition: i personally think the most important modern decision-making technology is actually ethics, and that if you want to learn something at a school somewhere to help you make decisions, ethics is the field to be studying. fun bonus: you will get to learn a lot about very smart people trying to work out how to put reins on game theory so it can be placed into real-life contexts to try to stop it from making us farm babies for organs or w/e, which is nice knowledge to have if you want to try applying things like game theory to situations in your own life.
Sounds like dual determinism is a useless "technology" since it adds no useful information to the decision and the user just ended up making an emotional decision (like 99% of people do anyways lol). By contrast any technology that attempts to introduce additional information or context to a decision is usually helpful even if it's not perfect.
@@tehy123I'm not sure that "adding new facts to the mix is usually useful" is actually as simple and correct of a claim as you make it out to be Let me give you a few facts: When you apply pressure to ice, it becomes water. Small amounts of water on smooth surfaces can make them very slippery. Ice is very slippery. Given these facts, you can probably create a pretty convincing model of why ice is slippery - and indeed, most physicists have actually believed the same theory for a while. It's profoundly rational - we have a logically consistent model of how the world works, connecting a whole bunch of true facts about how things behave. We've also got a lot of gaps - like the question "why is ice slippery?". So naturally, when we get new facts, we use them to fill in those gaps. Problem is, the theory that ice is slippery because of small layers of liquid water that form between it and other objects is actually totally wrong, and physicists who study this have known that the math for that explanation didn't line up with reality at all - but it wasn't until earlier this year that we actually found what appears to be the correct explanation. Nonetheless, even most physicists believed that explanation for the longest time, because most physicists don't study the slipperiness of water, so it doesn't matter. The reason I bring this up is because I stated a few totally true, simple facts about physics - and it turns out even physicists aren't qualified to extrapolate from them. but we all extrapolate anyways, because our brains hate gaps. This is pretty harmless when we're talking about why ice is slippery, but we do it all the time for things that have a real impact on the world and our lives. To bring this all back to the original point, this is (I believe) where jorbs' criticism of game theory lies. I don't think he's saying that game theory is useless or that dual determination is objectively better than it. What he's saying is that game theory produces facts that even most mathematicians aren't qualified to make decisions with. It's like my example, except instead of having the wrong explanation for something trivial, when people get to apply the prisoner's dilemma to real situations, they make real awful decisions that hurt real people, and usually themselves too. Emotional decisions, though, are a bit different. We are often led to believe that making decisions based on emotions is bad, but when you think about it, emotions evolved in conjunction with the human brain, and actually experienced some evolutionary pressure to make more good decisions than bad ones. I'd say the best thing is to temper emotions with rational decisionmaking - but I don't actually know if I can say that either, because if you tell someone the facts about how the math of the prisoner's dilemma works out and they use that fact to logically deduce that they should screw their partner in a shady operation, their rational, fact-based decision could lead to them getting shot in the head.
Just because emotions evolved in conjunction with the human brain doesn't mean that they're great at decision-making. At best they might be OK at it given time and brainpower constraints. At worst they might have other stuff going on under the hood. Speaking of getting shot in the head, most cases where criminals get shot and killed seem to be based on impulsive decision-making that doesn't stand up to basic cost/benefit analysis. I can't speak to whether or not criminals could benefit from game theory specifically - it's a niche field after all - but people involved in shady business deals are definitely not too rational and divorced from their emotions. And plenty of them get shot anyways! Either way, while I'm not an expert on game theory, I don't think it matches your example. More knowledge is usually good even if it can sometimes, very rarely, be misleading. If I was involved in a shady business deal and I learned game theory, I would probably just try to leave it behind if I couldn't trust my partner, which is definitely the right life decision.
I talked to my mum about this problem before. She understood the 100 door version instantly, the 99 version etc, but at 3, she thought it felt "different", like it was a special case when its 3, because she picked a door, one closed so shes only picking another. I wasnt able to explain it in a way that made intuitive sense for the 3 door problem.
I dunno, I think of it as taking the 1 in 3 shot (your original choice) or taking the 1 in 2 (the switch). Giving the problem more doors makes it easier to visualize, but essentially you get better odds as soon as an option is eliminated.
@@screamingliner The switch is not 1 in 2, it is 2 in 3, because the important thing is how often the switching door will be the winner option, not how many options there are left. So if your choice only contains the prize in 1 out of 3 attempts, the host will be who leaves it hidden in the other door that leaves closed in the 2 out of 3 times that you start failing.
I feel like people are prone to overly attaching themselves to their choices, and this is a consequence of that. You want your initial choice to be right, so you impart more value on it than it actually has and aren't willing to give up on it. It fits both people overvaluing sticking with their initial door and with people struggling to change their minds when given evidence refuting their beliefs. You could probably make it obvious that switching doors is the correct option by doing hundreds of trials and noting how often switching vs staying would win, but if you go so far to change someone's mind they'll probably just be annoyed with you more than anything.
The one thing people miss when explaining this to others is the critical factor in the scenario is that Monty MUST open a door, and CANNOT reveal the car. This is the crucial piece of information that must be clearly conveyed. I've talked to people who believe they understand the Monty Hall problem and still don't actually understand it properly because they fail to realise why this is the important bit.
As a someone who has acquired multiple advanced degrees in mathematics, published academic papers in mathematics, and has taught post-secondary mathematics for over a decade, I love the analogy to heart surgery - will happily and readily admit I am bad at statistics and game theory (in part because it is so far removed from my field of expertise, but also because statistics and game theory are so antithetical to every-day "common sense"). I think another point to raise with respect to stats is that they are only really meaningful in the context of a large number of repeated trials/experiments: as an example, the "expected value" from rolling a 6-sided fair die is 3.5 which is not a result you can ever actually achieve. Similarly, game theory makes assumptions about actors "behaving logically," which humans often won't do - just consider a particularly spiteful CEO who'd rather screw over a competitor than maximize their own gain.
Even in your last example, there is an implied assumption that the “logical” thing is more money is better than everything else so screwing over a competitor is a smaller gain for the CEO than maximizing money. But maybe, like Jorbs said, the thing you want to maximise might be making the people you love happy, or making the people you hate unhappy.
I once played a game called Zero Time DIlemma that made it easier for me. Instead of 3 doors, there were 10, so if you were going to blindly pick a door, there was a 1/10 chance that you picked the correct one. We then get rid of all the other 8 wrong answers, at this point we bet on the 9/10 chance that we picked the wrong door at the beginning, and trade it for the other one
My biggest problem with the monty hall problem is that most of the time they dont mention that the host knows which door is the right one before opening one and then you get treated like an idiot for saying it dosnt matter if you switch or not. I totaly agree that if he knows and has to choose a wrong door you should switch but when you dont mention it the solution becomes different
interesting also to think about this dual determinism in relation to the concept of the locus of control. by saying that there is this external force which determines the outcome but which you work in concert with and might be able to sway, it's like, encoding the component of chance in the language of effort and control. whereas thinking statistically, we have to contend with the idea that beyond a certain point we cannot improve our outcomes with more effort or desire and the outcome is entirely down to random chance. easy to say that's more logical, but human psychology or at least society is pretty foundationally illogical in some ways. it's all contextual but how a mode of decision-making shapes a person's locus of control is a meaningful consideration. to fit it to the statistical mode, you might need to adjust for the deleterious effect you would expect an external locus of control to have on the expected outcomes.
until i read your comment, the reason that jorbs chose to talk about the illiad, greek gods etc. kind of whooshed over my head; the perspective feels clearer to me now; thank you! have a nice day!
When I was studying economics, the thing I always found baffling is that new concepts would often start with "We're going to assume a dollar holds similar utility to whoever gets it." Which... Just... I'm starting to see why we have the problems we have.
I showed this problem to people a number of years ago (though with 2 empty doors rather than 2 goats), and was surprised by how many people got the math wrong. Even people that know a lot of math and that are pretty good at math usually gave the wrong answer. I think it's due to people that learn math sloppily applying heuristics for probability that either don't apply in this case or are applied at the wrong point in the problem. So most of them apply the heuristic without closely examining whether it applies or how to apply it. I agree that there are a lot of things in statistics that don't make intuitive sense. Bayesian statistics and quantum mechanics in particular stand out to me as not fitting my intuition. I also agree that many of the assumptions in probability problems are rigid. Often, in order to form the simple question with clear answers, you need to make enough assumptions that the question becomes divorced from reality. I see statistics as an important tool for evaluating risk and evidence in science. It can be used to inform decision making, but you have to be cognizant of the assumptions that are implicit in the framing of the problem.
i was 99% sure before the vid that switching was right but i was legit nervous that i was wrong and jorbs would laugh at me (in a nice professional way) as he showed why
The best explanation to simplify it that I have heard is, imagine there are 1000 doors and you choose, then the host eliminates 998. Obviously you’d want to switch, and the same is true for 997,996 etc. down to 3 doors. Really makes it click for me
I have always found the hardest thing to understand about the Monty Hall Problem to be why the sports car is assumed to be the desired choice. Thanks for acknowledging in your statement of the problem that this is not always a given lol Edit: okay I wrote that after only having seen the first slide lol I didn't expect it to be elaborated on later but I am all here for it
@@andrewzmorris But what if you wanted that goat, and you didn't want the paperwork and headache of buying the sports car? Like what if you just want a goat, bro?
The sports car is assumed to be a desired choice within a society that produces game shows because a) the majority of people would have a subjective preference for a sports car and b) the sports car has an objectively greater value than a goat. In fact, the subjective preference of the majority entirely rests upon objective material conditions. The exchange-value of the car far outstrips that of the goat as the car objectively requires more human labor to produce. Use-value is a bit of a different scenario, with subjectivity thrown into the mix, but even so, a number of people who have a use for a car far outstrips the number of people who have a use for a goat. All of this applies only to a market-based capitalist economy, but again, that is the kind of society that produces game shows and the society that we live in.
When I frame this problem, I usually remove the goat because I think the goat is confusing. I just say 1 door has a pile of money, and the other doors are empty. The value judgement of what's behind the doors isn't supposed to be a part of the problem
Data and statistics can be used to find vulnerable people. Often the goal is to exploit those people for profit, but imagine the world we could live in if that data was instead used to give help to the people who need it most
Like many of your videos this one was very helpful to me because I used to be really interested in game theory, but that was at a time where I was a much less empathetic person
Obviously I have no opinion because it was just uploaded but JORBS ON A POWERPOINT?!?! My weekend has been perfected. He always makes me grow as a person from these.
Slightly unrelated, but I was reminded of a seminar from a philosophy professor who believed that ancient Greeks had a less developed sense of self than we currently have, or even none at all, because they state so often that Gods influence their actions. He called it the "Greek zombies" theory. I wondered a lot how it would affect their sense of accountability. Anyway, nice video, thanks for making it!
Always was one of my favorite bits, saying "I want the goat though." Also stating that "I understand the monty hall problem, I just don't believe in it."
Honestly this was a very thought-provoking video for me. I've been spending a while deconstructing my hyper-rationalist tendencies, but struggling because those same tendencies are very good at feeling like they are sound, reasonable, logical arguments. There were a few things in the video, especially near the end, which gave me enough of a jolt that I think it's going to be much easier for me to shake those things free. So, thanks, I think that did a number on me in a good way! This isn't necessarily an argument that was new to me, it reminds me of things that friends have tried to help me internalize and process, but I think this video put it in words that were well-suited to my background/frame of reference, and I think that was very valuable to me. Keep on doing what you do, it's good stuff.
I don't like sports cars, I find them revolting. I do love Monty Hall. It's a great, simple example to demonstrate that our intuition can be wrong and needs calibration. Without proper reflection, you will make decisions contrary to your own interests. I think it shows when truth is better than conviction, even to folks far outside the spectrum. That goes far beyond game theory, it is a statement about philosophy, politics, UX, intrapersonal relations, whatever. The extension of it as "would society be better if people acted game-theory-optimally" I hadn't considered, ever at all. Sort of makes me think statistics is a sort of toxic subject (on account of culture, not substance). Well, that's my intuition anyway. Monty Hall as used to promote "anti-intuition" as opposed to promoting relentless exploitation via statistic optimization, I'm not too worried. Then *should* we be smarter and do math and go to space or should we live in the jungle and worship trees and trade rocks? It we end up there, it's gonna be a while. Now if they put Athena instead of a sports car, it would be a different story.
The last part of this reminds me a bit of a book i picked up recently called storythinking by angus fletcher. I'm not done with it and it's a bit of a strange thing, but it talks about stories as a fundamental mode of human thought and not just a tool subservient to the modern concept of logic. It discusses the difference between the transient "should" and eternal "is", which I've suspected is a little ethics 101 but is new to me and difficult to parse. Your video, I believe, has some overlap here and is helping me to digest some of my thoughts about it. Apologies if this seems like a bit of an unrelated comment, and thanks for an interesting presentation!
Hi Jorbs I love your Shirt and the content you make is always so mindful and interesting even your "lets play's" which are much more than that are always good to watch because of you!
Did you even listen until the end of his sentence before you started writing your "well actually" or do you really assume Jorbs doesn't know that Monty Hall was a real person? The reason I ask is that he acknowledges that the people posing the problem are also referring to a specific game show host _in the very sentence_ you linked to. That doesn't change the point that there are a whole bunch of assumptions and repercussions that stem from the theoretical game show host being male, when they didn't have to specify this.
I think people get caught up with Monty Hall because the maths seem easy to do at first and make intuitive sense, but don't actually work out intuitively. I describe it most simply as "you are most likely to choose wrong on your first pick, so you should most likely switch on your second pick given that Monty can NEVER reveal the correct door." That being said, it's RARELY mentioned that one should pray to Athena first for a sports car and a bountiful harvest before choosing, so I do think this video has a lot of merit still.
Not really, Hiller’s veganism wasn’t related to his bad behavior. Jorbs is saying the conception of stats was intimately related to a bad behavior and that should give us some pause. The linking is important.
hitler taking a census of minority households in germany and what he then did with it is a very relevant historical consideration when taking and storing data from censuses!
it's more that specifically creating databases of human beings is used to control those human beings. like that is it's purpose. this wasn't a general statement it was a hyperspecific observation of one of the main things statistics has been for in its brief histoy.
@@Reashu Statistics and game theory aren't knowledge, though, they're methods of data (knowledge) analysis. The knowledge is much more complicated, which is why it's sometimes useful to apply methods of analysis to it, but those methods are necessarily reductive and incorporate assumptions. It's dangerous to conflate the two things as you seem to be doing here. Both because there can be analysis methods that do incorporate evil assumptions, and because even if you don't have one of those, if you mistake the method for the knowledge you flatten your understanding of the world. There's a reason it's (in)famously easy to create deceptive statistics, sometimes without even meaning to. And there's a reason Jorbs pointed out the assumptions underlying game theory analysis and that we should maybe have problems with some of these assumptions.
unfortunately, the major problem here is that people are bad at keeping discourse / context straight (which makes sense, because they're usually co-created, not a feature of individual thinkers at one time). statistics generally, and to a lesser extent, game theory, are fine in the limited use cases where they are abstractions helping a particular agent make a choice about possible outcomes. sadly, people want to *generalize* from individual instances of statistics helping one make a decision---but those instances were possible because of generalized knowledge both scientific and of the sort mentioned in the video (Iliad style knowledge). seeking to generalize wisdom from individual uses of game theory is like trying to get LLMs/generative AI to "be creative." that's not going to happen, because LLMs aren't spontaneously-or-rationally reacting to a lifetime of encountering myriad complexity, they're predicting the most likely next token based on things they've already "seen" according to predefined rules. this tendency to not properly scope thought and discussion is historically used to manipulate people into acting, collectively, against their best interests (both collective and individual). it's evil, and that's one reason among many that smart people think game theory sucks. game theory is a science-y way of doing the kind of reasoning where someone systematically disadvantages one group according to a certain metric, then talks about the wholesale inferiority of that group in virtue of their failing to thrive, as a group, according to that metric. it's for stupid people and sociopathic abusers (and strategy gamers).
These videos bring out the best possible of all Jorbs. I chose the door with this video, not some bunkum possibly vaguely imagined sports car (that I couldn’t drive on roads here 9 months out of 12, and the insurance alone would require a second job - at least videos are entertaining and goats are delicious).
This spin about the problem being flawed because some people might value the goat higher than the sports car seems absolutely inane to me, it is absolutely clear to anyone with any kind of social experience living of planet earth that within the logic of the problem you are supposed to accept the car as desirable and the goat as undesirable for the problem to make sense, even if that doesn't match with your personal values you would still understand that this is what the problem expects from you and reason as such, any other interpretation is simply bad faith. The problem could be reworded as the two goat doors being empty and the car door instead having x arbitrary item that the person being asked values and it would be functionally the same.
I do believe the world would be a much better place with more people acknowledging „Real World Economics“ (or Plurale Ökonomie as we say in Germany). It’s a small field of economic research which really needs more attention!
Game theory and statistics have a profoundly different and younger history than math. Jorbs highlights their arbitrariness because they were designed initially for eugenics. Math is arbitrary too, as is basically every human endeavor, but there's an ill-will intent in the other case we can meaningfully trace. But yes, it's still important for people to learn math, and all 3 actually.
@ZaItan1 math is funny because it's both made up and an objective truth. What I mean is, the language of mathematics (i.e. 1, 2, +, -) is completely arbitrary, we just made it up. However, the concepts expressed using the language are objectively true. That is to say, if you have 1 of a thing, and you gain another, you now have 2 of that thing. That's objectively true regardless what words we chose to express the quantities
@@smitty215able "the concepts expressed using the language are objectively true" - just like any language it is both the approximation of the real world and it can be used to state some truths. What is troublesome is when people say things like "math is the purest science". That sentiment I find quite disturbing (purified of what, pray tell?)
@@gg829Purified of reliance on other fields. If you accept that math is a science (which seems strange to me), then it uniquely stands alone, physics requires a core of math, chemistry requires a core of physics, biology requires a core of chemistry, and similar statements can be made of everything else (and to the extent that I oversimplified things, unsimplifying them emphasizes the point). (minor edit)
I didn't like this presentation at all, it makes lot of assertions on game theory which have nothing to do with the field at all. The whole point of using utility values is that you can assign them however you want, if you value goats more, you can plug that in the equations and calculate how best to get the goat. Maybe you were more aiming to critique the application of it in e.g economics, instead of viewing statistics and game theory as subfields of math, which is what they actually are. But then this should have been made much clearer in the slides. As it stands, you might as well start discussing about how differential equations are a detriment to human civilization because they have been used to calculate ballistic weapon trajectories.
i mean, that's essentially my point. someone can't use calculus to create ballistic missiles and then argue that it's okay because it's justified by calculus, but we indefensibly allow people to use game theory to launch ballistic missiles while claiming it's justified by game theory
also, if people can say value is whatever they want it to be, a field is no longer knowledge-seeking and is now instead power-reinforcing. these are things often understood by people with developed understanding of game theory and statistics, but that describes a minuscule percentage of human beings.
@@Jorbs So all of mathematics is not knowledge seeking because it doesn't have the concept of (moral) value? That's completely backwards, half the reason fields like analytic philosophy based themselves on mathematical thinking was exactly to be able to reason about the world without abitrary value judgements. Logical positivism itself may have failed, but it led to pivotal advances in the philosophy of science and mathematics. If these are not knowledge seeking, I don't know what is.
Thanks. I dont know quite how to feel about this yet, i was really put off by the last part which felt like pretending game theory and statistics is hilarious and ridiculous to ever be used... But maybe i just need to rethink again and come clear on what exactly isnt clicking with me. Pointing out flaws in the applications of Science, specifically Game Theory & Statistics seem fine but it felt like it went too far and only in one direction? (This might be the part where i was getting impressions that werent even made). Pretending dual-determenism is such a cool thing for badasses and statistics is such a fake made up thing only used for horrible things such as eugenics just doenst seem like nothing jorbs or anyone in their audience would seriously believe
This comment is pretty bizarre. If game theorists - or any group of people trying to understand how something works - were to explicitly try to define what are good and bad outcomes, that would be "power reinforcing". Knowledge seeking means trying to understand how people can achieve their goals without judging which goals are good and bad.
I'm not sure if I missed something in the RPS stats, but I think that adds up to 89, which even if you rounded down all the numbers would max out at like... just shy of 92
I don't really get the part about "it assumes the car to be the desired outcome. But not everyone would think of it as the desired outcome". And by that I mean that I always thought you could just replace it with anything. To me, "sport car" and "goat" are just word put there instead of having a placeholder "good outcome" and "bad outcome". But if that's really a problem with this game problem... Then wouldn't it be easier to say "the desired outcome you get something like money, which can then be used to buy something you'd want. And the undesired outcome is you receive absolutely nothing." Or if you reeeeaaaally want to fit all scenarios including one were money doesn't have any value. Then we can just say "The desired outcome is something good, the undesired outcome is something bad." So we're all on the same page... Right ? It just feels like a weird thing to me... But I guess the problem itself is kind of whacky, so... Yeah. That tracks I guess.
Shoutouts to Zero Time Dilemma for illustrating the Monty Hall problem in a novel way. As an addendum to dual determinism, it always irked me that success was shared with the divine, but often failure was hung squarely on the shoulders of the mortal- a deliberate fallacy that we still use today in some circles.
I feel like you are arguing more against the people that use game theory and the way in which it is applied more than you are arguing against game theory itself. For example at the end there you talk about what game theory says about the climate crisis, but game theory itself does not make any assertions about the climate crisis, only people that try to apply game theory to the climate crisis do. I don't think it makes sense to look at situations where game theory and statistics are misused and then say that game theory and statistics are rubbish because of that. Though I could be taking your point wrong and all you are saying is that we should be aware of the problems with how game theory and statistics are commonly applied, not that we shouldn't use game theory and statistics, in which case I agree with you.
I think there's a great deal of both. Thinking in "Game Theoretical" and "Statistical" fashions can often lead to failed solutions to problems, to the confusion of those deploying such solutions. Stuff like the "effective altruism" movement, even ignoring its main controversies, comes up with very confusing answers as to what people should do to improve the world. Oftentimes when researchers in a rich country try to find solutions to problems in poorer countries in an abstract sense, without having lived in those places and understood what the people need, they develop solutions that are ill-suited to the actual populations for which they are created. Sure, you can chock this up to the "people" being bad rather than the process, but at some point you have to contend with the process of reducing problems and communities to sets of numbers can fail to accurately capture a problem space, as well as give a false sense of certainty as to the projected efficacy of the solution. There's also the issue that just this idea of game theory and statistics, or I think jorbs is saying, selfish optimization becomes a prominent part of casual reasoning and leads to bad places: e.g. most people I know "care" about the climate, but dismiss the idea of consider taking personal actions such as reducing consumption or growing a garden because it's big companies that are mostly responsible anyway, what does my personal consumption really mean. Is this true statistically? In a certain set of assumptions, yes. But it's really not helpful.
He’s not saying statistics and game theory are bad, he’s saying vast majority of people are bad at them so this obsession by people who are bad at these things to use them to support their decision making is ill informed and ill advised, and it wouldn’t yield any better results than the ancient tradition of double determination. Notably, double determination is still relied on by large swathes of modern humans: see, Christians. His point that these disciplines are formalised by bad people does matter to his main argument that they are implicitly biased by cultural factors from the creators and dominant societal thinking, but overall I think they distract us from his main points: via-a-vis this comment chain.
@@undeniablySomeGuy effective altruism is not even game theory or stats, it's a branch of ethics/philosophy. The people you mentioned who "care" about climate would use the same argument whether game theory or stats is invented or not, because well... tragedy of commons happen throughout history. It's very very very very easy to point out problems with any process in real life because it turns out real life is messy af. What's tough is prescribing an actual solution which this video haven't actually done so at all. It's a complete non sequitur to say that process A is worse/equally effective as another process B by pointing out problems A have. 1. "This thing is bad because of the problems it create under some scenarios" (stats/game theory) 2. "Therefore it's equally effective/worse than this other thing because of the problems in 1" (dual determinism, i.e. religious beliefs) 2 does not follow from 1.
I hate to break it to you, but the reason people are willing to yap about X cause but not make any changes to their comfortable life, is definitely not because of math or statistics.
I know you will likely dislike this comparison, but this video has a great number of similarities and narrative outcome with what Jordan Peterson lectured about in his Maps of Meaning course.
I cant believe my favorite video of the year is a 30 minute powerpoint presentation in which jorbs uses the monty hall problem as a spring board from which to make a case for greek polytheism
As someone who does statistics for a living... I like to think I've graduated from "bad at statistics" to "mediocre at best at statistics" ;P I will push back a bit on the "provably false" part for the rigid assumptions - it's a bit more that the axiomatic assumptions underpinning statistics are *unprovable* It's a subtle distinction, but it's more that proving Wonder Woman's Invisible Jet is neon yellow - if you could see the color it reflects/emits, it wouldn't be invisible.
Game theory can absolutely arrive at the Golden Rule. Look up the Prisoner's Dilemma. Veratasium did a great video on it a while back, and the conclusions that game theory leads to are entirely that being among the most cooperative and forgiving people in the room leads to the best outcome for yourself.
In my (not very😀) humble opinion, the video didn't realy explicitly name what I feel is the core issue. Math (including statistics and game theory) in its purest form exist only in an abstract world and it is meant to be used there. There are fields that use mathematic tools, but only as a tools. For example, to determine how far can you throw a stone, you would probably use physics to describe the motion of the stone and also probably biology to understand how you are throwing it (e.g. to determine force and angle your muscles and bones are capable of), maybe even psychology (if you are angry, you would probably throw with bigger force), etc. Yes, you are using (mathematicians would say applying) math to help with this task, but you would not be able to do it with just math. There are simply no stones or people to work with in the beautiful and (more importantly) abstract math world. One problem is, we are generaly very bad at navigating in this abstract math world, so we often tell ourselves little stories to help us feel we understand what things are. Goats, sport cars and doors in the Monty Hall Problem are great examples of them. But often these stories are actually hurting our ability to understand, rather than helping it. Notice when jorbs tried to explain the original correct solution to the Monty Hall Problem, he did it by making the problem more abstract, not less. He switched from doors and goats in a game show to squares, letters, numbers and arrows. The problem itself became a little harder to follow (asuming you don't have prior knowledge about the original formulation with goats behind doors etc.), but easier to solve. (I believe that making the problem more and more abstract would eventually made the problem totaly incomprihensible for humans while also self-evident). Making the problem less abstract by packing it with all the emotional and cultural baggage managed not only to make it less intuitive but, as jorbs argued in the video, it outright changed what the correct solution is. Game theory and statistics are also operating in the abstract world, as they are part of math. But making stories about inhabitants of the math world made lot of people (including some mathematicians) think they are real, as real as goats (or humans for that matter). But they are not. E.g. game theoretical "rational agents" are not people, they are abstractions and thinking about them as people is just wrong. It's like using a saw to hammer a nail, you can succed sometimes, but you can also fail spectacularly, it is just not a right tool for that. And if a saw is the only thing you have available, you should probably use it to make yourself a hammer (i.e. tool which is actually suitable for the task) instead of just repeatedly bashing the nail with it. TL:DR, blindly using math in the real world without factoring in context - bad, using context to know how (or if) you should use math - good. Anyway, thanks for listening to my TED talk😀. Disclaimer: Sorry for all potential spelling or gramar mistakes, english is not my first language, it's too hot outside today, I haven't slept well, yada yada...
Replacing "virgin" with "incel" in the "chad vs virgin" meme format is fundamentally genius and more accurate than relying on purity culture rhetoric like "virgin" and I approve 1000 fold.
Yo, I’ve rarely seen UA-cam-Jorbs this unhinged. Jorbs woke up and chose sarcasm 'n sh*t. Did people actually annoy you this much into making this video? I really enjoyed it, though.
What I took away from this is that you portrayed dual-determination as the handsome muscular gigachad and game theory as the incel virgin soyjak. Game theory is no more
Every time I try to understand what statistics is, and what it means to assign probabilities to things which already are (what is the probability I have COVID?) I find it a really mind fuck. I'd love to actually read more about what I guess is the philosophy of statistics, but I've no idea what books would be good.
my biggest grievance with the monty hall problem is that it frames the choice to stay with the first door as a continuation of the 1/3 choice. Lets assume we pick door A, door B opens and the desired car is not there, and then we are offered to switch to door C or keep door A. I see no reason to frame the choice to stay with A as a continuation of the same choice, when we have new information. We can't go back and change the previous choice, and it has no bearing on what we now know or now can do, so why is it framed as the same choice? It seems clear from a practical viewpoint that this is a seperate choice with new information, we have door A and C, and we can pick either one. EDIT: wrote goat instead of car, now fixed EDIT: Also nice shirt :)
The disparity comes because as the host must reveal a door that is not which you chose and neither which has the car, then when yours is the same that has the car he is free to reveal any of the other two, making it uncertain which he will prefer, but when yours has a goat, he is restricted to reveal specifically which has the only other goat. In that way, if you start choosing door A and he opens door B, we know that the revelation of door B was 100% mandatory in case the car were in door C, because he wouldn't have had another choice. But if the car were in door A (yours), it was only 50% likely that he would open door B, not 100% sure, because he could have preferred to open door C instead. Because of that, it is twice as likely that the reason why he opened B and not C is because C is which contains the car, rather than because A contains it. In the long run every door would tend to be correct with the same frequency, but once you start choosing A, for every two games that it is correct the host would use one to open door B and the other to open door C, on average. But for every two games that door C has the car, he would be forced to open door B in both, so twice as many.
@@RonaldABG im sorry, i was tired when i wrote that and wrote goat by mistake. I was referring the the original problem and assumptions that there is a desired prize behind one door, and nothing/an undesired prize behind the other two. i guess just replace goat with car in by comment and it is what I actaully wanted to say.
@@michaelcabrera6210 Anyway, what I put still applies. If your door A were correct you don't know if he would have revealed door B or door C, but if door C were correct, you know for sure he would have revealed door B.
I hate the goats vs car, to me it would be better answered by item you want, and two empty door, and the game show host is a all knowing being that has to do only one thing open all doors to nothing that are not currently picked and the one that has the item you want. To me we should use this as stepping stone to show that not everyone is out to stop us, and that an all knowing being is not always out to screw us over.
If you made your initial pick and then were asked if you would like to switch and can now open both remaining doors instead would that fix your intuition?
I come from a tradition of tearing down previous traditions for great good and great ill. The previous traditions either say that it is my destiny (inescapable) to self-destruct for violating the natural order or that people such as I should be shunned as my existence is corrosive to the society which is what the tradition lives in. What I am heir to is the result of the things that made this world what it is now. It says little about me directly, but tells me that I can aspire for more, and it cares not if that is for ill or good. I think I will choose instead to make something better out of all the flawed things that came before, and teach my children to do the same.
The thing that often isn't brought up in consideration of this problem is the psychological damage of having essentially "won" something and then giving that up (by changing your choice to the losing door) versus just picking the wrong door in the first place. I think this has a lot of sway over people who are in this position - imagine you'd picked the winning door and you were silly enough to swap it? Whereas if you'd picked the losing door, well, you only had a 1 in 3 chance right, so that's not on you? Mathematically the right choice is to switch (assuming you don't want a goat for some weird reason), but that's not really how brains work.
I don't know about you but I'd feel pretty darned silly if I lost a car because I decided to go with a 1/3 chance instead of a strictly better 1/2 chance.
@@Math.Bandit*2/3, not just 1/2. The choice of swapping is the choice between picking the contents of one door or the max of the contents of two doors.
Another western science assumption that has upset me ever since I first learned of it was the one about kids and marshmallows. You have a kid sitting alone in a room, but are being watched. You place a marshmallow on the table in front of them, but explain they will get more marshmallows if they don't eat it and wait instead. But what if the kid doesn't trust these candymen? They could very well withold marshmallows indefinitely, this might be the only chance at a marshmallow. But you don't know for sure, so you wait a bit. But notice the longer the kid waits, the more convinced they become the adults are not going to deliver. It takes some serious conditioning to not only overcome impulses, but also decide that most other people are not generally trying to hurt or steal from you. Is that really a prophecy of intelligence then? Or a coincidental indicator of supportive upbringing?
That's not a western science assumption. You are instead describing an experiment that was performed, and a correlation that was discovered. As you say, there are lots of ways to interpret that correlation. Science as a whole does not assume any particular interpretation to be correct - that's not the role of science. Instead that is used to construct theories, and those theories can be tested in other experiments and then scientists can argue over which theory is strongest.
I don't think that's so much a criticism of the experiment as a very important deduction. People who are conditioned to have low trust in authority also are less likely to defer gains, exactly because of situations like this. Lower stability goes hand in hand with a tendency to make decisions that will pay off short term.
The main issue here is that a lot of social sciences are pretty ass, the people making the study have a set of beliefs around what is and isn't true and will twist data into supporting their theorem without taking into account alternate viewpoints. In this example they could have progressed onwards to do more experiments to try and remove some elements and see what changes. For example, maybe you have a machine which automatically releases the marshmallows after X minutes or something similar. How does that change it?
I never realised that the Monthy Hall problem actually has very little to do with probability and everything to do with figuring out if the game host is adversarial towards you. If you correctly deduct that the host is adversarial, you always win. If you fail to identify an adversarial host, you always lose. One does not need to consider any probabilities to deduct this fact. Figuring out what happens after you correctly deduct that the host is not adversarial towards you feels like a secondary priority under these circumstances.
It's not, the premise of the problem states the rules the host acts under, and it's based on a real scenario of which we know the host wasn't acting adversarially. You can state a problem where you don't know whether the host is adversarial, but then it's not the Monty Hall problem.
Maybe in 11% of cases, the player fails to choose a definite answer and thus doesn't fall in either category. But if that was the case, it would probably have been said.
8:00 "the problem assumes you want a sports car more than a goat" Yes. This, 100 times, yes! I feel like this points out very eloquently why game theory is much more limited than it seems when it comes to real life applications. Essentially, what game theory is, is a function that takes as input the value hierarchies of one or more agents, applies them to a particular game, and outputs a winning strategy. The value hierarchies determine the winning strategy, yes, but game theory does not determine/prescribe the value hierarchies. It has nothing to say on the matter. Those are *givens* . To say that game theory can tell you what you should fundamentally want is equivalent to saying that when using the function f(x)=x^2, you should have x be equal to 6.
But who says "game theory tells you what you want"? It's always "game theory tells you how to get what you want" or (probably more commonly) "game theory tells you how people will act, if you can predict what they want".
1) you have valid points but the part about cultural value is a bit pointless. yes, it is true, but the problem could be rewritten with anything that would make sense for a given culture. or, given that it is mostly used as a toy problem nowadays, it could just be "bad thing, good thing" without specifying anything in particular. the statistics still apply. 2) and speaking about toy problem, this started as a real world "problem", not a toy problem. 3) the host is described as a "he" and called monty hall because monty hall is a real person and a man.
Yeah, I think this video is a little bit cooked. Most of the assumptions are based on a real life example and used to communicate an idea. That's a pretty good way to teach even if it's not 100% perfect
...Are you not aware that Monty Hall is a real person and Let's Make a Deal was a real gameshow in the 60s that really had sports cars and goats behind real doors? You're presenting this as if the Monty Hall problem was just made up from whole cloth by three white guys in the 90s.
the reason car and goat is used is the same reason the you're speaking english. why are you speaking english? is it not possible for this video to be done in so many other languages? does it have to be a video? why are you on the screen? why do you have a powerpoint? the purpose is communication and not just any communication to any group of people. the goat and car is used to communicate with a demographic that generally values cars over goats. the trivial case of broken car and really really nice goat that can get you exactly what you actually want in life is implicitly discarded. if you were speaking to a demographic that did not value cars over goats, you wouldn't use that example and if you speak to people who dont know what a game show is you won't use that example and if you speak to an audience that doesn't understand english or are unable you hear you, you would probably not use a video that relies on english audio to communicate with them. the reason why game theory is so good is because it doesn't actually have any implicit biases except for the ones you input into the system. game theory does not demand you use game theory all the way through. if im in some life or death situation and i have some being with other inherent decision making capabilities imparted to me through my evolutionary past and the situation is exactly what my other inherent decising making ability is trained for, then game theory says that the probability of me surviving if i use my evolutionary abilities trained to help me survive this situation is gonna be more useful a mechanism for making decision than trying to apply statistics in that situation where i have neither sufficient explicit knowledge or enough time or skill to apply statistics. the key here is if you want something out of game theory you need to put something into game theory. put it "how do i survive this life or death situation" and game theory might say "stop asking and follow your instincts you have no time." if your bias is "how do i make the most money" why are you surprised that game theory is telling you about making money? you're right to say that people are bad at game theory and bad at statistics and that's important to know. game theory is a tool in your toolbox. you pull out that tool when the tool makes sense. it is not a hammer to apply to everything (tho conveniently it is a hammer that with enough awareness, will often tell you to stop using it on the screw and go get your screwdriver.) why are you glorifying society btw? do you not look back on human history and see that maybe the culture of many different societies are fucked and flawed and picking ancient greek decision is a really really fucked argument?
The first time someone introduced me to the Monty Hall problem I got so angry. Not because I didn't get it, but because they didn't get it. The way they said it they never stressed the host HAD to show a false door after the contestant selected and before offering the option to switch. I realized the math was enticing to switch, assumed the game host was trying to entice the switch by showing a false door, and opted to not switch.
Somehow in the moment neither of us realized the disconnect -- I was arguing over the psychology of the host and they were arguing about the statistical incentive. We were only at odds because I assumed the host was a second rational actor, acting in opposition to the contestant, and they failed to explain this was a scenario with only one rational actor, the contestant.
It's annoying, isn't it? Feel like this is why a lot of people struggle at first - bad framing. I wonder if framing it as 'A friend of yours picks one of the two remaining doors, opens it, and reveals a goat' would make people better understand that the situation has changed. Naturally, the relationship with a game show host tends to be an adversarial one, but in the problem as stated Monty provides consistent and helpful information.
Also while I hate to just regurgitate wikipedia, the articles on the problem reference a newspaper interview in which Monty stated that the real show ("Let's Make a Deal") did not, in fact, work like this, and that he would deliberately play with the contestant's expectations. Sometimes his offer would help, and sometimes it would be a trap, and sometimes he would simply make them accept the door they picked - whatever seemed best for the show. "It all depends on [the host's] mood," he said.
Was this a 27 minute lecture about why shovel is the best relic in Slay the Spire?
DIG!
I believe it was a 27 minute lecture about why shovel is actually a block card
I feel like this is misrepresenting what you can do with game theory. Game theory isnt necessarily presctiptive. You dont have to take it as "this is what you should do". Its the question: "If everyone acts to maximize their own value, what happens?" and in a world where many people and corporations act that way, its usefull to understand why something is happening.
Things like: how do we design systems so they cant be exploited by greedy people? How to make better voting systems? How to share items that have different value to different people in some way that you could call "fair"? Or "how do you get corporations that only care about their profit to act for some common good (like fixing the climate or some equality notions)? Are all questions that gametheory can be helpfull to answer.
I totally agree that you should not live your day to day life "according to gametheory", but that doesnt mean the field is useless or bad. Otherwise great explanation of Monty Hall!
I feel like this runs into the same problem as "economics" because yes whether or not a field is descriptive or prescriptive isn't like, anything inherent to it; but it is shaped by people in that field and how they choose to behave. I think if we're going to be descriptive (as we should) we should notice that a lot of people are using game theory to justify and reinforce the "what happens" and even twist it into "and this is what should happen" rather than just understand it.
idk your comment is in no way wrong, nuanced disagreement.
@@FinetalPies yea, there definitly are people that look at game theoretic solution concepts and their takeaway is "everyone should act like this, anything else is wrong" or "acting like this is superior / how things should be", and those are really not people we should listen to. Those people are probably the reason that Jorbs shows such a negative view of the field in this video, and i totally get that, those people suck.
I just really wanted to highlight that these are not the only people doing game theory, and that there are more positive (in terms of social wellfare/common good) ways to use game theoretic ideas.
The problem with game theory presented here is how it is often presented as "the rational way to make decisions" without acknowledging that it's only rational within a huge range of cultural assumptions and arbitrary value judgements.
the video did not claim gametheory to be bad but I appreciate the explanation nontheless myself :>
The preoccupation with not whether or not we want the goat pays off in this video, but I always viewed the Sports Car/Goats as symbolic for "Desirable outcome/Undesirable outcome". To me it's pretty clear that those things were picked because that was an easy shorthand from the framers culture which fit context being constructed around the problem. It's irrelevant what's behind the doors to the theory itself, the only thing that matters if that you want to open a particular one...
Exactly! That bothered me too
The point is that it isn't an easy shorthand. It's an easy shorthand to YOU.
@@karma5390 by YOU do you mean the audience? the people who are going to have access to the show because it is the 1960s and not the internet. sure not everyone in the possible audience value the car over the goat, but would 95% of them? And also its not required for it to match the values of everyone who has ever lived, thats impossible, just enough of the people there.
@@MrMarson123456 The world has progressed since the 1960s. People use the same Monty Hall problem, but now it's widely been brought to have a larger meaning than it did in its original game show context in the same way that the word "broadcast" has come to harbor a more symbolic meaning, rather than the literal casting of seeds broadly. Yet, we still use the goat and the sports car.
The reason jorbs brings up the issue of why the rewards are or are not culturally relevant is because the Monty Hall problem, much like the Prisoner's Dilemma, is presented with the text that it is here, explicitly for the purpose of making it more readable to a certain audience for whom it was made. That makes it easier for certain people to intuit than others.
Nobody is claiming that it is required for every hypothetical to match the values of everyone who has ever lived, that's absurd, and I don't know what in the video made you receive that message.
This video is asking you to consider this typically theoretical problem and its implications on decision-making in the context of real-world decision-making, which has many more nuances and considerations. Yes, the goat and car are useful symbolic shorthands, but we are trying to think about more than just the symbols.
Don't worry Jorbs, ancient Greeks used dual-determination to justify all kinds of genocide. See origin of Sparta or myths of Dionysus,
Rationalize, not justify.
It seems like a nitpick, but it is actually very important.
@@gg829 I mean same with scientific racism, which is the point of comparison here. It wasn't that people invented statistics and decided they will be racist because of it, they invented transatlantic slave trade and the next several centuries of thought got defined by the need for convincing excuses.
@@stw7120 precisely - material reality shapes how we think and not the other way around.
@@gg829 jorbs uses the term justify so I used the term in my flippant comment. It's not that deep.
@@gg829 I don't think this is a convincing summary of what stw7120 said. Yes, material reality can shape how we think, but I think it is ridiculous to claim that the way people think doesn't shape material reality. Humans, more than any other creature, have the capability to impose their thoughts upon the world; to shape the world in a way befitting of on contingent on the way one thinks.
Trying to hold back the frustration I felt watching this video - my major takeaway is this:
Maybe this video is better suited for someone else who doesn't have a healthy separation between ideological tools and self.
I do not hold onto ideas like they are permanent augmentations of my self. As such, I can simultaneously value the Monty Hall game theory while also casting it aside when presented with information that is outside the bounds of its applicability.
Side note: Who the fuck has ever been presented with a real life monty hall problem? That shit ain't real, it basically only exists as a theoretical mathematical proof.
that is a very good takeaway. then why were you frustrated if you think the video is not for you fr
@@MysticalRefpanel Good question, I worded the beginning poorly - probably because I was processing as I was typing lol.
IICR; I was not frustrated so much after I had discovered the appropriate takeaway for me.
I felt frustration for the dogmatism and lack of nuance used in the video. The cope I needed was interpreting those tools as tactics to persuade certain listeners and not as heuristics all people should follow.
I feel like this is being unnecessarily harsh on statistics, for example you seem to dislike that statistics is "based on rigid assumptions which are knowably and provably untrue", but simplifying assumptions are important and useful in many cases. When you learn about projectile motion in physics, you are taught to assume that air resistance due to friction is 0 which is provably false but does that make the underlying equations bad? No, they are still very useful even though they're technically not "true". Relating this back to statistics, yes you're correct that we're assuming that the car/goat/whichever outcome you like best has an equally likely chance of being behind any of the three doors and that this assumption may not be correct, but isn't it still useful and cool that we can know what the right answer is IF that assumption were correct? And isn't it useful/cool that if we relaxed our assumption (and figured out the true probabilities), we could still use the framework of statistics to come up with an accurate answer? Idk I guess I'm just saying that from my layman point of view, statistics seems kind of amazing and I feel like you're being overly dismissive of it. Of course, your perspective is probably more informed than mine given that you've studied (and taught I think?) statistics so I acknowledge that I'm probably underlooking some things and missing some perspective.
Also fwiw, I'm not a huge fan of the argument that statistics is bad because it was originally used for bad things (eugenics in this case). Statistics can be used for good or bad things and what it's used for doesn't make it inherently good or bad, it's just a tool. It's our job as humans to try to use it for good things.
Anyway, thanks for the video, it was interesting and thought provoking at the very least even if I don't fully agree with everything presented.
I agree statistics is fascinating, and some results indubitably useful, even potentially wholesome, and shouldn't be totally discounted even if its original purpose was malicious. But the argument here wasn't that it's "bad," but that we shouldn't worship it the way it often is. It also helps put in perspective why these concepts might be so hard to grasp intuitively. Not because people are dumb, but because our brains and social concept developed with different sets of assumptions. It's politically and philosophically interesting to ask why those assumptions are so different.
Many statistical models leave out some aspect of reality with no clear statement or explanation of why, and there aren't always explicit assumptions and oversights presented with those models. Those kinds of details are often important when deciding what lessons are transferable.
From my personal experience, strategy gamers who use statistics for their own arguments tend to frame arguments as "here are numbers that you can't argue against because all of this math is correct", without recognizing that their fundamental assumptions don't align with the real life situation.
I agree that there's still a lot of value from statistics even from imperfect models, but I think that the average person shouldn't blindly use statistics with absolute confidence
@@ZaItan1 I definitely agree it's good to question assumptions and to think critically about our sources of knowledge. But I will also note that the last slide is heavily implying that trying to use statistics to solve problems is no better than praying to Athena which to me seems like the argument is indeed "statistics is bad". Although, maybe Jorbs is just saying something like "statistics is bad when you don't understand how to do statistics" which does sound a lot more defensible to me and I would probably agree with that. But then he's also saying stuff like "...[statistics] doesn't mean anything if you get it right" (third-to-last slide) which kinda sounds like a condemnation of the whole field.
I think it's like the weather forecast. It's unreliable; in earlier years because we didn't know how to calculate it, nowadays because we lack the data to make exact predictions. Just because the model is used without all the data filled in, that doesn't mean the model is wrong or useless
@reidmcinroy8629 if the point of the video was to demonstrate that statistcs sometimes ignores reality with no clear statement of why, then the monty hall problem is a horrible choice for an example.
That's my main ossue with the video. Every assumption made in the monty hall problem is clearly stated basically every time it's posed. Yes, you do want the car. Yes, Monty does have to open a door. Yes, the door which contains the car is completely random. The formal definition of this problem isnt vague at all.
There are plenty of times stastics leads to incorrect assumptions because they make unspoken assumptions. That is a good lesson to teach. Unfortunately, the example chosen just ... isnt one of them and it makes the message muddled.
Never knew that there were actually many written versions of the Iliad spread across multiple cultures. Probably the most interesting part of the video to me.
This could use a clarifying example of a problem that can be solved with game theory and with the ancient technology, yielding alternative solutions. As it is dual-determinism doesn't sound like a decision-making process. What uncertainty is Achilles trying to get over when he decides to work on being a warrior while favored by his deity? Would Achilles' approach to the Monty Hall problem surprise us with new insights the probability calculations couldn't get at?
Associating game theory with eugenics is suggesting something you don't say explicitly. What is it? It's not as if humanity's capacity for racism and reverse-justification reasoning for racism lay dormant until the advent of game theory. Anti-jorbs could harshly criticize dual-determinism by highlighting its role in some historic atrocity, while gesturing towards comparatively innocuous uses of game theory. That wouldn't illuminate much either. Hence the need for a like-to-like comparison of solutions to, well, compare the two decision-making methods
one thing about dual determinism is that it collapses your option space a lot. achilles chooses between becoming a famous warrior (because he is a strong warrior and favored by zeus) or becoming a farmer on an island (because he is good at farming and his mother is a sea goddess). he has clarity that he can achieve either, but few thoughts that he could do something else. a large benefit of dual determinism as a decision-making technology is that it eliminates many options immediately, so you're not seeing the payoff in the same part of the decision-making process as you do with game theory.
he ultimately makes his choice between the two options emotionally, which is absolutely another decision-making technology we have, and one a lot of us could be more honest about using regularly xD.
in all cases though, these are just decision-making technologies. none of them guarantee a better outcome than others in all situations, and a "best" life includes utilizing many different ones as needed. i'm not speaking to the idea that there are other types of decision-making technology which we should all adopt over game theory, but rather against the idea that game theory is a sensible way for everyone to make lots of decisions. it's a thought which nags at me whenever game theory enters public discourse, like with elections, collective action problems, etc.; i often want people to realize that not everyone makes decisions based on game theory, that they themselves don't have to, and that doing so doesn't necessitate better outcomes than using other decision-making technologies.
small addition: i personally think the most important modern decision-making technology is actually ethics, and that if you want to learn something at a school somewhere to help you make decisions, ethics is the field to be studying. fun bonus: you will get to learn a lot about very smart people trying to work out how to put reins on game theory so it can be placed into real-life contexts to try to stop it from making us farm babies for organs or w/e, which is nice knowledge to have if you want to try applying things like game theory to situations in your own life.
Sounds like dual determinism is a useless "technology" since it adds no useful information to the decision and the user just ended up making an emotional decision (like 99% of people do anyways lol). By contrast any technology that attempts to introduce additional information or context to a decision is usually helpful even if it's not perfect.
@@tehy123I'm not sure that "adding new facts to the mix is usually useful" is actually as simple and correct of a claim as you make it out to be
Let me give you a few facts: When you apply pressure to ice, it becomes water. Small amounts of water on smooth surfaces can make them very slippery. Ice is very slippery.
Given these facts, you can probably create a pretty convincing model of why ice is slippery - and indeed, most physicists have actually believed the same theory for a while. It's profoundly rational - we have a logically consistent model of how the world works, connecting a whole bunch of true facts about how things behave. We've also got a lot of gaps - like the question "why is ice slippery?". So naturally, when we get new facts, we use them to fill in those gaps.
Problem is, the theory that ice is slippery because of small layers of liquid water that form between it and other objects is actually totally wrong, and physicists who study this have known that the math for that explanation didn't line up with reality at all - but it wasn't until earlier this year that we actually found what appears to be the correct explanation. Nonetheless, even most physicists believed that explanation for the longest time, because most physicists don't study the slipperiness of water, so it doesn't matter.
The reason I bring this up is because I stated a few totally true, simple facts about physics - and it turns out even physicists aren't qualified to extrapolate from them. but we all extrapolate anyways, because our brains hate gaps.
This is pretty harmless when we're talking about why ice is slippery, but we do it all the time for things that have a real impact on the world and our lives. To bring this all back to the original point, this is (I believe) where jorbs' criticism of game theory lies. I don't think he's saying that game theory is useless or that dual determination is objectively better than it. What he's saying is that game theory produces facts that even most mathematicians aren't qualified to make decisions with. It's like my example, except instead of having the wrong explanation for something trivial, when people get to apply the prisoner's dilemma to real situations, they make real awful decisions that hurt real people, and usually themselves too.
Emotional decisions, though, are a bit different. We are often led to believe that making decisions based on emotions is bad, but when you think about it, emotions evolved in conjunction with the human brain, and actually experienced some evolutionary pressure to make more good decisions than bad ones. I'd say the best thing is to temper emotions with rational decisionmaking - but I don't actually know if I can say that either, because if you tell someone the facts about how the math of the prisoner's dilemma works out and they use that fact to logically deduce that they should screw their partner in a shady operation, their rational, fact-based decision could lead to them getting shot in the head.
Just because emotions evolved in conjunction with the human brain doesn't mean that they're great at decision-making. At best they might be OK at it given time and brainpower constraints. At worst they might have other stuff going on under the hood.
Speaking of getting shot in the head, most cases where criminals get shot and killed seem to be based on impulsive decision-making that doesn't stand up to basic cost/benefit analysis. I can't speak to whether or not criminals could benefit from game theory specifically - it's a niche field after all - but people involved in shady business deals are definitely not too rational and divorced from their emotions. And plenty of them get shot anyways!
Either way, while I'm not an expert on game theory, I don't think it matches your example. More knowledge is usually good even if it can sometimes, very rarely, be misleading. If I was involved in a shady business deal and I learned game theory, I would probably just try to leave it behind if I couldn't trust my partner, which is definitely the right life decision.
I talked to my mum about this problem before. She understood the 100 door version instantly, the 99 version etc, but at 3, she thought it felt "different", like it was a special case when its 3, because she picked a door, one closed so shes only picking another. I wasnt able to explain it in a way that made intuitive sense for the 3 door problem.
I dunno, I think of it as taking the 1 in 3 shot (your original choice) or taking the 1 in 2 (the switch). Giving the problem more doors makes it easier to visualize, but essentially you get better odds as soon as an option is eliminated.
@@screamingliner The switch is not 1 in 2, it is 2 in 3, because the important thing is how often the switching door will be the winner option, not how many options there are left. So if your choice only contains the prize in 1 out of 3 attempts, the host will be who leaves it hidden in the other door that leaves closed in the 2 out of 3 times that you start failing.
@@RonaldABG indeed. Your chances actually double (1/3 to 2/3) by switching. Not some marginal amount.
I feel like people are prone to overly attaching themselves to their choices, and this is a consequence of that. You want your initial choice to be right, so you impart more value on it than it actually has and aren't willing to give up on it. It fits both people overvaluing sticking with their initial door and with people struggling to change their minds when given evidence refuting their beliefs. You could probably make it obvious that switching doors is the correct option by doing hundreds of trials and noting how often switching vs staying would win, but if you go so far to change someone's mind they'll probably just be annoyed with you more than anything.
The one thing people miss when explaining this to others is the critical factor in the scenario is that Monty MUST open a door, and CANNOT reveal the car. This is the crucial piece of information that must be clearly conveyed. I've talked to people who believe they understand the Monty Hall problem and still don't actually understand it properly because they fail to realise why this is the important bit.
As a someone who has acquired multiple advanced degrees in mathematics, published academic papers in mathematics, and has taught post-secondary mathematics for over a decade, I love the analogy to heart surgery - will happily and readily admit I am bad at statistics and game theory (in part because it is so far removed from my field of expertise, but also because statistics and game theory are so antithetical to every-day "common sense").
I think another point to raise with respect to stats is that they are only really meaningful in the context of a large number of repeated trials/experiments: as an example, the "expected value" from rolling a 6-sided fair die is 3.5 which is not a result you can ever actually achieve. Similarly, game theory makes assumptions about actors "behaving logically," which humans often won't do - just consider a particularly spiteful CEO who'd rather screw over a competitor than maximize their own gain.
Even in your last example, there is an implied assumption that the “logical” thing is more money is better than everything else so screwing over a competitor is a smaller gain for the CEO than maximizing money.
But maybe, like Jorbs said, the thing you want to maximise might be making the people you love happy, or making the people you hate unhappy.
I’ve read explanations about this one and each time I fail to understand. Jorbs has another chance.
I once played a game called Zero Time DIlemma that made it easier for me.
Instead of 3 doors, there were 10, so if you were going to blindly pick a door, there was a 1/10 chance that you picked the correct one.
We then get rid of all the other 8 wrong answers, at this point we bet on the 9/10 chance that we picked the wrong door at the beginning, and trade it for the other one
Just imagine that instead of being asked to switch, you were asked to pick 2 Doors and leave one alone
So, what did you think? 😂
Imagine there's 100 doors and 99 goats. After you pick a door, I open 98 doors with goats. Do you switch?
The critical bit is the host doesn't pick a door at random. They always open a door with a goat.
My biggest problem with the monty hall problem is that most of the time they dont mention that the host knows which door is the right one before opening one and then you get treated like an idiot for saying it dosnt matter if you switch or not. I totaly agree that if he knows and has to choose a wrong door you should switch but when you dont mention it the solution becomes different
interesting also to think about this dual determinism in relation to the concept of the locus of control. by saying that there is this external force which determines the outcome but which you work in concert with and might be able to sway, it's like, encoding the component of chance in the language of effort and control. whereas thinking statistically, we have to contend with the idea that beyond a certain point we cannot improve our outcomes with more effort or desire and the outcome is entirely down to random chance. easy to say that's more logical, but human psychology or at least society is pretty foundationally illogical in some ways. it's all contextual but how a mode of decision-making shapes a person's locus of control is a meaningful consideration. to fit it to the statistical mode, you might need to adjust for the deleterious effect you would expect an external locus of control to have on the expected outcomes.
until i read your comment, the reason that jorbs chose to talk about the illiad, greek gods etc. kind of whooshed over my head;
the perspective feels clearer to me now;
thank you! have a nice day!
I was really hoping this was just a subtweet to the cedh draw discussion in some form
When I was studying economics, the thing I always found baffling is that new concepts would often start with "We're going to assume a dollar holds similar utility to whoever gets it." Which... Just... I'm starting to see why we have the problems we have.
Return to the video about breakpoints for elaboration on that!
I showed this problem to people a number of years ago (though with 2 empty doors rather than 2 goats), and was surprised by how many people got the math wrong. Even people that know a lot of math and that are pretty good at math usually gave the wrong answer. I think it's due to people that learn math sloppily applying heuristics for probability that either don't apply in this case or are applied at the wrong point in the problem. So most of them apply the heuristic without closely examining whether it applies or how to apply it.
I agree that there are a lot of things in statistics that don't make intuitive sense. Bayesian statistics and quantum mechanics in particular stand out to me as not fitting my intuition. I also agree that many of the assumptions in probability problems are rigid. Often, in order to form the simple question with clear answers, you need to make enough assumptions that the question becomes divorced from reality. I see statistics as an important tool for evaluating risk and evidence in science. It can be used to inform decision making, but you have to be cognizant of the assumptions that are implicit in the framing of the problem.
i was 99% sure before the vid that switching was right but i was legit nervous that i was wrong and jorbs would laugh at me (in a nice professional way) as he showed why
The RPS association first has to explain what people are throwing that other 11% of the time…
The best explanation to simplify it that I have heard is, imagine there are 1000 doors and you choose, then the host eliminates 998. Obviously you’d want to switch, and the same is true for 997,996 etc. down to 3 doors. Really makes it click for me
He basically gave this as an example 5:39
When was first told of this problem, I was disappointed to learn you didn't actually get to keep the goat, but it was just a loss state.
I have always found the hardest thing to understand about the Monty Hall Problem to be why the sports car is assumed to be the desired choice. Thanks for acknowledging in your statement of the problem that this is not always a given lol
Edit: okay I wrote that after only having seen the first slide lol I didn't expect it to be elaborated on later but I am all here for it
If you wanted a goat, you still want the sports car since you can sell that and buy a goat and have lots of money left over.
@@andrewzmorris But what if you wanted that goat, and you didn't want the paperwork and headache of buying the sports car? Like what if you just want a goat, bro?
I named my goat Greg 🐐
The sports car is assumed to be a desired choice within a society that produces game shows because a) the majority of people would have a subjective preference for a sports car and b) the sports car has an objectively greater value than a goat.
In fact, the subjective preference of the majority entirely rests upon objective material conditions. The exchange-value of the car far outstrips that of the goat as the car objectively requires more human labor to produce. Use-value is a bit of a different scenario, with subjectivity thrown into the mix, but even so, a number of people who have a use for a car far outstrips the number of people who have a use for a goat.
All of this applies only to a market-based capitalist economy, but again, that is the kind of society that produces game shows and the society that we live in.
When I frame this problem, I usually remove the goat because I think the goat is confusing. I just say 1 door has a pile of money, and the other doors are empty.
The value judgement of what's behind the doors isn't supposed to be a part of the problem
Data and statistics can be used to find vulnerable people. Often the goal is to exploit those people for profit, but imagine the world we could live in if that data was instead used to give help to the people who need it most
Like many of your videos this one was very helpful to me because I used to be really interested in game theory, but that was at a time where I was a much less empathetic person
2 things.
1. I need to know where I can get your shirt.
2. The RPS totals actually add up to 89, so I am very confused
2. that suggests 10%ish of the time the player fucks up the throw, a very serious sport
asdfmovie. The shirt's available only during pride month, though it seems to come back every year.
kind of unbelievable that the official rock paper scissors association website wasn't a reliable source for this information.
Perhaps 11% of the time the first move is to pray to Athena for victory
11% is shotgun
Obviously I have no opinion because it was just uploaded but JORBS ON A POWERPOINT?!?! My weekend has been perfected. He always makes me grow as a person from these.
Slightly unrelated, but I was reminded of a seminar from a philosophy professor who believed that ancient Greeks had a less developed sense of self than we currently have, or even none at all, because they state so often that Gods influence their actions. He called it the "Greek zombies" theory. I wondered a lot how it would affect their sense of accountability.
Anyway, nice video, thanks for making it!
Always was one of my favorite bits, saying "I want the goat though." Also stating that "I understand the monty hall problem, I just don't believe in it."
Its 50/50, either its a sports car or its a goat 🗿
🧠
Honestly this was a very thought-provoking video for me. I've been spending a while deconstructing my hyper-rationalist tendencies, but struggling because those same tendencies are very good at feeling like they are sound, reasonable, logical arguments. There were a few things in the video, especially near the end, which gave me enough of a jolt that I think it's going to be much easier for me to shake those things free. So, thanks, I think that did a number on me in a good way!
This isn't necessarily an argument that was new to me, it reminds me of things that friends have tried to help me internalize and process, but I think this video put it in words that were well-suited to my background/frame of reference, and I think that was very valuable to me. Keep on doing what you do, it's good stuff.
I don't like sports cars, I find them revolting.
I do love Monty Hall. It's a great, simple example to demonstrate that our intuition can be wrong and needs calibration. Without proper reflection, you will make decisions contrary to your own interests. I think it shows when truth is better than conviction, even to folks far outside the spectrum. That goes far beyond game theory, it is a statement about philosophy, politics, UX, intrapersonal relations, whatever. The extension of it as "would society be better if people acted game-theory-optimally" I hadn't considered, ever at all. Sort of makes me think statistics is a sort of toxic subject (on account of culture, not substance). Well, that's my intuition anyway. Monty Hall as used to promote "anti-intuition" as opposed to promoting relentless exploitation via statistic optimization, I'm not too worried. Then *should* we be smarter and do math and go to space or should we live in the jungle and worship trees and trade rocks? It we end up there, it's gonna be a while.
Now if they put Athena instead of a sports car, it would be a different story.
What a twist! From the thumbnail I did not expect this video to involve Jorbs coming out as a religious conservative.
Gotta love those "problems with applying pure mathematics to the real world because in real world we can't isolate for variables" bits
The last part of this reminds me a bit of a book i picked up recently called storythinking by angus fletcher. I'm not done with it and it's a bit of a strange thing, but it talks about stories as a fundamental mode of human thought and not just a tool subservient to the modern concept of logic. It discusses the difference between the transient "should" and eternal "is", which I've suspected is a little ethics 101 but is new to me and difficult to parse. Your video, I believe, has some overlap here and is helping me to digest some of my thoughts about it. Apologies if this seems like a bit of an unrelated comment, and thanks for an interesting presentation!
Every time I watch one of these, I always learn to think in ways I've never thought before.
Hi Jorbs I love your Shirt and the content you make is always so mindful and interesting even your "lets play's" which are much more than that are always good to watch because of you!
Conditional probability is a description of how much information you have.
14:47 The game show host is a he because Monty Hall was a real person and man that this scenario is based on.
Did you even listen until the end of his sentence before you started writing your "well actually" or do you really assume Jorbs doesn't know that Monty Hall was a real person? The reason I ask is that he acknowledges that the people posing the problem are also referring to a specific game show host _in the very sentence_ you linked to.
That doesn't change the point that there are a whole bunch of assumptions and repercussions that stem from the theoretical game show host being male, when they didn't have to specify this.
Does the problem change if the game show host is not a man?
I think people get caught up with Monty Hall because the maths seem easy to do at first and make intuitive sense, but don't actually work out intuitively. I describe it most simply as "you are most likely to choose wrong on your first pick, so you should most likely switch on your second pick given that Monty can NEVER reveal the correct door." That being said, it's RARELY mentioned that one should pray to Athena first for a sports car and a bountiful harvest before choosing, so I do think this video has a lot of merit still.
35% + 28% + 26% = 89%. Apparently, 11% of the time, people do not choose any of the 3 options available to them in the game
It‘s actually Rock, 35.4%, Paper 35%, scissors 29%. However men start with Rock at close to 50% and women tend to start with paper.
The video is entertaining but the last few slides are basically 'did you know HITLER was also VEGAN?'
Not really, Hiller’s veganism wasn’t related to his bad behavior. Jorbs is saying the conception of stats was intimately related to a bad behavior and that should give us some pause. The linking is important.
hitler taking a census of minority households in germany and what he then did with it is a very relevant historical consideration when taking and storing data from censuses!
Knowledge is power, and evil people do evil things with power. Doesn't mean that knowledge is evil.
it's more that specifically creating databases of human beings is used to control those human beings. like that is it's purpose. this wasn't a general statement it was a hyperspecific observation of one of the main things statistics has been for in its brief histoy.
@@Reashu Statistics and game theory aren't knowledge, though, they're methods of data (knowledge) analysis. The knowledge is much more complicated, which is why it's sometimes useful to apply methods of analysis to it, but those methods are necessarily reductive and incorporate assumptions. It's dangerous to conflate the two things as you seem to be doing here. Both because there can be analysis methods that do incorporate evil assumptions, and because even if you don't have one of those, if you mistake the method for the knowledge you flatten your understanding of the world. There's a reason it's (in)famously easy to create deceptive statistics, sometimes without even meaning to. And there's a reason Jorbs pointed out the assumptions underlying game theory analysis and that we should maybe have problems with some of these assumptions.
unfortunately, the major problem here is that people are bad at keeping discourse / context straight (which makes sense, because they're usually co-created, not a feature of individual thinkers at one time). statistics generally, and to a lesser extent, game theory, are fine in the limited use cases where they are abstractions helping a particular agent make a choice about possible outcomes. sadly, people want to *generalize* from individual instances of statistics helping one make a decision---but those instances were possible because of generalized knowledge both scientific and of the sort mentioned in the video (Iliad style knowledge). seeking to generalize wisdom from individual uses of game theory is like trying to get LLMs/generative AI to "be creative." that's not going to happen, because LLMs aren't spontaneously-or-rationally reacting to a lifetime of encountering myriad complexity, they're predicting the most likely next token based on things they've already "seen" according to predefined rules.
this tendency to not properly scope thought and discussion is historically used to manipulate people into acting, collectively, against their best interests (both collective and individual). it's evil, and that's one reason among many that smart people think game theory sucks. game theory is a science-y way of doing the kind of reasoning where someone systematically disadvantages one group according to a certain metric, then talks about the wholesale inferiority of that group in virtue of their failing to thrive, as a group, according to that metric. it's for stupid people and sociopathic abusers (and strategy gamers).
These videos bring out the best possible of all Jorbs. I chose the door with this video, not some bunkum possibly vaguely imagined sports car (that I couldn’t drive on roads here 9 months out of 12, and the insurance alone would require a second job - at least videos are entertaining and goats are delicious).
This spin about the problem being flawed because some people might value the goat higher than the sports car seems absolutely inane to me, it is absolutely clear to anyone with any kind of social experience living of planet earth that within the logic of the problem you are supposed to accept the car as desirable and the goat as undesirable for the problem to make sense, even if that doesn't match with your personal values you would still understand that this is what the problem expects from you and reason as such, any other interpretation is simply bad faith. The problem could be reworded as the two goat doors being empty and the car door instead having x arbitrary item that the person being asked values and it would be functionally the same.
I do believe the world would be a much better place with more people acknowledging „Real World Economics“ (or Plurale Ökonomie as we say in Germany). It’s a small field of economic research which really needs more attention!
Ancient gamer rants about not being in total control.
Magic twitter leading to great jorbs videos is a wonderful coincidence
If game theory is 'made up' then so is mathematics. That's a fine assertion but it's important for people to understand math.
Game theory and statistics have a profoundly different and younger history than math. Jorbs highlights their arbitrariness because they were designed initially for eugenics. Math is arbitrary too, as is basically every human endeavor, but there's an ill-will intent in the other case we can meaningfully trace.
But yes, it's still important for people to learn math, and all 3 actually.
@ZaItan1 math is funny because it's both made up and an objective truth. What I mean is, the language of mathematics (i.e. 1, 2, +, -) is completely arbitrary, we just made it up. However, the concepts expressed using the language are objectively true. That is to say, if you have 1 of a thing, and you gain another, you now have 2 of that thing. That's objectively true regardless what words we chose to express the quantities
@@smitty215able "the concepts expressed using the language are objectively true" - just like any language it is both the approximation of the real world and it can be used to state some truths.
What is troublesome is when people say things like "math is the purest science". That sentiment I find quite disturbing (purified of what, pray tell?)
@@gg829Purified of reliance on other fields. If you accept that math is a science (which seems strange to me), then it uniquely stands alone, physics requires a core of math, chemistry requires a core of physics, biology requires a core of chemistry, and similar statements can be made of everything else (and to the extent that I oversimplified things, unsimplifying them emphasizes the point).
(minor edit)
Math requires the science of language
I need to up my teaching game after seeing this. Thanks.
I didn't like this presentation at all, it makes lot of assertions on game theory which have nothing to do with the field at all. The whole point of using utility values is that you can assign them however you want, if you value goats more, you can plug that in the equations and calculate how best to get the goat. Maybe you were more aiming to critique the application of it in e.g economics, instead of viewing statistics and game theory as subfields of math, which is what they actually are. But then this should have been made much clearer in the slides. As it stands, you might as well start discussing about how differential equations are a detriment to human civilization because they have been used to calculate ballistic weapon trajectories.
i mean, that's essentially my point. someone can't use calculus to create ballistic missiles and then argue that it's okay because it's justified by calculus, but we indefensibly allow people to use game theory to launch ballistic missiles while claiming it's justified by game theory
also, if people can say value is whatever they want it to be, a field is no longer knowledge-seeking and is now instead power-reinforcing.
these are things often understood by people with developed understanding of game theory and statistics, but that describes a minuscule percentage of human beings.
@@Jorbs So all of mathematics is not knowledge seeking because it doesn't have the concept of (moral) value? That's completely backwards, half the reason fields like analytic philosophy based themselves on mathematical thinking was exactly to be able to reason about the world without abitrary value judgements. Logical positivism itself may have failed, but it led to pivotal advances in the philosophy of science and mathematics. If these are not knowledge seeking, I don't know what is.
Thanks. I dont know quite how to feel about this yet, i was really put off by the last part which felt like pretending game theory and statistics is hilarious and ridiculous to ever be used... But maybe i just need to rethink again and come clear on what exactly isnt clicking with me.
Pointing out flaws in the applications of Science, specifically Game Theory & Statistics seem fine but it felt like it went too far and only in one direction? (This might be the part where i was getting impressions that werent even made).
Pretending dual-determenism is such a cool thing for badasses and statistics is such a fake made up thing only used for horrible things such as eugenics just doenst seem like nothing jorbs or anyone in their audience would seriously believe
This comment is pretty bizarre. If game theorists - or any group of people trying to understand how something works - were to explicitly try to define what are good and bad outcomes, that would be "power reinforcing". Knowledge seeking means trying to understand how people can achieve their goals without judging which goals are good and bad.
I'm not sure if I missed something in the RPS stats, but I think that adds up to 89, which even if you rounded down all the numbers would max out at like... just shy of 92
Very interesting discussion. Enjoyed over my morning coffee while doing 'dles
I don't really get the part about "it assumes the car to be the desired outcome. But not everyone would think of it as the desired outcome".
And by that I mean that I always thought you could just replace it with anything. To me, "sport car" and "goat" are just word put there instead of having a placeholder "good outcome" and "bad outcome".
But if that's really a problem with this game problem... Then wouldn't it be easier to say "the desired outcome you get something like money, which can then be used to buy something you'd want. And the undesired outcome is you receive absolutely nothing."
Or if you reeeeaaaally want to fit all scenarios including one were money doesn't have any value. Then we can just say "The desired outcome is something good, the undesired outcome is something bad." So we're all on the same page... Right ?
It just feels like a weird thing to me... But I guess the problem itself is kind of whacky, so... Yeah. That tracks I guess.
It's not deep, bro just thought he cooked
@@tehy123 It's only as deep as you're willing to dig
Shoutouts to Zero Time Dilemma for illustrating the Monty Hall problem in a novel way.
As an addendum to dual determinism, it always irked me that success was shared with the divine, but often failure was hung squarely on the shoulders of the mortal- a deliberate fallacy that we still use today in some circles.
I feel like you are arguing more against the people that use game theory and the way in which it is applied more than you are arguing against game theory itself. For example at the end there you talk about what game theory says about the climate crisis, but game theory itself does not make any assertions about the climate crisis, only people that try to apply game theory to the climate crisis do. I don't think it makes sense to look at situations where game theory and statistics are misused and then say that game theory and statistics are rubbish because of that. Though I could be taking your point wrong and all you are saying is that we should be aware of the problems with how game theory and statistics are commonly applied, not that we shouldn't use game theory and statistics, in which case I agree with you.
I think there's a great deal of both. Thinking in "Game Theoretical" and "Statistical" fashions can often lead to failed solutions to problems, to the confusion of those deploying such solutions. Stuff like the "effective altruism" movement, even ignoring its main controversies, comes up with very confusing answers as to what people should do to improve the world. Oftentimes when researchers in a rich country try to find solutions to problems in poorer countries in an abstract sense, without having lived in those places and understood what the people need, they develop solutions that are ill-suited to the actual populations for which they are created. Sure, you can chock this up to the "people" being bad rather than the process, but at some point you have to contend with the process of reducing problems and communities to sets of numbers can fail to accurately capture a problem space, as well as give a false sense of certainty as to the projected efficacy of the solution.
There's also the issue that just this idea of game theory and statistics, or I think jorbs is saying, selfish optimization becomes a prominent part of casual reasoning and leads to bad places: e.g. most people I know "care" about the climate, but dismiss the idea of consider taking personal actions such as reducing consumption or growing a garden because it's big companies that are mostly responsible anyway, what does my personal consumption really mean. Is this true statistically? In a certain set of assumptions, yes. But it's really not helpful.
He’s not saying statistics and game theory are bad, he’s saying vast majority of people are bad at them so this obsession by people who are bad at these things to use them to support their decision making is ill informed and ill advised, and it wouldn’t yield any better results than the ancient tradition of double determination. Notably, double determination is still relied on by large swathes of modern humans: see, Christians.
His point that these disciplines are formalised by bad people does matter to his main argument that they are implicitly biased by cultural factors from the creators and dominant societal thinking, but overall I think they distract us from his main points: via-a-vis this comment chain.
@@undeniablySomeGuy effective altruism is not even game theory or stats, it's a branch of ethics/philosophy. The people you mentioned who "care" about climate would use the same argument whether game theory or stats is invented or not, because well... tragedy of commons happen throughout history.
It's very very very very easy to point out problems with any process in real life because it turns out real life is messy af. What's tough is prescribing an actual solution which this video haven't actually done so at all. It's a complete non sequitur to say that process A is worse/equally effective as another process B by pointing out problems A have.
1. "This thing is bad because of the problems it create under some scenarios" (stats/game theory)
2. "Therefore it's equally effective/worse than this other thing because of the problems in 1" (dual determinism, i.e. religious beliefs)
2 does not follow from 1.
I hate to break it to you, but the reason people are willing to yap about X cause but not make any changes to their comfortable life, is definitely not because of math or statistics.
I know you will likely dislike this comparison, but this video has a great number of similarities and narrative outcome with what Jordan Peterson lectured about in his Maps of Meaning course.
Didn't know Jorbs was a Tomska fan, but I am happy to know that factoid.
The moment I saw the title of this video, I just immediately heard 'BOOONNNNE!' in my head XD
IDK man, I'd just bet 1/3 on the flop.
I cant believe my favorite video of the year is a 30 minute powerpoint presentation in which jorbs uses the monty hall problem as a spring board from which to make a case for greek polytheism
Also, I'm now obsessed with the concept of the "inventing game theory" solution to the fermi paradox.
As someone who does statistics for a living...
I like to think I've graduated from "bad at statistics" to "mediocre at best at statistics" ;P
I will push back a bit on the "provably false" part for the rigid assumptions - it's a bit more that the axiomatic assumptions underpinning statistics are *unprovable*
It's a subtle distinction, but it's more that proving Wonder Woman's Invisible Jet is neon yellow - if you could see the color it reflects/emits, it wouldn't be invisible.
Such a good and strong ending slide, Truely blessed by Iris to convey this message with precision and excellence!
Game theory can absolutely arrive at the Golden Rule. Look up the Prisoner's Dilemma. Veratasium did a great video on it a while back, and the conclusions that game theory leads to are entirely that being among the most cooperative and forgiving people in the room leads to the best outcome for yourself.
> doors don't actually have names, they're doors
Well I'm naming my door just to prove you wrong è_é
In my (not very😀) humble opinion, the video didn't realy explicitly name what I feel is the core issue.
Math (including statistics and game theory) in its purest form exist only in an abstract world and it is meant to be used there. There are fields that use mathematic tools, but only as a tools. For example, to determine how far can you throw a stone, you would probably use physics to describe the motion of the stone and also probably biology to understand how you are throwing it (e.g. to determine force and angle your muscles and bones are capable of), maybe even psychology (if you are angry, you would probably throw with bigger force), etc. Yes, you are using (mathematicians would say applying) math to help with this task, but you would not be able to do it with just math. There are simply no stones or people to work with in the beautiful and (more importantly) abstract math world.
One problem is, we are generaly very bad at navigating in this abstract math world, so we often tell ourselves little stories to help us feel we understand what things are. Goats, sport cars and doors in the Monty Hall Problem are great examples of them. But often these stories are actually hurting our ability to understand, rather than helping it. Notice when jorbs tried to explain the original correct solution to the Monty Hall Problem, he did it by making the problem more abstract, not less. He switched from doors and goats in a game show to squares, letters, numbers and arrows. The problem itself became a little harder to follow (asuming you don't have prior knowledge about the original formulation with goats behind doors etc.), but easier to solve. (I believe that making the problem more and more abstract would eventually made the problem totaly incomprihensible for humans while also self-evident). Making the problem less abstract by packing it with all the emotional and cultural baggage managed not only to make it less intuitive but, as jorbs argued in the video, it outright changed what the correct solution is.
Game theory and statistics are also operating in the abstract world, as they are part of math. But making stories about inhabitants of the math world made lot of people (including some mathematicians) think they are real, as real as goats (or humans for that matter). But they are not. E.g. game theoretical "rational agents" are not people, they are abstractions and thinking about them as people is just wrong. It's like using a saw to hammer a nail, you can succed sometimes, but you can also fail spectacularly, it is just not a right tool for that. And if a saw is the only thing you have available, you should probably use it to make yourself a hammer (i.e. tool which is actually suitable for the task) instead of just repeatedly bashing the nail with it.
TL:DR, blindly using math in the real world without factoring in context - bad, using context to know how (or if) you should use math - good.
Anyway, thanks for listening to my TED talk😀.
Disclaimer: Sorry for all potential spelling or gramar mistakes, english is not my first language, it's too hot outside today, I haven't slept well, yada yada...
Really enjoying this type of video, Jorbs :)
omg that's totally a Monogatari reference
Replacing "virgin" with "incel" in the "chad vs virgin" meme format is fundamentally genius and more accurate than relying on purity culture rhetoric like "virgin" and I approve 1000 fold.
Yo, I’ve rarely seen UA-cam-Jorbs this unhinged. Jorbs woke up and chose sarcasm 'n sh*t. Did people actually annoy you this much into making this video? I really enjoyed it, though.
What I took away from this is that you portrayed dual-determination as the handsome muscular gigachad and game theory as the incel virgin soyjak. Game theory is no more
Every time I try to understand what statistics is, and what it means to assign probabilities to things which already are (what is the probability I have COVID?) I find it a really mind fuck. I'd love to actually read more about what I guess is the philosophy of statistics, but I've no idea what books would be good.
my biggest grievance with the monty hall problem is that it frames the choice to stay with the first door as a continuation of the 1/3 choice. Lets assume we pick door A, door B opens and the desired car is not there, and then we are offered to switch to door C or keep door A. I see no reason to frame the choice to stay with A as a continuation of the same choice, when we have new information. We can't go back and change the previous choice, and it has no bearing on what we now know or now can do, so why is it framed as the same choice? It seems clear from a practical viewpoint that this is a seperate choice with new information, we have door A and C, and we can pick either one.
EDIT: wrote goat instead of car, now fixed
EDIT: Also nice shirt :)
The disparity comes because as the host must reveal a door that is not which you chose and neither which has the car, then when yours is the same that has the car he is free to reveal any of the other two, making it uncertain which he will prefer, but when yours has a goat, he is restricted to reveal specifically which has the only other goat.
In that way, if you start choosing door A and he opens door B, we know that the revelation of door B was 100% mandatory in case the car were in door C, because he wouldn't have had another choice. But if the car were in door A (yours), it was only 50% likely that he would open door B, not 100% sure, because he could have preferred to open door C instead.
Because of that, it is twice as likely that the reason why he opened B and not C is because C is which contains the car, rather than because A contains it.
In the long run every door would tend to be correct with the same frequency, but once you start choosing A, for every two games that it is correct the host would use one to open door B and the other to open door C, on average. But for every two games that door C has the car, he would be forced to open door B in both, so twice as many.
@@RonaldABG im sorry, i was tired when i wrote that and wrote goat by mistake. I was referring the the original problem and assumptions that there is a desired prize behind one door, and nothing/an undesired prize behind the other two. i guess just replace goat with car in by comment and it is what I actaully wanted to say.
@@michaelcabrera6210 Anyway, what I put still applies. If your door A were correct you don't know if he would have revealed door B or door C, but if door C were correct, you know for sure he would have revealed door B.
zephyr noise at 10:44
I hate the goats vs car, to me it would be better answered by item you want, and two empty door, and the game show host is a all knowing being that has to do only one thing open all doors to nothing that are not currently picked and the one that has the item you want.
To me we should use this as stepping stone to show that not everyone is out to stop us, and that an all knowing being is not always out to screw us over.
Would you switch doors if the host was a bear?
This video is helping me quit smoking nicotine products
Or it isnt give me 30 min
Im sorry but i still chose the goat
I am confused, bemused, and abashed by this video . . . ty?
I was shitposting about monty hall problem literally last night in your Gloomhaven stream, and then this happens, what are the odds?
33% --> 50%
I like how it has to be specified that you prefer the sports car since a farmer in Tajikistan might actually prefer the goat, lol
If you made your initial pick and then were asked if you would like to switch and can now open both remaining doors instead would that fix your intuition?
I come from a tradition of tearing down previous traditions for great good and great ill. The previous traditions either say that it is my destiny (inescapable) to self-destruct for violating the natural order or that people such as I should be shunned as my existence is corrosive to the society which is what the tradition lives in. What I am heir to is the result of the things that made this world what it is now. It says little about me directly, but tells me that I can aspire for more, and it cares not if that is for ill or good.
I think I will choose instead to make something better out of all the flawed things that came before, and teach my children to do the same.
You didn't mention the respective dollar values of the car and the goat. Don't forget to take future petroleum shortages into account.
The thing that often isn't brought up in consideration of this problem is the psychological damage of having essentially "won" something and then giving that up (by changing your choice to the losing door) versus just picking the wrong door in the first place. I think this has a lot of sway over people who are in this position - imagine you'd picked the winning door and you were silly enough to swap it? Whereas if you'd picked the losing door, well, you only had a 1 in 3 chance right, so that's not on you? Mathematically the right choice is to switch (assuming you don't want a goat for some weird reason), but that's not really how brains work.
I don't know about you but I'd feel pretty darned silly if I lost a car because I decided to go with a 1/3 chance instead of a strictly better 1/2 chance.
@@Math.Bandit "Strictly better" in a world that is made up in your head; have you seen the video?
@@undeniablySomeGuy Yes. As it turns out 1/2 is always better than 1/3.
@@Math.Bandit*2/3, not just 1/2. The choice of swapping is the choice between picking the contents of one door or the max of the contents of two doors.
@@delta3244 Right, yes. That's what I meant.
Honestly, I don't really know physics, either.
I think I missed the most important part. Which pantheon god did I need to pray to to get success in Slay the Spire?
Another western science assumption that has upset me ever since I first learned of it was the one about kids and marshmallows. You have a kid sitting alone in a room, but are being watched. You place a marshmallow on the table in front of them, but explain they will get more marshmallows if they don't eat it and wait instead.
But what if the kid doesn't trust these candymen? They could very well withold marshmallows indefinitely, this might be the only chance at a marshmallow. But you don't know for sure, so you wait a bit. But notice the longer the kid waits, the more convinced they become the adults are not going to deliver. It takes some serious conditioning to not only overcome impulses, but also decide that most other people are not generally trying to hurt or steal from you. Is that really a prophecy of intelligence then? Or a coincidental indicator of supportive upbringing?
That's not a western science assumption. You are instead describing an experiment that was performed, and a correlation that was discovered.
As you say, there are lots of ways to interpret that correlation. Science as a whole does not assume any particular interpretation to be correct - that's not the role of science. Instead that is used to construct theories, and those theories can be tested in other experiments and then scientists can argue over which theory is strongest.
I don't think that's so much a criticism of the experiment as a very important deduction. People who are conditioned to have low trust in authority also are less likely to defer gains, exactly because of situations like this. Lower stability goes hand in hand with a tendency to make decisions that will pay off short term.
The main issue here is that a lot of social sciences are pretty ass, the people making the study have a set of beliefs around what is and isn't true and will twist data into supporting their theorem without taking into account alternate viewpoints. In this example they could have progressed onwards to do more experiments to try and remove some elements and see what changes. For example, maybe you have a machine which automatically releases the marshmallows after X minutes or something similar. How does that change it?
What school do I go to to get professor Jorbs?
and thats why you should always !dig
Gosh I love these videos.
I never realised that the Monthy Hall problem actually has very little to do with probability and everything to do with figuring out if the game host is adversarial towards you.
If you correctly deduct that the host is adversarial, you always win. If you fail to identify an adversarial host, you always lose. One does not need to consider any probabilities to deduct this fact. Figuring out what happens after you correctly deduct that the host is not adversarial towards you feels like a secondary priority under these circumstances.
It's not, the premise of the problem states the rules the host acts under, and it's based on a real scenario of which we know the host wasn't acting adversarially.
You can state a problem where you don't know whether the host is adversarial, but then it's not the Monty Hall problem.
I wonder what the 4th move is in RPS that makes up the other 11 percent, and what its interactions are with the base moveset.
Maybe in 11% of cases, the player fails to choose a definite answer and thus doesn't fall in either category. But if that was the case, it would probably have been said.
Love this content
8:00 "the problem assumes you want a sports car more than a goat" Yes. This, 100 times, yes! I feel like this points out very eloquently why game theory is much more limited than it seems when it comes to real life applications.
Essentially, what game theory is, is a function that takes as input the value hierarchies of one or more agents, applies them to a particular game, and outputs a winning strategy. The value hierarchies determine the winning strategy, yes, but game theory does not determine/prescribe the value hierarchies. It has nothing to say on the matter. Those are *givens* .
To say that game theory can tell you what you should fundamentally want is equivalent to saying that when using the function f(x)=x^2, you should have x be equal to 6.
But who says "game theory tells you what you want"? It's always "game theory tells you how to get what you want" or (probably more commonly) "game theory tells you how people will act, if you can predict what they want".
@@Reashu You must surround yourself with fewer weird people than me then
1) you have valid points but the part about cultural value is a bit pointless. yes, it is true, but the problem could be rewritten with anything that would make sense for a given culture. or, given that it is mostly used as a toy problem nowadays, it could just be "bad thing, good thing" without specifying anything in particular. the statistics still apply.
2) and speaking about toy problem, this started as a real world "problem", not a toy problem.
3) the host is described as a "he" and called monty hall because monty hall is a real person and a man.
Yeah, I think this video is a little bit cooked. Most of the assumptions are based on a real life example and used to communicate an idea. That's a pretty good way to teach even if it's not 100% perfect
...Are you not aware that Monty Hall is a real person and Let's Make a Deal was a real gameshow in the 60s that really had sports cars and goats behind real doors? You're presenting this as if the Monty Hall problem was just made up from whole cloth by three white guys in the 90s.
I mean, this is all just a theory. A _game_ theory.
...I'll see myself out
the reason car and goat is used is the same reason the you're speaking english. why are you speaking english? is it not possible for this video to be done in so many other languages? does it have to be a video? why are you on the screen? why do you have a powerpoint?
the purpose is communication and not just any communication to any group of people. the goat and car is used to communicate with a demographic that generally values cars over goats. the trivial case of broken car and really really nice goat that can get you exactly what you actually want in life is implicitly discarded. if you were speaking to a demographic that did not value cars over goats, you wouldn't use that example and if you speak to people who dont know what a game show is you won't use that example and if you speak to an audience that doesn't understand english or are unable you hear you, you would probably not use a video that relies on english audio to communicate with them.
the reason why game theory is so good is because it doesn't actually have any implicit biases except for the ones you input into the system. game theory does not demand you use game theory all the way through. if im in some life or death situation and i have some being with other inherent decision making capabilities imparted to me through my evolutionary past and the situation is exactly what my other inherent decising making ability is trained for, then game theory says that the probability of me surviving if i use my evolutionary abilities trained to help me survive this situation is gonna be more useful a mechanism for making decision than trying to apply statistics in that situation where i have neither sufficient explicit knowledge or enough time or skill to apply statistics.
the key here is if you want something out of game theory you need to put something into game theory. put it "how do i survive this life or death situation" and game theory might say "stop asking and follow your instincts you have no time." if your bias is "how do i make the most money" why are you surprised that game theory is telling you about making money?
you're right to say that people are bad at game theory and bad at statistics and that's important to know. game theory is a tool in your toolbox. you pull out that tool when the tool makes sense. it is not a hammer to apply to everything (tho conveniently it is a hammer that with enough awareness, will often tell you to stop using it on the screw and go get your screwdriver.)
why are you glorifying society btw? do you not look back on human history and see that maybe the culture of many different societies are fucked and flawed and picking ancient greek decision is a really really fucked argument?
This was a very poor video until the halfway point... which is when I stopped watching it.
Brooklyn 99 beat you to it
I wanted to see some goat pictures. very disappointing
great video! I learned a lot.