Timestamps: 0:00:02 - Introduction to talk on applied mathematics and thinking styles. 0:03:03 - Ronald Fisher's life and work at Cambridge University. 0:05:57 - Testing if milk affects tea taste with experiments. 0:08:42 - Designing experiments using combinatorics. 0:11:31 - Gary Neville statistic measures player performance after conceding a goal. 0:14:17 - Statistics used to rank top players, limitations of measurement. 0:17:00 - Explaining the limitations of statistics in 10 words. 0:19:46 - Eugenics, smoking, and misuse of statistics. 0:22:25 - Context matters in statistical significance and causation. 0:25:13 - Balanced chemical reactions and ecological models explained with math. 0:27:46 - Mathematical modeling of foxes and rabbits. 0:30:32 - Analysis of applause and social recovery in groups. 0:33:12 - Human behavior equation: non-smiling person + 2 smiley people = 3 smiley people. 0:35:55 - Using physics-based models to scout football players. 0:38:51 - Introduction to chaos theory and Margaret Hamilton. 0:41:29 - Weather simulation error due to decimal input mistake. 0:44:05 - Generating divergent numbers through a simple process. 0:47:06 - Chaos, Margaret Hamilton, and the importance of control. 0:49:55 - Finding balance between order and chaos in life. 0:52:33 - Simple rules create complex patterns in simulations. 0:55:21 - Capturing complexity in science through detailed descriptions. Follow me for more AI
His lecture is not really about ‘ways of thinking’ but mainly arguments to solve certain problems using techniques he expects his audience to know. Always the British Oxford snobbish people!
Between 0:17:00 & 0:19:46, what are the 10 words explaining the limitations of statistics? A) "Statistics are valid. It is our interpretation that is wrong." And Sumpter then uses the Grit study as an example. B) "Statistics are powerful but doesn't give you all the answers." And Sumpter then uses Gary Neville's point as an example. C) Something else, i.e. please enlighten me.
Exactly the reason why i chosed to study mathematics and physics that is to understand the world around me.And i love studying these two subjects.When i study these two subjects properly i get totally involved,you can say i live in those moments fully.
That's interesting. If I were smarter, I could have moved up a few notches. I, also, wanted to understand the world around me - reality driven - I started with psychology, (psychoanalysis) then later, added the financial markets, geopolitics and a bit of economics.I mostly enjoy my time spent moving around in these realms and feel totally involved and focused, daily. I'm happiest doing my "work". I feel fortunate to live as I choose.
Have you completed your degree(s)? If "Yes", then please confirm / deny the following *hypothesis*: The channel, Veritasium, has recently (on average) significantly declined in quality wrt the subjects explored (Note: ⬅ eval is distinct from the quality of editing, audio, camera, storyboarding, and so on). Answered separately: The quality decline is in part due to over-focusing on analytics and the business side of the channel. Therefore, in a circumstance where they lacked a solid video idea, rather than skipping the weekly scheduled video release, they prioritized making and releasing one anyways to keep up with the algo. In my estimate, this reason accounts for 30% of the quality decline. Of course, running out of feasible ideas would likely account for the lion's share, say 60%, but I want to know if you agree or disagree with the 30% estimate AND the reasoning supporting your stance.
The smoking bit reminds me of a silly anecdote when Bertrand Russell was asked if he was concerned about smoking impacting his health. He said that one time he was going on a flight, and they told him he couldn't smoke on the airplane, so he decided to delay his trip rather than go without smoking for a few hours, and the plane ended up crashing.
If you want to smoke on a plane, no problem. Just buy your own plane. I don't want to smoke on a plane including second hand smoke. If I sit next to you and drink whiskey, you don't get drunk. If you sit next to me a smoke, nicotine ends up in my blood.
This isn't what happened. On a flight he had to sit in the smoking section and the plane actually crashed. Everyone that was sitting in the non-smoking section died. He had to swim to safety. Therefore his life was saved by having to sit in the smoking section.
@@catloverJyes i never dissapointed by this youtube algorithm Iam happy that it doesn't promotes only ones who has million followers and post nonsense content ,obviously youtube shorts are 🤮 But This video algorithm is like batman 🦇
Nah, it's not random, it's maybe semi-random, selected from the pool adjacent to your interests. Have you seen the recommendations when you are unlogged? Ridiculous 😅
@@Leonhart_93 yes i have that's why i have 20 different accounts for youtube , 1 for astonomy , 1 for physics chem and maths , 1 for history , 1 for unreal engine animation , 1 for medical research purposes etc,and semi random isn't a proper term because you allow them to show videos you subscribe so either they are a random or seen like a schodinger cat
John Maynard Keynes 'principle of uncertainty' in economics states that we can predict the very near future with some degree of success, but the more distant, the more uncertain we become; thus, mathematical modelling in economics is not a science of absolutes; rather, it's a social science of human infinite variability-a perfect example of chaos theory. Daniel Kahneman later backed this up in behavioural economic theory. Also Adam Smith wrote, that the market in the long run will return to an equilibrium, to which Keyens replied, in the long run we're all dead! (ie the disaster that was 1930s and the 2010s, both unnecessary suffering, we can always afford a war, so let's win the peace).
@@jonasastrom7422@jonasastrom7422 Please cite your conclusion about the 1930s and, again, the post-war consensus. I'd be interested in where you get your information to draw your conclusion. I'll be happy to provide a rebuttal once I understand your source. Thanks for the reply.
@@jonasastrom7422, Please offer some empirical proof; for example, his policies weren't even enacted until the Attlee government of 1945. The 1930s were laissez-faire (Hayek, Smith, Marshall) do nothing and the market would correct, which it didn't! Thus, the Keynes quote, in the long run, we're all dead.
@@jonasastrom7422 Please offer some empirical proof; for example, his policies weren't even enacted until the Attlee government of 1945. The 1930s were laissez-faire (Hayek, Smith, Marshall) do nothing and the market would correct, which it didn't! Thus, the Keynes quote, in the long run, we're all dead.
Yin-yang is chaotic manifestations; taiji is the orderly existence of the whole... if I was to apply this side of classical Chinese philosophy. This of course in no way detracts from what a great talk this is. Thank you!
It strikes me as amusing that this only seems brilliant on the surface because it's really nothing more than a tautology - which is exactly why it's so concise itself, because it's not really saying anything at all, hence its own complexity is low. It's essentially saying, "A thing is as complex as it is complex." Okay. And? lol
@HorrorMakesUsHappy A very long pattern may look complex, but if it can be condensed into a repeating motif it can be considered simpler than something that cannot be condensed. Complexity is characterised here by resistance to compression. This definition distinguishes bloated patterns from complex ones. It's not a tautology.
My statistical way of thinking: of a mindset care for probabilities, historical data, an assessment of assets and liabilities before pursuits of actualization occurs. My interactive way of thinking: of a mindset care that awaits or hunts for societal circumstances that appears to 'door' or facilitate pursuits of actualization. My chaotic way of thinking: of a mindset care for rolling the dice-- a 'just do it' actualization pursuit that minds no statistical odds or favorable circumstances and may even lack certainty of what actualization looks like. It is a pleasant surprise when no sense of chaos follows an unstudied action. My complex way of thinking: of a mindset that utilizes statistical and interactive cognitive intelligence to facilitate pursuits of actualization. I do believe success is most likely when complex thinking is applied. Additionally, the way of thinking determines what the mind is inclined to focus upon; what we focus upon occurs within mindsets. If I am of a Will to think in a manner that invites the possibility of chaos, there would be no consideration of mindsets associated with statistical or interactive thinking. Often, it is the actualization objective and not the Will that dictates the mind's cognitive cascade options-- care of success demands management of one's Will, usually. There is a control 'care' in mathematical logic; through an understanding or a care of the answer, one controls the problem. The lecture's turn towards discerned marital controls that are implemented and sustained, if "care" exists, is of complex thinking that minds objective (statistical) and subjective (interactive) truths. The identified existential marital problem has a determined answer that is not natural within our lecturer-- conscious, willful "control" of his natural inclinations is the answer that he believes "care" will sustain. If a strong "care" becomes weakly felt, the answer and the "manipulations required" will feel a key to a prison versus a key to greater marital happiness; sustaining a marital reality that survives upon an unnatural answer, of little care, is difficult-- even grit requires passion to exist in its application. Addendum: What is in grit? A Will to persevere, despite any headwinds of statistical and/or interactive inform. What does perseverance require? A deeply held belief that a pursuit will see actualized success. What is in belief? A measured patience that is governed by an adjustable long view. When the long view dies, all that underpins grit dies; grit may require an ability to visualize and intensely feel an anticipated actualization of victory. What a grit-gifted victorious actualization satisfies is dependent upon the state of one's ego and other psychological influences; a satisfied ego (no circumspect psychological insecurities or of a felt need to proof/validate identity associations of gender, abilities, intelligence, etc.), scar-free interactive psychological truth (never felt derisively underestimated or had declared goals disparaged and/or discouraged) and/or of no passion to defy odds-- less likely to endure pursuits of actualization that requires grit. Some form of anticipated internal resolution, validation or balm-- beyond the externalized actualization pursuit-- exists in the long view of a grit-driven pursuit, I believe.
The more ways of thinking a person have, the more complete a person is going to be. In my perspective creativity plays a huge role when we apply different ways of thinking, as well as how we organize the language( numbers, symbols, information, knowledge, etc), how we understand the language, the information and how we apply.
A relatively minor point in the talk as a whole but the bit about grit being a good indicator of success is spot on in my experience. I've worked in a technical field for ~30 years so I've observed plenty of folk through those years. Grit is a much better measure of how good an engineer a person will make than GPA or similar.
🎯 Key Takeaways for quick navigation: 00:02 *🎓 Sumpter uses applied math to understand the world through stories from football, science, and life.* 02:48 *📊 Fisher pioneered powerful statistical methods but had problematic views on eugenics and smoking.* 11:28 *⚽ Statistics can measure player attitude and performance in football, with limitations.* 23:25 *🧪 Lotka used "unbalanced" equations to model exciting biological and social patterns.* 37:31 *🧩 But Lotka never found a grand theory. A key limitation was not knowing chaos theory.* 39:18 *🦋 Hamilton and Lorenz discovered chaos - small errors cause drastically different outcomes.* 47:54 *🚀 Hamilton left chaos for Apollo software, showing the need to eliminate errors in critical systems.* 50:00 *☯️ The yin-yang of randomness: control what matters, let go of the rest.* 52:03 *🧩 Complex patterns emerge from simple rules in cellular automata and tiny code tweeted graphics.* Made with HARPA AI
@@davidjohnston4240 60 million people can be wrong. Lord knows I am in America and see it all the time. We have two of them running for president. For other British faux pas, you have Nigel Farage and Brexit. Just because you, personally, are wrong does not make it right.
I suspect the tea will mix eventually, but not in the time between pouring and drinking. A better test today might be clear teacups and video cameras that record from all angles.
@@kooisengchng5283 Putting a small volume of cold milk into a larger volume of near boiling water scalds the milk and changes the flavour. Adding hot water to cold milk doesn't scald the milk.
Is it possible that Ronald Fisher did not know there is a difference between putting water to acid and putting acid to water? I stand with Prof Sumpter's opinion that Ronald Fisher is an asshole. He wanted to prove Dr Muriel wrong and devised an experiment that he thought would prove Dr Muriel wrong. Ronald Fisher tried to do the same thing with cancer and tobacco, and with Bayesian statistics. He is an arrogant jerk.
at 6:15, there is this milk-tea problem. i've done a highly scientific research of drinking coffee with vegetarian based (oat and soymilk) milk and because these conditions are a lot more extreme, it enabled me to do some practical tests with clear results. many of those vegetarian "milk" products curdle or accumulate to nasty bit and pieces easily in coffee. to prevent that, pour milk first so it is the strong coffee introduced to a lot of milk first vs. a wiff of milk in a strong coffee. in both cases, a vigorous mixing will help alot. also there has to be enough of milk to make the mix mild enough. yes, i know i missed the point :) so she was right, milk first! also the toilet paper should come out from top side.
@@MrLuigiFercottiWhen people talk about their own experience, it sounds more authentic than speculating what someone else might think or feel. "You" is 2nd person. He/she/they is 3rd person. A lot of people phrase it like you did, speaking to the reader, pondering. Some though try to speak for the world, as in "No one cares about..." Anyway, I agree and feel the same, still can't decide which hat...
i'm a "sugar before coffee" man myself. There is definitely a difference in milk before tea: the complex colloid solution of proteins and solutes in milk will react differently based on the relative concentration of milk and tea -- especially with regard to the relative concentration of heat in the tea. I always put two cubes of ice in my coffee after brewing to cool the solution so that the cream doesn't gunk up. The undesirable byproducts of heat on milk are less if there is more milk which allows it to heat more slowly.
this assumes that you aren't leveraging the introduction of variance into the values of experimental variables. but for the same reason that Fischer was wrong, your assumption that you can perfectly manage the experimental variables for four pairs of glasses is an oversimplification. it's better to embrace the randomness, quantifying the deviation of variables from norms and thus extracting more information.
GPA is an acceptable way to compare a cohort of students, but when some students don't fit into the variables a cohort controls, it can be a particularly terrible metric for success. You can easily find yourself measuring the ruler. Students representative of the cohort who focus their efforts on meeting the objectives of the classes and programs will typically have high GPA's. At least their effort should correlate to their GPA. Students who don't fit the cohort well, fail for lack of social support or try to extend themselves beyond the constraints of a program may not exhibit a GPA that correlates with time investment or resilience. It's the last one which can be particularly sad: by setting the expectations for a 4.0 GPA, you load the student with work and prevent them from accomplishing other things. Even the best students are then stovepiped into a singular field in the initial years of their career, unless they did the coursework before their college started. I almost failed out of community college: I had no one to study with for physics or statics and while I was simultaneously studying devops (no one to help study). So on paper, I'm a terrible student, but I'm okay with this since no one would understand me anyways.
@@DavidConnerCodeaholic I like to think I fall into that category. I did engineering with not a good GPA (< 3.0 GPA) because at some point I got really interested in the derivation of theorems in engineering and the underlying mathematical, physical and sometimes philosophical insights behind the engineering conclusions, than the applicability of the theorems themselves. Along with battling learning disability, unstable housing situation, poor social awareness, etc., there was a point where I had to pick a choice between 'actual education' of the concepts and the bending of the mind, or to follow blind and measure up to the standard. It is hard to say that I regret this, because I certainly feel like I have learned far more than I was given the chance to learn, but it may have limited, even in the slightest bit, for further learning, as GPA is used as a source of evaluation. Although perhaps other activities may be used to supplement my profile in the context of graduate level admissions or so, that is if I am even in interested in pursuing more academia. Since I have graduated my engineering program, I am eligible to start working, but post secondary is something in my mind to pursue potentially. Ideally I would like to get a job. It really is a tough world out there. To "map" the complexity of human capability to a single number 'GPA' is such a limiting way of looking at things, but at least it is 'a measure' of looking 'at something'. A world, that looked so objective and rational to me, upon further reading and studying different concepts, seems rather subjective, arbitrary and uncertain.
For the milk tea test. The difference with the pairwise option, is that there is extra information that 1 of those 2 cups is milk first and the other is tea first, that's why its more probability for her to get it right. For the 3rd option: 8 choose 4 = 70, as from 8 positions choose 4 for the milk first option.
Great talk. As regards the slight relationship between grit and success the terms 'statistically significant' vs 'practically significant' come to mind. The problem then is practically significant for who? As we move through life we meet individuals and it is a foolish person who makes any assumptions about an individual based on secondary characteristics. But what about those who are making evaluations in bulk? Suppose someone has found a similar slight relationship between 'grit' and programming skills and is charged with hiring 100 programmers for a project. The temptation might be to use 'grit' as an evaluative proxy when hiring. But what if 'grit' has some cultural factors that influence it? Then the hiring process might be biased against some groups and this bias might be incorrect. Bean counting micro-optimisation is (unfortunately) a common feature in business, moreover increasing use of 'AI' is likely to hide such biases behind a wall of pseudo-objectivity. Bringing us back to the Fisher problem.
One is then reminded of the chaotic and complexity ways of thinking, since programming ability (as well as football/soccer) are definitely not easily predictable at all
I am very naive and want to learn more, you made very good point, but my concern is the 4% variance explained by grit provides valuable information for researchers, educators, employers, and individuals themselves who are interested in understanding and fostering success. It highlights the importance of factors like perseverance and passion for long-term goals in achieving positive outcomes. Apart from cultural factor the other factors are not proven and measured yet, don't you think one should believe on proven statistics? If not, then how come one believe on other factors if they are not proven yet?
the art of the measurement becomes the reality to the measured ambitions of the narrowness of arrogance. i found David Sumpter's beauty in these 4 measures the protagonist toward the Long Term Capital Management story. really appreciate this opportunity of sublime serendipitous interaction the youtube algo sought out my desire for such. thank you.
A bit more intuition on the "Complex" segment would have made it a lot more worthwhile, it while being a brilliant talk without a doubt. The two examples were good, but I guess it needed a few more words from him. I suppose the time limitation at hand...
great! all these ways of thinking are from macro perspective to see the world. so that we can see the pattern, trending, development of organic system.
I've only seen thru 32 min of this video, am enjoying it, have skills and education in math, and suggest that there is a stronger correlation between the LOUDNESS of applause (coupled with the peak loudness correlation) and enjoyment of the "event" rather than the length of time of the applause. Just a spontaneous theory here. Could be wrong...
I write this comment around 9.00 minutes. The second method, "B" , isn't "fair". One has to keep in mind that humans use their MEMORY. The second test assumes that a human can instantly reset their pallette after a short taste-test, this is simply NOT the case. When the difference in taste is subtle, then all the interpretations become one big jumble for a human's brain if the taste tests are set too close after each other. Also , the first thing that sprung into my mind was, "Maybe it's due to a difference in TEMPERATURE, that the lady can tell the two variants apart with a significant amount of certainty." So, is there a difference between T(t) and T*(t)? T(t) is the temperature curve when milk is poured first, and T*(t) is the temperature curve when tea is poured first.
Stats are often suspect because people fail to remember that independence - essentially that "all else remains equal" - is usually a dubious, or just wrong, assumption. When a goal is scored, offensive and defensive strategies can change. There can be more emphasis, or less emphasis, on giving more, or less, responsibility to a player - or such a variation in how the other team chooses to handle that player. You tend to take bigger risks when you;re down, and so on. Did you start to play worse, or better, ... or did your teammates? Or did your opponent? It's no surprise that you're a more/less accurate passer if they're now running faster/slower and so are more/less open than before, or that their defenders are now playing better/worse than before. I go through life endlessly hearing arguments like this video's "stats measure which players have the most grit/heart", and I'm just endlessly unimpressed. It's not that it's necessarily wrong, it's just never as compelling as people seem to imagine. Because all else is never equal, usually. In other words, statistics, without scientific controls, aren't justifications for conclusions - they aren't science - they're just unanalyzed data. You need control groups, first and foremost. You need to account for placebo effects. You need reliability (objective standard and immediately recorded, not memory), and impartiality (double blind) in the data collection. You need to at least consider if not incorporate the considerations of previous related research. You need to understand all the ways your assumptions can be flawed... and do a numerical analysis those assumptions as well. And of course ultimately, you need to allow others to critique what you did. On and on it goes. Statistics - the data - doesn't speak. Only a scientific analysis speaks.
@48:18: But there was room for errors at NASA. During the first successful moon landing, Apollo 11, they encountered programming alarms 1201 and 1202. There was a question on whether to abort the landing mission or not, and they went ahead.
I wish he could have elaborated on why complexity is distinct from chaos, but he cut it short. Seems he was only describing the same thing - chaos is macro-behaviour that emerges from the complex interaction of micro-motives. So not sure why they merit a separate category.
Good question, they may be different in scale instead of category. Let's think about this a bit. You can have chaotic behaviour arise from something very simple and logical. On the other hand complexity interacts with the strength of the connections in various ways, when one of those ways is chaotic then complexity + tight coupling -> danger. So in a loosely coupled complex system you will not get chaotic macro behaviour, because the loose coupling prevents the small chaos from growing.
I think it relates back to the Kolmogorov quote, applied to the cellular automata example: If a chaotic-seeming thing can be described with just three simple rules, then the apparent chaos is merely complex. Indeed, one might argue that the fewer rules that can be in play in a system, the less likely it is that the outcome will be truly chaotic. If, however, you need many, many rules to describe the system, it can reasonably be called chaotic. This would tend to imply, though, that as we generate better descriptions of what we do (seem to) understand, we can turn some chaotic-seeming systems into (merely) complex systems. I guess the current work on modelling the impact of CO2e emissions might be an example. It's the mathematical equivalent of the 'God of the Gaps' argument. God was frequently invoked as the "reason" for something that was inexplicable, and then we explained it, and that was one less gap that God was needed as an explanation for.
@@Alan_Duval I think the line dividing chaotic systems from complex (but non-chaotic) ones, is if it’s possible to reproduce a process *exactly*. Some systems, due to nonlinearity, cannot be reproduced exactly due to roundoff errors, so they’re chaotic rather than complex, and would continue to be so regardless of how much we know about them, in my opinion.
Yes, Dr. Muriel Bristol, the subject of Ronald Fisher's famous tea-tasting experiment, was indeed able to tell whether the milk was poured into the tea or the tea was poured into the milk, according to the results of the experiment!!! Please don't leave us hanging.
Descriptions meaning and information content depend on the formula or algorithm evaluation it. It is always a sort of function giving the se of possible source data meaning in the result space.
It's pretty amazing how a whole field of mathematical inquiry began because a guy just couldn't believe that a woman knew how she liked her tea and had to - HAD TO - prove her wrong.
Incorrect @10:08 - No matter how many ways there are to present four cups of tea, there are always 2 cups 'milk first' and 2 cups 'tea first' on the table. So, the chance of getting the first cup right is 50% (there are 2 cups of each). If the first guess was right, there are 3 cups left on the table: 1 'milk first' and 2 'tea first'. Now, the chance of getting the second cup right is 33%. And, if the second guess was right, both other guesses are per definition also correct. (The only cups left are the 'tea first' ones.) So, the chance of getting your guesses right in the second set-up are 50% x 33% = 16,6% or 1 in 6.
15:30 - right right i remember her Ted talk about "Grit"....what a god awful way to do analysis. the gist, as you can plainly guess, was "well, those that succeeded had....grit...because they were the ones that...succeeded in the end" talk about water is wet. kind of like the survivor bias, amirite? show a bunch of people pictures of where bombers, who returned to base, were shot full of holes and you'd rightly assumed it would be "hey, let's strengthen those bits that are more likely to be shot up"...except they forgot that this ignored the ones that were shot up in the places not full of holes in the survivors...and didn't make it back to base.
I shouldn't have to ask this because credit should have been given in the slide deck, but can anyone identify the artists behind the Twitter art @54:19 ? Note: In 2024, the X, formerly known as Twitter, character limit is 280 not 240 as suggested in this video.
38:20 "Magnus opus"? Also, a lot of his complaints aren't shortcomings of the techniques he discusses--they're shortcomings of the people using the techniques.
I like Ronald Fischer, from being a Eug@nics advocate to going 80 years ageless teaching at Oxford about Himself in "Four Ways of Thinking: Statistical, Interactive, Chaotic and Complex"
On Tea and Cookies: I think there is a big difference between the two methods and she is right. Tea is acidic and milk is a base. As the milk is added to the tea, it heats up quickly and is overwhelmed by the acid and curdles. Adding tea to the milk slowly warms the milk and the acid is quickly neutralized by the milk which doesn't curdle (form a colloid) and that's better.
Not a stats expert, nor a futbol expert, but as a scientist I strongly suspect the explanation for these strikers' performances has a lot more to do with how the other players on the field react to a goal being scored. You can't separate a player from his team, at least not with such crude data
My way of thinking about medical science, engineering or mathematics in general =simple. Believe me I have spent a lifetime trying the most simple things that Nikola Tesla or Albert Einstein would have never thought would be so complex.
Concerning the tea cups, it probably also depends on the length of time the milk and tea had to mix, their different densities, temperatures of each liquid and cup etc.
How do you make your tea? That answers all your questions. I think there is a big difference between the two methods and she is right. Tea is acidic and milk is a base. As the milk is added to the tea, it heats up quickly and is overwhelmed by the acid and curdles. Adding tea to the milk slowly warms the milk and the acid is quickly neutralized by the milk which doesn't curdle (form a colloid) and that's better.
I feel it's easier to eyeball a more precise measurement of milk when added first. When getting a to-go coffee, the barista will typically ask if you need room. If milk is in first, then the question is not necessary.
I don't like football, but I admit my interest was at least disturbed when I found out that statistics is a BIG thing in football nowadays. Data Scientists and -Analysts must be pretty ecstatic working in that field.
Great talk. There seems to be some confusion between mathematics and psychology though. Some mathematical principles are wrongly generalized to the human behavior and sometimes he just puts out unsolicited claims about psychology. It is interesting to have both of these fields meet but this is obviously not what the talk is about nor what the professor is fluent in. So yeah, it might be another instance of what was pointed out in the talk; mathematicians getting caried away and trying to explain everything
The first example with the tea doesn’t seem like either of the experiments are measuring what the milk first is really conveying: whether the milk first gives a sense of ritual and a small pleasure of identity, or if the milk first created a different temperature variance for immediate consumption. ?
I plead guilty ... Once my then wife complained when I put the milk in her cup and then added the coffee. The next time I prepared her coffee out of sight and asked her which method I had been using. Actually when I suspect that the milk might be close to "expiry" I always put the milk first and test with a minimal amount of coffee. "Shaken, not stirred" ... some scientist did the test.
Maybe I just gotta go back and re-watch, but I missed how the interactive method of thinking was able to to predict things that statistical thinking could not.
interactive method detects factors or variables that can/do change patterns. He used examples from biology/chemistry. So statistical methods identify patterns based on what has already happened; interactive methods identify agents or actions that can change patterns as they are happening (i.e. infectious disease management).
I wonder what kind of thinking was at play to come up with an idea to persistently decrease the video's effective surface area by 20% in favour of displaying three distracting logos.
Hey, I got a math puzzle for you. How long will it take to ring a bell on the moon when you push the earth end of a stick that is within 1 inch of the moon end when you push it at the rate of 1 inch in 1 second?
Probably the length of the stick (around 380 thousand km) divided by the the speed of sound for the stick material - this becomes less of a math problem and more so a physics or engineering problem at the distance described. Without really digging into the problem, another reassurance that it would take longer than a second as suggested is that light (which we can consider a maximum possible speed) only reaches the Moon from the Earth in a little over a second.
One inch per second is much slower than the speed of sound. Not to worry, scientist can't wrap their head around this puzzle either. I believe one's IQ has to be greater than 120 to even grasp this puzzle and probable greater than 200 to solve it.
I take it that the stick is not rigid, and that it has mass. In that case, pushing on the stick causes a wave front that travels up the stick at a specific velocity which can be calculated from the physical properties. I do not think that relativistic issues play any role here in spite of the dimensions involved.
And how many cups did she identify? Actually hang on I'm gonna google this... According to her Wikipedia page, Dr. Bristol correctly identified all eight cups. Appropriately enough I found this right around 21:53 in the video, when Professor Sumpter observed, "I mean, what a dick." 🤣
In the real world, perhaps it is true to say there is no such thing as chaos? Orders born of differing magnitudes all relate back to the smallest factors, whether or not we perceive them. Inverting the Margaret Hamilton example imagine however, a system capable of infinite complexity designed to apply order from a macro perspective, but without the capability for seeing below a certain quantity... lets say in practical terms its design allows for '0' after other numbers, but not before them, and somehow requires the storing of a data in a minimum number of functional memory 'locations' to give a value in calculating. Suppose it's a system with a functionally infinite number of decimal places, that can't however 'hold' numbers starting under a value of 2 separate integers - in effect saying it can make 'two' by 1+1, but cannot make 'one' by 0+1. Designed to bring 'Order' and in abstract terms capable of perceiving division and multiplicity, it's attempting a 'real world' unity based on 'perfect' conception, but lacking the capability for a small yet important part of 'Loving' - the forgetting of self.
Well, to understand the world around you you'd better start watching it. As a physicist I do respect many mathematicians (most of them did some nice works last years), but they are not as lunatic as this speak.
OK I agree option B is the better way to design the trial but what about the answer? Could she tell if the milk was put in first or added after the tea?
The real probability of getting a lady to identify 4 cups out of 8 given two possibilities correctly is 1/256. The lady must evaluate each cup, and given only two outcomes, she would have to correctly label all 8, if she was guessing, no matter the orientation of the cups she would have only a (1/2)^8 or 1/256 chance of correctly identifying all 8 of the teacups. As the test is practically described, this is what happened. A complicating factor is if you decide that she can use process of elimination or observes the cups not all at the same time she could avoid observing all 8 cups and avoid categorizing all 8 cups. But the ladies inability or ability to taste 8 cups of tea at the same time is not part of how this experiment is described. What Fisher describes, when he says "1/70" maybe because he gets a kick out of making people repeat falsehoods, is really the probability of what would happen if she picked out the correct order without observing the cups individually.
He forgot a fifth option: use all seventy combinations but present Pair-wise. I.e. present some "no win" options, and some "no loose" options. If you don't gaslight your rat then are you really a scientist?
02:50 Statistical
23:20 Interactive
39:15 Chaotic
52:10 Complex
Timestamps:
0:00:02 - Introduction to talk on applied mathematics and thinking styles.
0:03:03 - Ronald Fisher's life and work at Cambridge University.
0:05:57 - Testing if milk affects tea taste with experiments.
0:08:42 - Designing experiments using combinatorics.
0:11:31 - Gary Neville statistic measures player performance after conceding a goal.
0:14:17 - Statistics used to rank top players, limitations of measurement.
0:17:00 - Explaining the limitations of statistics in 10 words.
0:19:46 - Eugenics, smoking, and misuse of statistics.
0:22:25 - Context matters in statistical significance and causation.
0:25:13 - Balanced chemical reactions and ecological models explained with math.
0:27:46 - Mathematical modeling of foxes and rabbits.
0:30:32 - Analysis of applause and social recovery in groups.
0:33:12 - Human behavior equation: non-smiling person + 2 smiley people = 3 smiley people.
0:35:55 - Using physics-based models to scout football players.
0:38:51 - Introduction to chaos theory and Margaret Hamilton.
0:41:29 - Weather simulation error due to decimal input mistake.
0:44:05 - Generating divergent numbers through a simple process.
0:47:06 - Chaos, Margaret Hamilton, and the importance of control.
0:49:55 - Finding balance between order and chaos in life.
0:52:33 - Simple rules create complex patterns in simulations.
0:55:21 - Capturing complexity in science through detailed descriptions.
Follow me for more AI
His lecture is not really about ‘ways of thinking’ but mainly arguments to solve certain problems using techniques he expects his audience to know.
Always the British Oxford snobbish people!
Between 0:17:00 & 0:19:46, what are the 10 words explaining the limitations of statistics?
A) "Statistics are valid. It is our interpretation that is wrong." And Sumpter then uses the Grit study as an example.
B) "Statistics are powerful but doesn't give you all the answers." And Sumpter then uses Gary Neville's point as an example.
C) Something else, i.e. please enlighten me.
Exactly the reason why i chosed to study mathematics and physics that is to understand the world around me.And i love studying these two subjects.When i study these two subjects properly i get totally involved,you can say i live in those moments fully.
That's interesting. If I were smarter, I could have moved up a few notches. I, also, wanted to understand the world around me - reality driven - I started with psychology, (psychoanalysis) then later, added the financial markets, geopolitics and a bit of economics.I mostly enjoy my time spent moving around in these realms and feel totally involved and focused, daily. I'm happiest doing my "work". I feel fortunate to live as I choose.
dont let how smart you are define what your work ethic can get you. @@bellakrinkle9381
ok, ok...
Have you completed your degree(s)?
If "Yes", then please confirm / deny the following *hypothesis*: The channel, Veritasium, has recently (on average) significantly declined in quality wrt the subjects explored (Note: ⬅ eval is distinct from the quality of editing, audio, camera, storyboarding, and so on).
Answered separately: The quality decline is in part due to over-focusing on analytics and the business side of the channel. Therefore, in a circumstance where they lacked a solid video idea, rather than skipping the weekly scheduled video release, they prioritized making and releasing one anyways to keep up with the algo. In my estimate, this reason accounts for 30% of the quality decline. Of course, running out of feasible ideas would likely account for the lion's share, say 60%, but I want to know if you agree or disagree with the 30% estimate AND the reasoning supporting your stance.
Ĺm@@pangeaproxima3681
Got me thinking already
The smoking bit reminds me of a silly anecdote when Bertrand Russell was asked if he was concerned about smoking impacting his health. He said that one time he was going on a flight, and they told him he couldn't smoke on the airplane, so he decided to delay his trip rather than go without smoking for a few hours, and the plane ended up crashing.
Wow...Bertrand Russell
If you want to smoke on a plane, no problem. Just buy your own plane. I don't want to smoke on a plane including second hand smoke. If I sit next to you and drink whiskey, you don't get drunk. If you sit next to me a smoke, nicotine ends up in my blood.
Smoking saves lives!
This isn't what happened. On a flight he had to sit in the smoking section and the plane actually crashed. Everyone that was sitting in the non-smoking section died. He had to swim to safety. Therefore his life was saved by having to sit in the smoking section.
You really didn't understand this story did you lol? Maybe watch the above video and get some critical thinking skills.@@chuckhall5347
i love this youtube algorithm which recommends random great video just like a random variable
are you sure it is random?
@@catloverJyes i never dissapointed by this youtube algorithm
Iam happy that it doesn't promotes only ones who has million followers and post nonsense content ,obviously youtube shorts are 🤮
But This video algorithm is like batman 🦇
Random sometimes Calculated when needed.
Nah, it's not random, it's maybe semi-random, selected from the pool adjacent to your interests.
Have you seen the recommendations when you are unlogged? Ridiculous 😅
@@Leonhart_93 yes i have that's why i have 20 different accounts for youtube , 1 for astonomy , 1 for physics chem and maths , 1 for history , 1 for unreal engine animation , 1 for medical research purposes etc,and semi random isn't a proper term because you allow them to show videos you subscribe so either they are a random or seen like a schodinger cat
This was one of the best lectures ive ever watched in my life. Loved it
John Maynard Keynes 'principle of uncertainty' in economics states that we can predict the very near future with some degree of success, but the more distant, the more uncertain we become; thus, mathematical modelling in economics is not a science of absolutes; rather, it's a social science of human infinite variability-a perfect example of chaos theory. Daniel Kahneman later backed this up in behavioural economic theory.
Also Adam Smith wrote, that the market in the long run will return to an equilibrium, to which Keyens replied, in the long run we're all dead! (ie the disaster that was 1930s and the 2010s, both unnecessary suffering, we can always afford a war, so let's win the peace).
Maynard Keynes caused the economic troubles of the 1930s and all the ones after it more than any other economist, that's rich of him to say.
@@jonasastrom7422@jonasastrom7422 Please cite your conclusion about the 1930s and, again, the post-war consensus. I'd be interested in where you get your information to draw your conclusion.
I'll be happy to provide a rebuttal once I understand your source. Thanks for the reply.
@@jonasastrom7422, Please offer some empirical proof; for example, his policies weren't even enacted until the Attlee government of 1945. The 1930s were laissez-faire (Hayek, Smith, Marshall) do nothing and the market would correct, which it didn't! Thus, the Keynes quote, in the long run, we're all dead.
@@jonasastrom7422 Please offer some empirical proof; for example, his policies weren't even enacted until the Attlee government of 1945. The 1930s were laissez-faire (Hayek, Smith, Marshall) do nothing and the market would correct, which it didn't! Thus, the Keynes quote, in the long run, we're all dead.
Yin-yang is chaotic manifestations; taiji is the orderly existence of the whole... if I was to apply this side of classical Chinese philosophy. This of course in no way detracts from what a great talk this is. Thank you!
55:11 A pattern is as complex as the length of the shortest description that can be used to produce it.
Andrei Nikolaevich Kolmogorov
The Kolmogorov complexity stikes again!
It strikes me as amusing that this only seems brilliant on the surface because it's really nothing more than a tautology - which is exactly why it's so concise itself, because it's not really saying anything at all, hence its own complexity is low. It's essentially saying, "A thing is as complex as it is complex."
Okay. And? lol
@HorrorMakesUsHappy A very long pattern may look complex, but if it can be condensed into a repeating motif it can be considered simpler than something that cannot be condensed.
Complexity is characterised here by resistance to compression. This definition distinguishes bloated patterns from complex ones. It's not a tautology.
@@Andrew.baltazar You've missed my point. My point is that the quote is a tautology, not any system it might refer to.
I like how the quote relates to Wolfram's term of computational irreducibility
My statistical way of thinking: of a mindset care for probabilities, historical data, an assessment of assets and liabilities before pursuits of actualization occurs.
My interactive way of thinking: of a mindset care that awaits or hunts for societal circumstances that appears to 'door' or facilitate pursuits of actualization.
My chaotic way of thinking: of a mindset care for rolling the dice-- a 'just do it' actualization pursuit that minds no statistical odds or favorable circumstances and may even lack certainty of what actualization looks like. It is a pleasant surprise when no sense of chaos follows an unstudied action.
My complex way of thinking: of a mindset that utilizes statistical and interactive cognitive intelligence to facilitate pursuits of actualization.
I do believe success is most likely when complex thinking is applied. Additionally, the way of thinking determines what the mind is inclined to focus upon; what we focus upon occurs within mindsets. If I am of a Will to think in a manner that invites the possibility of chaos, there would be no consideration of mindsets associated with statistical or interactive thinking. Often, it is the actualization objective and not the Will that dictates the mind's cognitive cascade options-- care of success demands management of one's Will, usually.
There is a control 'care' in mathematical logic; through an understanding or a care of the answer, one controls the problem.
The lecture's turn towards discerned marital controls that are implemented and sustained, if "care" exists, is of complex thinking that minds objective (statistical) and subjective (interactive) truths. The identified existential marital problem has a determined answer that is not natural within our lecturer-- conscious, willful "control" of his natural inclinations is the answer that he believes "care" will sustain. If a strong "care" becomes weakly felt, the answer and the "manipulations required" will feel a key to a prison versus a key to greater marital happiness; sustaining a marital reality that survives upon an unnatural answer, of little care, is difficult-- even grit requires passion to exist in its application.
Addendum: What is in grit? A Will to persevere, despite any headwinds of statistical and/or interactive inform. What does perseverance require? A deeply held belief that a pursuit will see actualized success. What is in belief? A measured patience that is governed by an adjustable long view.
When the long view dies, all that underpins grit dies; grit may require an ability to visualize and intensely feel an anticipated actualization of victory. What a grit-gifted victorious actualization satisfies is dependent upon the state of one's ego and other psychological influences; a satisfied ego (no circumspect psychological insecurities or of a felt need to proof/validate identity associations of gender, abilities, intelligence, etc.), scar-free interactive psychological truth (never felt derisively underestimated or had declared goals disparaged and/or discouraged) and/or of no passion to defy odds-- less likely to endure pursuits of actualization that requires grit.
Some form of anticipated internal resolution, validation or balm-- beyond the externalized actualization pursuit-- exists in the long view of a grit-driven pursuit, I believe.
Beautifully explained!!
Grit = "short" term control of what actually matters (to that individual). So explained by the third way of thinking.
The more ways of thinking a person have, the more complete a person is going to be.
In my perspective creativity plays a huge role when we apply different ways of thinking, as well as how we organize the language( numbers, symbols, information, knowledge, etc), how we understand the language, the information and how we apply.
A relatively minor point in the talk as a whole but the bit about grit being a good indicator of success is spot on in my experience. I've worked in a technical field for ~30 years so I've observed plenty of folk through those years. Grit is a much better measure of how good an engineer a person will make than GPA or similar.
🎯 Key Takeaways for quick navigation:
00:02 *🎓 Sumpter uses applied math to understand the world through stories from football, science, and life.*
02:48 *📊 Fisher pioneered powerful statistical methods but had problematic views on eugenics and smoking.*
11:28 *⚽ Statistics can measure player attitude and performance in football, with limitations.*
23:25 *🧪 Lotka used "unbalanced" equations to model exciting biological and social patterns.*
37:31 *🧩 But Lotka never found a grand theory. A key limitation was not knowing chaos theory.*
39:18 *🦋 Hamilton and Lorenz discovered chaos - small errors cause drastically different outcomes.*
47:54 *🚀 Hamilton left chaos for Apollo software, showing the need to eliminate errors in critical systems.*
50:00 *☯️ The yin-yang of randomness: control what matters, let go of the rest.*
52:03 *🧩 Complex patterns emerge from simple rules in cellular automata and tiny code tweeted graphics.*
Made with HARPA AI
brilliant and easy to understand. Thanks for publishing
Needs an Oxford comma in the title, of course.
Best comment of all.
Of, course
@@jonminton1878 Why, not, go, for, it, all, t,h,e,n?
Nope. He's British and he's in Britain. Oxford commas are not used in Britain. This is the issue I didn't cave on with the copy editors of my book.
@@davidjohnston4240 60 million people can be wrong. Lord knows I am in America and see it all the time. We have two of them running for president.
For other British faux pas, you have Nigel Farage and Brexit. Just because you, personally, are wrong does not make it right.
Great lecture, thank you for posting it!
How do you arrive at this conclusion?
He's verbose to a fault.
For the curious, Dr. Muriel aced the tea test and thereby hoisted Fisher on his own petard.
I suspect the tea will mix eventually, but not in the time between pouring and drinking.
A better test today might be clear teacups and video cameras that record from all angles.
What causes the difference?
@@kooisengchng5283 Putting a small volume of cold milk into a larger volume of near boiling water scalds the milk and changes the flavour. Adding hot water to cold milk doesn't scald the milk.
Karl Lew, through you, today I learned the word “petard”, thanks
Is it possible that Ronald Fisher did not know there is a difference between putting water to acid and putting acid to water? I stand with Prof Sumpter's opinion that Ronald Fisher is an asshole. He wanted to prove Dr Muriel wrong and devised an experiment that he thought would prove Dr Muriel wrong. Ronald Fisher tried to do the same thing with cancer and tobacco, and with Bayesian statistics. He is an arrogant jerk.
Thank you for wonderful lectures ❤
48:10 "No butterflies allowed."
Why not change the title to “three ways of thinking” ? One can hardly not feel abused (time wise) when told “go buy the book” for the fourth-complex!
5th one is left out, “complainers”
@@cdd7672these are the 4 ways how the presenter thinks. Not how complainers think 😊
I believe the answer to this question is too complex for a short description here.
Because he came there the 3rd time as he mentioned at the start!
The answer is complex
at 6:15, there is this milk-tea problem. i've done a highly scientific research of drinking coffee with vegetarian based (oat and soymilk) milk and because these conditions are a lot more extreme, it enabled me to do some practical tests with clear results. many of those vegetarian "milk" products curdle or accumulate to nasty bit and pieces easily in coffee. to prevent that, pour milk first so it is the strong coffee introduced to a lot of milk first vs. a wiff of milk in a strong coffee. in both cases, a vigorous mixing will help alot. also there has to be enough of milk to make the mix mild enough. yes, i know i missed the point :) so she was right, milk first! also the toilet paper should come out from top side.
This is one of those videos where you can't decide whether you feel smarter or dumber after watching.
ok, ok...
Let me fix that for you.
"This is one of those videos where I can't decide whether I feel smarter or dumber after watching."
@@BariumCobaltNitrog3n ok, ok.....
@@BariumCobaltNitrog3n Got a problem with 3rd person references?
@@MrLuigiFercottiWhen people talk about their own experience, it sounds more authentic than speculating what someone else might think or feel. "You" is 2nd person. He/she/they is 3rd person. A lot of people phrase it like you did, speaking to the reader, pondering. Some though try to speak for the world, as in "No one cares about..."
Anyway, I agree and feel the same, still can't decide which hat...
earphones on and my volume was up, the sound in the beginning almost made me deaf. great content btw.
i'm a "sugar before coffee" man myself. There is definitely a difference in milk before tea: the complex colloid solution of proteins and solutes in milk will react differently based on the relative concentration of milk and tea -- especially with regard to the relative concentration of heat in the tea. I always put two cubes of ice in my coffee after brewing to cool the solution so that the cream doesn't gunk up. The undesirable byproducts of heat on milk are less if there is more milk which allows it to heat more slowly.
but my guess is B, because the experiment yields more information for insight, though maybe i'm wrong.
this assumes that you aren't leveraging the introduction of variance into the values of experimental variables. but for the same reason that Fischer was wrong, your assumption that you can perfectly manage the experimental variables for four pairs of glasses is an oversimplification. it's better to embrace the randomness, quantifying the deviation of variables from norms and thus extracting more information.
GPA is an acceptable way to compare a cohort of students, but when some students don't fit into the variables a cohort controls, it can be a particularly terrible metric for success. You can easily find yourself measuring the ruler. Students representative of the cohort who focus their efforts on meeting the objectives of the classes and programs will typically have high GPA's. At least their effort should correlate to their GPA. Students who don't fit the cohort well, fail for lack of social support or try to extend themselves beyond the constraints of a program may not exhibit a GPA that correlates with time investment or resilience.
It's the last one which can be particularly sad: by setting the expectations for a 4.0 GPA, you load the student with work and prevent them from accomplishing other things. Even the best students are then stovepiped into a singular field in the initial years of their career, unless they did the coursework before their college started.
I almost failed out of community college: I had no one to study with for physics or statics and while I was simultaneously studying devops (no one to help study).
So on paper, I'm a terrible student, but I'm okay with this since no one would understand me anyways.
@@DavidConnerCodeaholic I like to think I fall into that category. I did engineering with not a good GPA (< 3.0 GPA) because at some point I got really interested in the derivation of theorems in engineering and the underlying mathematical, physical and sometimes philosophical insights behind the engineering conclusions, than the applicability of the theorems themselves. Along with battling learning disability, unstable housing situation, poor social awareness, etc., there was a point where I had to pick a choice between 'actual education' of the concepts and the bending of the mind, or to follow blind and measure up to the standard.
It is hard to say that I regret this, because I certainly feel like I have learned far more than I was given the chance to learn, but it may have limited, even in the slightest bit, for further learning, as GPA is used as a source of evaluation. Although perhaps other activities may be used to supplement my profile in the context of graduate level admissions or so, that is if I am even in interested in pursuing more academia. Since I have graduated my engineering program, I am eligible to start working, but post secondary is something in my mind to pursue potentially. Ideally I would like to get a job.
It really is a tough world out there. To "map" the complexity of human capability to a single number 'GPA' is such a limiting way of looking at things, but at least it is 'a measure' of looking 'at something'. A world, that looked so objective and rational to me, upon further reading and studying different concepts, seems rather subjective, arbitrary and uncertain.
For the milk tea test. The difference with the pairwise option, is that there is extra information that 1 of those 2 cups is milk first and the other is tea first, that's why its more probability for her to get it right. For the 3rd option: 8 choose 4 = 70, as from 8 positions choose 4 for the milk first option.
That was a well-coordinated crowd at the end.
Great talk. As regards the slight relationship between grit and success the terms 'statistically significant' vs 'practically significant' come to mind. The problem then is practically significant for who? As we move through life we meet individuals and it is a foolish person who makes any assumptions about an individual based on secondary characteristics. But what about those who are making evaluations in bulk?
Suppose someone has found a similar slight relationship between 'grit' and programming skills and is charged with hiring 100 programmers for a project. The temptation might be to use 'grit' as an evaluative proxy when hiring. But what if 'grit' has some cultural factors that influence it? Then the hiring process might be biased against some groups and this bias might be incorrect.
Bean counting micro-optimisation is (unfortunately) a common feature in business, moreover increasing use of 'AI' is likely to hide such biases behind a wall of pseudo-objectivity. Bringing us back to the Fisher problem.
Fantastic insight.
One is then reminded of the chaotic and complexity ways of thinking, since programming ability (as well as football/soccer) are definitely not easily predictable at all
Several layers of unwarranted assumptions in there.
I am very naive and want to learn more, you made very good point, but my concern is the 4% variance explained by grit provides valuable information for researchers, educators, employers, and individuals themselves who are interested in understanding and fostering success. It highlights the importance of factors like perseverance and passion for long-term goals in achieving positive outcomes. Apart from cultural factor the other factors are not proven and measured yet, don't you think one should believe on proven statistics? If not, then how come one believe on other factors if they are not proven yet?
the art of the measurement becomes the reality to the measured ambitions of the narrowness of arrogance. i found David Sumpter's beauty in these 4 measures the protagonist toward the Long Term Capital Management story. really appreciate this opportunity of sublime serendipitous interaction the youtube algo sought out my desire for such. thank you.
Thanks for the talk! I enjoyed watching it with my kid!
A bit more intuition on the "Complex" segment would have made it a lot more worthwhile, it while being a brilliant talk without a doubt. The two examples were good, but I guess it needed a few more words from him. I suppose the time limitation at hand...
great! all these ways of thinking are from macro perspective to see the world. so that we can see the pattern, trending, development of organic system.
I've only seen thru 32 min of this video, am enjoying it, have skills and education in math, and suggest that there is a stronger correlation between the LOUDNESS of applause (coupled with the peak loudness correlation) and enjoyment of the "event" rather than the length of time of the applause. Just a spontaneous theory here. Could be wrong...
Very interesting, highly recommended
Interesting, but I prefer the six thinking hats, simpler and yet more complex!
Many Thanks for Right Explain.............! Wo ist der Zürich geschnetzen Menüs..............................................?
Wonderful lecture!
I write this comment around 9.00 minutes. The second method, "B" , isn't "fair". One has to keep in mind that humans use their MEMORY. The second test assumes that a human can instantly reset their pallette after a short taste-test, this is simply NOT the case. When the difference in taste is subtle, then all the interpretations become one big jumble for a human's brain if the taste tests are set too close after each other.
Also , the first thing that sprung into my mind was, "Maybe it's due to a difference in TEMPERATURE, that the lady can tell the two variants apart with a significant amount of certainty." So, is there a difference between T(t) and T*(t)? T(t) is the temperature curve when milk is poured first, and T*(t) is the temperature curve when tea is poured first.
Extremely interesting to me
Mr. Kolmogorov is the favourite one
Stats are often suspect because people fail to remember that independence - essentially that "all else remains equal" - is usually a dubious, or just wrong, assumption.
When a goal is scored, offensive and defensive strategies can change. There can be more emphasis, or less emphasis, on giving more, or less, responsibility to a player - or such a variation in how the other team chooses to handle that player. You tend to take bigger risks when you;re down, and so on. Did you start to play worse, or better, ... or did your teammates? Or did your opponent? It's no surprise that you're a more/less accurate passer if they're now running faster/slower and so are more/less open than before, or that their defenders are now playing better/worse than before.
I go through life endlessly hearing arguments like this video's "stats measure which players have the most grit/heart", and I'm just endlessly unimpressed. It's not that it's necessarily wrong, it's just never as compelling as people seem to imagine. Because all else is never equal, usually. In other words, statistics, without scientific controls, aren't justifications for conclusions - they aren't science - they're just unanalyzed data. You need control groups, first and foremost. You need to account for placebo effects. You need reliability (objective standard and immediately recorded, not memory), and impartiality (double blind) in the data collection. You need to at least consider if not incorporate the considerations of previous related research. You need to understand all the ways your assumptions can be flawed... and do a numerical analysis those assumptions as well. And of course ultimately, you need to allow others to critique what you did. On and on it goes.
Statistics - the data - doesn't speak. Only a scientific analysis speaks.
23:28 where did this come from?
i'd rather learn more about this way of thinking
@48:18: But there was room for errors at NASA. During the first successful moon landing, Apollo 11, they encountered programming alarms 1201 and 1202. There was a question on whether to abort the landing mission or not, and they went ahead.
I wish he could have elaborated on why complexity is distinct from chaos, but he cut it short. Seems he was only describing the same thing - chaos is macro-behaviour that emerges from the complex interaction of micro-motives. So not sure why they merit a separate category.
Chaos may produce randomness, but complexity is something which may look very random, yet be completely deterministic and hence exactly reproducible.
Good question, they may be different in scale instead of category.
Let's think about this a bit.
You can have chaotic behaviour arise from something very simple and logical.
On the other hand complexity interacts with the strength of the connections in various ways, when one of those ways is chaotic then complexity + tight coupling -> danger.
So in a loosely coupled complex system you will not get chaotic macro behaviour, because the loose coupling prevents the small chaos from growing.
I think it relates back to the Kolmogorov quote, applied to the cellular automata example:
If a chaotic-seeming thing can be described with just three simple rules, then the apparent chaos is merely complex. Indeed, one might argue that the fewer rules that can be in play in a system, the less likely it is that the outcome will be truly chaotic. If, however, you need many, many rules to describe the system, it can reasonably be called chaotic.
This would tend to imply, though, that as we generate better descriptions of what we do (seem to) understand, we can turn some chaotic-seeming systems into (merely) complex systems. I guess the current work on modelling the impact of CO2e emissions might be an example.
It's the mathematical equivalent of the 'God of the Gaps' argument. God was frequently invoked as the "reason" for something that was inexplicable, and then we explained it, and that was one less gap that God was needed as an explanation for.
@@Alan_Duval I think the line dividing chaotic systems from complex (but non-chaotic) ones, is if it’s possible to reproduce a process *exactly*. Some systems, due to nonlinearity, cannot be reproduced exactly due to roundoff errors, so they’re chaotic rather than complex, and would continue to be so regardless of how much we know about them, in my opinion.
7:21 oh, good methodology 21:42 clever isn't the same as smart. 33:56 where does the thinking come in?
Yes, Dr. Muriel Bristol, the subject of Ronald Fisher's famous tea-tasting experiment, was indeed able to tell whether the milk was poured into the tea or the tea was poured into the milk, according to the results of the experiment!!! Please don't leave us hanging.
It's funny how he tried to use statistics to really prove her wrong :)
Descriptions meaning and information content depend on the formula or algorithm evaluation it. It is always a sort of function giving the se of possible source data meaning in the result space.
It's pretty amazing how a whole field of mathematical inquiry began because a guy just couldn't believe that a woman knew how she liked her tea and had to - HAD TO - prove her wrong.
One way to interpret the story. When I was told the story, they said everyone enjoyed the experiment.
Incorrect @10:08 - No matter how many ways there are to present four cups of tea, there are always 2 cups 'milk first' and 2 cups 'tea first' on the table. So, the chance of getting the first cup right is 50% (there are 2 cups of each). If the first guess was right, there are 3 cups left on the table: 1 'milk first' and 2 'tea first'. Now, the chance of getting the second cup right is 33%. And, if the second guess was right, both other guesses are per definition also correct. (The only cups left are the 'tea first' ones.)
So, the chance of getting your guesses right in the second set-up are 50% x 33% = 16,6% or 1 in 6.
thank you so much
Vielen Dank allen Politiker Opas Omas Präsentiert & Dokumentation Strategien
15:30 - right right i remember her Ted talk about "Grit"....what a god awful way to do analysis. the gist, as you can plainly guess, was "well, those that succeeded had....grit...because they were the ones that...succeeded in the end" talk about water is wet.
kind of like the survivor bias, amirite? show a bunch of people pictures of where bombers, who returned to base, were shot full of holes and you'd rightly assumed it would be "hey, let's strengthen those bits that are more likely to be shot up"...except they forgot that this ignored the ones that were shot up in the places not full of holes in the survivors...and didn't make it back to base.
I shouldn't have to ask this because credit should have been given in the slide deck, but can anyone identify the artists behind the Twitter art @54:19 ? Note: In 2024, the X, formerly known as Twitter, character limit is 280 not 240 as suggested in this video.
He is very smart spoken.
38:20 "Magnus opus"? Also, a lot of his complaints aren't shortcomings of the techniques he discusses--they're shortcomings of the people using the techniques.
What would that phrase be in your native language?
I like Ronald Fischer, from being a Eug@nics advocate to going 80 years ageless teaching at Oxford about Himself in "Four Ways of Thinking: Statistical, Interactive, Chaotic and Complex"
He didn't give the fourth option: "I don't know which one is better", for what the effect is of the way you present the experiment.
17:21 Bird shot variance.
On Tea and Cookies: I think there is a big difference between the two methods and she is right. Tea is acidic and milk is a base. As the milk is added to the tea, it heats up quickly and is overwhelmed by the acid and curdles. Adding tea to the milk slowly warms the milk and the acid is quickly neutralized by the milk which doesn't curdle (form a colloid) and that's better.
Not a stats expert, nor a futbol expert, but as a scientist I strongly suspect the explanation for these strikers' performances has a lot more to do with how the other players on the field react to a goal being scored. You can't separate a player from his team, at least not with such crude data
The chaos example with yin and yang hit home so hard i might be able to sleep tonight 😂
How many maths were you good at?
That was very good. Thank you.
My way of thinking about medical science, engineering or mathematics in general =simple. Believe me I have spent a lifetime trying the most simple things that Nikola Tesla or Albert Einstein would have never thought would be so complex.
Concerning the tea cups, it probably also depends on the length of time the milk and tea had to mix, their different densities, temperatures of each liquid and cup etc.
How do you make your tea? That answers all your questions. I think there is a big difference between the two methods and she is right. Tea is acidic and milk is a base. As the milk is added to the tea, it heats up quickly and is overwhelmed by the acid and curdles. Adding tea to the milk slowly warms the milk and the acid is quickly neutralized by the milk which doesn't curdle (form a colloid) and that's better.
Always pour the added liquid down close to the inner edge of the cup, like a good Chemist, and then it doesn't matter much which went first.
I feel it's easier to eyeball a more precise measurement of milk when added first. When getting a to-go coffee, the barista will typically ask if you need room. If milk is in first, then the question is not necessary.
i'm exited about this presentation
You should exit this presentation yeah we all agree!
How can I Win this Game...................?
I don't like football, but I admit my interest was at least disturbed when I found out that statistics is a BIG thing in football nowadays. Data Scientists and -Analysts must be pretty ecstatic working in that field.
Confusing lecture, but interesting. I hope his books are more clear on demonstrations
Great talk. There seems to be some confusion between mathematics and psychology though. Some mathematical principles are wrongly generalized to the human behavior and sometimes he just puts out unsolicited claims about psychology. It is interesting to have both of these fields meet but this is obviously not what the talk is about nor what the professor is fluent in. So yeah, it might be another instance of what was pointed out in the talk; mathematicians getting caried away and trying to explain everything
The first example with the tea doesn’t seem like either of the experiments are measuring what the milk first is really conveying: whether the milk first gives a sense of ritual and a small pleasure of identity, or if the milk first created a different temperature variance for immediate consumption. ?
These are great stories. But what about the tea: which was better? And what did explain success better than grit & IQ at 4A% each?
What number Butterfly Top Model Please
0:08 the most funniest part is how they didn’t use the Oxford comma here despite this video literally being published by Oxford Mathematics 😂
Not the Oxford English Dept lol
Thx
Maths is magic and magic is maths
46:26 What happens if you choose 50 as your number? ( It's neither smaller than nor greater than 50 😂)
Simulation malfunction. Reboot human.
Freeform Creative - Extreme Synthesis ?
I plead guilty ... Once my then wife complained when I put the milk in her cup and then added the coffee. The next time I prepared her coffee out of sight and asked her which method I had been using.
Actually when I suspect that the milk might be close to "expiry" I always put the milk first and test with a minimal amount of coffee.
"Shaken, not stirred" ... some scientist did the test.
Maybe I just gotta go back and re-watch, but I missed how the interactive method of thinking was able to to predict things that statistical thinking could not.
interactive method detects factors or variables that can/do change patterns. He used examples from biology/chemistry. So statistical methods identify patterns based on what has already happened; interactive methods identify agents or actions that can change patterns as they are happening (i.e. infectious disease management).
I wonder what kind of thinking was at play to come up with an idea to persistently decrease the video's effective surface area by 20% in favour of displaying three distracting logos.
I still don’t understand what 4% of the variance means😢
on 🔛 to be Thninker in life amazing gift !
Hey, I got a math puzzle for you. How long will it take to ring a bell on the moon when you push the earth end of a stick that is within 1 inch of the moon end when you push it at the rate of 1 inch in 1 second?
tell us
I believe 1 second, but I am trying to get a scientist to respond.
Probably the length of the stick (around 380 thousand km) divided by the the speed of sound for the stick material - this becomes less of a math problem and more so a physics or engineering problem at the distance described. Without really digging into the problem, another reassurance that it would take longer than a second as suggested is that light (which we can consider a maximum possible speed) only reaches the Moon from the Earth in a little over a second.
One inch per second is much slower than the speed of sound. Not to worry, scientist can't wrap their head around this puzzle either. I believe one's IQ has to be greater than 120 to even grasp this puzzle and probable greater than 200 to solve it.
I take it that the stick is not rigid, and that it has mass. In that case, pushing on the stick causes a wave front that travels up the stick at a specific velocity which can be calculated from the physical properties. I do not think that relativistic issues play any role here in spite of the dimensions involved.
Fisher designed the tea tasting experiment. But did Dr Bristol actually do the experiment with him???
And how many cups did she identify? Actually hang on I'm gonna google this... According to her Wikipedia page, Dr. Bristol correctly identified all eight cups.
Appropriately enough I found this right around 21:53 in the video, when Professor Sumpter observed, "I mean, what a dick." 🤣
In the real world, perhaps it is true to say there is no such thing as chaos? Orders born of differing magnitudes all relate back to the smallest factors, whether or not we perceive them. Inverting the Margaret Hamilton example imagine however, a system capable of infinite complexity designed to apply order from a macro perspective, but without the capability for seeing below a certain quantity... lets say in practical terms its design allows for '0' after other numbers, but not before them, and somehow requires the storing of a data in a minimum number of functional memory 'locations' to give a value in calculating.
Suppose it's a system with a functionally infinite number of decimal places, that can't however 'hold' numbers starting under a value of 2 separate integers - in effect saying it can make 'two' by 1+1, but cannot make 'one' by 0+1. Designed to bring 'Order' and in abstract terms capable of perceiving division and multiplicity, it's attempting a 'real world' unity based on 'perfect' conception, but lacking the capability for a small yet important part of 'Loving' - the forgetting of self.
Fisher is one of the encyclopedia examples of “that guy.”
How about just getting to the point?
Rubics cube, paths to solution, I had to watch 55min 31s to ... we all want the power of action at a distance.
Well, to understand the world around you you'd better start watching it.
As a physicist I do respect many mathematicians (most of them did some nice works last years), but they are not as lunatic as this speak.
Cowboy hat being labeled "chaotic" caught me off guard not gonna lie
OK I agree option B is the better way to design the trial but what about the answer? Could she tell if the milk was put in first or added after the tea?
1 to 99? Why not 0 to 99, or 1 to 100?
because both 0 and 100 (given the specific set of rules) would cause a loop of zeros
Cute, intuition , feeling, sensation, and thinking
19:20
The real probability of getting a lady to identify 4 cups out of 8 given two possibilities correctly is 1/256. The lady must evaluate each cup, and given only two outcomes, she would have to correctly label all 8, if she was guessing, no matter the orientation of the cups she would have only a (1/2)^8 or 1/256 chance of correctly identifying all 8 of the teacups. As the test is practically described, this is what happened.
A complicating factor is if you decide that she can use process of elimination or observes the cups not all at the same time she could avoid observing all 8 cups and avoid categorizing all 8 cups. But the ladies inability or ability to taste 8 cups of tea at the same time is not part of how this experiment is described.
What Fisher describes, when he says "1/70" maybe because he gets a kick out of making people repeat falsehoods, is really the probability of what would happen if she picked out the correct order without observing the cups individually.
He forgot a fifth option: use all seventy combinations but present Pair-wise. I.e. present some "no win" options, and some "no loose" options. If you don't gaslight your rat then are you really a scientist?
Is language a medium of chaos?
Didn't some bloke called de Bono come up with Six Thinking Hats already? Which hat is the one for ripping off ideas?
I do not believe the title describes the content of the video.
I think that’s the name of his book