The Model Thinker: What You Need to Know to Make Data Work for You
- Downloads：4370
- Type：Epub+TxT+PDF+Mobi
- Create Date：2022-02-28 09:55:02
- Update Date：2025-09-06
- Status：finish
- Author：Scott E. Page
- ISBN：0465094627
- Environment：PC/Android/iPhone/iPad/Kindle
- Download
Summary
Work with data like a pro using this guide that breaks down how to organize， apply， and most importantly， understand what you are analyzing in order to become a true data ninja。
From the stock market to genomics laboratories， census figures to marketing email blasts， we are awash with data。 But as anyone who has ever opened up a spreadsheet packed with seemingly infinite lines of data knows， numbers aren't enough: we need to know how to make those numbers talk。 In The Model Thinker， social scientist Scott E。 Page shows us the mathematical， statistical， and computational models—from linear regression to random walks and far beyond—that can turn anyone into a genius。 At the core of the book is Page's "many-model paradigm，" which shows the reader how to apply multiple models to organize the data， leading to wiser choices， more accurate predictions， and more robust designs。 The Model Thinker provides a toolkit for business people， students， scientists， pollsters， and bloggers to make them better， clearer thinkers， able to leverage data and information to their advantage。
From the stock market to genomics laboratories， census figures to marketing email blasts， we are awash with data。 But as anyone who has ever opened up a spreadsheet packed with seemingly infinite lines of data knows， numbers aren't enough: we need to know how to make those numbers talk。 In The Model Thinker， social scientist Scott E。 Page shows us the mathematical， statistical， and computational models—from linear regression to random walks and far beyond—that can turn anyone into a genius。 At the core of the book is Page's "many-model paradigm，" which shows the reader how to apply multiple models to organize the data， leading to wiser choices， more accurate predictions， and more robust designs。 The Model Thinker provides a toolkit for business people， students， scientists， pollsters， and bloggers to make them better， clearer thinkers， able to leverage data and information to their advantage。
Reviews
Vlad Ardelean
，
Really good book。 A little too many formulas （formulae） to be fun to listen in audio format。 Also a little too academical。 Still， it's a book for smart people， or （like me）， people who want to look like they're trying to be smart :p Really good book。 A little too many formulas （formulae） to be fun to listen in audio format。 Also a little too academical。 Still， it's a book for smart people， or （like me）， people who want to look like they're trying to be smart :p 。。。more
Samuel
，
Initially interesting but found it hard to understand the examples abd the application of models toward the end。 Also some useless bible quotes here and there。
D
，
Models have not been this fashionable since Linda Evangelista， Naomi Campbell， and Claudia Schiffer were strutting down catwalks in the 90's。 The rise of big data， and high-profile debate over models for climate change and COVID have made mathematical models increasingly important。 The Model Thinker does an admirable job explaining many different models， making a convincing case that using many models to investigate issues from different perspectives is a sensible approach。Page provides 29 chapt Models have not been this fashionable since Linda Evangelista， Naomi Campbell， and Claudia Schiffer were strutting down catwalks in the 90's。 The rise of big data， and high-profile debate over models for climate change and COVID have made mathematical models increasingly important。 The Model Thinker does an admirable job explaining many different models， making a convincing case that using many models to investigate issues from different perspectives is a sensible approach。Page provides 29 chapters， most of which deal with a separate set of models， moving from concepts familiar to high school stats students like power laws and normal distributions through Lyapunov functions to game theory and rugged landscapes。 If you've ever wondered what pseudo-intellectual writers mean when they mention entropy or Markov models， this book is an excellent place to start。Each chapter neatly explains the concept behind each model， the underlying maths， and examples of how they work in practice。 The text heavy approach means you need to keep your wits around you - the more complex models require less mathematical-minded readers to re-read and check the text carefully。 More liberal use of graphs and illustrations would have been beneficial， but the avoidance of too much maths will be welcomed by many。 Given the speed with which some techniques are fine-tuned and adopted， this is a wise approach lest the book becomes too swiftly outdated。The Model Thinker deserves a place on the desks of anyone working with or interested in data。 The final chapter shows why modelling is vital to understanding epidemics， drug addiction， and inequality。 We need people with a practical， nuanced understanding of how to best use data to understand the world and society。 No single book can provide all the answers， but The Model Thinker is a valuable starting point。 。。。more
Kilian Murphy
，
Really good introductory text for many aspects of data analysis。 Quite a comprehensive suite of statistical models explained and with a variety of applications。 For the models I’m familIar with （e。g Agent-Based simulations） the author does well to reference appropriate literature for further investigation into more comprehensive texts， which gave me confidence that the same care was taken with models that I am not familiar with - thus trusting the author to point me in the right direction。 One a Really good introductory text for many aspects of data analysis。 Quite a comprehensive suite of statistical models explained and with a variety of applications。 For the models I’m familIar with （e。g Agent-Based simulations） the author does well to reference appropriate literature for further investigation into more comprehensive texts， which gave me confidence that the same care was taken with models that I am not familiar with - thus trusting the author to point me in the right direction。 One aspect that was missing was pointing the reader to appropriate software， packages and libraries to deploy these models and offer applied examples。 For instance， R stats， python and Netlogo are all great software with an active online community full of learning resources and inbuilt packages for getting started in using these models to analyse your data in real life scenarios。 Finally， certainly not a “sit down and read” book， I would in future go back to this prior to getting familiar with a new class of model as a reference to understand the concept before diving into the computations。 Great book。 。。。more
Landon Wall
，
Too technical / math focused to be entertaining， not technical enough to be useful for professionals。
emma
，
His class is one of the most interesting I have ever taken， and I would take it again。 I do recommend that to everyone: https://www。coursera。org/learn/model-。。。His book is less illustrative， since it lacks the visuals helping in understanding models like Conway's Game of Life。 It might have been better as a full textbook， with example problems for the reader to walk through like he does in his online course。 Libby and Kindle did the book dirty though， in that all of the included equations were i His class is one of the most interesting I have ever taken， and I would take it again。 I do recommend that to everyone: https://www。coursera。org/learn/model-。。。His book is less illustrative， since it lacks the visuals helping in understanding models like Conway's Game of Life。 It might have been better as a full textbook， with example problems for the reader to walk through like he does in his online course。 Libby and Kindle did the book dirty though， in that all of the included equations were illegibly formatted。 。。。more
Zhijing Jin
，
Strong recommendation！Key takeaway: Either） Smartly choose the one out of many models that applies to the current scenario。 Or） Ensemble all the models， and get collective intelligenceIntuitions: - Ensemble is more stable than putting all eggs into one basket。- In real world， the conditions are more complicated， so be prepared to be surprised。Toolbox of different models:- Forest fire （bigger fires can be prevented by allowing smaller fires to get rid of some dry woods）- Gaussian distribution vs。 Strong recommendation！Key takeaway: Either） Smartly choose the one out of many models that applies to the current scenario。 Or） Ensemble all the models， and get collective intelligenceIntuitions: - Ensemble is more stable than putting all eggs into one basket。- In real world， the conditions are more complicated， so be prepared to be surprised。Toolbox of different models:- Forest fire （bigger fires can be prevented by allowing smaller fires to get rid of some dry woods）- Gaussian distribution vs。 power law （e。g。， through how interconnected each element is）- Value of a talent （which can be understood as how replaceable that person is）- Network model （big nodes in the network）- And many more :）Related readings:Echoing with the point of the book:- "More Is Different" （Nobel laureate in physics， P。 W。 Anderson， 1972）- The wisdom of the crowdsBasis that this book builds on/improves upon:- Expert Political Judgment- "The Unreasonable Effectiveness of Mathematics in the Natural Sciences"， a 1960 article by the physicist Eugene Wigner 。。。more
Justohidalgo
，
This is an amazing and totally necessary book for any of us who need to apply models to our decision making。 Scott E。 Page， whom I first new about with his great Coursera course on models， takes one step （or many！） further and builds a tour de force with twenty four different models and mathematical constructs that aim to explain a huge number of real-world problems。 At least for me， the book has been an increíble journey but also a very though one。 I started reading it at the end if August and This is an amazing and totally necessary book for any of us who need to apply models to our decision making。 Scott E。 Page， whom I first new about with his great Coursera course on models， takes one step （or many！） further and builds a tour de force with twenty four different models and mathematical constructs that aim to explain a huge number of real-world problems。 At least for me， the book has been an increíble journey but also a very though one。 I started reading it at the end if August and just finished it in December。 Sixty-three pages of notes that I will to need to review and re-review。 Models that I will need to study deeper。 But the base knowledge I got from this book is invaluable。 。。。more
☘Misericordia☘ ⚡ϟ⚡⛈⚡☁ ❇️❤❣
，
The good: the outline， the breadth， the sheer volume of material。 The reader gets bits of everything。The bad: skimming and no really deep dives。 The bits are way too tiny。The not-nearly detailed enough: - math and the know-how。 I don't think many newbies will manage to do many （or even any） of these models basing on just this one skinny volume。 But then again， a girl can dream of a book that has it all， right？The really good take-outs driver: 29 model groups are conversationally discussed， most The good: the outline， the breadth， the sheer volume of material。 The reader gets bits of everything。The bad: skimming and no really deep dives。 The bits are way too tiny。The not-nearly detailed enough: - math and the know-how。 I don't think many newbies will manage to do many （or even any） of these models basing on just this one skinny volume。 But then again， a girl can dream of a book that has it all， right？The really good take-outs driver: 29 model groups are conversationally discussed， most of these are a mix of some more models that are really broadly used in a lot of industries。 A lot of motivation for usage is quite creatively and readably layered out。 Nice work！ 。。。more
Ben Stenhaug
，
looks really good！
Max
，
Nice collection of models。 If you have some behavioural science modelling experience a lot will be familiar。 A rough sample: game theory stuff， epidemiological models， Markov models， some statistical distributions， learning models， network models。 I found the explanations well done， the chapters are relatively short and lean more in the direction of intuition and application， but the basic equations are always featured and explained as well。 I'd recommend skipping the first two chapters， at leas Nice collection of models。 If you have some behavioural science modelling experience a lot will be familiar。 A rough sample: game theory stuff， epidemiological models， Markov models， some statistical distributions， learning models， network models。 I found the explanations well done， the chapters are relatively short and lean more in the direction of intuition and application， but the basic equations are always featured and explained as well。 I'd recommend skipping the first two chapters， at least for me they were a drag and I feel like they can be summarized like this kinda obvious point: Don't rely on any single model， they almost always simplify too much， at least use multiple models。I really liked this quote from Jean Piaget， apparently from 1968: Knowing reality means constructing systems of transformations that correspond， more or less adequately， to reality。And I think I found a mistake in his definition of Simpson's paradox？ Logic can also reveal paradoxes。 Using models we can show the possibility of each subpopulation containing a larger percentage of women than men but the total population containing a larger percentage of men， a phenomenon （Simpson’s paradox）。 Shouldn't the average number of women of the total population just be a weighted average of the average numbers of women in all subpopulations， and therefore not be able to fall below 50%？ Anybody can explain what I'm missing？ 。。。more
May
，
Book is just what you would know but don’t know， it may prove one thing by the end its title， book
Jung
，
In an increasingly complex world， we need models to make sense of the perplexing systems around us。 They can help us to explain the world， to create new designs， and to predict what’s coming next – but only if we’re careful to apply the right ones。 In fact， to optimize our results， we should try to tackle a problem using as many diverse and relevant models as we can。---The key message here is: Modeling humans is a thorny endeavor。People can be troublesome: if that’s true when it comes to real li In an increasingly complex world， we need models to make sense of the perplexing systems around us。 They can help us to explain the world， to create new designs， and to predict what’s coming next – but only if we’re careful to apply the right ones。 In fact， to optimize our results， we should try to tackle a problem using as many diverse and relevant models as we can。---The key message here is: Modeling humans is a thorny endeavor。People can be troublesome: if that’s true when it comes to real life， then it’s doubly true when it comes to modeling。Unlike bowling balls or weather cycles， humans have agency。 We have minds of our own and we use them to make decisions。 We experience social pressure， we have different preferences， and we often make mistakes。 Sometimes， we might even learn from those mistakes。That’s what makes us fascinating – but from the point of view of modeling， it also makes us frustrating。 So does that mean it’s hopeless to try to model human behavior？Whenever we try to model human behavior， there are some choices and assumptions that we can’t avoid。 The first decision we have to make is between two different ways of picturing it – as rule-based or as rational。Rule-based behavior can be broken down into two main types: fixed and adaptive。 Fixed rules don’t evolve as time passes and circumstances change。 If we were to formulate a fixed rule in words， we might come up with something like “If the conversation lulls for longer than 20 seconds， change the topic。”An adaptive rule， on the other hand， changes and evolves in response to new circumstances and novel information。 An adaptive rule might allow silences to go on for longer than 20 seconds if it became apparent that these conversational lulls ultimately led to better discussions。By contrast， the rational-actor model of human behavior assumes that people make rational decisions in order to achieve optimal outcomes。 Instead of following set rules， they calculate what the best action is in any given situation and act on that information。Think of someone buying a house and calmly weighing up their options: the number of bedrooms each house has， the view out the kitchen windows， even the schools in each neighborhood。Neither rule-based nor rational-actor models work for every situation。 When choices are simple or made by sophisticated decision-makers， they're likely to be rational ones。 But when a choice is fairly low-stakes， like what color coat to buy， it’s likely that people will apply fixed rules。 In others， like deciding who to trust in a delicate negotiation， people might apply adaptive rules。When it comes to modeling humans， we can rarely hope for complete accuracy。 But choosing the right models can make our predictions， designs， and explanations far more accurate。 。。。more
Nhi
，
The great abstract of data modelling that I would highly recommend for the armchairs in this area。Not kind of mind-changing （for me only）， but it synthesizes the dots and makes them clear & concise enough ^^Key models mentioned:- Basic distribution （bell curve）- Power laws （long-tailed） - upgrade of 80/20 ^^- Linear regression- Concave & convex functionsBasic distribution & 80/20 have been 2 of my favorite topics with guys so far （idk why huhu）， but maybe I will add power laws then =）））Aw and I The great abstract of data modelling that I would highly recommend for the armchairs in this area。Not kind of mind-changing （for me only）， but it synthesizes the dots and makes them clear & concise enough ^^Key models mentioned:- Basic distribution （bell curve）- Power laws （long-tailed） - upgrade of 80/20 ^^- Linear regression- Concave & convex functionsBasic distribution & 80/20 have been 2 of my favorite topics with guys so far （idk why huhu）， but maybe I will add power laws then =）））Aw and I miss Nielsen a little bit when jotting down these words 🥺An article of the author with the related topic， published in HBR: https://hbr。org/2018/11/why-many-mode。。。 。。。more
Sid Johnson
，
The agenda of this book is to convince the reader that using multiple models to attempt to understand a situation beats using a single model for most problems of interest。 He does so in a way that lets you enjoy the beauty of models and the surprise of the unexpected results。 Very little math is involved， and what there is could be skipped easily。 The book as a whole will only be of interest， I think， to people who love to see how systems in the world work and don’t work。 I was engrossed in it i The agenda of this book is to convince the reader that using multiple models to attempt to understand a situation beats using a single model for most problems of interest。 He does so in a way that lets you enjoy the beauty of models and the surprise of the unexpected results。 Very little math is involved， and what there is could be skipped easily。 The book as a whole will only be of interest， I think， to people who love to see how systems in the world work and don’t work。 I was engrossed in it immediately but， after several days， set it aside and didn’t come back to it for a few months。 It was no problem to pick it up and carry on。 It’s semi-cookbook structure lends it to that。 If you liked Freakonomics， or any book of that sort， this is that with more depth and breadth。 The seven-model examination of the causes of wealth/income inequality is extremely enlightening and worth the price of the book。 It takes only 11 pages。 It will give you a very different perspective and is worth reading， even if you have to get the book from the library or read it at your local bookstore。 Starts on page 343！ Not one mention of politics。 。。。more
Arvilla
，
Very informative with lots of real-world examples。 I admit I had issues getting through it （I just paged through the latter half） but will keep it on my shelf as a reference book for future data endeavors。
Chrisly
，
Great book on modelling！
Jamil KASSAB
，
I read along with the course scott is offering on coursera 。 So insightful book with lot of useful models
Max
，
After using Scott Page's previous work on path dependence models during my bachelor thesis， I was really excited when this book was suggested during my master。 A beautiful book！ It took me some time to go through all the introduced models but it was worth it！ After using Scott Page's previous work on path dependence models during my bachelor thesis， I was really excited when this book was suggested during my master。 A beautiful book！ It took me some time to go through all the introduced models but it was worth it！ 。。。more
David Kritz
，
Dr。 Page is one of my academic heroes。 I have taken his courses online and watched his videos on the Great Courses Plus。 I cannot endorse his books highly enough。
Humam Fauzi
，
Explain well about many model and its application。 Not too deep but not to shallow either so a layperson could understand without making the one who is familiar with the field bored。Maths does not too hard but on few model require to write things down to make it more clearer。 Very satisfying。 Recommended for someone who want to understand analysis， causality， data and statistics。
Robert
，
A good addition to the subject matter if you have some experience with modeling。
Walter
，
This book accomplishes what it sets out to do: to expose the reader to a wide range of mental models in a very accessible way， without the need for any specialized training in mathematics。 Of course， a little math goes a long way in extracting the most here。 For those in this category， some of the models will be familiar。I subtract one star because though it opened very strong and remained so for 75% of the book， it kinda lost steam in the later chapters。 Recommended。
Min
，
This books offers insight into so many disciplines by giving a primer of so many models and concepts， developed by different people for different problems encountered in the real world。 While it can only be an introductory reading due to the vast number of models presented， it is still very dense and at times very technical。 So be prepared to do some skimming if you want to get through it。 But your horizon will be broadened。
Jaume Sués Caula
，
Walkthrough of the up-to-date mathematical models， with real-life applications。 I missed more in-depth description of neural networks， basis of artificial intelligence。 Great dense lecture， in any case。
Silvio
，
Interesting and revealing， a guide to help think on today's complex problemsWhat's a model？ How to use a model？ Why to use a model？ Any qualitative approach to apply a model？ Only use 1 model or various？ These are the questions that these book look to help to answer。 It is fascinating as depending your area of expertise some model will be recognizable and others not。 Maybe someone can say that many other models are not included， or not clear the difference what a model is， is the Standard Atomic Interesting and revealing， a guide to help think on today's complex problemsWhat's a model？ How to use a model？ Why to use a model？ Any qualitative approach to apply a model？ Only use 1 model or various？ These are the questions that these book look to help to answer。 It is fascinating as depending your area of expertise some model will be recognizable and others not。 Maybe someone can say that many other models are not included， or not clear the difference what a model is， is the Standard Atomic model a "model"？ Not described here however it is。 Same with many others。 In any case it is very informative and overeating to understand the why: the more diverse approach to think about a problem， the better answers or guidelines you can get。 Some are most difficult to apply unless further study or practice like bandit models or NK models in my case。 What about machine learning？ Algorithms are also a model？ It is a super interesting book to start this thinking approach， and I recommend to do also his course in Coursera: Model Thinking。 Ideally to read paired with Algorithms to Live by: different style， different Algorithms （or model？） But completes this vision of learning to apply formal models to analyze a problem and think on different potential solutions。 。。。more
Rory Fox
，
An Interesting and thought provoking read， but certainly not an easy book。 Whilst this is not an academic text book， it does assume a familiarity with mathematics and statistics。 Readers who prefer not to engage with formulas and notation will find the book challenging。The earlier chapters explain the concepts of models and modelling， then the book works through 29 chapters which apply the models， chapter by chapter。 Around 25% of the book is footnotes。 Interspersed amongst the models are anecdo An Interesting and thought provoking read， but certainly not an easy book。 Whilst this is not an academic text book， it does assume a familiarity with mathematics and statistics。 Readers who prefer not to engage with formulas and notation will find the book challenging。The earlier chapters explain the concepts of models and modelling， then the book works through 29 chapters which apply the models， chapter by chapter。 Around 25% of the book is footnotes。 Interspersed amongst the models are anecdotes and examples which show the application and significance of the models。 This makes the book enjoyable with the authors wide knowledge leading to frequent surprises。 We see， for example， that a better understanding of Standard Deviation would have avoided the mistaken policy decisions which saw many countries chasing ‘small schools，’ as an educational panacea in the 1990s （Kindle location 15%）。 One of the author’s insistent points is that people need to be prepared to apply multiple models， and to apply models from different areas。 He illustrates this creatively。 For example， he uses a model of radioactive decay to illustrate how memories also ‘decay’ （Kindle 22%）。Whilst the models are thoughtfully presented， there were occasional questions of over-simplification。 For example， in Chapter 26 on Learning models， we hear about how rewards and incentivisation can improve learning （65%）。 This is certainly true， but there are also well-known sets of biases which can counteract some of the implications。 For example ‘incentivisation biases’ show how performance can actually get worse when it is rewarded， because it can replaces an inner， intrinsic incentive with an extrinisic motivation。 It would have been helpful if the book could have acknowledged some of these additional complexities。On a more fundamental level， I also wonder if the core insight is quite right。 The philosophy of the book is that multiple models is the answer to converging on truth。 Our thinking will be better proportionate to our ability to select and apply multiple models。But is this really true？ Our selection of models takes place inside an overarching single model of rationality。 Our ability to compare models is because we have a single， more fundamental model which can be used to judge all the other models。 For example， our over-arching model of rationality means that we do not accept contradictions， and that is a criterion which we apply to all models。 This suggests that although multiple models may indeed be helpful， they are helpful within a single model of rationality， not as a pluralistic alternative to rationality。 Ultimately， there are some important philosophical questions raised by this book， but they are left tantalizingly incomplete。 However， I enjoyed the opportunity to refresh my knowledge of statistics and I appreciated the creative way which the models are presented and related to each other。 。。。more
Jenny
，
Maybe one of my favorite books for this topic， though it doesn’t make for a particularly enjoyable linear read。 Bookmark and come back to it。 Re-read things multiple times。 Revisit it annually， maybe。
Bill Rand
，
Fantastic tour-de-force of a wide variety of models， with an emphasis on many model thinking。
Denis Vasilev
，