ALT Video still, lecturer in front of blackboard holding chalk, caption: demonstrate to you today is that game theory is incredibly stupid

ALT Cover of 2nd edition of Statistical Rethinking

ALT Winners of the DeGroot Prize 2021 Nicolas Chopin and Omiros Papaspiliopoulos (2020). An Introduction to Sequential Monte Carlo. Springer. 2019 Subhashis Ghosal and Aad van der Vaart. Fundamentals of Nonparametric Bayesian Inference. Cambridge University Press. 2017 David Banks, Jesus Rios, and David Rios Insua. Adversarial Risk Analysis. CRC Press. 2015 Andrew Gelman, John B. Carlin, Hal S. Stern, David B. Dunson, Aki Vehtari, and Donald B. Rubin. Bayesian Data Analysis, Third Edition. CRC Press. 2013 Noel Cressie and Christopher Wikle (2011). Statistics for Spatio-Temporal Data. John Wiley and Sons. Kevin P. Murphy (2012). Machine Learning: A Probabilistic Perspective. MIT Press. 2011 Jay Kadane (2011). Principles of Uncertainty. CRC Press. 2009 Giovanni Parmigiani and Lurdes Inoue (2009). Decision Theory: Principles and Approaches. John Wiley and Sons. Carl Edward Rasmussen and Christopher K. I. Williams (2006). Gaussian Processes for Machine Learning (freely availa

ALT A Mathematical Model of Quitting Academia Academia has problems, you might have heard. Many normative practices in the sciences benefit careers at the price of cumulative scientific knowledge. These problems could in principle be addressed through institutional reform. However many academics resist reform, and those who participate in reform behavior may be disadvantaged in competition for jobs and resources. Simultaneously, academics may find non-academic outside options more appealing, especially if they believe in reform. This is especially true for the most talented and skilled individuals, who have better outside options. When can reform spread within such a system, even though those who believe in it are passed over or are incentivized to leave? I construct and solve a dynamic model of this depressing scenario.

ALT Table with 3 rows and 3 columns. Rows: (1) Algorithm purist (mimic human cognition), (2) Algorithm netural (learns & generalizes), (3) Algorithm rebel (method irrelevant) Cols: (1) Ability purist (exceeds human ability), (2) Ability neutral (makes task easier), (3) Ability rebel (usefullness questionable) Cells: [1,1] "Terminator is AI" [1,2] "C3PO is AI" [1,3] "WALL-E is AI" [2,1] "AlphaGo is AI" [2,2] "XGBOOST is AI" [2,3] "Tinder is AI" [3,1] "A metal detector is AI" [3,2] "Bubble sort is AI" [3,3] "Magic 8 Ball is AI"

ALT MIT leaders describe the experience of not renewing its largest journal contract as overwhelmingly positive. MIT has long tried to avoid vendor lock-in through big deal contracts and, in 2019, maintained individual title-by-title subscriptions to approximately 675 Elsevier titles. In 2020, they took the significant step of canceling the full Elsevier journals contract – all 675 titles – leaving users with immediate access to only pre-2020 backfile content. Since the cancellation, MIT Libraries estimates annual savings at more than 80% of its original spend. This move saves MIT approximately $2 million each year, and the Libraries provide alternative means of access that fulfills most article requests in minutes.

ALT Science Before Statistics: Introduction to Bayesian Causal Inference September 25 2024 URL for details: https://statisticalhorizons.com/seminars/science-before-statistics-introduction-to-bayesian-causal-inference/

ALT Terminator 2 phone call meme. 1: What's your mom's job? Statistician. 2. Is it okay to replace missing values with zeroes? 3. Sure honey that sounds fine 4. Your foster parents are dead

ALT Left: Cat washing his nether regions, leg cocked behind his head, tongue out, looking at camera. Subtitle: Generalized Linear Model Right: Tiger doing same. Subtitle: Generalized Linear Mixed Model

ALT Screenshot from a youtube science video, showing caption that defines a p-value of "3-sigma": "something has a 1 in 400 chance of just being randomness in your data"

ALT Arrays Some languages start arrays at 0, which can be unintuitive for beginners. Some languages start arrays at 1, which isn't representative of how the code actually works. DreamBerd does the best of both worlds: Arrays start at -1. const const scores = [3, 2, 5]! print(scores[-1])! //3 print(scores[0])! //2 print(scores[1])! //5

ALT In the past, we have designed experiments and gathered data on the hope that the results would lead to insight and possible publication—but then the actual data would end up too noisy, and we would realize in retrospect that our study never really had a chance of answering the questions we wanted to ask. Such an experience is not a complete waste— we learn from our mistakes and can use them to design future studies—but we can often do better by preceding any data collection with a computer simulation. Simulating a study can be more challenging than conducting a traditional power analysis. The simulation does not require any mathematical calculations; the challenge is the need to specify all aspects of the new study. For example, if the analysis will use regression on pre-treatment predictors, these must be simulated too, and the simulated model for the outcome should include the possibility of interactions. Beyond the obvious benefit of revealing designs that look to be too noisy to de

ALT Sometimes contemporary evolutionary theory is called “Neo-Darwinian.” But it would be better to call it “Post-Darwinian.” Darwin employed no formal mathe- matics, got some big things wrong, and most of contemporary evolutionary theory was created from mathematical scratch afterwards, in the 20th century. 5 The cult of prestige surrounding Darwin is misleading. For example, Darwin argued that four postulates make evolution by natural selection adaptive. These postulates are: (1) Individuals vary in their characteristics (2) There is competition to survive and reproduce 10 (3) Variation has consequences for this competition (4) Offspring resemble parents If these four things are true, the argument goes, then natural selection occurs and produces adaptations. There are several problems here, as any contemporary evo- lutionary biologist knows. The postulates are still taught in introductory courses, 15 but we all know they are a kind of white lie. Post-Darwinian evolutionary theory is muc

ALT The Portuguese man o’ war (Physalia physalis) is neither Portuguese nor a war- ship. It looks like a jellyfish, but it’s not. It’s not even an animal. Instead it’s a colony of individual animals. Each individual has a specialized job, fusing together in cooperation with the others to sail and sting their way through the ocean. Like 5 a pirate ship made of jelly. Each man o’ war colony can last for a year or more. So the evolutionary biolo- gist must ask, how did this cooperation evolve and how is it stable against cheaters? It is the same kind of question we have to ask about more familiar relationships among animals, like cooperating lions or breeding penguins. Or actual pirates. 10 In all of these examples, the evolutionary game to analyze is unlike those in the previous chapters. In the previous chapters, every example was a group of individuals who met, interacted once, and then never met again. There are animal societies like that. But primate groups and man o’ war colonies are no