The hottest Statistics Substack posts right now

And their main takeaways
Category
Top Technology Topics

GDP per capita or lying with statistics for advanced practitioners

Klement on Investing • 3 implied HN points • 20 Nov 25
💰 Finance Statistics
  1. GDP per capita is a poor proxy for living standards and doesn’t tell you how well people actually live, so blunt comparisons (like Europe vs a US state) are misleading.
  2. Which exchange rates or base years you use (current dollars, constant dollars) can swing GDP comparisons a lot, letting statistics be used to support very different narratives.
  3. GDP per capita adjusted for PPP better reflects what people can buy with their income and usually narrows gaps with the US, but it’s more complex and rarely used in media headlines.

Series B Activity: October 2023

Magid and Co • 19 implied HN points • 08 Nov 23
💼 Business Statistics
  1. The post provides data on Series B deals done in October 2023
  2. The data focuses on Series B deals worldwide (excluding China) that raised more than $5M and were not in the therapeutics sector
  3. The post offers a summary of statistics for Series B deals

Don't Run Coibion-Gorodnichenko Regressions with Micro Data

inexactscience • 39 implied HN points • 27 Mar 23
💰 Finance Statistics
  1. Running Coibion-Gorodnichenko regressions with individual data can lead to misleading results. It's important to use appropriate data types to avoid confusion in the findings.
  2. Individual forecasts tend to produce negative results compared to positive results in average forecasts. This means that the insights from these regressions can differ significantly based on the data used.
  3. The methodology is sensitive to noise and measurement errors, which can skew results. Researchers need to be cautious and robust in their approach to ensure accurate interpretations.

Housing Starts Decreased to 1.311 million Annual Rate in October

CalculatedRisk Newsletter • 33 implied HN points • 19 Nov 24
💰 Finance Statistics
  1. Housing starts in October dropped to 1.311 million, which is lower than both September and October of last year. This shows a continued decrease in new home construction.
  2. Single-family housing starts also fell by about 6.9% compared to the previous month, suggesting potential challenges in the market for individual homes.
  3. Multi-family housing starts saw a trend of weakness over the past year, indicating it may be harder to get those types of buildings underway, while single-family starts have been improving recently.
Get a weekly roundup of the best Substack posts, by hacker news affinity:

Zero is the Universal Bayesian Prior

Bram’s Thoughts • 19 implied HN points • 14 Jul 23
🔬 Science Statistics
  1. Estimating values in multiple dimensions can be more accurate by making them slightly smaller towards zero.
  2. Using a Bayesian prior of zero in analysis can be counterproductive and arbitrary.
  3. Consider using realistic Bayesian priors, like room temperature, for more reasonable estimates.

Gapminder

Barn Lab • 19 implied HN points • 16 Jun 23
🚌 Education Statistics
  1. Gapminder is an independent Swedish foundation focused on fact-based global statistics.
  2. Their bubble chart software, Trendalyzer, was acquired by Google in 2007.
  3. Gapminder provides educators with free resources and offers a 'Worldview Upgrader' quiz to challenge misconceptions.

How I once won the lottery

inexactscience • 19 implied HN points • 06 Sep 23
🚌 Education Statistics
  1. Sticking to one choice in a lottery doesn't change your odds, which stay at 1 in 24 no matter what. It seems like it should matter, but it really doesn't.
  2. If a lottery is unfair and avoids streaks, choosing the same number can actually be a better strategy because it decreases your risk of never winning.
  3. Many people fall for the gambler's fallacy, thinking just because a number hasn't won in a while, it should win soon. But in a fair lottery, each draw is independent and has the same odds.

There’s a 110 Percent Chance That Our Discomfort With Statistics Is Costing Us

I Might Be Wrong • 10 implied HN points • 08 Jul 25
🚌 Education Statistics
  1. Many people struggle to understand statistics and probability, which can lead to poor decision-making. For example, weather forecasts often misrepresent low probabilities to avoid backlash from confused audiences.
  2. Vague language used in warnings, like 'possible' or 'could', can create misunderstanding and result in people ignoring important alerts. It would be better to provide specific probabilities so that people can respond appropriately.
  3. Improving statistical education could help people make better choices and improve communication between authorities and the public. Clear numbers would be more useful than unclear descriptions.

Math for Computer Science [Math Mondays]

Technology Made Simple • 59 implied HN points • 26 Apr 22
🕹 Technology Statistics
  1. Focus on Calculus for software development: Understand precalc topics like functions, transformation, and algebra well.
  2. Importance of Probs and Stats: Learn to think in a Bayesian context, focus on probabilistic thinking.
  3. Value of Linear Algebra: Grasp foundational concepts, computational side less important for traditional software development.

OSD 261: Murder rates are determined by your level of zoom

Open Source Defense • 63 implied HN points • 19 Feb 24
🇺🇸 U.S. Politics Statistics
  1. Violence in society is like bubble tea - the bubbles, not the milk. It's not evenly distributed, but lumpy.
  2. Murder rates can vary widely within the same city, with certain neighborhoods being safer than others. High-level averages can be misleading.
  3. Your personal risk of violence is heavily influenced by factors like geography, behavior, and demographics, rather than just general statistics. Adjust your level of focus accordingly.

Welcome, and what this is

The Counterfactual • 59 implied HN points • 17 Jul 22
🚌 Education Statistics
  1. The newsletter will cover topics like language, statistics, and AI, mixing research with personal thoughts. Expect both solid research reviews and imaginative columns about the future.
  2. Posts will be written in a clear, clean format using Substack. This platform helps catch mistakes easily and connects with a larger community of writers and readers.
  3. The author aims to write about things that are interesting and useful, hoping to share knowledge and insights that spark curiosity in readers.

The most important assumption in statistics- IID [Math Mondays]

Technology Made Simple • 39 implied HN points • 01 Aug 22
🔬 Science Statistics
  1. The most important assumption in statistics is IID, which stands for Independently and Identically Distributed
  2. IID assumption is crucial for statistical analysis - it helps in making accurate deductions and avoiding mistakes, like the gambler's fallacy
  3. Understanding IID involves recognizing independent and identical distributions in data samples, which are essential for various statistical techniques

Case-Shiller: National House Price Index Up 2.7% year-over-year in April

CalculatedRisk Newsletter • 9 implied HN points • 24 Jun 25
💰 Finance Statistics
  1. The national house price index is up by 2.7% over the past year, showing a general increase in home prices.
  2. However, there was a month-to-month decrease of 0.4% in home prices in April, indicating some fluctuation in the market.
  3. Certain regions are seeing lower gains or even declines, suggesting a shift in real estate trends across the country.

Probability Distributions 101

Olshansky's Newsletter • 68 implied HN points • 20 Feb 23
🔬 Science Statistics
  1. Probability is about the likelihood of events happening.
  2. Probability distributions are functions describing random events over a sample space.
  3. Key concepts include random variables, probability mass function, probability density function, cumulative distribution function, Gaussian distribution, and Bernoulli distribution.

How Modern AI Arose Accidentally

Erik Explores • 61 implied HN points • 21 Apr 23
🕹 Technology Statistics
  1. Modern AI systems like ChatGPT and Stable Diffusion arose from solving small, narrowly defined problems.
  2. Today's AI systems are more of an emergent phenomenon rather than meticulously planned for general intelligence.
  3. The key to AI advancements lies in simple, iterative processes based on statistics and probabilities.

World Series!

Political Currents by Ross Barkan • 16 implied HN points • 22 Oct 24
🎾 Sports Statistics
  1. Playing softball can be just as fun and competitive as baseball, even if it's not at a professional level. It offers a chance to enjoy the game and connect with friends, despite any frustrations on the field.
  2. Baseball has become very strategic and data-driven, making it even harder for players to succeed. However, the unpredictability of the game keeps it exciting, as even the best plans can fail.
  3. The upcoming World Series between the Yankees and Dodgers is historic and highly anticipated. It brings together two legendary teams with rich histories, making it a big deal for baseball fans everywhere.

House COVID report claims the COVID vaccines saved millions of lives by citing statistical models as the proof

Steve Kirsch's newsletter • 13 implied HN points • 02 Dec 24
🏥 Health Politics Statistics
  1. The House report claims COVID vaccines saved millions of lives, relying on statistical models rather than actual data. This raises questions about the reliability of their conclusions.
  2. The vaccines were said to save lives by either reducing infection risk or lowering death rates from COVID. Evidence suggests these claims may not be accurate.
  3. There are reports showing vaccinated individuals might have been more likely to get COVID and experience worse outcomes, contrary to the vaccine's intended purpose.

Where inflation comes from

The Works in Progress Newsletter • 12 implied HN points • 14 Nov 24
💰 Finance Statistics
  1. Inflation measurement is complex and changes in how we calculate it can significantly impact our understanding of the economy. Small adjustments can make us feel either richer or poorer.
  2. The Consumer Price Index (CPI) is crucial for economic policies, affecting everything from Social Security benefits to tax brackets, which makes its measurement a point of political debate.
  3. Various methods and indexes, like the PCE and the chained CPI, aim to capture price changes accurately, but they each have strengths and weaknesses in addressing real-life consumer experiences.

The Science of Science

The Palindrome • 3 implied HN points • 30 Jul 25
🔬 Science Statistics
  1. Thinking in terms of probabilities helps us make better judgments when we are not certain. Unlike absolute truths, we can measure how likely something is to be true instead.
  2. Bayes' theorem allows us to update our beliefs based on new evidence. This means we can make smarter decisions by adjusting our understanding as we gather more information.
  3. To figure out causes from effects, we can use conditional probabilities. This helps us connect symptoms, like a headache and sore throat, to possible underlying issues, like the flu, in a more accurate way.

API Preview and Signal Stats

Dataplane.org Newsletter • 19 implied HN points • 07 Mar 22
🕹 Technology Statistics
  1. The API Preview allows users to search data with a subset of current data and provides documentation for various queries and returns.
  2. The Signal Statistics page offers graphical insight into daily event summaries, aiding both the organization and the community.
  3. The commentary on UA and RU signals highlights the organization's operational status in these regions amidst recent events, emphasizing the importance of monitoring and preparedness.

An analysis of Berkson's Paradox

Of Aurochs and Angels • 1 HN point • 14 May 24
🔬 Science Statistics
  1. In statistics, a collider can affect the relationship between variables, leading to false impressions. For example, controlling for a collider can change the true effect of one variable on another.
  2. Berkson's Paradox shows how relationships between variables can change when selecting samples based on certain traits, leading to counterintuitive results.
  3. The impact of weighting test scores in selection processes can alter the relationship between test scores and outcomes. Higher weighting can sometimes lead to negative correlations, despite the underlying positive relationship.

What Full Body Composition Scales Teach About Machine Learning

The Palindrome • 3 implied HN points • 16 Jun 25
🕹 Technology Statistics
  1. Not all body composition scales are accurate, but some of them are less wrong than others. It's important to understand how bias and variance affect their readings.
  2. Bias refers to a consistent error in measurements, while variance relates to the randomness of measurement errors. Both play a role in how reliable a scale's readings can be.
  3. When choosing a scale, it's better to prioritize low variance over low bias if you're only interested in tracking trends rather than precise values.

Data Science Weekly - Issue 451

Data Science Weekly Newsletter • 19 implied HN points • 14 Jul 22
🕹 Technology Statistics
  1. Many people believe that data scientists today often do tasks very similar to data analysts. They're not just creating charts; there's a concern that their work lacks deeper statistical analysis.
  2. There's a lively debate about what it means to be a data scientist. While some argue the role has become too diluted, others believe that practical application in companies differs from academic definitions.
  3. Data science is evolving, with new techniques and applications emerging, like the importance of understanding datasets and using principles from various fields to improve intelligence in AI.

The Candidates in 13 Graphs

Chess Engine Lab • 1 HN point • 25 Apr 24
🎾 Sports Statistics
  1. Pragganandhaa performed well with Black, while Firouzja struggled with Black.
  2. Gukesh displayed strong gameplay, being rarely in a worse position throughout the tournament.
  3. There are plans to explore more stats and graphs in future tournaments, including novelties in openings and move times.

The Advent of “Feather Alerts” is a Great Time to Reflect on How Racist Antiracism Has Gotten

I Might Be Wrong • 15 implied HN points • 20 Feb 24
🇺🇸 U.S. Politics Statistics
  1. California has implemented race-specific alert systems like Ebony Alerts and Feather Alerts, which raise questions about the approach to addressing racial disparities.
  2. The use of separate but equality-named systems like Ebony Alerts and Feather Alerts in California highlights flaws in current antiracist practices.
  3. Lawmakers embracing race-based thinking by passing the Ebony Alerts and Feather Alerts bills reflects a concerning trend of ineffective responses to racial disparities.

Data Science Weekly - Issue 434

Data Science Weekly Newsletter • 19 implied HN points • 17 Mar 22
🕹 Technology Statistics
  1. Understanding NLP is important. It involves tokenization and encoding, which helps to improve how machines understand language.
  2. Performance in deep learning can often feel random, but reasoning from first principles can help simplify the process. Focus on compute, memory, and overhead to improve performance.
  3. There is a growing need for data product managers as data teams modernize. These managers bridge the gap between data science insights and product development.

Data Science Weekly - Issue 431

Data Science Weekly Newsletter • 19 implied HN points • 24 Feb 22
🕹 Technology Statistics
  1. Vector databases are important for storing and searching data in various applications like image search and drug discovery.
  2. Statistics may not be the best path to becoming a data scientist; other fields could be more relevant and useful.
  3. Teaching and practicing reproducible workflows in data science helps ensure that research and findings can be verified and built upon.

Lost science

The Works in Progress Newsletter • 14 implied HN points • 19 Jan 24
🔬 Science Statistics
  1. Scientific papers can be forgotten but later become highly influential.
  2. Sleeping beauties in science are more common than expected.
  3. Technology, access to findings, and interdisciplinary collaborations play a role in awakening dormant scientific knowledge.

Data Science Weekly - Issue 427

Data Science Weekly Newsletter • 19 implied HN points • 27 Jan 22
🕹 Technology Statistics
  1. Using offline replay experimentation can help predict results faster, cutting down the time usually needed for online experiments.
  2. Bad data can seriously affect business operations, and understanding how it breaks is crucial for fixing dashboards and reports.
  3. Shapley values can explain machine learning models by distributing how each feature contributes to predictions, making the model's decisions clearer.

Data Science Weekly - Issue 412

Data Science Weekly Newsletter • 19 implied HN points • 14 Oct 21
🕹 Technology Statistics
  1. Machine learning is much more than just nonparametric statistics. It involves complex principles that go beyond what you learn in basic statistics.
  2. The State of AI Report 2021 highlights important areas like research, talent supply, industry applications, politics, and future predictions for AI. It's a comprehensive look at how AI is evolving.
  3. Self-supervised learning is becoming a major player in AI research. It allows models to learn from data without needing labeled examples, which can lead to significant advancements.

Data Science Weekly - Issue 410

Data Science Weekly Newsletter • 19 implied HN points • 30 Sep 21
🕹 Technology Statistics
  1. When looking for a job in data science, different companies suit different career stages, so it’s important to evaluate what works best for you.
  2. Advanced techniques in weather prediction are being developed to predict rain within the next couple of hours, showing a real-life application of data science.
  3. The effectiveness of deep learning is facing challenges as researchers approach the limits of what can be achieved, raising concerns about future improvements.

Data Science Weekly - Issue 401

Data Science Weekly Newsletter • 19 implied HN points • 29 Jul 21
🕹 Technology Statistics
  1. Open-ended play can help train AI agents to perform well on different tasks without needing direct human input. This means they can learn and adapt quickly to new challenges.
  2. Time-weighted averages are useful for getting accurate averages from data that isn't collected on a regular schedule. They help in making sense of messy time-series data.
  3. Triton is a new programming tool that makes it easier for researchers to write efficient GPU code, allowing even those without deep technical skills to optimize their computations effectively.

Considerable overlap

Klement on Investing • 3 implied HN points • 24 Jan 25
🔬 Science Statistics
  1. Baboons can learn to assess scatterplots and identify trends, similar to humans. This shows that different species can have overlapping skills in certain tasks.
  2. In general, smart baboons can outperform less intelligent humans in recognizing relationships in data. This highlights that intelligence can vary greatly, even within our own species.
  3. The study suggests that education plays a big role in developing skills, like understanding statistics. Humans often learn these concepts in school, while baboons do not.