Blog

Field notes — one a day since May 2023. Things we shipped, things we broke, and numbers we went and checked.

2026 2025 2024 2023

1,200 posts. , or browse by year and tag below.

Aug 8, 2026

The overall average hides 8 different numbers

Real group means for one column across 8 segments, and what the pooled average conceals.

statistics practice Analytics practice

Aug 7, 2026

A essentially none correlation, and what it is not

A measured correlation of -0.00 between two columns, what its square says, and the diagnostic that costs one line.

statistics Statistics

Aug 6, 2026

What 3,500 job postings say about city

One column of a postings dataset, and what the shape of it means for anyone reading listings.

career practice Analytics practice

Aug 5, 2026

Correction: As-of feature computation

The page claimed more than it should have. What it says now, and where the pattern stops applying.

quality practice Machine learning

Aug 4, 2026

How much statistics do I actually need?

Less theory than you fear, and more discipline than you would like.

statistics career Statistics

Aug 3, 2026

Why there is no analytics on this site

No analytics, no tag manager, no cookie banner, because there are no cookies to consent to.

practice Data platform

Aug 2, 2026

Charting revenue_usd by device

Two facts about the columns decide the chart. Cardinality and skew, in that order.

visualization Visualization

Aug 1, 2026

9.9%, plus or minus what?

A point estimate of 9.9%, the interval around it, and what a tighter one would cost.

statistics Statistics

Jul 31, 2026

A five-minute use for Experiment Calculator

Sample size before you launch, and a confidence interval after you stop. Both directions, with the assumptions written on the screen rather than buried in a footnote.

visualization Visualization

Jul 30, 2026

SaaS subscriptions, in one paragraph

Twenty-five hundred B2B subscription accounts with signup dates, plans, seat counts, MRR and churn dates.

quality practice Data quality

Jul 29, 2026

Cutting a lesson down — Designing an experiment that can succeed

Most experiments fail before launch, because they were never large enough to detect the effect anyone hoped for. The fix costs ten minutes and happens before you write code.

practice Experimentation

Jul 28, 2026

How the scheduled posts work

The build refuses to publish a page whose date has not arrived, and a daily job does the rest.

engineering Data engineering

Jul 27, 2026

Pattern: Slowly changing dimension, type 2

Keep every version of a row so "what was true in February" stays answerable after March changes it.

warehousing modelling Analytics engineering

Jul 26, 2026

The time coverage of grid-energy-load

365 populated days across a 365-day span, and what that allows you to compute.

engineering quality Data engineering

Jul 25, 2026

The category that eats the chart

4 values with 44.3% concentrated in one of them, and what that does to every chart downstream.

visualization practice Visualization

Jul 24, 2026

A running total that does not lie about ties

running total against a real schema, and the clause in it that is doing the work.

sql Analytics engineering

Jul 23, 2026

The total came out 8.2× too high

8.2 rows per key on one table, and the totals that quietly inherit the multiplier.

sql quality Analytics engineering

Jul 22, 2026

Counting rows is not testing grain

900 distinct combinations against 900 rows, and what that answers about the table's key.

sql quality warehousing Analytics engineering

Jul 21, 2026

57.8% of churned_at is missing

1,444 blanks in one column, and the one-line check that decides whether dropping them is free.

quality statistics Statistics

Jul 20, 2026

5,383 zeros in revenue_usd, and what they mean

89.7% of one column is exactly zero. Whether that is a measurement or an absence changes every aggregate built on it.

quality statistics Data quality

Jul 19, 2026

Before you clip the outliers

A max 1.4× the 99th percentile, and the three different things that could be.

statistics quality Statistics

Jul 18, 2026

The average distance km is not the typical one

A 51.0% gap between mean and median on one column, and which of the two answers the question you were asked.

statistics practice Statistics

Jul 17, 2026

Reading reopened before trusting it in support-tickets

4,000 rows, 2 distinct values, and the one fact about reopened that changes how you query it.

quality practice Data quality

Jul 16, 2026

Reading 3 columns instead of 12

Column pruning times partition pruning on a real schema, and the function call that quietly defeats both.

performance warehousing Data platform