Blog, page 2

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.

Jul 15, 2026

gdp_per_capita_usd by region: a 105% spread

Real group means for one column across 5 segments, and what the pooled average conceals.

statistics practice Analytics practice

Jul 14, 2026

r = 0.51 between distance_km and tip_usd

A measured correlation of 0.51 between two columns, what its square says, and the diagnostic that costs one line.

statistics Statistics

Jul 13, 2026

Correction: Fan-out join

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

quality practice Analytics engineering

Jul 12, 2026

My experiment came back flat. Now what?

First, work out whether it was ever capable of detecting the effect you cared about. Compute the minimum detectable effect for the sample size you act

experiments statistics Experimentation

Jul 11, 2026

How scheduled posts work here

Posts carry a date in their front matter. Anything dated after the build date is written but not published: no page, no sitemap entry, no RSS item, no

engineering production Data platform

Jul 10, 2026

Charting baseline_score by arm

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

visualization Visualization

Jul 9, 2026

The interval is the result

A point estimate of 7.1%, the interval around it, and what a tighter one would cost.

statistics Statistics

Jul 8, 2026

Grid energy load, in one paragraph

A year of hourly electricity demand, renewable generation, temperature and wholesale price for one balancing area — including the hours when price goes negative..

quality practice Data quality

Jul 7, 2026

What Data Explorer will not do

A fast table with per-column profiling: type inference, missing counts, distinct values, quantiles, and a sparkline distribution for every numeric column.

visualization Visualization

Jul 6, 2026

Cutting a lesson down — The first hour with an unfamiliar dataset

A repeatable routine for finding out what you are actually holding, before you compute anything anyone will act on.

practice Data quality

Jul 5, 2026

saas-subscriptions.plan: 4 values, 94.0% in the top three

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

visualization practice Visualization

Jul 4, 2026

Zero is not a small number

54.2% 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 3, 2026

Date spine, in short

Join to a generated calendar so days with no activity appear as zero instead of vanishing.

sql Analytics engineering

Jul 2, 2026

posted_on runs 2023-01-01 to 2025-09-27

966 populated days across a 1,001-day span, and what that allows you to compute.

engineering quality Data engineering

Jul 1, 2026

Comparing a period to the one before it

period comparison against a real schema, and the clause in it that is doing the work.

sql Analytics engineering

Jun 30, 2026

What wind_mw actually contains in grid-energy-load

8,760 rows, 6,314 distinct values, and the one fact about wind_mw that changes how you query it.

quality practice Data quality

Jun 29, 2026

The total came out 4.2× too high

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

sql quality Analytics engineering

Jun 28, 2026

Is ticket_id + channel the grain of support-tickets?

4,000 distinct combinations against 4,000 rows, and what that answers about the table's key.

sql quality warehousing Analytics engineering

Jun 27, 2026

21.8% of salary_max_usd is missing

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

quality statistics Statistics

Jun 26, 2026

The arithmetic behind a cheap query

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

performance warehousing Data platform

Jun 25, 2026

The average tip usd is not the typical one

A 59.6% gap between mean and median on one column, and which of the two answers the question you were asked.

statistics practice Statistics

Jun 24, 2026

The top of bmi is 1.1× the 99th percentile

A max 1.1× the 99th percentile, and the three different things that could be.

statistics quality Statistics

Jun 23, 2026

revenue_usd by variant: a 48% spread

Real group means for one column across 3 segments, and what the pooled average conceals.

statistics practice Analytics practice

Jun 22, 2026

r = 0.14 between load_mw and temp_c

A measured correlation of 0.14 between two columns, what its square says, and the diagnostic that costs one line.

statistics Statistics