June 24, 2026
SQL got stats and the crowd got loud
Statistics that live in your SQL
This SQL stats tool dropped a comeback update—and the comments instantly split
TLDR: the-stats-duck just shipped a new update that lets people do advanced analysis and charts right inside SQL, even in a web browser. Commenters were split between “finally, this is useful” and “nice demo, but not enough for serious work,” with bonus drama over security warnings and copycat vibes.
A niche data tool just staged a very online comeback. the-stats-duck, an open-source add-on for DuckDB, released version 0.6.0 with the gloriously needy subtitle “i-m-not-dead”—and honestly, the community took that as an invitation to talk about everything except the release notes. The actual news is pretty wild: it lets people do serious number-crunching, model fitting, resampling, and even charts without leaving SQL, the query language many people already use to poke at databases. It even runs in a browser, which is the sort of thing that makes data people say “oh no, I actually want this.”
But the comments? That’s where the fun started. One camp was instantly onboard, saying this is exactly the missing piece for people who want to stay inside DuckDB instead of juggling R, Python, or Stata. Another camp slammed the brakes, arguing this smells like a quick-analysis toy rather than something you’d trust for dashboards or deep research. Then came the classic internet energy: one commenter politely corrected a technical claim about DuckDB’s built-in summarize feature, while another raised a red flag after a browser extension screamed about a possible cross-site scripting issue on the demo site. Suddenly the vibe went from “cool penguin charts” to “is this a breakthrough or a browser jump scare?”
There was also a little side-eye over how similar the plotting system looks to Posit’s ggsql alpha, giving the whole thread a faint whiff of “haven’t we seen this somewhere before?” In other words: a classic tech launch—part applause, part nitpick Olympics, part security panic, and 100% comment-section theater.
Key Points
- •the-stats-duck released version 0.6.0, an MIT-licensed DuckDB extension for running statistical analysis and plotting directly in SQL.
- •The extension provides a `meta()` table function for column profiling and `lm()`/`lm_summary()` functions for ordinary least squares regression using formula syntax.
- •Its regression implementation uses a Cholesky solve of X'X, applies complete-case filtering, and was cross-verified against R datasets to four decimal places.
- •The extension includes a `bootstrap()` function for resampling with replacement and supports reproducible percentile intervals when paired with `list_aggregate()` and optional seeds.
- •Version 0.6.0 expands charting through ggsql with violin marks, 2-D facets, and per-layer `STAT` modifiers including `smooth`, `summary`, and `identity`, compiling to Vega-Lite.