// StatPlay
An interactive tool for working through statistical ideas by moving them around with your own hands. It isn't meant to replace a textbook — it's there for the moments when a definition makes sense on the page but no picture comes with it.
Drilling calculations and studying a subject systematically still belong to textbooks and problem sets. What this tool can do is build the picture: drag a slider, watch the curve respond, and see what the formula was describing. If it leaves you thinking statistics is more interesting than you expected, that's plenty.
The content draws on the following:
- Introduction to Statistics, Yasuhiko Torii, Nikkei Publishing
- A Completely Self-Taught Introduction to Statistics, Hiroyuki Kojima, Diamond, Inc.
- Statistics Is the Most Powerful Discipline, Hiromu Nishiuchi, Diamond, Inc.
- Japan Statistical Society Certified: Tokei Kentei Grade 2 Official Problem Collection [CBT Edition], edited by the Japan Statistical Society, Jitsumu-Kyoiku-Shuppan
- Japan Statistical Society Certified: Tokei Kentei Grade 2 Official Problem Collection 2018–2021, edited by the Japan Statistical Society, Jitsumu-Kyoiku-Shuppan
- Japan Statistical Society Certified: Statistical Practice Workbook for Tokei Kentei Pre-Grade 1, edited by the Japan Statistical Society, Jitsumu-Kyoiku-Shuppan
Feel free to quote the content and use it for teaching.
The code is open source, so adapting and reusing it is welcome too.
Commercial use falls under the CC BY-NC 4.0 licence — get in touch first if that's what you have in mind.
© 2026 Sasai Lab
// How This Tool Came to Be
StatPlay started at a sticking point during self-study: the page on standardizing to the normal distribution. The explanation parsed fine — but no picture came with it.
The hunch behind the project is simple: an interactive version of that page, something you can drag and watch respond, might be the thing that makes the idea visible. So that's what got built.
Building interactive visualizations on the frontend, though, isn't a skill that comes with a background in directing and managing projects. Not long ago, that would have been the end of it — a nice idea, no way to make it. AI changed that: with it as an option, the project became something that could actually be shipped.
The AI writes the code. What gets built stays a human decision. A directing background helps here in a specific way — shaky implementations tend to show themselves just from watching the terminal: is the structure sound, will it hold up when more topics get added? And there's a harder constraint underneath that one. Without actually understanding the statistics, there's no way to answer the real design question — which variable should the slider move so the picture does its job? You can't build an explanation of something you don't understand yourself. StatPlay is a fairly direct record of working that out in public.
Building it turned out to be the deepest part of the learning.
And the further the learning goes, the longer the list of things worth building gets — and the trickier it gets to decide what should move and how.
Statistics, it turns out, is genuinely interesting.
// Dev notes
Notes from building StatPlay are also published on Qiita (Japanese only): Qiita @sasailab ↗
// Creator
Jumpei Sasai / 笹井淳平
Started out at NTT as a network systems engineer, then moved into directing the development of games, web services, and apps at Square Enix and Drecom. Currently a non-degree (credited auditor) student at the University of Tsukuba Graduate School of Business Sciences, alongside the day job. Statistics turned out to be more interesting than expected while studying it independently, and StatPlay grew out of that. There's still plenty left to learn — learning by building seems to be the way it sticks.
This site is operated as a personal project and does not represent the views of the author's employer.