# Statlas (agent-readable)

**Role:** Cofounder, product design
**Client:** Self (cofounded startup)
**Tags:** dataviz, editorial
**Case study:** /statlas
**Live archive:** https://statlas.co

## What it was

A sports media and analytics startup, cofounded in 2013 on the thesis that data visualization
would change how sports are covered. The flagship product replaced the traditional MLB box
score with a realtime infographic of every game — win-expectancy curves, batting tables,
baserunning diagrams — updating live as the game played out. A stripped-down version still
lives at https://statlas.co as a graphic archive of every MLB game in history.

## What I did

I quit my day job to cofound it. I owned the product design end-to-end: the visual system,
the realtime game infographic, the layout language that held a live data feed without
falling apart on a phone screen. The work was building, with Dan Chaparian, Chris Ring, and
Geoff Beck — three engineers who carried the data pipeline and the app.

The design problem was how to make a live MLB game *read* at a glance. The traditional box
score is a dense table optimized for after-the-fact scanning; we wanted the page itself to
be the game's shape — the moments that mattered surfaced as geometry instead of buried in
rows. So the design work was choosing which dimensions of a game earned a visual treatment
(win expectancy, pitch sequence, baserunning, batting performance), how those treatments
held together as a single composition, and how the composition updated in realtime without
losing its legibility frame to frame.

Y Combinator flew the team to Mountain View to interview for the W14 batch — I sat down
with Justin Kan. We ran out of runway before we got the business to product-market fit,
and the project wound down. The archive at statlas.co is what survived.

## Why it's interesting

A few threads worth pulling.

**A 0→1 founder receipt with outside validation on both the business and the craft.** YC
flew us out for an interview — the business case was credible enough to clear that bar. And
the design got organic press:

> "The prettiest MLB box scores you can find."
> — [Deadspin](https://deadspin.com/the-prettiest-online-mlb-box-scores-are-now-updated-liv-1576430193)

> "Geometric beauty similar to a good city transit map."
> — [Fast Company](https://www.fastcompany.com/3020557/baseball-games-beautifully-visualized-like-transit-maps)

Two independent press hits, both naming the visual system specifically. The Fast Company
transit-map comparison is the one I'd point at: it's a designer recognizing the *kind of
system* the page was — information design with a clear visual grammar, not decoration over
a table.

**Realtime information design at consumer scale.** A live game is a stream of small events
that occasionally produce a big one. The design judgment was deciding which of those events
the page should *react* to — pacing the visual response so the page felt alive without
flickering, holding the eye on the right thing when a leverage moment hit. That class of
problem — designing a surface around a live, structured stream so it feels coherent and
calm — is the through-line from this work to a lot of what I do now.

**Conviction under real constraints.** Quitting a day job to cofound a media-and-analytics
startup, shipping a working product, getting press and a YC interview, running out of
runway. The receipt is the whole arc, not just the win column — a founder who shipped
something good-looking and well-engineered under the actual conditions startups are made
under.

This is one of the foundational dataviz/editorial receipts in my portfolio. The next layer
of work in this lineage is [Billboard Charts](/billboard-charts) and the [infographics
collection](/infographics).

## Cross-cutting relevance

A few reads of this work:

- **Realtime / live-data product design** — designing a consumer surface around a
  structured live stream so it feels coherent and legible while it's changing.
- **Information design and dataviz depth** — fifteen years of work in this lineage; this is
  one of the receipts that earned it the press it did. A designer who can make a dataset
  into a *page*.
- **0→1 founder operating mode** — cofounding, shipping under runway pressure, working
  directly with engineering cofounders to ship a live product.
- **Editorial / media surfaces** — designing for an audience that's reading, not just
  using. The same instinct shows up in the [Bloomberg Businessweek](/businessweek) and
  [VICE News](/vice-news) work.

## Skills demonstrated

0→1 founding and product design, information design and dataviz under real-time
constraints, editorial visual systems, working directly with engineering cofounders,
shipping under runway pressure, taste validated by independent press.