Data Dash logo

Data Dash

Archives
Subscribe
August 12, 2026

Which details matter?

The answer can't be all of them

Welcome to Data Dash

Compressing an avalanche of thoughts about data + learning into byte-sized chunks. In your inbox every Wednesday and sometimes Fridays.

The alley you didn't see

There's a moment in The Last of Us video game most players don't experience. In the first playable section post-apocalypse, everything about the game design points the player to the right. But if you turn left you see a comical number of basketballs, soccer balls, and footballs littering an alley.

There's nothing quite wrong about these items, but they're not right either. There's no kids running around who just finished a pickup game. Other sections of the game don't display abandoned sports equipment to frame fun as a luxury of the past.

These details don't serve the game.

A coherent underground

This all might sound like a ridiculous nitpick. Except in the remake of The Last of Us the team redid these details. They kept pointing players to the right, so they knew most people would never see this update. And they still refined this alley from “every sport you can imagine” to “one soccer ball overtaken by the environment.”

I couldn't get the attention to this specific detail out of my head. I worry this next leap channels “and here's what this taught me about B2B sales.” But I wonder how many codebases stay littered with piles of random, littered sports equipment. Not because the code or project is bad — Last of Us and its sequel redefined what video games could be — but because we can't care about every detail.

We're proverbially pointing our users + potential contributors to the right, and asking them to ignore what's going on in the alley.

Our projects (probably) aren't The Last of Us

There's a version of this post where I lambast everyone to pore over every detail of their code. This story points in the opposite direction.

The Last of Us team crafted a genre-defining masterpiece without nailing every detail. When their project was already beloved, they leveraged that to go over every scene with a fine-toothed comb.

So should every codebase pay this much attention to every detail? No! Nail the important details people will see on their intended paths + common edge cases. If our project gets beloved enough we can redo every alley in the code, we should consider ourselves lucky.

A data thing I liked

Are open OCR models good enough to unlock historical knowledge?

A not-data thing I liked

The source for all this Last of Us remake lore - What Naughty Dog quietly fixed in the Last of Us remake

Don't miss what's next. Subscribe to Data Dash:
← Newer Find the boring wins Older → The unreasonable effectiveness of drop-in replacements
Share this email:
Share on LinkedIn Share on Bluesky
GitHub
Bluesky
LinkedIn
Powered by Buttondown, the easiest way to start and grow your newsletter.