Jonathan Blow's new game outsold The Witness by 45% in week one

TL;DR: Jonathan Blow says his new game, Order of the Sinking Star, sold 45% more Steam copies in its first week than The Witness did in 2016, at a 25% lower price. In a Steam that now gets dozens of new games every single day, a $30 puzzle game with no live service and no battle pass just had a better launch than one of the biggest indie hits of the last decade. Quality still cuts through. It just has to be absurdly good.

The number nobody expected in 2026

On Oct 15 at 4:20 PM ET, Blow posted that Order of the Sinking Star had been out a week, so he "decided to run some stats on Steam where it's easy."

The result: Order sold 45% more units on Steam in its first week than The Witness did. By about 7:40 PM ET the post had 39K views and 2.7K likes.

Quick context if you missed 2016: The Witness was Blow's open-world puzzle game, seven years in development. In its first week it made over $5 million across PC and PS4, about 100,000 sales, per Blow's own numbers reported by MCV at the time. That was a monster launch for an indie studio.

So "beat The Witness" is not a low bar. It's the bar.

Cheaper game, more copies, same money

Blow is upfront about the catch. Order is 25% cheaper: The Witness was $40, Order is $30 (Steam lists it at $29.99), with some bundle discounts, plus ten years of inflation.

His words: "you would expect more units at the lower price."

Napkin math (mine, not his): 1.45x the copies at 0.75x the price is roughly 1.09x the gross on Steam. Before the bundle discounts and before inflation. So in real dollars it's roughly a tie, which is exactly what he says next.

That's an honest post. No "record breaking launch!!" marketing voice. Just a dev reading his own dashboard out loud. Lowkey that's why people trust him.

Players are backing it up

The Steam page lists the release as Oct 8, 2026, developer Thekla, Inc., at $29.99. When we checked Steam's review API at about 7:40 PM ET Oct 15, the game sat at about 97% positive across all languages, rated Overwhelmingly Positive.

That matters because units in week one are mostly hype and name recognition. Reviews are the part a famous name can't buy. 97% means the people who showed up for "the Braid guy" stayed for the game.

The Steam flood it launched into

Here's the part that makes this interesting for every indie dev. Notebookcheck, citing SteamDB counts, reported that Steam is averaging almost 70 new games a day in 2026, up from roughly 58 a day in 2025.

Seventy a day. Your launch week is competing with around 490 other launches, plus every older game on sale.

The common take is that discovery is cooked and only giant marketing budgets break through. Order's week is a data point against that, but a complicated one.

Is this proof quality wins, or proof fame wins?

Honest answer: nobody can separate those from one game.

The case for "fame did it": Blow has two of the most famous indie games ever, a big following, and he's a loud, divisive voice online. Every argument he starts is free marketing. A no-name studio shipping the exact same game probably doesn't sell the same week one.

The case for "quality did it": fame gets you the first week of curiosity. It doesn't get you about 97% positive from thousands of reviewers. And reviews are what feed Steam's algorithm after the launch spike fades, which is where Blow's "depending on how the tail goes" comes in.

My bet: fame bought the week, quality buys the year. The tail is the real verdict, and we'll know a lot more by Christmas.

What this means if you ship games (or anything)

Three things I'd steal from this launch:

1. Premium still works. A $30 one-time purchase, no microtransactions, did better than a $40 one did a decade ago. The "everything must be free-to-play" crowd should look at this.

2. Post your real numbers. Blow's post did more for the game than a trailer would have. A real stat, a caveat, a thank-you. That's it.

3. The tail is where you win. Week one is the launch. Reviews, word of mouth and updates are the business.

Our own Steam launch is in 5 days

This one hits close to home. Our game, O'BLOCK, a sneaker-flipping sim built by one person with AI agents, launches free on Steam on Oct 20. 70-a-day crowd, no famous name, no budget. Basically the opposite experiment from Blow's.

The game is free. The only money it will ever make is a $4.99 Complete Supporter Pack, on sale from Oct 20. If you want to back a tiny dev in the same flood, wishlist it now and grab it on launch day: O'BLOCK Supporter Pack on Steam.

The numbers so far

Order vs The Witness, Steam units, week one: +45% (Jonathan Blow, Oct 15 2026)

Price: $29.99 vs $40 at launch, 25% cheaper

The Witness week one, 2016: over $5M, about 100,000 sales across PC and PS4 (MCV, Feb 2016)

Order Steam reviews: about 97% positive (Steam, ~7:40 PM ET Oct 15)

New Steam games per day in 2026: almost 70 (Notebookcheck citing SteamDB)

Next: we check back on Order's reviews and our own O'BLOCK numbers after Oct 20, every number real.

So, your call: if a no-name studio shipped Order of the Sinking Star under a different name, does it still beat The Witness's week one? Tell me in the replies.

Sources: Jonathan Blow on X (Oct 15 2026), Steam store page and review API for Order of the Sinking Star (app 499170), MCV "The Witness makes $5m in a week" (Feb 3 2016), Notebookcheck on 2026 Steam release counts.