The Setup
Balatro was, by every honest measure, the indie phenomenon of the last eighteen months. The poker-themed roguelike from a single Canadian developer working under the pseudonym "LocalThunk" sold north of five million copies in its first year, won game-of-the-year nods from outlets that don't usually give game-of-the-year nods to indies, and quietly rewired conversations about deck-building games. We spoke over Zoom for ninety minutes earlier this month. The transcript below is condensed and edited for clarity.
On the Beginning
GAMEPULSE: When did you start working on Balatro?
LocalThunk: Late 2021. The first prototype was three weeks of work, written in Lua because that's what I knew. It was just poker hands plus a multiplier system. I was making it for myself. I didn't think about releasing it for the first eight months.
What was the prototype like to play?
Bad. [laughs] It was bad. The numbers didn't escalate fast enough. It was almost a math problem. I think a friend played it and was very polite about it.
When did it start to feel like the game it became?
There's one specific moment. I added the joker system in early 2022. The moment I had three jokers active at the same time and watched my multiplier explode in a way I hadn't planned, I realized: oh — this is the thing. This is what's actually fun. The poker is the disguise. The synergy is the game.
On the Slow Years
You mentioned earlier that 2022 was hard. What does that mean concretely?
I had a day job until the third trailer dropped. I'd come home, eat dinner, and work on Balatro from 9 PM to maybe 12:30. I did that for about thirty months. There were stretches where I'd post a screenshot on Twitter and get four likes. Friends would ask how the game was going and I'd say "it's going" and we'd both pretend that was a real answer.
Did you ever consider quitting?
Yeah. Not in a dramatic way. More like — I'd budget six more months, and at the six-month mark I'd reassess. Around month thirty I had a real conversation with myself about whether to ship a half-finished version or try to do another year. I chose to do another year.
On the Day the Demo Hit
When did you know it was working?
Steam Next Fest, October 2023. I'd had previous demos that did okay. This one was different from the first hour. The peak concurrent number was something I hadn't seen on any of my own games before. I refreshed the page maybe forty times.
And then.
And then I got an email from a publisher I won't name asking if I'd consider acquisition. I said no. Then another. I said no. Playstack reached out and was — I want to be careful here — they were not pushy. That was the moment I realized this was happening with or without me.
On Design Decisions
The "no internet, no daily challenge" choice. Walk us through that.
I made a list early on of features I would refuse to ship. Daily challenge was on the list. Account login was on the list. Always-online was on the list. I wanted Balatro to be a game you could play on a plane. I'd been so frustrated with games that demanded an account login to do anything. I didn't want to make one.
What about leaderboards?
Same calculus. Leaderboards turn the game into a comparison machine. Balatro is supposed to be the opposite of that. It's a game you sit alone with.
There's a Joker called "LocalThunk" — the meta-joker. Why?
Honestly? I added it as a joke after the demo blew up. I figured if anyone was going to be smug about my own game's success it should be me, not the audience. It's a little bit of an in-joke. It's also the most balanced joker in the game, which is a separate accomplishment.
On the Aftermath
What does five million copies feel like?
Strange. There's a lag. The numbers come in over weeks. You get used to seeing larger numbers than you've ever seen, and then they keep climbing. I've stopped looking at them. I check the bank account once a month and that's enough.
What did you buy first?
[Laughs] A coffee machine. Like — a really good one. Then I paid off my parents' mortgage. Then I went on vacation. In that order.
What are you working on next?
I shouldn't say, but it's not Balatro 2. I'm aware of how every interview ends with me being asked that. I don't have it in me to do another deck-builder right now. The next thing is something small, in a different genre. I think I owe myself the right to try something nobody is expecting.
On Advice
What would you say to a solo dev working on the version of this you were working on in 2022?
Two things. First: don't fall in love with the early prototype. It's almost certainly not the game. The game is in there somewhere, but the way to find it is to keep playing it and noticing what surprised you. Second: ship the demo earlier than you think you should. The feedback you get from a public demo will reshape the game in ways you can't anticipate from inside your own head.
Anything you wouldn't say?
I won't tell anyone "you can do this too." I think Balatro hitting was 60% the game, 30% timing, 10% luck, and the luck part isn't optional. People who tell you they "earned" their breakout success are usually lying. I'm just one of the lucky ones who happened to also have a game that worked.
The Closing Note
What struck me most about LocalThunk's account, in retrospect, is the quietness of it. He never sounds like a man who took the world by storm. He sounds like a person who spent thirty months working on something at the kitchen table, woke up one day to find the world had decided to pay attention, and is still trying to figure out what to do with that fact.
The next thing he ships might or might not work. He sounded clear-eyed about that. The remarkable part is that he gets to ship a next thing at all — that's the privilege Balatro bought him, and the only kind of privilege that should matter to anyone watching this from the outside. Buy the game. Tip the developer. Recommend a friend. Repeat.