Paul Giovannini shipped two games that sold under 700 copies between them, worked out exactly why, and then sold 200,000 copies of the third in a fortnight

Two games. Fewer than seven hundred copies between them. That is the part of Paul Giovannini's story that is worth your attention, because almost nobody makes a third one.

He made a third one. Gambonanza came out on the first of May and sold more than two hundred thousand copies in a fortnight.

He went and found out why

OVERWHELMED was a twin-stick shooter. Sunstone War was an action tower defence. Both shipped, both landed in silence, and the ordinary response to that is to decide the market is random and go again on feel.

He treated it as a question with an answer instead. The first two games reached their launch days with roughly 115 and 2,000 wishlists. Gambonanza had 3,500 within five days of the store page opening and cleared 10,000 in three weeks. The trailers for the first two were seen a few thousand times each; the Gambonanza reveal did around 580,000. None of that is mysterious once it is written down, and writing it down is the work. He then published the whole breakdown so other developers could have it, which is a strange thing to do with the only advantage you have.

The two earlier games are not the failures that set up the win. They are two finished games, shipped and launched by one person, and the studio that made Gambonanza could not have existed without them.

A chess game by someone who does not play chess

The premise sounds like a disqualification. He says it plainly: "I'm not a chess expert. In fact, while I deeply admire the game's richness and elegance, I don't naturally feel at home playing it."

That is exactly why the game is good. Chess punishes you for not having memorised things, and a chess game built by someone fluent in openings would have been a chess game for people fluent in openings. Not being that person, he built for improvisation instead: pieces that pick up their own personalities as you go, so a run rewards reading the board in front of you rather than recalling a line from a book. When playtesters kept reaching for Queens, he turned that pull into Gambits, the mechanic the whole game now hangs on. The name is Gambit and Bonanza stitched together, deliberately not a chess word, because he was not making a chess game for chess people.

The three years he spent on hyper-casual mobile games before any of this are all over it. That is a school of design where a player who does not understand the screen in four seconds is gone forever, and it turns out to be exactly the training a roguelike about a 1,500-year-old game needs.

Stray Fawn and Sidekick came in to publish something he had already built. It went to SXSW Sydney and picked up a Game of the Year shortlist nomination. His own summary of the two hundred thousand is "it still feels a bit random honestly", which, from the person who published the wishlist numbers proving it was not, is just modesty.

Go and play it before his fourth game, when nobody will remember there were two that sold seven hundred.

I'm not a chess expert. In fact, while I deeply admire the game's richness and elegance, I don't naturally feel at home playing it.

Paul Giovannini, on the qualification he did not have for the game he was making