Two years on Web3, and what I got wrong with 21BJ and Sp3llbound
Twenty people, hundreds of calls, events across half of Europe and a pre-seed that never came. What actually sank us, without the comfortable version.
In April 2023 I stopped studying physics to give startups a hundred per cent of my time, and started building a game with two childhood friends. In March 2025 we shut down 21BJ. Two months later we shut down Sp3llbound Studios, the studio we had built around it.
Twenty three months on the first, eleven on the second. No salaries, no round, and twenty people to tell that none of it was happening.
It was not a technical mistake. The game worked, the builds shipped, the beta was solid. It was accumulated inexperience, and the honest thing to do is list all of it instead of telling it as bad market luck.
What we had actually built
21BJ was a fantasy casino for mobile with no real betting: blackjack in a fantasy setting, in 3D, on Unreal Engine 5. The stated ambition was to become a sort of Zynga Poker with a more interesting model.
We got there from three directions that overlapped. All three of us had spent years studying Bitcoin and blockchain, as enthusiasts before anything else. All three of us were gaming nerds. And in 2023 play to earn looked like a real opportunity, with the sector being funded heavily.
What we put together in two years, without a euro of outside capital:
- A team of over twenty people
- Between three and four hundred calls to reach those twenty people
- A solid beta, with a public gameplay teaser
- Two years funded entirely out of our own pockets
And then the fundraising: London, Paris, events across half of Europe. We learned to build a pitch, to talk to VCs, to understand what to say and above all what not to say. It was the most stressful and the most exhilarating part at once, and I don't disown it.
Laid out in order, the run looks like this.
Sp3llbound was not a pivot
This is the thing people get wrong every time, so let me be clear: Sp3llbound Studios is not the project that came after 21BJ. It started in June 2024, while 21BJ was very much alive, for two very practical reasons.
The first was presenting better to investors: a development studio carries differently than a single game. The second was that the team had grown large, and turning it into a production house looked like the natural way to use it.
That is where the first real money of our whole run arrived:
- A Taiko grant
- A second grant from Immutable
- An acceleration with Helika, a gaming intelligence company based in Toronto
- Over 100 brand partnerships worldwide
That is not nothing, for three guys with no round. It's also why the story isn't "we failed to do anything": we did some of it, and it wasn't enough.
The four things that sank us
There was no single cause, and I distrust anyone who tells a failure with a single cause. There were four, and they held each other up.
- The product was too big for what we had.
- Investors raised the bar every round.
- The market stopped being funded.
- Our own money was running out.
The first two held each other up. A 3D mobile game on Unreal Engine 5 takes enormous resources, and in 2023 there was no AI to compress the cost: everything was slower and more expensive than planned. Meanwhile investors needed the beta first, then certain user numbers, then other targets. Always before the pre-seed, never after. At some point you notice you aren't getting closer to a finish line, you're chasing one that keeps moving.
Then the market went quiet. In the quarter we shut down Sp3llbound, the second of 2025, over 300 web3 games closed and funding to the sector was 73 million dollars, down 93% year on year. Studios far bigger than ours closed in the same window. Looking around and seeing we weren't alone was one of the few consolations.
The last one deserves its own section. But the practical point is this: with no round, the promise made to twenty people was not one we could keep. That is where the shutdown came from.
The part nobody writes about: personal money
A founder needs a financial floor. Not wealth: the certainty of being able to eat and live decently while building.
When that floor is missing it isn't just uncomfortable. You start making worse decisions. You take the call that makes no sense because you can't afford to say no. You chase the wrong grant because you need cash now. You pick the road that pays in two months over the one that makes sense in two years.
Running out of personal money is both a symptom and a cause of a project's decline. I watched it happen to me in the last months of 21BJ, and at the time I didn't notice.
It's also why today I recommend the opposite of what we did. We went all in, and it exposed us. With the tools available now you can launch a side project without jumping off a cliff, keeping a floor under your feet. It only applies to people with the itch to build something of their own, obviously: I have software engineer friends who earn well working little and are perfectly happy.
I would do it again a hundred thousand times
I could have spent those two years in big tech, on a high salary with my evenings free. I wouldn't trade.
That is where I learned what running a company actually means: not the theory, the part where you have to decide on incomplete information and then live with it. Everything I can do now, from SEO to the way I talk to people, I paid for that way.
What changed is the constraint. The constraint was never the code, it was the team: twenty people to coordinate, hundreds of calls, a technical cofounder we didn't have and kept looking for. And that is exactly the constraint that drops away today, which is why I started again alone instead of assembling another group.
A few months after shutting down, with the same head but without the same weight, we built and sold Bitzuma in five months. That is the proof the two years did return something.
Frequently asked questions
Why did 21BJ end?
From accumulated inexperience, not from a single cause. The product was too big for what we had, investors raised the bar every round, web3 gaming stopped being funded and our own money was running out. With no round we couldn't keep the promise made to over twenty people.
Did 21BJ and Sp3llbound close together?
No, and the confusion is fair because they died two months apart. 21BJ closed in March 2025, Sp3llbound in May. Sp3llbound wasn't even the follow up project: it was born inside 21BJ, while that was still running.
Do you regret working on Web3?
No. The sector was a reasonable bet with 2023 information, and the reasoning that took us there wasn't wild. What I'd do differently is the structure: start with someone who knows how to build, and pick a product proportionate to the resources I actually have.
What would you do differently today?
I wouldn't go all in. I'd keep a financial floor under my feet, because without it your decisions get worse, and I'd pick something I can put in front of a first user in weeks rather than months. I write about it in from co-founder to solo founder.
Guglielmo Vaccaro
Startup founder and solopreneur. I build products from zero, alone and with no investors. I share the process as it happens, mistakes included.
About me
