Blood, Sweat, and Pixels — and why I make games

I started reading for pleasure again. For the past year it's been mostly product and engineering culture books. This time I picked up Blood, Sweat, and Pixels by Jason Schreier — ten chapters, ten games, each one a behind-the-scenes story of how the game got made. Indie to AAA. The common thread is simple: making games is brutally hard and everyone underestimates how hard.

The Witcher 3 chapter is the one I kept coming back to. CD Projekt Red started as a CD distribution company in Poland. They took Andrzej Sapkowski's novels — deeply Polish, Slavic folklore, morally grey, politically complex — and didn't water any of it down for a global audience. They just made it well. The Witcher 3 won over 800 awards and sold tens of millions of copies. That's the lesson: don't imitate what's already on the market. Tell the story only you can tell. If it's good, people will find it.

Stardew Valley hit differently. Eric Barone built the whole game alone — code, art, music, design — over five years. No salary, ten to fifteen hour days, girlfriend covering the bills. Nobody around him really understood what he was doing. Then it sold over a million copies in two weeks. One person, one game, five years of stubbornness. Not a realistic path for most people, but proof that it's possible.

Good timing to finish this book. I just wrapped up migrating my prototype to the new comic design system — all legacy components are gone, everything runs on shared primitives. Codebase is clean. Now I'm moving into playtesting. Larithea is a passion project and it's been through five different tech stacks at this point. AI coding tools have made me dramatically more productive — I can rebuild things in days that used to take weeks. But that's also the trap. It's never been easier to scratch the engineering itch, tear everything down, and start fresh with a shinier stack. The hard part isn't building it again. The hard part is stopping. Shipping is still a long way off, and I already know the temptation to rewrite will come back. The question is whether I can resist it long enough to actually finish.