Game Developers Say It Helps When Fans Realize They’re Human

When Cyberpunk 2077 was delayed in October 2020, no one was more disappointed than the developers. Already working 100-hour weeks and crunching hard, the developers at CD Projekt Red were suddenly flooded with an internet-wide freakout topped off with death threats.

This is a companion discussion topic for the original entry at

Any programmer will tell you that working longer than like, 50 hours a week is asking for punishment. You get tired. You get sloppy. You make very dumb mistakes that are that much harder to spot because you’re too tired to see them. You can fight some of this with caffeine, but that’s the physical equivalent of taking out a payday loan (to say nothing of how this pairs with the stress). Your managers are dragging you off to daily (or sometimes multiple times a day) status meetings so you can context switch just long enough to tell your boss’s boss’s boss that “no, we don’t have a fix yet I was just working on it until you interrupted me for your stupid status meeting” and then have to context switch back to work again.

Start stacking these and best case scenario you’ll need a substantial vacation (after three months of 70 hour weeks I needed a solid three weeks vacation and I’m still not sure that was enough). Anybody who piles on in this scenario doesn’t understand. And I hope for their sake that they develop enough empathy independently that they never will.

  1. The week Cyberpunk came out, I was reading articles about it and the development process and it seemed very clear that they had a lot of talented people on the team and almost all of the problems were leadership’s fault. I think that’s also evident in the game itself, individual pieces are well done (e.g. the writing, the adaptation of the TTRPG system into the skills), but there are also a lot of things that seem rushed or unfinished. I can’t imagine threatening the developers over that game when you know they have been overworked, screwed over, and the major problems aren’t their fault.

  2. As a person who works for a software company, it’s always frustrating to me that gamers think they know a lot more software development than they actually do. My main pet peeve is when a game doesn’t run well on one person’s computer so they say “Ugh, lazy devs didn’t even bother to optimize the game.” It seems like gamers generally think that optimizing performance just means a developer clicking a “go fast” button after they finish writing their code, which is not at all accurate. Optimizing performance means rewriting your code to give the exact same output without doing as much work. Sometimes, there are some relatively simple fixes you can make when you first finish a thing, but in general improving performance is complex and risky and there are limits to what you can do without removing functionality. When developers put out an update that improves performance after a game comes out, it’s not “Oops, forgot to add this before I hit submit”, it’s the result of lots of hard work after reviewing most if not all of the code for the game.


And good optimization is hard. Can you afford to shortcut your garbage collection routine or are you too hard-up for memory? Can we get rid of those desks so the water-filling algorithm the AI is using to navigate through this tiny room doesn’t melt the CPU? How many of those fight animations can you keep in-memory when you’re cruising in the open world? Etc.


As I understand it, only some optimization is rote. Even if it’s rote, it’s still often pretty detailed work. You can generate LoD’s automatically, but you an often do better by having a 3D artist hand tailor the low poly versions for a particular usage.

The bulk of optimization is often coming up with extremely creative ways to not do things the “real” or hard way, essentially fooling the user into thinking something complex is going on when it’s actually quite simple.

This means that optimization is often extremely tied in with the game design and game systems. Once “optimized”, the game gets harder to change. So what can seem like a simple parameter change to a player, might actually involve re working tons of assets because the performance strategy has to change.

The lack of empathy/gratitude toward dev’s is a hellish part of this hobby/medium. It’s nice to hear about some positive change there.