It's been a long time since I've seen any specific stats on this, but I think there was something like over half of game developers leave the industry within their first five years, and some insane number (like 80-90%+) were out by their early 30s.
That's anecdotally accurate to my circle of AAA devs as well. There's a handful that stuck around and they are nearly all now lead/senior managers at well known AAA studios. But the majority have either left the industry entirely or moved to mobile development where there is significantly less (and less harsh) issues that come with AAA development.
And the drought of senior quality developers sounds about right as well, seeing as a year removed the industry I still get a handful of recruiter pings/job offerings each month from various sources.
I also feel it's worth stating for all the people that have said they gave up on games:
1) I love my job. Making games is simultaneously invigorating and inspiring.
2) I think I get paid pretty well.
3) I have never worked a 100 hour week.
4) I have never done 7 day weeks for months on end.
5) The last extended crunch I went through was over 5 years ago, and we've been making real efforts to get better.
1) Yep.
2) Hell no. With cost of living, I was making about the same as someone on minimum wage where I live now. And I know that because after leaving the industry that's basically exactly what I did, yet maintained the same basic standard of living. And I was consistently paid industry average. Sure, the numbers tend to
sound good, but when you compare figures to non-industry relevant salaries
(I could have made an instant 25-30% more taking a lower title at business software developer) and consider COL expenses, it's hard to argue the industry pays well. You don't get rich being a game developer.
3) I once worked 100 hour weeks, back to back, for 4 months straight.
4) Those were also 7 day weeks. There were a few other times I worked 80+ hour weeks, 7 days a week, for several months just before release.
5) The last real project I worked on before outright quitting the industry starting crunching 18 months
before release and we never actually left the crunch cycle until more than 12 months
after launch because of post-release support. Yes, literally more than 2 1/2
years of crunch. On one project. That I didn't get a launch bonus for because we missed both our sales and Metacritic targets, if only barely.