Tempering expectations to avoid hype and eventual disappointment. Important due to sensitivity to loss.
One of the most visible types of failure in game development is expectation management. Game devs have a persistent problem with promising far more than they can deliver:
No Man’s Sky released an early demo and described gameplay that was not consistent with the release. Fallout 76 promised to give players a multi-player experience where they become the NPCs, but didn’t develop any of the mechanics to make that feasible, and packaged it in a buggy mess of a game. Atlas promised intense pirate-themed combat in huge multiplayer worlds. Instead they delivered a grind-filled, unpolished mess that was nothing but a reskin of their previous title, Ark: Survival Evolved. And Star Citizen, in my personal opinion, may prove to be the biggest example of failed expectation management and Dunning-Kruger to date: For the sake of backers, I hope I am wrong.
The moral of the story is don’t promise more than you can deliver. Better yet, play it safe: Under promise and shoot to over deliver. That way if you miss, you still have a good chance of meeting your original expectations, at least. Know that humans are loss averse, so losing out on promised features is going to be treated much more severely than proportionate gains.
As some final food for thought, my kickstarter research revealed, out of successful kickstarter campaigns between 2009 and 2012, 70% of developers underestimated the amount of time it would take to develop their titles. On average it took 3 times longer than expected! Furthermore, a full one-third completely failed to deliver the game they promised backers.
No Man’s Sky