Nothing is more boring than playing a computer game "perfectly," i.e. never failing.
There are big failures, that can only be recovered using a "start-over" or "load game" option, and little failure, that require small a series of small corrections.
Game designers need to spend a whole lot of time thinking about failure. What are different types of failures? What do they look like? How can one learn from them?
Most training people, meanwhile, focus on one brittle path to success. Focusing just on success is faster, after all. One can cover more ground. It is easier and more "positive" as well. Failures are harder to research as well, as most SME's from sales people to engineers, hate thinking about and dwelling on failure.
Ironically, of course, is that by sticking to the brittle path of success, so many training programs fail to achieve real behavioral transformation.