They did what most "lean startups" did. They made an MVP, and iterated from there. The only difference is that their target audience were gamers who hivemind way more than most other markets.
A lean startup, done right, plans the MVP and promises only that, as a first iteration. Then, if necessary, they can pivot in a new direction with another MVP v2, or use the credibility from the execution of their first, viable, release, to raise support for a major expansion or three.
Clearly they did not create a “minimum viable” product if their customers were so unhappy with the end result. As a startup, they failed completely at determining what that “minimum viable” was.