'RimWorld' hit Steam in summer 2016 with remarkable success, despite being developed by a tiny team and entering a genre littered with failures. But how? This talk looks at how, as a developer, to find unique value by doing things that are commonly assumed to be wrong, impossible, or ridiculous. Specific focuses include: How Ludeon Studios defined 'RimWorld' not as a game, but as a story generator, and how forcing themselves into this frame opened up entirely new mechanisms for creating compelling play. How, by strategically leaving out features, they made players engage with features and story elements that aren't actually there. How they made better decisions by not planning despite everyone wanting them to plan. And finally, how they shipped with many seemingly-critical features missing, but nobody cared, because they used a special methodology for selecting features that actually matter instead of the ones that commonly assumed to be necessary.