It’s Friday night and you’re making your character for a fun evening of gaming. You’ve rolled your abilities and even got two 18s. But what class should you pick given your rolls? If you pick that class, is your character playable? What kind of character have you created? Or maybe you’re a game master and you’ve been running a game for some time. Do you need to know how much treasure that dragon hoard should have? Or are you wanting to figure out how many encounters your players will make it through in an evening so you can prepare enough material? These important questions can all be answered using machine learning. Many developers want to make use of machine learning in their applications but aren’t sure what sorts of problems can be solved with it. This talk will explain the sorts of problems that can be solved, what data is required to solve them, and what the results look like. And, we’ll explore it using fun and geeky examples. We will cover five major types of problems that machine learning can solve: regression, classification, anomaly detection, forecasting, and impact analysis. When we’re done, you’ll have a basic understanding of what machine learning can do and what you might want to use it for. It might even be something other than role-playing!