By now, you probably know why we need to get the right data and prepare it properly—a forecast or analysis is only as good as the data that went into making it. The garbage in/garbage out maxim really is true in the data science world, and the chance that data in its rawest form will be good in a machine learning model without processing is very slim. That doesn’t mean that the data doesn’t contain massively useful info, but it’s not inherently ready for prime time. Simone Biles didn’t turn into a world-renowned gymnast overnight—it took years of prep, finding and honing particular strengths, and filling in gaps meaningfully.