For applied ML, stats comes more handy more frequently, and you can gloss over the calculus, but for theoretical ML or research in ML methods, or having a good grasp of how these algorithms work, you’ll need both. And advanced mathematical stats also depends on calc BTW.