However, if you stuff them into an int at the last minute, you can get that effect.

Under the hood, it's floats. On the output, it's ints.

It's obvious and silly with small examples. The problem can creep in when you are using larger libraries or frameworks.