Yeah, I do apologize - I’m somewhat simplifying my explanation because when you start going into the full detail, it just brings up more questions.

So yes, like the other comment says, the particles are constantly bouncing into other things.

• If they’re bounded in by something - walls of a container, or even just more gas surrounding the specific sample you’re looking at - they’ll bump into that, and transfer some of their energy to that.
• If they don’t have something to bump off of and the particles are free-floating, they’ll take off in any given direction. If they only have something to bump off of in a limited number of directions, they’ll take off in the other direction. (For instance, in a rocket engine, we make a lot of molecules really, really hot and then surround them with barriers in every direction except the one we want them to zoom out in.)
• In some cases, the molecules have electromagnetic bonds with each other, which take more energy to break than the energy contained in their “bouncing around”. So they’ll stay stuck, just bouncing off each other, even in a vacuum, (Or at least, until they radiate away their heat via electromagnetic energy… another whole story.)