That's kind-of where compatibalism comes in. For most definitions of free will, it is fighting the feeling that we would like to have it because it feels like we have it, and the evidence that says that the feeling that we have it does not provide any proof.

Sapolsky takes the approach to come up with a definition that can't be met and declare that it can't be met. Compatatablisim is more about finding a definition that is consistent with the feeling because without the feeling there isn't really anything to anchor the idea to anything meaningful. It doesn't use the feeling of making decisions as proof of any power as such, it is treating the feeling more like a measurement of the concept. Doing what we feel like doing is free will. What we feel like doing can be caused by anything and it still wouldn't matter.

Considering the inverse situation makes it seem like any other definition of freedom would be intolerable. If you 'freely' chose to do X but consistently had the perception of wanting to do not-X, it seems like you would not have a happy life. Similarly considering the alternative to determanism for your decisions seems like literal chaos. If something has no cause, it is literally random. Any pattern you can discern would indicate that it has a cause.

Then of-course you get into the nature of what is a cause and consequently considering the nature of time itself. Something travelling backwards in time interacting with something that you see appears uncaused because there was literally no preceding event that indicated it was going to happen (because it wasn't preceding, it came from the other direction)

That's one of those weird things I wonder about when people ask why is there more matter than antimatter, or why does times arrow mainly point in one direction. It feels like riding the crest of a wave wondering why all the water is going the same direction.