>And so he claims they are compatible, but to do so he redefines free will, then claims he hasn't and that was the definition of it we were working with all along. It just isn't convincing to me. I'd like to be convinced, for a long time I thought I was missing something, but I'm now begining to believe I'm not missing anything.

I feel it's you who have redefined what free will means, you are using a definition that doesn't exist, is incoherent and no one outside of amateur philosophers actually use.

Most professional philosophers most are outright compatibilists.

>https://survey2020.philpeople.org/survey/results/all

Lay people have incoherent ideas around free will, but when properly probed the majority have compatibilist intuitions.

>In the past decade, a number of empirical researchers have suggested that laypeople have compatibilist intuitions… In one of the first studies, Nahmias et al. (2006) asked participants to imagine that, in the next century, humans build a supercomputer able to accurately predict future human behavior on the basis of the current state of the world. Participants were then asked to imagine that, in this future, an agent has robbed a bank, as the supercomputer had predicted before he was even born. In this case, 76% of participants answered that this agent acted of his own free will, and 83% answered that he was morally blameworthy. These results suggest that most participants have compatibilist intuitions, since most answered that this agent could act freely and be morally responsible, despite living in a deterministic universe.
>
>https://philpapers.org/archive/ANDWCI-3.pdf

​

>Our results highlight some inconsistencies of lay beliefs in the general public, by showing explicit agreement with libertarian concepts of free will (especially in the US) and simultaneously showing behavior that is more consistent with compatibilist theories. If participants behaved in a way that was consistent with their libertarian beliefs, we would have expected a negative relation between free will and determinism, but instead we saw a positive relation that is hard to reconcile with libertarian views
>
>https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0221617

Then finally most/all court and justice systems use the compatibilist definition of free will. >It is a principle of fundamental justice that only voluntary conduct – behaviour that is the product of a free will and controlled body, unhindered by external constraints – should attract the penalty and stigma of criminal liability.
>
>https://scc-csc.lexum.com/scc-csc/scc-csc/en/item/1861/index.do

In the case of R. v. Ruzic

>The accused had been coerced by an individual in Colombia to smuggle cocaine into the United States. He was told that if he did not comply, his wife and child in Colombia would be harmed.

The Supreme Court found that he didn't smuggle the cocaine of his own free will. He didn't do it in line with his desires free from external coercion. Hence they were found innocent.

So no, the compatibilist definition is what we have been using all along, it's you who have redefined it to be incoherent.