Let's imagine one particle hitting another. At the moment of "collision", assuming no quantum randomness is afoot, how many paths can that particle end up taking? Exactly one.
Given some state and some "rules" (physics in this case), any system can only ever have a specific result from a specific set of inputs.
If I were to observe the neurons of a person choosing chocolate ice cream, I would see that due to their current brain state (the weights & connections of the neurons, produced by DNA and memory/past events), that they would have always chosen chocolate given the same inputs (emotional state, location, relative costs of the flavors, etc).
Suppose we have three neurons A, B, and C:
A
|
v
C -> "I'll have chocolate, please"
^
|
B
If A and B both fire, C will fire. If C fires, it triggers the process for the person's mind to decide "chocolate". If not, they do not decide on that flavor.
For the sake of this being a though expiriment, let's imagine that we have perfect knowledge of this system. I can see that A and B are the only things that affect if C fires.
Let's rewind that person's decision. If we get to the point where A and B are, again, firing, we will, again, see C fire as well, and the person will therefore pick chocolate again.
This is the crux of the issue. Each decision is a combination of some things:
(determined causes + randomness)
It seems most likely to me that randomness does not play into consciousness, but ultimately it's irrelevant. I only include it here because it's a philosophical possibility.
"Free" will would have to look something like this:
(determined causes + randomness + ?)
Because, the determined part of the will is not "free", but neither is a will influenced by randomness.
But no one can define this "?". It can't be the part of the decision that is based on any causal influences (I like chocolate more than vanilla, the chocolate is cheaper, etc) - those fall under "determined causes". But it also can't be any un-caused part of the decision, because that's literally the definition of random:
the definition of random
Not only can we not coherently define true "freedom", we don't even see anything that looks like it in nature. After all, even quantum randomness shows a probablistic pattern.
Lastly, I do not see any part of my own decisions that are free from cause. I never feel "forced" to make the decisions I make, but nevertheless, they always, always have causes.