Tough decision. I think I would call here most of the time. I don't like folding aces and there is really a slim chance he has the queen.
There is some amount of variation based on how often he calls vs 3bets, but once he calls on the flop, I think there are more queens in his range than you might assume.
If he has a relatively standard preflop calling range, say something like:
[88-22
A9s-A2s, K2s+, Q8s+, J8s+, T7s+, 97s+, 86s+, 75s+, 65s, 54s
ATo-A2o, KTo+, QTo+, JTo]
We can assume that he calls a flop bet with all his Kx and Qx hands. There are 13 combinations of Qx to 48 combos of Kx on this turn. That's more than 20% of the time that he shows up with a Queen. And we have to assume he is betting all his Kx hands (which is terrible). Note that we don't even get 21%
equity here, because the Kx hands could hit a King on the river and suck out. So we lose 25% of the time. If we have him calling wider, we add more Qx hands faster than Kx ones. If he never 3bets, he does have AK, but he also has AQ and KK, which means we don't really do much better. In that case, they still have a Q more than 20% of the time, but our equity is lower.
I just mean there are more Qx combos (13-15) than you might expect even in a fairly tight calling range. If this person plays like 40% of their hands they can have like 25 combos or more (still is about a 4:1 ratio of Kx:Qx).
Now, we just need to know more about stakes, table dynamics, and the player. Against the vast majority of players that I am around, they are going to show me a queen the vast majority of the time I call here. If I have seen that they don't understand relative hand value (maybe they reraise all-in with the bottom end of a straight), I will just have to call. But, without that knowledge, I strongly lean towards folding.
Yeah, it sucks that we have AA. From a GTO perspective, we probably call here with AA. We need to be calling with some full houses, to not be overfolding against good players, and this would be the top of the full houses. But, that assumes our opponents are
bluffing this turn correctly, and I just don't think that's the case for most lower limits. The only way we are good here is when the person overvalues Kx.