Can You Ever Fold Pocket Aces?

No. Across 3053 solved tournament spots the solver throws aces away exactly zero times — not at a final table, not on the bubble. But “always shove” is just as wrong: in 1835 of those spots it does not move all in with them.

The number

We went through every spot in our library and asked what the solver does most often with aces. The answer is clear, and in one direction boring:

The highest fold share aces reach in any of the 3053 spots is 0.0 %. So there is not merely no spot where folding is right — there is none where it appears even as a small part of a mixed strategy.

Why people ask anyway

Because it is nearly right. Tournaments do produce spots where you fold very strong hands: jacks fold in 107 of our spots, queens in 14, AKo in 45. Anyone who has seen that is right to wonder where the line runs.

It runs at aces, and the reason is simple: against every other hand, aces are ahead. There is no opposing range they trail — and ICM can shrink an edge, but it cannot flip one.

Where the real question starts

The interesting half is the other one: in 60 % of the spots the solver does not shove with aces. It raises smaller, or just calls. Not out of fear, but because a smaller bet keeps more worse hands in the pot — and aces make their money when somebody comes along.

So if you want to take something away from this question, it is not “never fold aces”. You knew that already. It is this: how much you bet with aces is the decision — and across 3053 spots, 1835 times that decision is not all in.

Where this answer stops

Every number here comes from preflop tournament spots with a payout ladder. After the flop you certainly can fold aces — when four cards to a straight are out and your opponent jams, aces are often just one pair. That is a different question and a different data set.

Try it yourself

In the trainer you do not only see whether you were right — you see which action the solver picked and how often. That is where “aces are strong” turns into a decision.