Node locking
A solver technique where you fix (lock) one player's strategy at a decision point to match how a real opponent plays — for example, 'this player never bluff-raises here' — and let the solver find the best exploit against it.
Frequently asked questions
What is node locking?
A solver technique where you fix the opponent's strategy at a particular decision — making them fold too often to a river bet, say — and let the solver re-solve everything else to find the maximum exploit against it.
Why is it useful?
Because real opponents are not balanced, and a standard solve answers the wrong question. Node locking is how you turn a read into a strategy: it tells you not just that over-folding is a mistake, but exactly how much more you should bluff and with which hands.
What is the risk?
Garbage in, garbage out. The exploit is only as good as the assumption you locked in, and the answer is often radically unbalanced — fine against the player you modelled, badly exploitable against anyone else.