I noticed something about myself at the table. The cards hit, I look, and a decision locks in: fold, or come in. Then the action goes - a raise, a call, a 3-bet from the tight player in the cutoff. By the time it's on me, the spot is nothing like the one I decided in. I play my decision anyway.

The real leak

The decision wasn't wrong. It was right for a table that no longer exists. My cards didn't change; everything around them did, and I didn't move with them. The read was fine. Holding it too long is the leak.

Why it's so hard to change a decision

Once you form a plan, your brain defends it against the exact information that should kill it. Four forces drive that, and all four are documented.

None of this is stupidity - it's efficiency. Deciding fresh is expensive; running an old decision is nearly free. Your job is to spot the pots worth paying full price for, and pay it.

The fix: don't look until it's your turn

The fix is mechanical: don't look at your cards until the action reaches you. No hand to decide about means no decision to defend. You watch the table build the picture first - who limped, who raised and how much, who's been folding for an hour and just woke up - then you look once, with everything already in front of you. No stale anchor, because you never set one.

One honest tradeoff: a late glance can leak tells, and the table is watching you peek. Some strong players look early for exactly that reason, then refuse to touch the decision until it's their turn. Either way the rule holds - decide when the action is on you, with everything you know by then. If you look early, treat that first glance as a guess you confirm or drop when it's your turn.

The other half of the leak: after you commit

The leak doesn't stop before the flop. Two versions show up once you've committed. You 3-bet a strong hand, you apply pressure, you win the pre-flop exchange - and now you keep firing on the flop even when the board turns against you, because you already acted like the hand was yours. Or you flop a draw, it bricks, and you can't let go, because it already feels like a pot you should have won.

This one isn't only in your head. The moment you commit, your chemistry shifts into a state built to keep pursuing - exactly the wrong state for folding.

Because the in-the-moment you is chemically loaded to pursue, the release has to be decided while you're still cold. Before you 3-bet, name the boards and actions that make you shut down. Before you take a draw, decide what you do when it misses. Then you're not fighting your own hormones live - you're following a call you already made.

Update fast - at the table and off it

The table is a fast version of what you do everywhere. You take a position in the first minute of a meeting and defend it for an hour. You decide you're right before you've heard the argument out. The skill is the same in both places: change your action the moment the picture changes. Train it like this.

1. Separate the read from the decision. A first impression is fine. Just don't let it become the decision. The read is a guess; the decision happens only when you must make it.
2. Ask "what's changed?" before you act. Name what's different now versus when you first formed the view. A lot changed? The decision changes with it.
3. Default to re-evaluate, not execute. Your turn is a decision point, not a delivery slot for an old plan.
4. Say the anchor out loud. "First instinct was to call, still right after the 3-bet?" Naming it drops it from the answer to one input.
5. Change your mind faster. Good players update the second the reason changes, and don't defend the old plan. Slow updating costs chips and time.
6. Practice off the table. Take low-stakes moments - a work call, a plan with a friend, a first read on someone - and hold your first take as provisional. New info lands? Re-decide out loud. Same muscle that saves your stack when the tight player wakes up.

The first impression: when it's hard to change your mind about someone

The same leak works on reading people, not just cards. At the table you slap a label on a player early - "maniac," "nit," "fish" - and the label sticks. Even when they play differently, you keep reading every move through that first impression. Anchoring sets the label (Tversky and Kahneman), and confirmation bias makes sure you notice what confirms it and ignore what contradicts it (Nickerson). Asch showed that the first information we take in about a person weighs the most, and colors everything that follows.

The cost at the table: you call because "he always bluffs," long after he switched gears, or you fold because "she never bluffs," right as she starts. In life it's the same - a first impression of a colleague, a date, someone who got off on the wrong foot, sticks, and you keep reading them through it, missing who they actually are now.

The fix is the same: treat the read as a hypothesis, not a truth. Instead of looking for what confirms the label, look for what contradicts it - "what have they done that doesn't fit what I thought?" - and let yourself read the player, and the person, fresh. At the table, update the read every orbit. In life, give people a second chance to surprise you.

Which way do you get stuck?

Knowing which way you lean is where the work starts.

Sources

• Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185(4157), 1124-1131.
• Edwards, W. (1968). Conservatism in human information processing. In B. Kleinmuntz (Ed.), Formal Representation of Human Judgment. New York: Wiley.
• Arkes, H. R., & Blumer, C. (1985). The psychology of sunk cost. Organizational Behavior and Human Decision Processes, 35(1), 124-140.
• Continuation bias ("get-there-itis"). SKYbrary Aviation Safety; U.S. Federal Aviation Administration.
• Coates, J. M., & Herbert, J. (2008). Endogenous steroids and financial risk taking on a London trading floor. Proceedings of the National Academy of Sciences, 105(16), 6167-6172.
• Clark, L., Lawrence, A. J., Astley-Jones, F., & Gray, N. (2009). Gambling near-misses enhance motivation to gamble and recruit win-related brain circuitry. Neuron, 61(3), 481-490.
• Schultz, W. (2016). Dopamine reward prediction error coding. Dialogues in Clinical Neuroscience, 18(1), 23-32.
• Nickerson, R. S. (1998). Confirmation bias: A ubiquitous phenomenon in many guises. Review of General Psychology, 2(2), 175-220.
• Asch, S. E. (1946). Forming impressions of personality. Journal of Abnormal and Social Psychology, 41(3), 258-290.