Poker players have a better vocabulary for losing control than traders do, and it has always bothered me a little that we never borrowed it. A poker player who spews off three buy-ins after a bad beat does not say he got emotional. He says he tilted, and if he is any good he can tell you which kind of tilt it was, what triggered it, and what his early symptoms look like. Most traders I know can quote their win rate to a decimal place and cannot describe the first two minutes of their own meltdown. That gap is fixable, and the nice part is poker already did most of the work.
Why poker got there first
Reps, mostly. A live poker player sees maybe thirty hands an hour. An online player running six tables sees hundreds. Every one of those hands can deliver a brutally unfair outcome in seconds, with the money visibly leaving the stack. Poker compresses more emotional variance into a single session than many traders see in a quarter, so that community was forced to study the problem properly while trading psychology was still at the level of keep a journal and breathe.
The result is an actual taxonomy. Jared Tendler's work on the mental game of poker is the best known version, and it splits tilt into distinct types with distinct triggers, which matters because the fix depends on the type. Reading it as a trader is a strange experience, because nearly every category maps straight across. If anything trading needs the framework more, since we have more places to hide. You can blame the news, the fill, the venue, the market maker. A poker player who punts a stack has only the hand history staring back, which is a big part of why their thinking on this is so clean.
The taxonomy, translated
These are the types I see most often in trading, converted from their poker originals.
- Entitlement tilt. You did the work, so you deserve the win. You spent the whole weekend building the thesis, the trade goes against you on Monday, and you add size, because effort feels like it should be rewarded. Markets pay for being right rather than for hours logged, and adding to a loser because you worked hard on it is one of the most reliable ways good analysts lose money.
- Injustice tilt. The loss was unfair, so you are owed one back. You got stopped out by a wick that reversed instantly, or your order lagged at the worst possible moment. The unfairness feels like a debt the market now has to repay, so the next trade goes on bigger and faster than the plan allows, and the market, unhelpfully, keeps no ledger of grievances.
- Desperation tilt. You are down on the day or the month and getting back to even becomes the only trade that matters. This is the account killer, because the rational response to drawdown, which is getting smaller, is the exact opposite of what desperation demands. Nearly every blown-account story I have heard has this phase in it somewhere.
- Revenge tilt. Aimed at one specific instrument. The coin or the ticker that hurt you, and you keep coming back to it with size, outside your usual universe and off your usual setups, because beating that one specifically feels like it would prove something. Nothing about the next trade knows or cares about the last one.
- Mistake tilt. You broke your own rule or fat-fingered an order, and the anger at yourself quietly degrades the next ten decisions. This one is sneaky because the trigger was internal, so there is no external event to warn you that it fired.
The three-tier protocol
Naming the types is step zero. The working system has three tiers, and each one exists because the tier before it will sometimes fail.
Tier one: learn your tilt signature. Tilt shows up in behavior before it shows up in P&L, and the behaviors are personal and weirdly consistent. Mine is checking positions on shorter and shorter timeframes, plus a sudden interest in instruments I never normally touch. For other people it is cancelling and re-placing the same order, picking fights about markets online, or a feeling that a trade is urgent when nothing about it actually is. The workflow is simple. After any bad session or broken rule, write down what you were doing in the thirty minutes before things went wrong. Do that for a month or two and the same two or three behaviors keep appearing. That cluster is your signature, and it typically fires well before the expensive mistake does.
Tier two: cut size at the first symptom. The moment you catch a signature behavior, halve your size for the rest of the session. You do not need to be certain you are tilting, and that is the whole point, because the bet is asymmetric. If you were actually fine, trading half size for a few hours costs you a small amount of expected value. If you were tilting, you just cut the damage roughly in half at the exact moment your judgment was most inclined to argue against it. Poker players run the same play by dropping down in stakes when they are unsure of their own state, and I have never heard of anyone going broke from playing slightly too small.
Tier three: remove yourself at pre-committed thresholds. Write down, while calm, the conditions that end your session with no debate. Down a defined percentage on the day. A third rule violation. A second unplanned entry into the instrument that just hurt you. The exact numbers matter less than the fact that you chose them in advance, because the person hitting the threshold is the least qualified person on earth to decide whether the rule should apply today. When a condition trips, you flatten and close the platform. Sometimes the market then rips in your direction, and the rule still worked, because you are grading the process over hundreds of sessions rather than one exit.
What makes it stick
The honest weakness of all this is that it relies on self-observation from a person whose self-observation is the thing degrading. Tilt lies to you in real time, so I have come around to anything that shows you your own behavior from the outside. A journal you fill in yourself can be gamed by the person doing the tilting, and an execution log cannot. Pull your last fifty trades and look at the timestamps around your worst days. Entries clustered minutes apart, sizes creeping up, the same symbol over and over. It reads like someone else's mistakes, which is exactly the distance you want. A fair amount of how we built trade history and scorecards at Blockcircle comes from this, because the raw record is the one witness that was never tilted.
If you only take one thing, take the signature exercise. Skip the poker books if you want, but the next time a session goes wrong, write down the thirty minutes that came before it. The whole protocol rests on knowing what your own first symptom looks like, and in my experience almost nobody knows theirs until they go looking.