Stop-Loss and Stop-Win Rules in Poker: What Actually Works

Stop-loss rules get sneered at by exactly two groups: players so rolled that a losing session is rounding error, and players who've never tracked what happens to their game after losing three buy-ins. For everyone in between — which is most of us — a stop-loss is one of the few rules that measurably protects both the roll and the decision quality. Here's the honest case for and against, and how to actually implement it when you can't just close a lobby and log off.

What a stop-loss does (and what it doesn't)

A stop-loss does not change the expected value of any hand. The cards don't know you're stuck. What it changes is *which hours you play*.

Your win rate is not constant across a session. It's highest in hours one to three, degrades with fatigue, and falls off a cliff when tilt enters the picture. If your fourth hour after losing three buy-ins is a -EV hour — and for most players, measured honestly, it is — then declining to play that hour is worth exactly as much as winning it. The stop-loss doesn't make you money; it stops you donating during the hours when you're not yourself.

The classic number is 2–3 buy-ins per session. Two for players who tilt fast or play deep-stacked games where single pots are huge; three for steady players in normal 100bb games. Pick yours once, while winning, and write it down with the rest of your bankroll rules.

The case for stop-win

Stop-win rules are weaker, but not worthless. The anti-stop-win argument is clean: if the game is good and you're playing well, every hour is +EV, so quitting costs money. True. The pro argument is situational: live game quality is time-dependent. At 3 AM the drunk businessman went home, the lineup is four tired regs and a guy on a heater, and your +8 buy-in night is about to be played out against the worst possible version of the pool.

A workable compromise: no fixed stop-win number, but a review trigger. Up more than five buy-ins, or past hour seven? Stand up, take ten minutes away from the table, and honestly answer: is this game still good, and am I still sharp? If both are yes, sit back down and keep printing. If either is no, cash out and go home a big winner. The stop-win becomes a decision point rather than an autopilot.

Tilt indicators you can catch in real time

Stop rules only work if you notice the state they're designed for. Tilt rarely announces itself as rage; it announces itself as tiny deviations from your normal game. The reliable tells:

• Loosening preflop. Playing hands you'd fold on any normal night, "to see a flop."
• Calling to find out. River calls where you know you're beat but need confirmation.
• Faster rebuys. Reaching for the wallet within seconds of losing a stack, no pause, no breath.
• Targeting one player. Shaping your whole session around the guy who sucked out on you.
• Time-checking. If you're checking your phone for the time, some part of you wants permission to leave. Grant it.

Any two of these and the session is over regardless of the number. The stop-loss is the floor, not the trigger — tilt can end a session at one buy-in down, or none. The downswing survival guide covers the longer-timescale version of the same self-monitoring.

Implementing stop rules in live games

Online, a stop rule is a button. Live, the casino is literally designed to keep you seated — free coffee, no clocks, a cage twenty meters away. You need mechanics, not intentions:

• Bring session money, not roll money. If your stop-loss is three buy-ins, take three buy-ins plus €50 for tips and food. The ATM across the floor is the enemy of every stop rule ever made.
• The rack rule. When you hit the number, you rack up immediately — not after the orbit, not after the button passes. "One more orbit" is how a three buy-in stop becomes five, and it is never one more orbit.
• Announce it to a friend. If you play with regulars, tell one of them your number before the session. Social accountability beats willpower at 2 AM.
• Pre-commit the exit ritual. Cash out, log the session in the parking lot or on the train, eat something, go home. Logging matters here: a poker session tracker entry made at the stop number becomes hard data the next time you're tempted to extend "just this once."
• No re-entry same night. Hitting your stop and re-buying into the same game an hour later is not a loophole; it's the tilt winning.

When stop rules are wrong

Two honest caveats. First, if you're a strong player in an exceptional game — the kind of lineup that appears twice a year — a rigid stop-loss costs real expectation. Some pros suspend stop-losses for documented "game of the month" conditions and tighten them everywhere else. Second, a stop-loss can become a psychological trap if you start playing "get-out mode" as you approach the number, making scared decisions to avoid hitting it. The fix is the same as everywhere in bankroll management: the number is set in advance, and hands near the number are played exactly like hands far from it.

And note what stop rules are *not*: a substitute for stake discipline. If every week features a session hitting your stop-loss, the problem isn't the rule — it's that you're outrolled or outclassed at the stake, and the stop-loss is just capping the bleeding. Also worth remembering: swings of this size are structurally normal — see what variance actually looks like — so a stopped session is a cost of doing business, not a personal failure.

FAQ

Is a stop-loss +EV or -EV? Neither, applied to hands. Applied to hours, it's +EV whenever your degraded-state play is worse than the game is good — which for most players, after three lost buy-ins, it is. If you can honestly play your A-game while stuck, the rule costs you nothing because you'll rarely need it.

Should my stop-loss be bigger for bigger buy-in games? Keep it in buy-ins, not currency: 2–3 buy-ins at any stake. The absolute amount scales automatically, which is the point — the rule protects a fraction of your roll, not a fixed dollar figure.

What if I hit my stop-loss but the game is amazing? Leave anyway. An amazing game plus your tilted C-game is a mediocre game at best, and the habit of honoring the rule is worth more than any single night. Exceptional-game exceptions exist, but they should be rare enough to remember by name.

Does a stop-win hurt my long-term win rate? A rigid one, slightly. A review-trigger version, almost not at all — you only quit when the game or your state has degraded, which is where your win rate was falling anyway.

How do stop rules interact with moving down in stakes? They're the same idea at different timescales. The stop-loss protects today; the move-down threshold protects the season. Players who honor the small rule rarely need the big one.