The trip plan

Every session needs a structure, decided in advance, that governs entry, depth, and exit.

A recreational player can turn up at a casino, buy in for whatever's in their pocket, play until they're tired or out of chips, and call that a session. Someone playing with a structural edge and a finite bankroll doesn't have that luxury. The bankroll is a long-term instrument; each session is a withdrawal of risk against it. The decisions you make before you sit down (how much to bring, how long to play, when to leave regardless of the count) determine whether your bankroll survives long enough to reach the hands that statistically establish your edge. Planning is not caution; it's the framework that makes the edge real rather than theoretical.

The 200-Unit Bankroll: Structure and Practical Reality

The rule of thumb in counting literature is to maintain a total bankroll of at least 200 times your maximum bet. At a 1-to-12 spread with a £25 unit, the maximum bet is £300 and the recommended bankroll is £60,000. At a £5 unit with a £60 maximum bet, the recommended bankroll is £12,000. These are the numbers the spreads and bankroll lesson derives in detail; they come from risk of ruin calculations, not from convention.

Most part-time counters don't maintain a £60,000 dedicated blackjack bankroll. This is fine, but it requires an honest recalibration of unit size. If your available dedicated bankroll is £5,000, your unit size should not exceed £25 (200 units at maximum bet) at a 1-to-8 spread, and probably should be closer to £15-20 to allow for the spread. Using a £25 unit against a £5,000 bankroll at a 1-to-12 spread produces risk of ruin well above 20%, which most players find psychologically unsustainable when the drawdown arrives, and it will arrive. The answer is to reduce the unit, not to hope the drawdown doesn't happen.

The session bankroll (what you bring to one visit) is different from the total bankroll (your full dedicated capital). A session buy-in of 20-30 units at the maximum bet level is a reasonable session allocation. At a £25 unit maximum bet of £300, a session buy-in of £1,500 (30 max-bet units, or 60 min-bet units at 1 unit = £25) puts you at a reasonable session depth. This is not a stop-loss; it's an initial purchase. The stop-loss is a separate decision about when you walk regardless of buy-in remaining.

Session Length and Venue Rotation

Two to three hours is the practical upper bound for a single session at one London venue before heat risk starts to accumulate. This isn't a precise figure; it depends on the busyness of the floor, the shift change schedule of pit staff, and your specific bet behaviour. A two-hour session at the Hippodrome on a Friday evening is less visible than a two-hour session at Aspers Westfield Stratford on a quiet Tuesday afternoon when there are three other players at your table and the pit manager has nothing else to watch.

The floor manager shift at most UK casinos changes every 8 hours, sometimes more frequently in senior positions. Playing across a shift change means the incoming manager has no prior observation of your session, which resets some of the accumulated profile. Playing into a second shift change means you're building two profiles rather than one continuing one. This is one reason why short, well-separated sessions protect operational longevity better than marathon sessions.

Venue rotation amplifies this protection. If you play once a month at each of three London venues (the Hippodrome, Aspers Stratford, and a third option), each venue is accumulating observations at one-third the rate it would if you visited weekly. The heat threshold at each venue is a function of total observation time, not calendar time. Spreading that time across venues extends your operational lifespan significantly. The heat and cover lesson covers the specific behaviour patterns to manage during those sessions.

Walk Rules: Stop-Loss and Win Goals

A stop-loss is the maximum you're willing to lose in a single session, decided before you arrive. Once you've lost that amount, you leave, regardless of the count, regardless of the remaining shoe, regardless of any narrative about needing to "get it back." A stop-loss of 20-30 maximum-bet units is a reasonable range. At a £300 maximum bet, that's £6,000-£9,000. If that number is larger than what you're comfortable losing in one session, reduce your unit size until it isn't.

Stop-losses are primarily a bankroll preservation tool. They prevent the situation where a counter, experiencing a bad variance run, increases their bets emotionally to recover losses. That behaviour, called "steaming" in the literature, takes a temporarily unlucky session and compounds it into a potentially ruinous one. The stop-loss removes the decision from the emotional state and places it with the pre-session rational self, who designed the plan when not under the pressure of being down 20 units.

Win goals are more controversial. From a purely mathematical standpoint, there's no reason to leave a positive-expectation game because you're ahead; the expected value of continuing is positive regardless of your current session result. The argument for a win goal is operational, not mathematical: a very large win in a single session is memorable and visible. Walking away after an unusual upswing reduces the salience of the win to the pit and avoids the scenario where a dramatic session result triggers a retroactive review of your play history. A win goal of 50-80 units is sometimes used for this reason. It's not mathematically necessary; it's cover.

The Hand-Tracking Notebook

Professional advantage players keep records. The minimum useful record for a counter is: date, venue, session length in hands (estimated), starting and ending chip count, buy-in amount, average bet size, and any notable pit interactions. This record serves two purposes.

First, it gives you data to evaluate your actual performance against your theoretical expectation over time. The math appendix lesson covers how many hands are needed before that comparison is statistically meaningful (approximately 23,000 at a 0.75% edge). Before that threshold, the results are mostly noise. A notebook that shows you 5,000 hands with a 3% positive result doesn't confirm a 3% edge; it's within the range of variance at any realistic edge. But the record is still useful for detecting systematic errors in your game (wrong basic strategy decisions that you then notice on review) and for tracking venue exposure.

Second, the notebook is useful if your counting results are ever questioned by a casino or, in an extreme scenario, by a regulatory body. A documented record of sessions showing realistic variance is a more credible presentation of legitimate play than no records at all. UK casinos don't routinely demand that players account for their results; the notebook is primarily for your own analytical use. The card counter trainer logs your training session performance automatically, which complements the live notebook with data on your accuracy rate in practice conditions.

Key numbers

Sources: our risk of ruin analysis, Blackjack Forum Online camouflage, UKGC general betting rules, Hippodrome Casino.