Bankroll Management: The Arithmetic Between Your Budget and the Table

Tier one — the ceiling — is already built: the planner's monthly figure, computed from income minus essentials minus savings, installed as a hard deposit limit; nothing in this guide ever argues with it, because bankroll management is how the ceiling gets spent well, never how it gets raised. Tier two — the session bankroll: the ceiling divided by the month's planned sessions — four Friday evenings from a £60 ceiling means £15 sessions, arithmetic so simple it's mostly a decision to do it — with each session's figure treated as fully spendable entertainment money: a session bankroll is priced to zero, and zero is an acceptable, common, pre-approved outcome rather than a problem needing another deposit. Tier three — the stake: the session bankroll divided by the rounds it should survive, which is where the variance guide becomes practical: gentler games and bets need fewer reserve rounds; spikier ones need many more to reach their distribution's decent outcomes — the ratios in the next section. The structure's whole point in one sentence: every number flows downhill from a ceiling that never moves mid-month — sessions can't borrow from next week, stakes can't borrow from next session, and the one-deposit rule enforces the walls exactly where impulse tests them.

The cold rule is already written: the session bankroll's zero is the stop — pre-approved, entertainment fully consumed, no second deposit (the architecture's firmest wall) and no borrowing from next session's tier; a cold night ending on schedule is the system working, not failing. The hot rule needs writing in advance, because heat melts judgement: pick a simple win threshold — double the session bankroll is the classic — at which the original stake exits to the withdrawal queue and play continues (if at all) on winnings only; the exit reflex is the single behaviour separating players whose good nights survive from players whose good nights fund longer bad ones, and it only fires reliably when the threshold was set before the cards were. The long rule: time-boxed sessions with the reality-check timer on — variance's traps all strengthen with fatigue, and the third hour's decisions are measurably worse than the first's at any balance. And the accounting rule that makes "winning" mean something: the session log's one line per session, with the monthly net read against the ceiling — because mid-hobby, "winning" honestly means the entertainment cost what the plan said, wins banked and losses bounded, and the log is the only scoreboard that doesn't flatter.

Every long-running bankroll eventually meets a month that breaks it — overspent tiers, a chased loss, structure abandoned mid-evening — and what happens next decides whether the system was real. Step one: close the month early. A broken plan doesn't get patched mid-cycle — play stops until the next calendar month, because the remaining weeks of a breached month are where single breaches become new baselines; the pause is the protocol's spine, and it is deliberately unglamorous. Step two: write the breach while it's sharp. One honest ledger paragraph — what broke, at which trigger, what the evening felt like from inside — because next month's defence gets designed against this record, not against the softened memory. Step three: restart one tier smaller. The new month opens at a reduced allocation — not as penance but as recalibration: a plan that broke was mis-sized for the player who actually showed up, and earning back the previous tier over two clean months rebuilds the structure's authority along with its numbers. Step four: convert the breach into architecture. Whatever failed gets a structural answer — operator limits lowered to the new tier, the specific trigger (payday, late hours, a particular room) given its own guard — per the principle that resolve broke once and will again, while settings don't. The protocol's honest coda: a second broken month in a season is no longer a bankroll problem — it's the play-safe page's territory, and moving there promptly is the protocol working, not failing.