CS2 gambling6 min read

Bankroll management: budgets, bet size and how long your money lasts

How big a gambling budget should be, and what bet size and the multiplier do to the chance of losing it all, worked out exactly for a 1% edge.

Fact-checked 27 October 2026 · 8 sources

A bankroll is the money you set aside to gamble with. Managing it can't change the house edge, which takes the same share of every coin you wager on average, but it decides how much you spend, how fast it goes and how hard the swings hit. Here is one public-health answer to how big a gambling budget should be, then exact numbers for what bet size and the multiplier do to it.

Set the budget before the bets

In 2021 the Canadian Centre on Substance Use and Addiction (CCSA) published the Lower-Risk Gambling Guidelines, built from survey data on more than 60,000 people who gamble in eight countries. They set three limits, meant to be followed together:

  • How much: no more than 1% of household income before tax in a month. Their example is a household earning $70,000 a year, which should gamble no more than $58 a month.
  • How often: no more than 4 days a month, roughly once a week.
  • How many: no more than 2 types of games played regularly.

The money limit reflects how reported harm rose with spending: in the pooled data, 1.9% of people who spent up to 0.1% of their monthly household income before tax on gambling reported financial harm, against 8.1% of those spending 1.1% to 2% and 26% of those spending more than 5%. The guidelines add that anyone with alcohol or drug problems, anxiety or depression, or a personal or family history of gambling problems or substance use disorders should consider gambling less than this or not at all.

A 2024 online panel of 7,960 adults in Canada, reported by CCSA in November 2025, found that 93.8% of those who had gambled online in the past year exceeded at least one of these limits, against 9.4% of lottery-only players. Whatever the number, it should be money you can afford to lose, set aside before you play. When it's gone the session is over, and refilling it from other money is chasing losses.

What a bankroll buys

The Wizard of Odds puts the central fact plainly: what you can expect to lose is the total you bet times the house edge, and "It does not make any difference how you break up your total amount bet." House edge and RTP works out the edge game by game, and cs2.casino shows each game's house edge and bet limits. Over time the house wins.

Turned around, a bankroll buys a predictable amount of wagering. At a 1% edge every coin wagered costs 0.01 coins on average, so losing 100 coins takes 10,000 coins of wagering on average, in bets of any size. Winnings you bet again count towards that total. The guidelines warn that fast-paced games with frequent betting can lead to problems more quickly.

Keep playing with no stopping point and the bankroll runs out for certain: a gambler at a game with non-positive expected value eventually goes broke, whatever the betting system. On a 2× bet that wins 49.5% of the time, 100 coins last 10,000 one-coin bets on average, but half of all bankrolls are gone by bet 6,758. At 10 coins a bet the average is 1,000 bets, and half are gone within 178.

Bet size decides how fast it can go

Take a 100-coin bankroll, a 2× dice bet that wins 49.5% of the time (a 1% edge) and a plan to stop after 500 bets or when the money runs out. Only the size of each bet changes:

BetShare of bankrollLost all 100Ended aheadAverage loss
11%under 0.01%39.4%5.0
22%4.1%39.4%9.9
55%44.8%36.5%19.3
1010%71.9%24.5%24.6
2020%86.2%13.2%27.6

Bigger bets don't buy a better chance of finishing ahead. Scaling every bet up scales wins and losses alike, so over 500 bets it can't raise the odds of ending in profit; it only adds the risk of hitting zero first, from almost nothing at 1 coin to 86% at 20.

The average loss is exactly 1% of the coins wagered in every row, and it grows more slowly than the bet only because busted sessions stop wagering. The bankroll figures here are exact, computed over every possible balance, and a 2-million-session simulation of each table row agreed within 0.1 percentage points.

The multiplier decides how hard it swings

Now keep the bet at 1 coin and change the target. Every row has the same 1% edge, because the win chance is 0.99 divided by the multiplier:

TargetWin chanceLost all 100Ended aheadEnded on 200 or more
1.1×90%under 0.01%25.5%none
2×49.5%under 0.01%39.4%under 0.01%
10×9.9%14.1%43.2%7.0%
100×0.99%66.7%26.0%26.0%

At 100×, missing the first 100 bets in a row empties the bankroll, and that alone happens in 37% of sessions (0.9901^100). Low multipliers almost never bust but rarely end far from the average, while high ones push results out to both ends.

A row with a better chance of ending ahead isn't a better deal. Each loses 1% of what it wagers, and the high-multiplier rows wager less on average only because so many bankrolls run out early. As the Wizard of Odds notes, for the same game and total amount bet, money management can affect variance but not the expected loss.

Why bet-sizing formulas say don't bet

Advantage gamblers such as card counters size bets with the Kelly criterion, described by John Kelly of Bell Labs in 1956, or a fraction of it. The Wizard of Odds calls it "a bet-sizing technique which balances both risk and reward for the advantage gambler": on a bet with two outcomes, the share of your bankroll to stake is the advantage divided by what the bet pays to one.

Edward Thorp, who introduced it to gamblers in Beat the Dealer in 1962, opens his paper on it with the condition that comes first: "The central problem for gamblers is to find positive expectation bets."

A 2× bet at a 1% edge has an advantage of −1% and pays 1 to 1, so the Kelly stake is −1% of your bankroll. In general it works out to minus the edge divided by (multiplier − 1), negative for every bet with a house edge. A player can't take the house's side, so the best stake is zero.

Betting a fixed share of what's left doesn't escape this. Stake 5% of your current balance on that 2× bet for 1,000 bets and the typical (median) result is 17% of where you started, with about three runs in four ending below half. Thorp's paper gives the general rule: if betting a fixed fraction has a negative growth rate, meaning a typical bankroll shrinks bet by bet, the bankroll tends to zero. Against a house edge every fraction's growth rate is negative.

Stop points that hold

A bankroll only works as a limit if the stopping point is set before the first bet. The loss stop is the bankroll itself, and the guidelines' 4 days a month caps how often a new one starts. A win goal is fine if it helps you leave, but as the betting systems article shows, no rule for when to stop turns bets that lose on average into a session that wins on average.

A plan is easier to keep when it's written into your account. On cs2.casino, your account limits cap deposits, losses and wagers over a rolling day, week or month. Lowering one applies at once, while raising or removing one waits a cooling-off period, and self-exclusion runs from 24 hours to permanent. Responsible gambling limits explains how each tool works and lists free helplines.

Sources

Read next

All posts