Poker Bankroll Management: How Many Buy-ins You Need, with a Downswing Simulator

Your bankroll is the money you set aside for poker and for nothing else. Bankroll management is the set of rules that decides which stakes you play with it. The reason it matters: even a winning player loses for long stretches, and a player who sits at stakes that are too high for the bankroll can go broke before the edge has time to show.

How long and how deep those stretches get surprises most players. Below you can simulate 2,000 players with the win rate and the stakes you choose and see their results hand by hand. After that we show our own numbers for the question every player asks: how many buy-ins are enough?

24
Buy-ins for a 5% risk of going broke, at a win rate of 5 bb/100 in 6-max
19 BI
Biggest downswing of the median winning player within 100,000 hands
28,500
Hands in a row without a new high, median winning player
28.9%
Of winning players are still behind after 10,000 hands

1. Downswing simulator

Choose a win rate, a standard deviation (how much your results swing), the number of hands and your bankroll in buy-ins. One buy-in is 100 big blinds, a full stack in No-Limit cash games. The simulator plays every hand total for 2,000 players and shows 20 of them as lines, the middle 90% of all players as a blue band and the players who lose the whole bankroll in red.

Try a win rate of zero: every player goes broke sooner or later, however big the bankroll. Then try 2.5 bb/100 with 20 buy-ins and see how many of the winning players still lose everything.

2. Win rate and standard deviation

Two numbers describe a poker player’s results, and the simulator needs both.

  • Win rate (bb/100): how many big blinds you win on average per 100 hands, after the rake. At 5 bb/100 in a game with blinds of €0.50/€1 you win €5 per 100 hands on average. A solid winner in online cash games at small and mid stakes often wins between 2 and 7 bb/100; above 10 bb/100 over a large sample is rare.
  • Standard deviation (bb/100): how far a typical 100 hands land from the average. Tracking software shows it for your own games. Rough values: around 70 to 80 in full ring games with nine players, 85 to 100 in 6-max and 120 or more heads-up or with a very aggressive style.

The standard deviation is many times the win rate. That is the whole problem: after 100 hands at 5 bb/100 with a standard deviation of 90, you are behind almost half of the time. Only over tens of thousands of hands does the win rate start to dominate.

3. How many buy-ins do you need?

There is an exact answer for a given win rate and standard deviation. If you play forever and never move down in stakes, the risk of losing a bankroll of B big blinds is e−2 × win rate × B / SD². We checked the formula with a simulation of 4,000 players over 2 million hands each: at 5 bb/100 and 30 buy-ins it gives 2.46%, the simulation 2.43%.

Win rateBuy-ins for 5% riskfor 1% riskRisk with 20 BIwith 30 BIwith 50 BI
2.5 bb/100497529.1%15.7%4.57%
5.0 bb/10024378.5%2.5%0.21%
7.5 bb/10016252.5%0.4%<0.01%
10.0 bb/10012190.7%0.06%<0.01%
Risk of losing the whole bankroll with no time limit, standard deviation 90 bb/100 (6-max). 1 buy-in (BI) = 100 big blinds. Source: GamblingSociety calculation.

The table shows why one fixed number of buy-ins cannot fit everyone. A strong winner at 10 bb/100 is safe with about 12 buy-ins. A small winner at 2.5 bb/100 needs about 49 for the same safety, and with only 20 buy-ins loses the bankroll in 29% of cases.

Standard deviationTypical forBuy-ins for 5% riskfor 1% risk
70full ring1523
80full ring, looser1929
906-max2437
1006-max, aggressive3046
120heads-up4366
At a win rate of 5 bb/100. The bankroll you need grows with the square of the standard deviation.

4. Downswings are normal

We simulated 20,000 players over 100,000 hands for each win rate. The result: big downswings are not bad luck that hits a few players, they happen to almost everyone.

Win rateBiggest downswing, median1 in 20 playersLongest stretch without a new high, median1 in 10 players
0.0 bb/10032 BI63 BI62,40097,600
2.5 bb/10024 BI47 BI41,80084,800
5.0 bb/10019 BI36 BI28,50060,000
7.5 bb/10016 BI28 BI20,40041,500
10.0 bb/10014 BI24 BI15,40029,700
Within 100,000 hands, standard deviation 90 bb/100. “1 in 20 players” had a downswing at least this big. Source: GamblingSociety simulation, 20,000 players per win rate.

At 5 bb/100 the typical player loses 19 buy-ins in a row at some point within 100,000 hands, and one player in 20 loses 36 or more. The longest stretch without a new high lasts 28,500 hands for the median player. Playing four tables at about 75 hands per table and hour, that is about 95 hours of play without making money.

Still behind after 10,000, 50,000 and 100,000 hands
10,000 hands50,000 hands100,000 hands
Share of winning players who are still below zero after 10,000, 50,000 and 100,000 hands, at win rates of 2.5 to 10 bb/100 2.5 bb/100: 10,000 hands 39.1%, 50,000 hands 26.7%, 100,000 hands 19.0% 2.5 bb/100 39% 27% 19% 5.0 bb/100: 10,000 hands 28.9%, 50,000 hands 10.7%, 100,000 hands 3.9% 5.0 bb/100 29% 11% 3.9% 7.5 bb/100: 10,000 hands 20.2%, 50,000 hands 3.1%, 100,000 hands 0.4% 7.5 bb/100 20% 3.1% 0.4% 10.0 bb/100: 10,000 hands 13.3%, 50,000 hands 0.6%, 100,000 hands 0.0% 10.0 bb/100 13% 0.6% 0.02% 0% 10% 20% 30% 40% 50%
Exact share of players below zero, standard deviation 90 bb/100. Source: GamblingSociety calculation.

A player who wins 5 bb/100 is still behind after 10,000 hands in 29% of cases. And a break-even player, who neither wins nor loses on average, has a biggest downswing of 32 buy-ins within 100,000 hands. With 20 buy-ins, 48% of break-even players lose the whole bankroll in that time.

5. How well do you know your win rate?

Every number above depends on your real win rate, and that is hard to measure. Your results after a few thousand hands say very little about it:

Hands playedYour real win rate lies within
10,000± 17.6 bb/100 of your result
50,000± 7.9 bb/100 of your result
100,000± 5.6 bb/100 of your result
500,000± 2.5 bb/100 of your result
95% range at a standard deviation of 90 bb/100. Source: GamblingSociety calculation.

After 10,000 hands at a measured 5 bb/100, your real win rate could be anything from about −13 to +23 bb/100. Even after 100,000 hands the range is still ± 5.6. So plan your bankroll with a lower win rate than your results show, especially when you move up to a stake where the players are better.

6. Rules for your bankroll

  • Keep it separate: the bankroll is money only for poker. Rent, bills and savings are never part of it, and you do not top it up from them after a bad run.
  • Common rules of thumb: many players use at least 20 to 30 buy-ins for full ring cash games and 30 to 50 for 6-max. Our numbers show that this is enough for a solid winner, but thin for a small winner.
  • Tournaments need far more: the swings in tournaments with many players are much bigger, because most of the money goes to the last few places. The usual rule of thumb is 100 buy-ins or more. Our simulator is built for cash games and does not cover tournaments.
  • Move down in time: decide before you play at which bankroll you go back to the lower stake, for example when you fall below 20 buy-ins of the current stake. Make the decision when you are calm, not in the middle of a downswing.
  • Take shots with a fixed amount: to try a higher stake, set aside a few of its buy-ins. If you lose them, go back down and rebuild.
  • Count the rake: your win rate is what is left after the rake. At small stakes the rake can be larger than the win rate of a good player.

Poker should stay entertainment. Play only with money you can afford to lose, and stop if you find yourself chasing losses. Our bankroll guide for sports betting covers staking plans for bets, and the Kelly criterion calculator shows how the size of an edge and the right stake fit together.

Frequently asked questions

How many buy-ins do I need for poker?

For No-Limit cash games, a winning player at 5 bb/100 in 6-max needs about 24 buy-ins for a 5% risk of going broke and about 37 for 1%. A smaller win rate or a game with bigger swings needs more. For tournaments, 100 buy-ins or more is the usual rule of thumb.

What is a normal downswing in poker?

Within 100,000 hands, a player who wins 5 bb/100 in 6-max typically has a biggest downswing of about 19 buy-ins. One player in 20 loses 36 buy-ins or more in a row.

What is risk of ruin?

The chance of losing your whole bankroll before your edge pays off. It depends on your win rate, the size of your swings and the size of your bankroll. A bigger bankroll lowers it quickly: each extra buy-in multiplies the risk by the same factor.

What does bb/100 mean?

Big blinds won per 100 hands. It makes win rates comparable across stakes: 5 bb/100 is €5 per 100 hands at blinds of €0.50/€1 and €50 at €5/€10.

Can a losing player avoid going broke with a big bankroll?

No. With a win rate of zero or below, every bankroll is lost sooner or later. A bigger bankroll only makes it take longer.

How we calculated the numbers

  • Model: the result of every block of 100 hands is drawn from a normal distribution with the chosen win rate and standard deviation. Over many hands this describes cash game results well. It ignores that win rates change over time and with the stakes.
  • Downswings and breaks: 20,000 simulated players per win rate, 100,000 hands each, standard deviation 90 bb/100.
  • Risk of ruin: the formula e−2 × win rate × bankroll / SD² for play without a time limit, checked against a simulation of 4,000 players over 2 million hands each.
  • Simulator: runs in your browser with 2,000 players. Its numbers move a little each time you press “New players”; the tables above use the larger simulation.

Related guides

About the author: GamblingSociety is an independent site run by Andreas, who has operated casino and betting guides since the mid-2000s. No bookmaker or casino owns or pays for our content. More about us.