You require far more than you believe you do. You require far more than successful professionals suggest. You require far more than statisticians estimate.
You require a lot.
Let’s start with some fundamental statistics. Pretend you are Seo Awsum, a seasoned poker player. Seo earns a living by playing online for $1-$2 6-max. He plays eight or nine tables at once, buys in with a full stack of $200, examines stack sizes for each table before entering a pot, knows the regulars, analyzes stats, limits his playing time, and rarely tilts. Last year, Seo played a million hands, with an average of $12 every 100 hands and a standard deviation of $200.
If Seo plays another million hands this year in the same manner and under the same game conditions, he may expect (with 99 percent confidence) his total earnings at the end of the year to be between $70,000 and $170,000. That is a $100,000 difference due to pure randomness. And Seo’s gains over any given 80,000 hand sample—about a month’s worth of work—could be anywhere from around -$5,000 to
$25,000 (again, with a 99 percent confidence interval). His monthly income can vary by up to $30,000. This represents a monthly variance of 150 buy-ins owing to chance alone.
Now let’s talk about reality, because you’re not Seo Awsum. If you are an online poker professional, your situation is probably worse. For starters, your win rate may be less than $12 per 100 hands. Furthermore, you may not have the stamina to play a million hands in a year. Most importantly, we haven’t yet addressed one of the most serious dangers to bankrolls.
We calculated Seo’s variance numbers using basic Stats 101 formulas. Traditionally, statisticians have used a player’s historical winrate and standard deviation to anticipate bankroll requirements, which they plug into these equations. Some make calculated changes for stepping down and other considerations.
These adjustments are generally optimistic in that they believe you will not tilt and will step down as necessary. But there is a much bigger problem. Stats 101 calculations are based on the assumption that you will continue to play comparable games. This is known as sampling from the same distribution. Reality is really different. Games change.
Games change. It’s that simple. Past outcomes do not predict future performance. Or, in statistical terms, the underlying distribution can shift rapidly. One day, you’re up against Andy, Bill, Caroline, David, and Eric, who have a 3bb per 100 hands winrate with a standard deviation of $200 per 100 hands. The next day, two important events occur, lowering your winrate. Andy discovers out that raising the turn increases your chances of having a legitimate hand compared to your average opponent. He exploits your weakness by folding more frequently when you raise.
It’s a minor change, perhaps one more accurate fold every 600 hands, but if that fold costs you 12bb, you now win 2bb per 100 less. Meanwhile, Felicia takes over for Eric. Felicia reads hands more accurately, which affects you in a variety of ways. She is less likely to lose chips to David, the weakest player in the game. She also occasionally places a low-value bet against you, which Eric would not do. Again, it’s a minor change that costs 2bb per 100 hands.
Overnight, your win rate dropped from 3bb per 100 to -1bb per 100, while your standard deviation remained around $200 per 100 hands. You’re now a loser. You just will not figure it out until tens or hundreds of thousands of hands are involved.
The truth is that games change quickly, and minor alterations might derail your expectations. This aspect significantly raises bankroll requirements.
As a professional player, you will go through downswings that you never imagined possible. Sunny played several hundred thousand hands of $1-$2 6-max before writing this book. He took the job seriously, reviewed his play on a regular basis, slanted minimally, and won with a high percentage. However, at one time, he saw a 40-buy-in drop. Forty! That’s $8,000.
You may have heard about successful pros who ascended to the top with 30- or 40-buy-in bankrolls. So guess what? It is a case of selection bias. They got lucky. Hundreds other players who may have been equally successful did the same thing and went broke. If you play on 30 buy-ins, you’ll eventually bust out.
So what can you do? First, assume solid 6-max pros almost never suffer 30 buy-in downswings due to chance alone. This is not true, but it’s a great assumption. If you find yourself in a big downswing, ask how much you are tilting and stubbornly sticking to tough games. Then ask whether the game has changed. And don’t forget collusion. This is one of those delicate topics that all the “in” people don’t like to talk about. But the fact is, if Andy and Danny get on instant messenger and play best-hand against you, you will take a major hit right in the winrate.
The best way to deal with a small downswing is to tighten up a little while you determine whether you have played well. Run hands by your poker friends, your coach, or the online forums. Review stats like VP$IP, PFR, 3-Bet Percentage, and C-Bet Percentage, and analyze them by position—not just as averages—to help figure out if you are making consistent mistakes.
The best way to deal with a big downswing is to take a few days off. Relax, exercise, sleep, and get your head back in the game.
After a few losses, you should step down. There comes a point in a downswing where you may be playing badly and not know it. In general, after losing 20 buy-ins, not only should you take a break and analyze your recent game, you should also seriously consider dropping down a limit or two.
So how big a bankroll do you need? Clearly—and perhaps surprisingly—it is more than 40 buy-ins. But what is the magic number for an online pro who doesn’t have much room to move down and still make a good living?
This number depends on a great many factors, such as how much you think games might change, how aggressive you are, your risk tolerance, your winrate, your living expenses, how much you tilt, and how good you are at stepping down. It also assumes you don’t have a job or a trust fund to replenish your bankroll. No matter how you slice it, it’s just an estimation. No one can know your true number.
But who are we to deprive you of our recommendation? If forced to give you a specific bankroll number, ours is:
100 buy-ins
That recommendation may not be to your liking. That is okay. Our goal with this tutorial has been to challenge you to see situations in fresh light—to see all 64 squares, if you will—and to assist you in surviving and thriving in the online poker arena.
If our proposal convinces you to keep 60 buy-ins in your bankroll rather than 30 buy-ins, we’ve done our job. We have significantly boosted your chances of surviving, and, more importantly, we have made you think.
