The good news: it’s not your fault, and there’s a measurable framework for getting through it without going bust mentally or financially.
This is a piece for grinders who keep checking the graph at the end of every session, hoping the line finally tilts up. Patrick Howard at Mobius Poker ran the simulations on what extended breakeven runs actually look like at the population level. Jared Tendler, who wrote The Mental Game of Poker, mapped what happens to your decisions during them. Together their work answers the question most coaches dodge: how long is “too long” to run breakeven, and what should you actually do during it?
What is an extended breakeven stretch?
An extended breakeven stretch is any sample of hands where the profit at the last hand is no greater than one buy-in more than the first hand. In simple terms: you played a chunk of poker and ended where you started.
Howard and his coauthor Hansen He defined it that way in their Mobius variance research so the concept could be measured cleanly across thousands of simulated samples. The standard online assumption is that breakeven for a few weeks means you’re slumping. The data says otherwise.
How common are extended breakeven runs?
Howard simulated 1,000 theoretical players, each with the same win rate and standard deviation, each playing 1 million hands. He then measured the longest unprofitable sample every player endured. Here’s the result for a 115 bb/100 standard deviation, which is realistic for online cash:
| Win rate (bb/100) | Avg max breakeven sample | Probability of 100K-hand stretch |
|---|---|---|
| 2.5 | 291,000 hands | 97% |
| 5.0 | 141,000 hands | 72% |
| 7.5 | 86,000 hands | 27% |
| 10.0 | 58,000 hands | 5% |
Source: Howard & He, Beyond Downswings: Why Extended Breakeven Runs Are Inevitable, Mobius Poker, December 2024. Heuristic note: simulations assume constant win rate. Real players tilt and play worse during breakeven phases, so observed runs are typically longer than simulated ones.
Read the first row again. A 2.5 bb/100 winner has a 97% chance of running breakeven for 100,000 consecutive hands at some point in a 1 million hand career. That’s not a downswing. That’s roughly two months of full-time play with nothing to show for it.
Even at 7.5 bb/100, which is a strong online cash win rate, the average longest breakeven streak is 86,000 hands. If you’ve been breakeven for less than that, the math says you don’t have a strategy problem. You have a sample size problem.
Why does the variance feel worse than the calculator says?
Here’s the part most pros never get told: the standard variance calculator lies to you, in the less scary direction.
In an earlier Mobius article, Howard worked with Robert Wells on the selection bias problem. The setup is simple. You plug a 50-buy-in downswing into a variance calculator. The calculator says it’s a 1-in-14,000 event for a 7 bb/100 winner over 50,000 hands. You think you got cosmically unlucky.
“The odds of this particular downswing really are 1 in 14,000, but only if you play exactly one 50,000 hand sample. Obviously the players I talked to have played way more than 50,000 hands in their lifetimes.” — Patrick Howard, Mobius Poker.
A pro who plays 1 million hands a year gives variance many more chances to produce ugly outcomes. Each new hand is, statistically, the start of another sample. Ask “what are the chances of going on a 50-buy-in downswing at any point in my career,” not “what are the chances over the next 50,000 hands.” The first answer is uncomfortable. The second is wrong.
What happens to your decisions during a long breakeven stretch?
This is where most variance discussions stop. They show you the math, tell you to be patient, and assume your strategy stays intact. Tendler’s Boom & Bust shows what really happens, and it’s the part that turns a survivable stretch into a career-ender.
