The Kelly criterion answers one question for a tournament poker player: what fraction of your bankroll should be in a single event so that your long-run geometric growth is maximised? The closed-form answer is g = μ − ½σ². For top-heavy MTTs, this means risking less per buy-in than most rules of thumb suggest — and selling action above a threshold markup whenever you can.
The Kelly criterion is the closed-form answer to a question every serious tournament poker player faces: how much of your bankroll should be in a single event? The answer maximises long-run geometric growth — the only growth rate that survives sequential, compounding bets — and it has a simple form: at small bet sizes, log-bankroll growth equals expected return minus half the variance, g = μ − ½σ².
Two things follow from that identity that most poker bankroll advice misses.
First: variance is a tax on growth, not just a source of drawdowns. A player with 20% nominal ROI but very high variance can grow their bankroll more slowly than a player with 12% ROI and modest variance, even though the first looks like a “better” player on paper. The arithmetic ROI is the wrong number to optimise; the certainty-equivalent growth rate is the right one. The Kelly framework forces you to look at both.
Second: in top-heavy MTT payout structures, the variance term σ² is enormous. A single tournament concentrates most of its expected value in the top 1–2% of finishes — a winner-takes-25% structure means your realised return is overwhelmingly determined by whether you make a deep run, not by how often you cash. This is why even a strong tournament player with a 15% true ROI cannot rationally play a single $1,000 event with a $30,000 bankroll. The Kelly fraction at that variance level says no.
The Kelly logic also tells you when to sell action. If a buyer offers markup above the threshold where the certainty-equivalent of a sold piece exceeds the certainty-equivalent of holding the piece yourself, you should sell. The threshold isn’t fixed — it depends on your bankroll, the event’s variance, and your confidence in your own ROI. But the structure is closed-form: there’s a sell-fraction that maximises long-run growth for any combination of those inputs, and selling above that fraction at fair markup makes you wealthier despite reducing arithmetic EV, because variance is reduced more than expected value.
The practical workflow:
This is what the MOTA tier of the platform does for any single tournament you’re about to register: enter buy-in, field size, ROI, ITM rate, and get the Kelly stake plus the optimal sell-percentage curve as a function of markup. The free variance simulator shows the underlying picture — what your bankroll trajectory actually looks like under the assumptions you put in.
The math from this page, applied to your real numbers.
Introduction If you’ve ever delved into discussions about optimal betting strategies—whether for investing, poker, or other forms of gambling—you’ve likely come across the Kelly Criterion. Most treatments of the topic focus almost exclusively on the “fraction” of your bankroll to bet. In poker circles, you’ll often hear about “fixed bets” or “all-in” approaches that don’t…
Every Bankroll Management Strategy is a Fractional One In the world of poker tournaments, bankroll management is a topic that every player encounters, regardless of skill level or experience. While strategies and approaches may vary widely, there's a fundamental principle that underpins every decision you make: every bankroll management strategy is inherently a fractional betting…
The "Variance Delta": Selling for Stability and Faster Compounding – A Deeper Look Consider that $1k tournament (a player with a 20% ROI) and a market offering a 1.10 markup. Instead of just asking "Is 1.10 < 1.20, so I'm 'losing' EV on this piece?", MOTA asks: "What fraction of my action, if sold at…
Tournament poker bankroll management isn't 'keep 100 buy-ins'. It's a question about geometric growth under variance — what fraction of your bankroll you can risk per event without going broke and without growing slower than you should. The honest answer is fractional, calibrated to your edge, and almost always smaller than the conventional rule of thumb.
Tournament variance dwarfs cash-game variance at the same nominal ROI. Top-heavy MTT payouts mean a 1,000-runner event concentrates most of its EV in finishes that happen 1–2% of the time, so your realised year is shaped by a handful of deep runs that may or may not arrive. Most simulators get this wrong; here's what an honest variance picture looks like.
Poker staking is a contract market built on top of a high-variance game. A staker buys a fraction of a player's tournament action at a markup; the player gets variance reduction and capital, the staker gets exposure to a player they couldn't buy on the open market. The economics are tighter than they look — fair markup, makeup, swap pools, and audit are all consequences of the same incentive math.
Mystery bounty tournaments concentrate even more variance into the post-bubble field than standard MTTs. The bounty pool's lottery structure rewards survivors disproportionately, distorts ICM relative to standard payouts, and changes optimal bankroll sizing for any player who runs them seriously. Here's how the math actually works.