What tournament poker variance actually looks like.

Tournament poker variance is dramatically higher than cash-game variance at the same nominal ROI, and most players underestimate this by a wide margin. The reason is structural: top-heavy MTT payouts concentrate the realised return in a small fraction of the field. In a 1,000-runner GTD MTT with a typical PokerStars or GG payout structure, the top 1% of finishes account for roughly half the prize pool. The bottom 85% receive nothing.

This means a tournament player’s yearly return is shaped by a small number of deep runs that may or may not arrive. A 20% true-ROI player can have a losing year. A 5% true-ROI player can have a winning one. The number of tournaments needed to reliably distinguish a winning player from a break-even one — what statisticians call the sample-complexity of the ROI estimate — is much larger than poker culture acknowledges. A few hundred MTTs is not enough; a few thousand often isn’t either.

Most variance simulators on the internet get this picture wrong in one of two ways:

• Flat-ITM approximations. Assuming every cash returns the same multiple of the buy-in throws away the entire shape of the variance — the long thin tail of huge wins is what makes MTT variance so high in the first place.
• Flat-payout assumptions. Assuming a 15% ITM rate with uniform cash sizes ignores that ITM cash sizes range from min-cash (~1.5× buy-in) to first place (~250× buy-in in a typical 1,000-runner field).

A real variance simulation needs the actual realised payout structure of the events you play. The MUCHO·MOTA free Monte Carlo simulator runs 10,000 forward bankroll trajectories on the real payout structures from PokerStars / GG / ACR / iPoker — not flat-payout approximations. You enter your year (number of MTTs at a given buy-in) and your ROI, and the tool gives you the median trajectory, the 90% range, and the probability of busting before the year ends.

What honest variance pictures show, almost without exception:

• The median trajectory is below the expected trajectory (this is purely a feature of compounding multiplicative outcomes — the “median ≠ mean” gap).
• The 90% range is much wider than intuition suggests. A 15% ROI player over 1,000 MTTs at $100 has roughly a one-in-ten chance of being underwater at year end, even with no leaks.
• Risk of ruin is non-negligible at “safe” buy-in counts. The 100-buy-in rule provides far less protection than the folklore implies, especially at high-variance event types.

Variance is the dominant fact in tournament finance. Bankroll plans, sell decisions, and game selection are all variance-management problems disguised as edge-maximisation problems. The platform exists to put real numbers behind those decisions.