Skip to content
Partner placement Ad · BBZ Poker training

Tournament Variance Calculator

Sample paths across your tournament mix. Returns downswing distribution, bankroll requirement, and sample-size significance.

1

Enter Tournament Parameters

Running simulation...
Expected Value
70% Confidence
95% Confidence
Best Run
Worst Run
Sample Paths

The black line is your expected profit based on the ROI you entered. The green bands show where 70% and 95% of all simulated outcomes landed. The wider the bands, the more variance you should expect. The colored lines are 20 individual sample runs to show how wildly any single path can deviate from expectation. If your actual results fall outside the 95% band, your true win rate is likely different from the ROI you entered.

Variance in Numbers

Expected value
Simulated profit (average)
ROI
Standard deviation
Skew
Excess kurtosis
Probability of loss
70% confidence interval
95% confidence interval
99.7% confidence interval
ITM % (weighted average)

Result Distribution

This histogram shows the distribution of final profit across all simulations. Each bar represents a range of outcomes and how frequently they occurred. The curved overlay approximates the normal distribution. A tall, narrow peak means outcomes are clustered tightly around the expected value. A wide, flat shape means high variance. If the peak sits left of $0, the majority of simulations lost money at this sample size.

Cumulative Profit
Drawdown from Peak

The blue line tracks cumulative profit over the median simulated run. The red shading shows drawdowns: how far below your all-time high you dropped at any given point. Use the slider to zoom in on shorter stretches. If the red shading is deep and sustained, you need a larger bankroll to survive the swings without going broke or being forced to move down in stakes.

Downswing Statistics

PercentileDepth (Buy-ins)Depth ($)Duration (Tournaments)
USD

Kelly + Simulator use the ROI, field size, and buy-in from your first tournament type above. Edit those parameters to update the analysis.

Kelly Stake Sizing

Optimal tournament buy-in for your bankroll under the Kelly Criterion adapted for MTT variance. Half-Kelly is the practical default. Tournament edge estimates carry large standard errors (your real ROI may be half what your sample shows), so the sizing buffer matters more than in cash games.

Half-Kelly max buy-inlargest tournament you should regularly fire --
Bankroll in buy-insat your typical buy-in above --
EV per tournamentexpected $ at your typical buy-in --

Buy-in levels for your bankroll

Buy-in Kelly utilization Recommendation

Growth rate comparison

Strategy Buy-in Growth rate Drawdown risk

For shareable URLs and method notes, see the standalone MTT Kelly Calculator.

Risk of Ruin Simulator

Monte Carlo simulation of 5,000 independent tournament series using validated payout structures. Returns risk of ruin, profit distribution, and required bankroll by risk level for the buy-in and field size you set above. Series length defaults to 500 tournaments.

For full schedule modeling and the bankroll gauge, see the standalone Tournament Bankroll Simulator.

Am I a Winning Player?

Tournament poker has enormous variance. A profitable player can lose for thousands of games, and a losing player can run hot for just as long. This tool uses the variance profile from your simulation to calculate the statistical probability that your actual results reflect genuine skill rather than luck.

Run a simulation first so this tool knows your tournament's variance profile.
—%
Probability you're a winner
95%
90%
75%
60%
Observed ROI
95% confidence range
Sample progress

Tournament Variance Runs Higher Than Cash

Tournament Results Are Not Normally Distributed

Cash game results approximate a bell curve. Tournament results do not. You lose your entire buy-in 80 to 85% of the time. The remaining 15 to 20%, most cashes return 1.5 to 3x your entry. The meaningful profits come from a handful of deep runs returning 20x, 50x, or 100x+. Remove your top 5 results from 1,000 tournaments and you may be breakeven or negative over the other 995. That concentration of returns is the defining feature of tournament variance.

Observed ROI vs. True ROI

The ROI you enter above is the single most important input in this calculator, and the easiest to misjudge.

Players overestimate their ROI through several compounding mechanisms. You remember the $15,000 Sunday score. You forget the 200 buy-ins that preceded it. You calculate ROI from your last big win forward instead of over two full years. You benchmark against publicly reported numbers from 2012, when the games were softer by an order of magnitude. Chapter 8 of the variance guide breaks down what ROIs actually look like in 2026 across every stake level.

SharkScope tracks results, not edge. A player on a 500-tournament heater with one large bink will show a 40% ROI that has almost nothing to do with their true win rate. A player in a 500-tournament downswing will show a negative ROI despite having a genuine edge. At 500 tournaments, the 95% confidence interval for a 20% ROI player spans roughly -140% to +180%. The data cannot distinguish a strong winner from a losing player at that sample.

Downswings Change Your Effective ROI

ROI is not static. During a downswing, most players' actual edge decreases. Tilt creeps in. You start avoiding spots that build stacks because the pain of losing buy-ins is fresh. You bubble more because you're playing to cash instead of playing to win. You move down in stakes where the fields are tougher per dollar invested. Each of these responses is rational in isolation, but collectively they erode the very edge the bankroll math depends on. The ROI that justified your bankroll at the start of the year may not be the ROI you are producing six months into a cold run.

This is why the bankroll simulator matters alongside this tool. Variance tells you how wide the swings get. Bankroll simulation tells you whether your roll survives them.

Reading the Confidence Intervals

The green bands on the simulation chart are confidence intervals. A 95% confidence interval means that 95 out of every 100 simulated runs finished within that band at the given tournament count. A 70% interval captures the middle of the pack. A 99.7% interval captures nearly everything.

A wider confidence level means a wider band, because you are demanding more certainty that the true outcome falls inside it. Think of it as a net. If you accept a 30% chance the result slips outside (70% CI), the net can be tight. If you only accept a 5% chance (95% CI), you have to widen it. If you want near-certainty (99.7% CI), the net has to stretch to cover almost the entire distribution.

The expected value is identical across intervals; the confidence level determines how much of the outcome distribution the band captures. A player with 15% ROI over 1,000 tournaments might see a 70% CI of +$2,000 to +$12,000 and a 95% CI of -$4,000 to +$18,000 from the same inputs. The 95% range is wider because it accounts for more extreme heaters and downswings. If your actual results land outside the 95% band, the model is telling you that either you ran exceptionally hot or cold, or the ROI you entered does not match your real edge.

Confidence intervals narrow as sample size increases and widen as the field size and payout structure get more top-heavy. A 180-player freezeout with 15% paid produces much wider bands than a 45-player turbo with 30% paid, even at identical ROIs. This is why small-field tournaments converge faster and why grinders in large-field MTTs need thousands of events before their results say anything meaningful about their true win rate.

Rake and Fees

Tournament rake is a flat percentage of the buy-in, paid once at registration. The range across major rooms is 7% to 15%. At $109, a 9% rake means $9.81 goes to the house before the first card is dealt. Over 1,000 tournaments, that is $9,810 in fees alone. A player with a 15% gross ROI and 9% rake is keeping 6% after costs. A player with a 10% gross ROI and 12% rake is paying more in fees than they earn in profit.

Rake differences between rooms compound over volume. The difference between 8% and 12% rake on a $55 buy-in is $2.20 per tournament. Over 2,000 events per year, that is $4,400 in additional fees, enough to turn a marginal winner into a breakeven player. Compare rake across 19 networks at your buy-in level before committing to a schedule.

PKO and Mystery Bounty Variance

Progressive knockout tournaments have lower variance than equivalent freezeouts. The bounty component pays out throughout the event rather than concentrating all returns at the final table, which narrows the distribution of outcomes. A player who busts in 50th place in a PKO may still collect several bounties worth 2 to 5x their entry, while the same finish in a freezeout returns nothing or a min-cash.

Mystery bounty formats sit between the two. The envelope draw introduces a secondary layer of randomness that can widen variance beyond even freezeout levels when the top envelopes are large relative to the buy-in. A $109 mystery bounty with a $50,000 top envelope has a fundamentally different variance profile than a $109 mystery bounty where the top envelope is $5,000.

The PKO variance calculator models all three formats side by side using PrimeDope's power-law bounty capture model, a published correction to the proportional assumption most PKO tools use.

What This Calculator Cannot Tell You

This tool assumes constant ROI across every tournament in the simulation. In reality, your edge varies by field size, structure speed, starting stack depth, and time of day. It does not model ICM deal-making at final tables, which can significantly alter your actual dollar outcomes. It does not account for late registration, re-entries, or the psychological erosion of a downswing. Use it to understand the mathematical boundaries of your results, not as a prediction of what will happen.

For the full treatment of tournament variance, bankroll management, and the psychology of downswings, read The Ultimate Guide to Poker Variance, a 15-chapter reference covering everything from the core formulas to realistic ROI benchmarks at every stake.

Methodology

25,000 Monte Carlo simulations per calculation, each using one of 32 real payout structures matching standard online formats. The distribution is not normal: it has a long right tail (large wins) and a hard floor (losing one buy-in). The model assumes constant ROI and does not account for ICM deal-making, late registration, or format-specific edges.

Where to Play Tournaments

GGPoker Largest MTT schedule and guaranteed prize pools
CoinPoker Low-rake tournaments with crypto payouts
ACR Poker US-facing with large Sunday tournament series

Leave a Comment

Poker Room Ad · 18+ · Play responsibly