The Truth About Variance in MTTs: Lessons from BBZ’s Toy Game

Variance in multi-table tournaments is brutal. Not because it exists, but because players consistently underestimate how much of it there is. You can play well for months and have nothing to show for it. You can run deep three Sundays in a row and feel like you have solved the game. Neither version is the full picture, and your sample size is almost certainly too small to tell you which one is closer to reality.

Jordan “BBZ” Drummond illustrated this with a toy game he shared on Discord. It is one of the clearest explanations of MTT variance you will find anywhere, and it starts with 1,000 monkeys flipping coins.

The Coin Flip Toy Game

Imagine 1,000 players entering a winner-take-all tournament with a $100 buy-in. Every round is a 50-50 coin flip. When the tournament starts, your stack is worth exactly $100. You lose the flip, no big deal. You are out $100.

But if you win that first flip, 500 players remain and your stack is now worth $200 in expected value. Win again and there are 250 left. Your stack is worth $400. Keep winning and you reach heads-up, where your stack is worth $50,000 and you are flipping for the entire $100,000 prize pool.

You lose that final flip. You just lost a $50,000 coin flip. Then you enter the next tournament and the same thing happens. And the next one. If you lose five of these high-value flips in a row, your results graph goes straight down for 5,000 games, and you start wondering if the site is rigged.

It is not rigged. That is just what variance looks like when payouts are top-heavy and the moments that matter most are the ones with the least certainty.

Why Your Sample Size Is Smaller Than You Think

Here is where most players get tripped up. You play 2,000 tournaments and assume that is a meaningful sample. It feels like a lot. Two thousand of anything sounds statistically significant.

But as Drummond pointed out in a follow-up post: “Your 2000 MTT sample is small and might rely on 20 key deep runs. That is it. And there are all kinds of ways for that to go bad. Poor card distribution, unlucky runouts, setups, etc.”

Twenty deep runs. That is the number driving your entire results graph. Everything else is noise. If those 20 spots go well, you look like a crusher. If they go poorly, you look like a losing player. Neither conclusion is reliable because the sample that actually matters is tiny.

This is fundamentally different from cash games, where every hand contributes roughly equally to your win rate. In tournaments, the distribution is wildly uneven. A single final table can represent more expected value than the previous 500 buyins combined. When your results depend on a handful of high-leverage moments spread across thousands of entries, variance has an enormous window to operate in.

Why It Feels Personal

The emotional response to variance is almost always disproportionate to the mathematical reality. Drummond addressed this directly: “But there is no magic. And you are not cursed. And no one is cheating you to the degree you cannot win.”

When you lose five critical all-ins in a row at final tables, it does not feel like statistics. It feels targeted. The toy game explains why. Each of those losses represents a massive chunk of expected value. Losing a flip when your stack is worth $50,000 hurts exponentially more than losing a flip in round one when your stack is worth $100. Your brain registers the magnitude of what you lost, not the probability of it happening.

And it is entirely normal. Run the math on any 1,000-player tournament and sequences of five consecutive losses at high-value moments are not rare. They are expected to happen regularly across a population of players. You are not the exception. You are part of the distribution.

How to Survive Variance in MTTs

You cannot eliminate variance. You can only put yourself in a position where it does not eliminate you.

Bankroll management is the foundation. The swings in the toy game demand a cushion that most players underestimate. If you are playing large-field MTTs, 100 to 200 buyins is a starting point, not a ceiling. The deeper your bankroll, the longer you can weather losing stretches without making desperate decisions.

Volume is the other lever. Drummond’s point about 2,000 tournaments being a small sample means you need to play more, not less, to let your edge express itself. Skill only becomes visible over thousands of tournaments, and even then the signal can be noisy.

Beyond the numbers, the mental side matters. When you lose a $50,000 flip at the final table, your job is to show up the next day and play your best game. Not because variance owes you anything, but because every session played at your peak is another data point where your edge has a chance to compound.

Tools help with perspective. A tournament variance calculator lets you plug in your field sizes, ROI, and sample to see the realistic range of outcomes. When you can see that a 1,000-tournament downswing is within the expected range for a winning player, it takes some of the sting out.

Variance Is the Price of Admission

MTTs exist because of variance. The same top-heavy payout structures that create brutal downswings are what make tournaments worth playing in the first place. If payouts were flat, there would be no life-changing scores, no excitement, and no reason for weaker players to enter.

Drummond’s toy game strips away the complexity of poker and reveals the core truth: you are always flipping coins at some level, and the coins that matter most are the ones you flip last. The skill is in getting to those flips more often than your opponents and having the bankroll and mental game to survive the ones you lose.

Your 2,000 tournament sample is a snapshot, not a verdict. Keep grinding, keep studying, and trust that the math is on your side even when the results are not.