5 BB/hour in a swingy game: 1 in 5 runs are still losing after 1,000 hours
The setup is a 5 BB/hour win rate with a standard deviation of 200 BB/hour (the "very high" preset) over 1,000 hours. The expected result is +5,000 BB, yet 21.5% of the simulated runs finish below zero. None of those runs played badly: with this much variance, 1,000 hours is simply too small a sample.
A typical downswing: over 5,000 big blinds
In the same setup, the typical worst stretch is -5,367 BB, more than 50 buy-ins of 100 BB. The unluckiest 5% of runs see -10,301 BB or worse. With a bankroll of 30 buy-ins, a perfectly normal downswing is enough to put you out of the game.
To be 95% sure you're a winner: about 6,000 hours
The simulator puts it at about 6,150 hours. At 20 hours a week, that's roughly six years. Until then, your results graph says more about luck than about skill.
20% ROI, 1,000-player fields: 1 in 3 still losing after 500 tournaments
The expected result is +100 buy-ins, but 33.2% of runs are still behind after 500 tournaments. Tournament profits come from a few deep runs, so the swings are far bigger than in cash games.
With an 18% cash rate, going 24 tournaments in a row without a cash is normal, and the unluckiest 5% go 37 or more. Next time you're on a long dry spell, don't start doubting yourself straight away.
What these numbers mean
Short-term results say very little about how well you play. Rather than staring at last month's graph, focus on the quality of your decisions, keep a bankroll that can absorb a downswing, and wait for a large sample before drawing conclusions. Every player's win rate and variance are different, so the best way to know yours is to run your own numbers.