胜率与手数 → 估算置信区间与下行波动
Variance describes the degree to which short-term results deviate from long-term EV. Even +EV players experience downswings; the larger the sample size, the closer the results are to the true winrate.
Context: Tool guide: variance
Variance describes the fluctuation of short-term results relative to +EV: the smaller the sample, the greater the deviation. MTT and deep-stack PLO have higher variance than low-volatility cash games; bankroll multiples should be adjusted upward. ROI and ITM require sufficient sample size; combine with orbit-cost to check if "playing enough" truly reflects positive EV. Use the calculator on this page to compute the break-even win rate instantly, then cross-reference with opponents' ranges, position, and turn/river plans to avoid simply memorizing formulas without applying them at the table.
Changing to a winning strategy after a losing streak; moving up in stakes without BRM after a win. Mistaking luck for skill. Ignoring the impact of rakeback and level selection on variance. Also, do not treat Solver frequencies as immutable truths without pot odds verification; when information is incomplete, ICM and SPR remain the first principles of tournament play.
Set psychological expectations using the length of a downswing. Separate evaluation of decision quality from tournament-roi and EV analysis. Establish both bankroll and emotional stop-loss limits, and track execution rate. Each week, pick five real hand histories and recalculate using the same tool, then revisit the glossary to review pot odds, ICM, SPR, and other concepts that are automatically linked. Cross-referencing with push-fold and MDF can help build a complete preflop and postflop mathematical framework faster.
Scenario：Cash game winrate 5bb/100, played 50,000 hands.
Standard deviation changes with √(number of hands) 2. 50k hands can still have a swing of ±10bb+ 3. Context: Tool example step: variance
Need 100k+ hands to reliably estimate winrate
Conclusion：Short-term losses do not represent -EV; stick to +EV and sufficient sample.
Scenario：NL25 lost 5 buy-ins in a row, thinking about moving up to NL50 to win it back.
Check if BRM still satisfies 20+ buy-in. 3. Context: Tool example step: variance
You should first lower bets or stop playing and review.
Conclusion：Raising bets after consecutive losses violates BRM; lowering bets or taking a break is a better choice.