Check yourself

Every question is settled by arithmetic or by a figure on the base-rates page. None of them ask you to remember what a pattern is called, because remembering a name is not the skill that protects an account.

1. 1 / 10Break-even

   A position falls 40%. What gain on the remaining money brings it back to where it started?

   A+40%B+50%C+66.7%D+80%
   Why 0.40 ÷ (1 − 0.40) = 0.667. The loss is taken on the whole amount; the recovery is earned on the 60% that survived it, so it always takes a bigger percentage than it took to get there.
2. 2 / 10Break-even

   An account gains 10% one month and loses 10% the next. Where does it finish?

   AExactly evenBDown 1%CUp 1%DIt depends which came first
   Why 1.10 × 0.90 = 0.99. Percentage changes multiply rather than add, and multiplication does not care about order — reversing the two months gives the same 99%.
3. 3 / 10Cost drag

   A round trip costs 0.10% all in. After 250 round trips, roughly how much of the account have costs consumed, before any profit or loss on the trades?

   AAbout 0.25%BAbout 2.5%CAbout 22%DAbout 40%
   Why 1 − 0.999²⁵⁰ = 22.1%. Each charge lands on what is left after the last one, so costs compound exactly the way returns do — and every term is negative.
4. 4 / 10Expected value

   A method's average winner is twice its average loser. What win rate does it need just to break even?

   A25%B33.3%C50%D66.7%
   Why p* = 1 ÷ (1 + R) = 1 ÷ 3 = 33.3%. At 2:1 a method can be wrong two times out of three and still not lose money — which is why a win rate quoted without a reward-to-risk ratio says nothing at all.
5. 5 / 10Position sizing

   Risking 2% of the account per trade, how deep is the drawdown after ten consecutive losses?

   AExactly 20%BAbout 18%CAbout 22%DAbout 10%
   Why 1 − 0.98¹⁰ = 18.3%. Slightly less than the 20% you get by adding, because each 2% is charged on a smaller balance than the one before it. Compounding works in your favour here and against you everywhere else.
6. 6 / 10Streaks

   A method wins half its trades. Over 250 trades, how often should a run of seven straight losses be expected?

   AEssentially neverBAbout twiceCAbout twenty timesDOnly if the method is broken
   Why 0.5⁷ = 1 in 128, and 250 ÷ 128 ≈ 2. A seven-loss row is what a fair coin looks like at close range, not evidence that anything has stopped working.
7. 7 / 10Evidence

   An indicator is right 60% of the time on the 500 stocks you tested it on. Which of these most weakens that result?

   A60% is not much above halfBYou tested 40 variations and reported the best oneCThe sample was only five yearsDIt does not work on commodities
   Why Testing many variations and reporting the winner guarantees a good-looking result even from pure noise — the objection known as data snooping, and the reason the published tests of technical rules correct for it. The other three are real limitations; this one can manufacture the finding from nothing.
8. 8 / 10Base rates

   A trading course advertises that 60% of its graduates who are still trading after two years are profitable. What is the main problem with that claim?

   ATwo years is too short a windowBProfitable is not definedCThree quarters of day traders quit inside two years, so the survivors are not the cohortDNothing — it is a measured outcome
   Why Survivorship. In the Taiwanese data 44% were still trading after one year and 24% after two, so a statistic about survivors silently discards most of the people the course took money from. The honest denominator is everyone who started.
9. 9 / 10Base rates

   Of everyone who day traded Brazilian equity futures for more than 300 days, what share lost money?

   AAbout 50%BAbout 70%CAbout 85%DAbout 97%
   Why 97%, with 1.1% earning more than the minimum wage and 0.5% beating a bank teller's starting salary. Persisting past 300 days selected for commitment, and commitment did not improve the odds.
10. 10 / 10Evidence

    On this site, an entry tagged “studied” means what?

    AThe claim has been tested in published research, often with weak resultsBThe tool worksCProfessionals use itDIt has been back-tested by the author
    Why Only that a peer-reviewed test of the claim exists. Across that literature the edges are typically small, unstable between markets and periods, and frequently gone once costs are taken out. A tag is a pointer to evidence, not a verdict.