Trading Academy · Module 5 — The long game · Lesson 1/4

Ten trades tell you nothing

Lesson 19 of 22 in the curriculum · ~6 min read

Ten trades is the sample most opinions about trading are built on — a screenshot month, a hot streak, a bad first week. This lesson runs one fixed hypothetical system thousands of times and shows what ten trades can actually prove: almost nothing. Judging a method — or a person — on ten trades is astrology with a spreadsheet.

The system, invented for this page: it wins 4 trades in every 10 on average, winners pay +2R, losers cost −1R (R as defined in lesson 1.3 — the amount risked). Its expectancy is 0.4 × 2R − 0.6 × 1R = +0.2R per trade: a genuinely profitable method, which over ten trades should average +2R. Now watch what "should" means. Five consecutive ten-trade stretches, simulated from that same unchanging system:

Stretch	1	2	3	4	5
Ten-trade total	−4R	+11R	+2R	−7R	+8R

Simulated draws from one fixed hypothetical system — same rules, same edge, every column. The five average exactly +2R; no single stretch is obliged to.

Nothing changed between stretch 2 and stretch 4 — not the rules, not the edge, not the discipline. The +11R stretch would screenshot beautifully; the −7R stretch feels like proof the method is broken; both are the same coin thrown ten times. Across all possible ten-trade stretches, this profitable system finishes negative 38.2% of the time. More than a third of its own two-week windows look like a losing method, forever, by construction.

38.2% of all ten-trade stretches from this same profitable, unchanging system finish negative — exact binomial arithmetic, not a sampled estimate.

Histogram of 2,000 simulated ten-trade R totals from one fixed hypothetical system that wins 4 trades in 10 at +2R and loses at −1R. X axis: sum of 10 trades in R, with totals landing only on −10, −7, −4, −1, +2, +5, +8, +11, +14 and +17. Y axis: count of simulated ten-trade stretches. A dashed vertical line marks the mean at +2.0R, and an annotation reads: same system, every bar — spread is luck alone — 2,000 simulated ten-trade stretches from ONE fixed hypothetical system. Every bar — the −10R disaster on the far left, the +17R miracle on the far right — is the same method, obeyed perfectly. The spread is luck and nothing else.

The corollary reaches everything you read online: anyone showing you hand-picked ten-trade windows can harvest the right-hand tail of exactly this spread, with no edge at all. Small samples are not weak evidence — they are raw material for whichever story the seller prefers.

The full lesson shows the whole distribution — 2,000 simulated stretches in one histogram — puts honest numbers on how many trades a verdict really needs (hundreds, not tens), counts the losing streaks a working system routinely serves, explains how survivorship turns luck into advertising, shows why our own track record publishes losers and sample counts, and ends with an exercise in reading the histogram without flinching.

Full lesson

The rest of this lesson — the worked examples, the exercises, and how our own system applies it — is included with any paid NoGuru subscription. Members receive a personal access link by Telegram DM; open it once and every lesson unlocks on this device.

See channels & pricing Back to the curriculum

← What technical analysis can and cannot do The journal is the edge you control →