OF Options Force

Open the interactive version →

Number of Occurrences

Number of Occurrences

Research-finding entry: the statistical bridge between a theoretical edge and a realized one — and one of the most-repeated ideas in the entire premium-selling canon.
Sourcing note: this topic lives across a learn/concepts page and a long run of video segments. Page titles, dates, and URLs were verified via search; the episode bodies are JavaScript-rendered and could not be scraped verbatim, so in-segment wording is reconstructed from the platform's consistent phrasing and corroborated by an independent third-party replication (Option Alpha). Claims from general probability theory rather than a retrieved source page are graded C.

---

The Claim

A trade's theoretical probability of profit only shows up in your realized win rate after a large number of trades. The method calls this "number of occurrences": by the law of large numbers, as occurrences grow, the proportion of winners converges toward the position's theoretical POP — so the edge is real only across many small, repeated trades, not any single one.

---

What the Method Teaches

The rule is stated almost identically wherever it appears: "The greater the number of occurrences, the greater the chance that the probabilities will play out closer to our expectations."

The reasoning chain has four links:

1. Probability is a long-run frequency, not a per-trade guarantee. An 80%-POP trade does not mean "this trade wins 80%." It means that over a large sample of such trades, roughly 80% finish as winners. Any individual outcome is a single draw and tells you almost nothing.

2. Small samples are dominated by variance ("the random acts of the market"). With only a handful of trades, an 80%-POP strategy can easily produce a string of losers (or winners) purely by chance. The realized win rate over 10 trades is wildly variable; it is not evidence about the strategy.

3. The law of large numbers does the work — but only with a big enough N. As occurrences accumulate, the realized win rate tightens around the theoretical POP. The research puts a number on "big enough": roughly 1,000 occurrences, which corresponds to being within about ±3% of the true average at ~95% confidence.

4. *Small size is the enabler, not a separate rule. You can only afford to take hundreds-to-thousands of trades if each one is tiny relative to the account. The stock guidance — keep risk per trade small, frequently cited as under ~5% of net liquidity — is what makes a large N survivable*, so "trade small" and "maximize occurrences" are presented as two halves of one idea.

The slogan that compresses all four links is the house motto "trade small, trade often." Small keeps any one loss from mattering; often accumulates the occurrences that let probability assert itself.

---

Original Source(s)

The concept is anchored in a dedicated learn/concepts page and a sequence of named video studies (full URLs in Sources):

---

Supporting Evidence

The coin-flip simulation (the cleanest demonstration). The research runs a fair-coin (50%) Monte-Carlo as a stand-in for trades. The realized heads-proportion is volatile early, begins visibly converging toward 50% around 500 flips, tightens meaningfully near 1,000, and shows little further dramatic improvement beyond ~2,000 — the empirical basis for treating ~1,000 as the threshold where results become a reliable read on the underlying probability.

The SPY-strangle sample-size study. The 2013 founding segment ran an actual short-premium strategy (SPY strangles) and showed small samples producing erratic win rates while larger samples settled toward the expected win rate — the law of large numbers is not just a coin-flip abstraction but holds for real positions.

The convergence table (canonical teaching example). Across the occurrences segments the same illustration recurs:

Independent third-party replication. Option Alpha ran its own analysis ("How Many Trades Does it Take to be Successful?") and reached the same practical conclusion: temporally diversified, randomly selected trades converge on the desired probability with sufficient confidence after approximately 1,000 trades. Two independent groups landing on the same ~1,000 figure raises confidence that the benchmark is not a quirk of one shop's methodology.

Theoretical backing. The claim is a direct application of the (weak) law of large numbers: the sample mean of i.i.d. trials converges in probability to the expected value as N grows; the standard error of a proportion shrinks like 1/√N, which is exactly why gains slow after ~1,000 (halving the error again would require ~4,000). This is textbook probability, fully consistent with the published numbers.

---

Contradicting Evidence & Nuances

1. The i.i.d. assumption is the soft underbelly. The law of large numbers assumes occurrences are independent and identically distributed. Real portfolios violate both: simultaneous positions in SPY, QQQ, and large-cap tech are heavily correlated, so 50 open trades during a market shock are not 50 independent occurrences — they can all lose together. Correlation slows or breaks the convergence the law promises. The method addresses this with diversification and uncorrelated underlyings, but the assumption is a real limitation, not a footnote.

2. "Identically distributed" assumes a stable edge. Convergence is toward the true long-run probability — which is only stable if the volatility risk premium and your strategy's edge persist unchanged across all 1,000 trades. If the regime shifts (e.g., a prolonged low-IV environment compresses the premium), you are averaging over a moving target, and the realized number need not converge to the POP shown at entry.

3. High win rate ≠ profitability — occurrences amplify whatever the EV is. The law of large numbers makes your win rate converge; it does not guarantee profit. If the per-trade expected value is negative (e.g., losses are too large relative to the credit, or sizing is wrong), more occurrences converge you toward a reliable loss. Occurrences are a multiplier on edge, not a substitute for it — they must be paired with positive EV and disciplined sizing.

4. The retail-feasibility gap. Even Option Alpha — which endorses the ~1,000 figure — flags that it is unrealistic for a retail trader to place ~1,000 well-diversified occurrences "in a few short months"; the number must be distilled to something achievable. In practice the threshold is reached over years, meaning a trader operates in the high-variance pre-convergence zone for a long time.

5. ~1,000 is a convention, not a hard line. The required N depends on how tight a confidence band you demand. ~1,000 buys ±3% at 95%; tolerate ±5% and you need far fewer, demand ±1% and you need many more. The figure is a reasonable rule of thumb, not a law of nature.

---

Frequency of Mention

This is foundational philosophy, not a niche study. It appears in the permanent learn/concepts library and recurs across at least four show franchises spanning 2013 → 2021+, with multiple dedicated episodes sharing nearly identical titles. It is inseparable from the platform's two most-repeated mottos — "trade small, trade often" and "the probabilities play out over time" — and underpins the entire short-premium, high-occurrence business model. Among these research claims, few are more entrenched.

---

Practical Implementation

How a trader actually operationalizes "number of occurrences":

For the underlying probability machinery (POP, delta-as-probability, EV), see the companion treatment in ../02_probability/.

---

Limitations & Caveats

---

Verdict

Overall evidence grade: A−. The core claim — that realized win rate converges to theoretical POP as occurrences grow — is a direct, uncontroversial application of the law of large numbers, taught consistently for over a decade, demonstrated with both a coin-flip simulation and a real SPY-strangle study, and independently replicated by a third party that arrived at the same ~1,000-trade benchmark. The qualitative claim is essentially settled. Research-backed, with High confidence.

The grade is shy of a clean A only because the load-bearing quantitative benchmark (~1,000 occurrences ≈ ±3% at 95%) rests on a specific tolerance choice and on an i.i.d. assumption that real, correlated portfolios violate — so the precise number is Medium-confidence and partly heuristic in practice. The single most important caveat for a trader: occurrences make your win rate honest, not your account profitable — that still requires positive EV and small, consistent size.

---

Sources

_Evidence-labeled per the Project Charter. Education only, not financial advice._