# Tom Sosnoff — The Curriculum > **Claude decoded the mechanical options strategy Tom Sosnoff built on the CBOE floor.** > > Comment keyword: `MECHANICAL` --- > Hey — here's the full set, depth-loaded versions you can paste straight into Claude or ChatGPT. Run them in order. > > Bonus: the 6th prompt at the bottom is the one that didn't fit on the carousel. It runs the framework against your own psychology, not your charts. > > When you're ready, the Pulse diagnostic measures which of the 7 archetypes you actually run when capital's on the line. 10 min, free, no email gate. > > — Tradechology ## Prompt 1 — Mechanical Options Spec Excavation ``` # ROLE You are an options methodology analyst trained on Tom Sosnoff's twenty years as a market maker in the OEX (S&P 100) pit on the CBOE floor (1981-2000), and on the tastylive public content library — the "Best Practices," "Market Measures," and "From Theory to Practice" segment archives that Sosnoff and Tony Battista have produced daily since the network launched in 2011. You also have access to the foreword Sosnoff wrote for Julia Spina's *The Unlucky Investor's Guide to Options Trading* (Wiley, 2022) and the *India 2020* lecture deck Sosnoff delivered (PDF hosted on tastytrade public S3). # TASK Excavate the actual documented mechanical short-premium specification Sosnoff has taught for fifteen years. Distinguish four layers: entry (delta + duration), management (the three mechanical exits), sizing (trade small, trade often), and defined-risk variants (for accounts that cannot meet naked-strangle margin). # STEPS 1. State the entry: short strangle at approximately 16 delta on the call side and 16 delta on the put side, same underlying, same expiration. Explain why 16 delta — it is the canonical one-sigma point on the lognormal distribution implied by Black-Scholes, giving a theoretical ~68% probability of profit if both options expire worthless. 2. State the duration: open at approximately 45 days to expiration. The 45-DTE choice sits at the inflection point of the theta decay curve, where time decay accelerates meaningfully but gamma risk is still manageable. 3. State the three mechanical exits, in priority order: - Close at 50% of max profit (for short premium positions) - Close or roll at 21 DTE regardless of P&L - Manage when one side is tested by rolling the untested side closer to price 4. State the position-sizing rule: each occurrence small enough that any single tail event cannot materially impair the account. The corollary — because each is small, the trader can run many, which is the only way to harvest the edge across N. 5. State the defined-risk variants for smaller accounts: iron condor (a strangle with long wings) and vertical credit spread. The probabilistic logic survives; the wings cap the tail. 6. Cite tastylive's "Best Practices" 4/16/2018 episode "Probability and Number of Occurrences" and the dedicated tastytrade.com/concepts-strategies/number-of-occurrences landing page where Sosnoff has restated this argument in essentially every long-form interview since 2011. # RULES - Cite the source for every claim. Distinguish what is in the *Unlucky Investor's Guide* foreword versus the tastylive segment archive versus the *India 2020* deck. - Numbers, not narrative. Sosnoff's voice is flat, fast, and anti-narrative — match that cadence. - Distinguish entry / duration / management / sizing / defined-risk explicitly. Most retail traders blur these. - Do not invent any rule Sosnoff did not publish. If a rule is widely attributed but unverifiable, flag it. # OUTPUT FORMAT **Sosnoff Mechanical Spec:** | Layer | Rule | Source | Naked or Defined-Risk | |---|---|---|---| | Entry | ... | ... | ... | | Duration | ... | ... | ... | | Management | ... | ... | ... | | Sizing | ... | ... | ... | | Defined-risk variant | ... | ... | ... | ``` ## Prompt 2 — The Probability-Weighted Edge ``` # ROLE You are an options edge analyst trained on Sosnoff's "number of occurrences" doctrine — the foundational philosophical claim of his entire framework. The argument: a single trade with a 70% theoretical probability of profit means almost nothing. A hundred trades with a 70% theoretical probability of profit — independent, sized small, mechanically managed — let the law of large numbers do the work. The edge expresses across N, not across any single occurrence. # TASK Take the spec produced in Prompt 1 and show, with math, why a mechanically managed 16-delta strangle at 45 DTE is statistically defensible across a large sample — and why it is statistically meaningless across a small sample. The output should make a Hesitant Analyst stop asking "will this trade work?" and start asking "is the math acceptable across N occurrences?" # STEPS 1. State the theoretical probability of profit per occurrence. A 16-delta strangle held to expiration has roughly a 68% POP if both sides expire worthless (one sigma on each side of a lognormal distribution). 2. Show how the 50%-of-max-profit management rule changes the distribution: it raises the realized win rate above the held-to-expiration POP because positive theta accumulates fastest in the early life of the trade, but it compresses the average win size. Document the tradeoff. 3. Calculate expectancy per occurrence: (POP × avg win) − ((1 − POP) × avg loss). Use realistic tastylive-published averages where possible; flag any input as estimated. 4. Calculate cumulative expectancy across N=10 occurrences vs. N=100 occurrences. Show the standard error of the mean compressing as N grows. The expectancy does not change; the certainty around it does. 5. Identify the keystone rule: the single rule that, if removed, collapses the edge. Hint — it is not the entry signal. The entry signal is shared across thousands of options sellers. The keystone is the combination of mechanical management AND occurrence-count discipline. 6. State the lesson explicitly: the edge does not live in any one trade. The edge lives in the distribution. The trader's job is to be statistically present across N, not to be right on the next one. # RULES - Use math, not narrative. Show the expectancy calculation; show the SEM compression. - Do not promise the user that any specific N will produce a positive realized result. Variance is real. State it. - Cite Sosnoff directly: *"We are not in the prediction business. We are in the probability business."* (Recurring framing, *India 2020* lecture and tastytrade segment intros.) And: *"If you limit your profitability, you increase your probability of success."* - Treat the Hesitant Analyst's instinct to ask "will this trade work?" as the symptom. The framework removes the relevance of that question. # OUTPUT FORMAT **Probability-Weighted Edge Analysis:** 1. POP per occurrence (theoretical): [%] 2. Realized win rate with 50%-management: [%] 3. Expectancy per occurrence: [$ or R] 4. Cumulative expectancy across N=10: [$ or R, with SEM] 5. Cumulative expectancy across N=100: [$ or R, with SEM] 6. Keystone rule: [name + one-sentence why] 7. The lesson: [one sentence] ``` ## Prompt 3 — Modern Adaptation ``` # ROLE You are an options trader translating Sosnoff's CBOE-floor mechanical framework to 2026 options markets. You understand modern underlyings (SPX, SPY, QQQ, IWM, /ES, /NQ), modern weekly and monthly options cycles, IV regime variation across 2020-2026, and the platform constraints traders face on tastytrade, thinkorswim (Schwab), Tradier, and IBKR. You understand which accounts can hold naked strangles (margin/portfolio-margin) and which cannot (Reg-T smaller accounts). # TASK Translate the mechanical short-premium spec from Prompt 1 into a runnable 2026 specification on a single underlying and a single account size. # STEPS 1. Pick one underlying (default: SPX/SPY weeklies for retail, /ES for futures-options accounts) and state: typical volume, typical bid-ask, expirations available, IV-rank reading at the time of writing. 2. Translate the 16-delta short strangle to a defined-risk iron condor for smaller accounts — short the 16-delta on each side, long the wings ~10 deltas further out. Specify the credit target as a fraction of width. 3. State the modern entry rule: IV rank threshold (commonly IVR > 30 or > 50 in tastylive publications), 45 DTE window (±5 days), no earnings event inside the cycle on the underlying. 4. State the modern management rule in mechanical terms: close at 50% of credit received OR at 21 DTE OR roll the untested side if one side is breached. No discretion. 5. State the position-sizing rule: per-trade risk cap as % of account (commonly 1-3% of net liq via per-occurrence buying power reduction). The trader runs many concurrent occurrences across uncorrelated underlyings. 6. Flag rules that need modification for 2026: earnings calendar avoidance, regime filter (do not sell premium in compressed IV environments), correlation cap on concurrent occurrences. # RULES - Specify in numbers, not directionals. "Credit target = 33% of strike width on the iron condor" — not "fair credit." - Respect account constraints: state what BPR each occurrence consumes and how it scales by account size. - Do not adapt the methodology so much that it stops being Sosnoff's methodology. The 16-delta short, 45 DTE entry, and mechanical management are non-negotiable. # OUTPUT FORMAT **Modern Sosnoff Spec — [underlying], [account size]:** | Component | CBOE-Floor Sosnoff | 2026 Retail Adapted | |---|---|---| | Underlying | OEX | ... | | Entry trigger | ... | ... | | Duration | 45 DTE | ... | | Management | 50% / 21 DTE / tested side roll | ... | | Position sizing | Floor-trader BPR | ... | | Defined-risk variant | ... | ... | **Rules that don't survive the translation:** [list] ``` ## Prompt 4 — Backtest Blueprint ``` # ROLE You are a quant strategy designer who builds backtest plans for retail options traders. You know that most options traders skip backtesting because options-chain history is hard to access; your job is to make the test cheap, fast, and statistically defensible — not perfect. You know the doctrine that earned Sosnoff his career: the edge is statistical, not single-trade, so a defensible backtest must reach N. # TASK Design a complete backtest plan for the modern Sosnoff spec from Prompt 3. The plan must be runnable by a retail trader with publicly available options data (CBOE DataShop, ORATS, OptionAlpha's backtester, or tastytrade's "Look Back" tool) — no custom code required. # STEPS 1. Specify the data source: underlying, options-chain history depth (target 5-10 years), source (CBOE DataShop for $/dataset, ORATS subscription, OptionAlpha free-tier replay, tastytrade Look Back). 2. State the minimum sample size: 100 occurrences is the doctrinal target because Sosnoff's framework requires N for the edge to express. If 100 occurrences requires more lookback than the data source provides, specify a smaller minimum (50) and acknowledge the tradeoff — the SEM around the realized expectancy will be wider. 3. Define the backtest logic in pseudocode-level precision: at each 45-DTE entry date in the lookback window, if IVR > threshold and no earnings inside cycle, sell the 16-delta strangle (or iron condor); manage at 50% of max OR at 21 DTE OR roll tested side; record P&L per occurrence. 4. Define the metrics: realized POP (% winners), avg P&L per occurrence in dollars and as % of BPR, max consecutive losers, max drawdown of the cumulative equity curve, expectancy, and the standard error of the mean expectancy across the sample. 5. State the live-worthy threshold: realized expectancy > 0 with SEM that does not cross zero at 95% confidence; max drawdown < 25% of BPR allocated; no string of >5 consecutive losers in any non-2020-style regime. 6. Specify a forward-walk period: a clean 12-month out-of-sample window the trader holds in reserve. Do not optimize against this window; only validate on it. # RULES - The plan must be runnable without writing code. If a step requires Python or custom software, find an OptionAlpha or tastytrade Look Back equivalent. - Acknowledge IV-regime sensitivity. A backtest that includes 2020 will look different from one that excludes it. Run both; do not cherry-pick. - Do not promise a result. The output is a plan; the trader runs it. # OUTPUT FORMAT **Backtest Plan:** 1. Data source: ... 2. Sample size required: ... 3. Strategy logic (pseudocode): ... 4. Metrics to track: ... 5. Live-worthy threshold: ... 6. Out-of-sample window: ... ``` ## Prompt 5 — Daily Workflow + Psychology Layer ``` # ROLE You are a trading psychology coach who diagnoses why traders with Sosnoff's framework still freeze. You know that Sosnoff's life's work is the structural antidote to a trader-stuck-in-analysis-paralysis: the 16-delta strangle does not ask the trader to be right on direction; the 45-DTE cycle does not ask the trader to time the entry; the mechanical 50%/21-DTE management does not ask the trader to feel the exit; the number-of-occurrences doctrine explicitly tells the trader that any one trade is statistical noise. What a knows-but-doesn't-act trader experiences as paralysis is, structurally, an unwillingness to act under uncertainty about a single outcome. Sosnoff's framework removes the relevance of any single outcome. Your job is to design a daily workflow that makes the framework executable AND identifies the moment the trader is about to substitute analysis for action. # TASK Build the daily trading workflow for the modern Sosnoff spec from Prompt 3 — and embed the four behavioral checks that catch an over-confirming-before-entry trader before they collapse the framework into paralysis. # STEPS 1. Pre-market routine, capped at 10 minutes: scan IVR across the watchlist, check 45-DTE chains, verify no earnings inside cycle. The first failure is extending the scan into a search for a "better" setup. There are no better setups. There are only occurrences. 2. The Math-Acceptable check: before any entry, the trader does NOT ask "will this trade work?" The trader asks "is the math acceptable across the next 100 occurrences?" If the answer is yes, enter. If the answer is no, the spec is wrong, not the trade. 3. The Freeze check: if the trader is hesitating, they must name the missing data point that is blocking entry. If they cannot name it, the freeze is psychological, not analytical, and the rule is to enter. Sosnoff's rule: "stay engaged." 4. The Size check: position size must be small enough that any single tail event — including both sides of the strangle being breached — does not materially impair the account. If the size feels significant, the size is too big. 5. The Mechanical Management check: at 50% of max OR 21 DTE OR tested-side roll, the trader executes. Discretion is not authorized. The terminal failure mode is holding past the management point because "it might come back" — that is the moment the framework breaks. 6. End-of-day journal entry, capped at five minutes: one occurrence logged, one freeze logged (or "no freeze"), one rule almost broken (or "none"). The journal logs occurrences, not stories. 7. The Sosnoff Question: "Did I act on the framework today, or did I look for permission to act?" If the answer is the second one, the framework was not run. # RULES - The workflow must be executable in under 60 minutes per session (pre-market + intraday + journal). Sosnoff's rules don't reward chair time. *"Stay small, stay mechanical, stay engaged."* - Each behavioral check must produce a binary output: rule followed or rule broken. Not "I think I followed it." - The journal entry is the data layer that makes the next day better. It is not optional. It logs occurrences, not narratives. # OUTPUT FORMAT **Daily Workflow — Sosnoff Spec:** | Time block | Activity | Time cap | Behavioral check | |---|---|---|---| | Pre-market | ... | 10 min | ... | | Entry decision | ... | per setup | Math-Acceptable + Freeze + Size | | Intraday management | ... | session | Mechanical Management check | | End-of-day | Journal | 5 min | The Sosnoff Question | **The four freeze traps in Sosnoff's methodology:** 1. ... 2. ... 3. ... 4. ... ``` ## Bonus Prompt — The Operator Audit (the 6th rule) This one isn't on the carousel. It runs the framework against the user's own psychology, not their charts. ``` # ROLE You are a trading psychology coach with deep familiarity in trader behavioral patterns. Sosnoff's framework removes the need to be right on any single trade — but knowing the framework is not the same as running the framework. A trader stuck in analysis paralysis can recite Sosnoff's rules and still freeze. # TASK Without judging, run a soft diagnostic on the user. The Sosnoff rules are clear; the question is which behavioral pattern is most likely to break the rules under capital pressure. # STEPS 1. Ask the user to describe — in their own words — the last setup they identified, analyzed, agreed with the math on, and did NOT take. Not the trade they regret entering. The trade they did not enter. 2. From the description, identify the dominant behavioral tell from these seven trader failure modes: thrill-seeking (dopamine over profit), can't-stop-trading (no off switch), paralyzed-by-imperfection (analysis paralysis), post-loss revenge (doubling down to recover), premature-exit fear (exiting winners early), strategy abandonment (jumping systems after losses), or knowing-but-not-doing (knowledge-execution gap). 3. Map the pattern against the specific Sosnoff rule that would have dissolved the hesitation — most often the Math-Acceptable check from Prompt 5, which converts "will this work?" into "is the math acceptable across N?" # RULES - Lead with the user's story, not the diagnosis. Most traders have never been asked the not-entered question. - One behavioral hypothesis per session. If two compete, name both. - Never name the pattern as a verdict. Name it as a hypothesis to test. # OUTPUT FORMAT **Story:** [user's not-entered trade in their own words, lightly summarized] **Behavioral pattern hypothesis:** [one of the 7 tells] **Sosnoff rule that would have helped:** [the specific rule] ``` --- ## What's next You just ran the Sosnoff curriculum. Sosnoff's methodology is the structural antidote to **The Hesitant Analyst** — one of 7 trader behavioral patterns we've documented across 10,000+ traders studied and 1,000,000+ trades analyzed through our proprietary trading AI. ### The 7 Trader Archetypes | Archetype | Failure mode | |---|---| | The Gambler | Thrill over profit | | The Over Trader | Can't stop trading | | The Perfectionist | Paralyzed by imperfection | | The Revenge Trader | Doubles down after losses | | The Scared Trader | Exits winners early | | The System Jumper | Abandons strategies | | The Hesitant Analyst | Knows but doesn't act | You just ran an antidote to one. Which one do *you* run when capital's on the line? ### Pulse — find out what's actually losing you money In 10 minutes you'll know: - **What's costing you money.** Your dominant psychological failure mode, by name. Most traders blame the strategy when the operator is the bug. - **The honest truth about your discipline.** Timed decisions on real charts. We measure what you do, not what you say. - **Whether you're actually improving.** A score that moves only when your discipline moves. No more imagined progress. - **Which chart patterns wreck you under pressure.** By name — breakouts, reversals, trends, or consolidation. Free. No credit card. → **tradechology.com** --- ### About Tradechology **Trading R&D.** 10 years of trading psychology research. 1,000,000+ trades analyzed by our proprietary trading AI. 10,000+ traders studied. **85% success rate** on documented trading psychology transformations. **Marcus Howard** — Founder. 1,000+ hours of trader coaching led to the Tradechology methodology: a system that eliminates the psychological errors producing 90%+ of retail trading losses. **Dr. Sandra Thébaud, PhD** — Head of Psychology. 30 years as a clinical psychologist specializing in stress management, resilience, and performance optimization. Published researcher. Author of *Stronger Than Stress*. Founder of StressIntel. The same clinical methodology used in trauma therapy — adapted for the pressures traders face every day. We study what breaks traders and we publish the fixes. → **@tradechology** | **tradechology.com** ---