Trading Decision Course: 30 Lessons

By Eliza. Runtime 6:21. Opening: Two Separate Ledgers [0:00] Whether a trade makes or loses money is up to the market. Whether you should take that trade is up to you. Those are two separate ledgers. Keep them apart. Lesson 1: One result can't tell you if a decision was good. A Pair of Aces, First One Out (a hypothetical hand) [0:12] Picture a game of Texas Hold'em: you and one opponent across the table. You hold pocket aces and go all-in preflop; your opponent calls with just any random hand. Right now you win about 85% of the time. Play that exact spot 20 times and you win about 17 and lose only 3. Most of the time, this hand takes every chip your opponent has. Tonight, on the very first hand, you're dealt pocket aces. You shove all-in preflop, and get called. The board runs out, and your opponent makes two pair. You're the first player out of the whole game. Going all-in preflop with pocket aces is the least controversial good decision in Texas Hold'em. The same decision wins for you most of the time and tonight sends you home first. Skill vs. One Hand [0:49] On the way home you wonder: in one night, did my game drop from the top to the bottom? Of course not. Losing one hand tells you almost nothing about how well you decide. Why Trading Is Harder [0:59] In trading it hurts even more: when you lose at poker, at least you get to see your opponent's cards. When you lose on a trade, the market shows you nothing. So you guess—and it's easy to guess "I'm just not good enough" or "the market is rigged." Both guesses lead you astray. The Soup Problem: Three Things Mixed Together [1:14] Any single trade's result is really three things mixed together: the method's long-run distribution, how you executed this time, and this time's random noise, which you can't control or predict. You only see one blended number, with no way to separate the parts. Grading a decision on its outcome alone is like judging a chef's knife skills by tasting the soup— after someone's thrown in a handful of salt. The Coin Game [1:33] Let's play a coin game: heads, I pay you $250; tails, you pay me $100. Would you play? Of course. Every flip has an expected value of +$75. But nobody knows whether the next flip lands heads or tails. The question a professional trader answers every day: if this same spot came up 1,000 times, would I take it every single time? That question has an answer. "Will this trade make money?" doesn't. Expected Value [1:56] As a formula: expected value = win rate × average win, minus loss rate × average loss, minus trading costs. It describes the average result of making the same kind of decision many times, not a prediction for the next trade. A positive-EV setup can still lose five in a row. And win rate and payoff can only be estimated from backtests, sim and live logs, and they shift as the market changes. So "positive EV" isn't a permanent business license. It's more like a health checkup—you need to recheck it regularly. The Four Boxes [2:24] Separate outcome from decision and you get four combinations: Rules followed and won, rules followed and lost, rules broken and won, rules broken and lost. The most dangerous is "broke the rules, made money." Beginners celebrate it; professionals find it almost embarrassing. Because money is a reward, and rewards train behavior. It's like running a red light, and the cop doesn't stop you—he hands you a hundred bucks. At the next intersection, you'll run it again. "Followed the rules, lost money" feels the most unfair, but it's part of your sample, not your fault. Changing your rules because of it is like losing once with pocket aces and folding aces forever after. Five Pre-Trade Checks [2:54] Decision quality should be judged before entry, on five criteria, each using only information available at the time. Setup definition: spelled out in observable conditions— "it feels strong" doesn't count. Evidence: backed by a continuous sample of trades— a few winning screenshots don't count. Risk cap: before entry, write down your invalidation, max loss, and position size. Execution consistency: enter and exit per plan. Widen your stop on the fly, and it's a violation even if you make money. Context fit: is today's volatility, liquidity and event risk similar to the conditions in your sample? Three Questions, Result Set Aside [3:25] To judge whether a trade should have been taken, set the result aside and ask just three things. Is it on the setup list you wrote down? Did your entry and management stay on plan? Is the risk you took the size your own records support? All five pass and it loses? Log it as a "good loss." A key item fails and it wins? Log it as a "bad win." All it takes is honesty. Right Direction, Still a Loss: Options [3:45] You can be right on direction and still lose money. You pay $30 for a right: at expiry, you can buy a bicycle for $800. On expiry day the bike is worth $815. You were right on direction, but the right is only worth $15. After your cost, you're still down $15; the bike has to reach $830 just to break even. Now look at a 0DTE option: At 10 a.m. you buy one for $80. The index chops sideways all day. By 1 p.m. it's worth $55, by 3 p.m. $25, and at the 4 p.m. close, zero. Simply because time kept ticking away. The Interview Puzzle [4:15] An interview puzzle: a candidate shows up 20 minutes late with no notice, and is rejected on the spot. It later turns out someone had given him the wrong interview time. Was rejecting him a bad decision? The "be on time" rule is right most of the time. This time it wrongly rejected a candidate—the outcome was wrong, but the decision was based on reasonable assumptions. Whether to change the rule depends on whether what it saves overall outweighs the cost of the occasional miss. A Real Case: Avis (CAR), November 2, 2021 [4:38] Avis (CAR), November 2, 2021: a veteran US equities day trader lost seven figures in just over 40 minutes. In his review he split the trade in two: the setup passed— he'd trade the same setup again; the size failed, far beyond what his own rules allowed. Setup: pass. Position size: fail. That's what it means to grade a decision piece by piece. The Rush to Get Rich [5:00] Then there's a common kind of beginner: small account, trying to cram ten years of skill into six months. Size up, blow up, redeposit, blow up again, then revenge trading, strategy-hopping, buying indicators... years later, they're still stuck in the same place. The irony: the more you want to get rich fast, the slower you tend to learn. Because every outcome trains them: a win reinforces risk-taking, a loss reinforces panic. Putting It All Together [5:23] Let's tie the whole lesson together. Four things you can set before entry: setup definition, evidence sample, risk cap, and context fit. Add in-trade execution, and together they set your decision score. A single trade's P&L comes from decision quality plus random noise. Cross them into four boxes: bad wins are the most dangerous—fix the behavior. Good losses go into your sample—don't change the rules. Three Takeaways & Homework [5:45] Finally, remember three things. One: a single result mixes method, execution and luck, so it can't prove whether a decision was good or bad. Two: judge a decision by its expected value— if the same situation came up 1,000 times, would you still take it? Three: score decisions and outcomes separately—bad wins mean fix the behavior, good losses mean keep the rules. Homework: Find one "made money but broke the rules" trade and one "lost money but followed the rules" trade. Hide the price action after entry, score the decision 0–10, and record the result only in R-multiples. Individual P&L and record figures have been adjusted. Illustrative figures are the author's own. For educational purposes only; not financial advice. Trading involves risk of loss.