Demo trading has two failure modes and almost everyone picks one. The first trader opens a demo, doubles the fake money in a lucky week, declares the apprenticeship complete and funds a live account that is gone in a month. The second trader stays on demo for two years, endlessly comfortable, learning nothing that only stakes can teach. Simulation is a genuinely powerful tool that most people use wrong in one of exactly these two directions. Used correctly, it is a structured phase with entry criteria, exit criteria, and a defined graduation date.
What simulation teaches well
Demo is unbeatable for mechanics. Order types, attaching stops and targets, modifying positions, closing partials: these need to be reflexes before real money is at stake, because a fat-finger that buys 1.0 lots instead of 0.10 is a free lesson on demo and a $500 lesson live. Demo is also the right place to rehearse your written plan end to end: does the setup appear often enough, is the 2-hour window realistic, can you actually compute 1% sizing in five seconds. Most valuably, demo builds a sample at zero cost. One hundred trades of a single setup, fully journaled, tells you its approximate win rate and average payoff before a real dollar is exposed. Traders skip that sample and then discover, with rent money, that the setup wins 35% of the time.
What simulation cannot teach
Nobody has ever felt their pulse spike over demo money, and that is the whole problem. The defining skill of live trading is executing your rules while your body objects, and demo never triggers the objection. Losses that cost nothing also teach nothing about loss: demo traders routinely oversize, skip stops and average down precisely because the consequences are abstract, which means a careless demo phase actively trains bad habits. Fills are kinder too: demo executions ignore the queue and much of the slippage of real order books, so scalping strategies especially look better in simulation than they ever will live. Demo measures your strategy. It cannot measure you.
A graduation plan with real gates
Treat demo as a course with a final exam you set in advance. Trade the demo exactly as you intend to trade live: same instrument, same session window, same 1% risk on the same nominal $10,000, full journal, no do-overs. Then graduate only when the record, not the mood, says so. Reasonable gates for 100 consecutive trades:
- Risk rule followed on at least 95 of the 100 trades, verified in the journal, not from memory.
- The daily loss limit never breached, not once, because live it will be tested on your worst day.
- Positive expectancy over the most recent 50 trades, with at least a 40% win rate or winners meaningfully larger than losers.
- Every trade matched a written setup: zero improvised entries in the final month.
Crossing the bridge without burning it
Passing the gates does not mean jumping to full size with real consequences. The bridge is stakes that matter but cannot wound: the smallest real exposure available, or structured low-cost pressure, while keeping every rule from the demo phase unchanged. Expect your metrics to dip on the crossing, a lower win rate for a few weeks, hesitation on entries, early exits. That dip is the real curriculum, the part demo could never deliver, and it is normal. The mistake is responding to it by resizing upward to make results feel meaningful again. Let the numbers stabilize back toward your demo baseline at small stakes first. Consistency, then size: never the reverse order.