Paper trading for day traders: what simulators get wrong about fills
Paper trading trains decisions and mechanics, but simulated fills are usually kinder than real ones. Learn what simulators miss, how to add a slippage penalty, and how to move from sim to small live size.
Start in plain English
Paper trading, also called sim trading, means placing simulated orders against real or replayed market data without real money. For a beginner scalper it is the cheapest place to practise a plan, hotkeys and the habit of logging every trade. Its weakness is specific: simulated fills are usually better than real ones, and in scalping, where the edge is a few cents, that difference can turn a losing plan into a winning-looking log.
The core idea: a sim tests your decisions; it does not test your fills. Use it to practise the plan and its mechanics, then correct the results for what the simulator could not model before you believe them.
The order types themselves belong to the Market vs. limit orders lesson in US Day Trading Fundamentals. This lesson covers how a simulator treats them.
Two paths from the same start
All prices, sizes and costs are Illustrative. Gannet Rail is fictional.
The setup at 10:31 ET. Gannet Rail has pulled back to $3.20, a level it held twice this morning. A trader practising a passive entry plans to buy on the bid rather than pay the ask.
Gannet Rail (fictional), Level 2 at 10:31, ET (Illustrative)
| Side | Price | Size | Source |
|---|---|---|---|
| bid | 3.20 | 8,000 | Venue A |
| bid | 3.19 | 5,000 | Venue B |
| bid | 3.18 | 6,000 | Venue C |
| ask | 3.21 | 4,000 | Venue B |
| ask | 3.22 | 7,000 | Venue A |
| ask | 3.23 | 5,000 | Venue C |
8,000 shares already bid at $3.20, ahead of any new order at that price. Displayed size only.
The plan. Buy limit 625 shares at $3.20, joining the bid. Stop $3.12; target $3.36 (2R); time stop 10 minutes after the fill. Planned risk $0.08 per share, $50. Costs $0.005 per share round trip (Illustrative).
Many simulators fill a resting limit order as soon as the price touches it. A real order at $3.20 joins the back of a queue: it fills only after the 8,000 shares ahead of it trade or cancel, or when a seller trades through the level.
Path AWhat the simulator records
At 10:32 the price touches $3.20 and the sim fills all 625 shares. The stock bounces and the $3.36 target fills at 10:36.
Path BA live fill: the queue reaches you when sellers push through
A seller sends 12,000 shares into the $3.20 bid. The 8,000 ahead of you fill, then yours, and the next print is $3.19. Price keeps falling; the $3.12 stop fills at $3.10.
Path CA live non-fill: price touches and leaves
Price touches $3.20 twice; about 3,000 shares trade there and the rest of the queue stays ahead of you. The stock bounces to $3.36 without you.
Review note: "Passive entry at $3.20; plan followed; live outcome B or C, sim outcome A or B." The asymmetry is the lesson: a passive order tends to fill when price moves through it (Path B) and to miss when price bounces off it (Path C). A simulator that fills on the touch records Path A in the second case, so it overstates Path A. The rest of this lesson lists what simulators miss, shows how to correct a sim log, and sets out how to move to small live size.
What simulators model poorly
| Real-market feature | What many simulators do | Effect on the sim log |
|---|---|---|
| Queue position for limit orders | Fill on the touch | More Path A, fewer missed trades |
| Partial fills | Fill the whole order | Larger winners on thin stocks |
| Your own size moving the price | Ignore it | Better average prices on larger orders |
| Stop fills in thin books | Fill at or near the stop price | Smaller Path B losses |
| Halts and LULD pauses | Keep filling, or skip the halt in a replay | Missing the worst outcomes |
| Hard-to-borrow stocks and locates | Allow shorting anything | Short trades that could not have been placed |
| Fees and rebates | Omit or simplify them | Results that look gross, not net |
| Latency and order handling | Instant | Fills at prices already gone |
Settings differ widely between platforms and are worth checking one by one. Some simulators let you choose queue modelling or a fill delay; turn them on where they exist. For short-side practice, see FennEdge's article on stock locates.
Adding a slippage penalty
The simplest correction is to make every simulated fill worse by a fixed amount, chosen before the block starts:
- Entries: 1 cent worse than the sim fill (more on wide or thin stocks).
- Stops: 2 cents or more worse than the sim fill.
- Passive entries: count a share of touches as non-fills, for example every second touch where fewer shares traded at the level than were displayed ahead of you.
- Fees: your broker's real per-share costs; see ECN fees.
Penalties are blunt. They are still better than none, and they can be checked later against your own small live trades. Slippage shows how to measure the live side.
Observe vs. infer
At 10:31 in Gannet Rail (fictional; all figures Illustrative):
| What the screen shows (observed) | What a trader might infer | Why it is not established |
|---|---|---|
| 8,000 shares bid at $3.20 | "Strong support; my order is safe there" | Displayed orders can cancel; your order is at the back of the queue |
| The level held twice this morning | "It holds a third time" | Earlier holds do not set the next outcome |
| The sim filled on the touch | "That is the fill I would get live" | Live fills depend on the queue and on who trades through |
| A +1.94R sim result | "The plan works" | One result, with an optimistic fill model |
Odds box
The plan, the decisions and the price paths are identical. Only the fill model changed, and the sign flipped. The non-fill share is a teaching assumption; a measured version comes from comparing your own passive orders in sim and in small live trading.
What shifts the odds
These are hypotheses to record, not findings:
- Entry type. Passive (joining the bid) vs. marketable (paying the ask). Passive entries are most affected by the queue.
- Displayed size ahead of you against volume traded at the level.
- Stock liquidity. Thin books widen the gap between sim and live fills; see Is it liquid enough to scalp?.
- Order size. Larger orders move further from the sim's assumptions.
- Halts during the practice block, and whether the replay showed them.
Cost of being wrong
Gannet Rail plan, 625 shares (Illustrative):
| Outcome | Exit | R after costs | Dollars after costs |
|---|---|---|---|
| Sim Path A: fill on the touch, target | $3.36 | +1.94R | +$96.88 |
| Live Path B: filled as sellers push through, stop | $3.10 | −1.31R | −$65.63 |
| Live Path C: no fill | — | 0R | $0 |
| Live partial fill: 200 of 625 shares, then target | $3.36 | +0.62R | +$31.00 |
| Live halt after the fill, reopening at $2.95 | $2.95 | −3.19R | −$159.38 |
The partial-fill row shows the other side of passive entries: when they work, they may work for fewer shares. The cost of trusting an unadjusted sim log is moving real money onto a plan whose positive result came from the fill model.
Deliberate practice blocks
Practice works better in blocks with a single focus than as open-ended sim sessions. Ericsson and colleagues (1993) described deliberate practice as focused work on a specific skill with feedback, rather than repetition alone.
- One plan per block. 20 trades or more, definition fixed.
- One skill in focus. Entry timing, stop execution, hotkey correctness or logging.
- Feedback after each block, not after each trade: Path shares, average R after penalties, plan adherence.
- Hotkey and order checks before the first trade; the DAS course lesson Testing and recovery covers this for DAS users, and Hotkeys for day trading covers the principles.
Moving from sim to small live size
- Start small. A fraction of the planned dollar risk, for example a quarter, keeps the lesson cheap while real queues, fees and nerves arrive (Illustrative starting point, not a rule).
- Compare like with like. Run the same plan live that you ran in sim, and compare live results with the penalty-adjusted sim log, not the raw one.
- Expect worse fills, not just more nerves. If live results are much worse than the adjusted sim, check fills first, then plan adherence.
- Real trades go in the Journal. FennEdge's Trading Journal syncs real fills from supported brokers (as of 2026-10-05); practice trades stay in the practice log.
How many trades before this means anything
Comparing sim and live results needs counts on both sides. If 20 live passive orders fill 12 times (60%), the 95% interval for the true fill share runs from about 39% to 78% (Wilson interval, computed on these hypothetical counts). That is enough to see that many touches do not fill, not enough to set a precise penalty. How many trades before your numbers mean anything explains the method.
What this lesson does not claim
- It does not rank or name simulators or brokers. Fill models vary; check your platform's settings.
- The non-fill share, penalties and splits are Illustrative.
- Practice in a simulator does not show that a plan is profitable live.
- Gannet Rail and its quotes are fictional.
Go deeper
- FennEdge first.
- The Trading Journal holds real broker-synced trades only, not simulated or practice entries. Use it once you trade live.
- Level Up (Plus and above) is a private challenge scored on your Journal net P&L, with an optional daily max loss per level. It does not track tags or entry timing, so keep process goals such as days with every trade logged in your own log.
- In-Play Days shows 1-minute charts of past active days for replay practice. Without Max, including as a guest, you can open each ticker's most recent active day in full; earlier days and the two-year history need Max (as of 2026-10-05). Replaying many past days needs Max.
- The DAS course lesson Testing and recovery covers testing hotkeys and orders in a simulator.
- Replay vs. live sim. Replay tools let you practise more setups per hour, but you know the day already happened; a live sim keeps the uncertainty.
- Day-trader outcomes. Barber, Lee, Liu and Odean (2014) studied day traders in Taiwan and found that only a very small share earned predictably positive returns after fees. The SEC's investor guide on day trading describes the risks.
- Related lessons: Slippage, Marketable limit orders, A trading plan as a hypothesis, Reviewing scalps, Capstone: a measured 50-trade sim block.
Exercises
Five exercises follow this lesson: three concept checks, one calculation (apply a slippage penalty to a sim log) and one practice task. Exercises, answers and progress are available after signing in.
Practice task: a corrected 20-trade sim block. Run 20 trades of one written plan in a simulator and log them with every column of the FennEdge practice log: download the practice log (CSV), with a column guide. Write your penalties before the first trade, record the raw sim fill and the penalty separately (slippage_penalty_usd), and mark passive entries where fewer shares traded at your price than were displayed ahead of you. Compute expectancy with and without the penalties. Keep the log separate from the Trading Journal, which holds only real broker-synced trades.
Primary sources
- Ericsson, K. A., Krampe, R. T. and Tesch-Römer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review 100(3).
- Barber, B. M., Lee, Y.-T., Liu, Y.-J. and Odean, T. (2014). The cross-section of speculator skill: Evidence from day trading. Journal of Financial Markets 18.
- SEC. Day Trading: Your Dollars at Risk.
Sources checked 2026-10-05. Gannet Rail and all quotes, fills, splits, penalties and costs in this lesson are fictional and Illustrative. This lesson is education, not a recommendation to buy or sell any security. Past patterns do not indicate future results.