Most traders treat chart replay as one thing: you run price forward and see what happens. That's accurate the way "going to the gym" is accurate. Technically true, but it doesn't say anything about whether you're doing bench press, swimming, or sitting on a bike reading a magazine.
What you're trying to improve determines how you should practice. The same chart replay tool produces fundamentally different skill outcomes depending on how you structure the session. I've spent a lot of time doing all three types I'm about to describe, and I'd have saved months if someone had explained the distinction earlier.
Type 1: setup recognition drills
This is the most common form of replay practice, and when done correctly, it builds genuine pattern recognition. When done lazily, it builds the illusion of pattern recognition.
The goal is simple: run price forward and, before each candle prints, name the setup category you're looking at. Not "I think it's going up." Something more specific. "This is a pullback to the 20 EMA in an uptrend on the 15-minute chart." Or "Price broke structure, retested the breakout level, and this candle is the confirmation bar."
The naming step is the training step. Without it, you're just watching. Your brain will still learn things from watching, but the learning is slow and less transferrable. With explicit naming, you're forcing your pattern recognition system to make a definite commitment before the outcome is revealed, which is exactly what live trading requires.
How to structure setup recognition drills
Pick one setup and practice it for at least a full session before moving to another. If you're practicing breakouts from consolidation, replay 30-50 examples, naming each breakout setup as you identify it. Was it genuine? Was the volume there? Did price follow through?
After each example, ask: did I call it correctly before the outcome? Not after, not with hindsight. The discipline of calling it before the reveal is the whole exercise.
Track your hit rate. If you're correctly identifying genuine breakouts 70% of the time before price confirms, your recognition is working. If you're at 40%, your identification criteria need work. The number tells you something. Random chart browsing tells you nothing.
Speed note: run setup recognition drills at a moderate pace, roughly one candle per 1-2 seconds. Fast enough to create some decision pressure, slow enough that you're actually thinking before the next candle appears.
Type 2: full trade simulation
This is what most people mean when they say they're doing "paper trading with replay." You step through a session exactly as you would in live trading: you identify setups, place simulated orders, manage the position, and exit. One trade at a time, with full attention on execution from entry to exit.
Full trade simulation is harder than it looks. Setup recognition is just the first step. Then you have to actually decide whether to take the trade. Then size it. Then monitor it, which means watching for your stop to be hit or your target to be reached while new candles continue to print and new information comes in. Then you have to exit properly, not just when price touches your target but with the kind of decisive click that live trading requires.
The part traders most often skip in full simulation is proper position management. They'll take a trade, get close to their target, then just note mentally that "I would have exited here" without actually going through the execution step. That shortcut is a problem. The execution decision under real-time pressure is a trained reflex, not an intellectual exercise. You have to practice actually clicking.
What full simulation builds
Sequentially, full simulation trains five distinct skills: setup identification, entry decision (do I take it or not), order placement mechanics, position monitoring under time pressure, and exit execution. These are all separate skills. Watching charts builds the first one. Full simulation builds all five.
The simulation also builds a realistic relationship with trade randomness. Some of your best setups will lose. Some mediocre setups will work. Over 50+ trades, the distribution of outcomes teaches you what a statistical edge actually feels like, which is very different from what a guaranteed winning system feels like. Most beginners don't have realistic expectations about how often correct setups result in losses even with a positive expectancy. Full simulation gives them that data.
Log every simulated trade as discussed in this guide. The log is how you turn the experience of full simulation into usable feedback.
Type 3: post-mortem replay
This is the most underused type, and in some ways the most educational. Post-mortem replay means going back to charts you've already traded, either in practice or live, and replaying them again specifically to find what you missed.
The goal isn't to feel bad about losing trades. It's to find the earlier signals you could have used, the exits you left too late, the entries that looked right in real time but had warning signs visible in hindsight. Post-mortem replay converts hindsight into foresight, because if you can identify a specific visible signal that preceded a repeated outcome, you give yourself something to look for earlier next time.
How post-mortem replay works in practice
Take a trade you recently lost. Go back to the chart at the start of that day or session. Replay it forward without knowing what you already know about the outcome. Pay attention to what the chart was actually showing in the lead-up to your entry. Were there warning signs you glossed over? Was the volume pattern off? Was market structure actually against you even though the setup looked clean?
Do the same for your winning trades. Find the ones where you exited early and then price continued in your direction. What was the chart showing at the point you exited? Were there continuation signals you could have read as a reason to hold?
The specific insight you're hunting for in post-mortem replay: patterns that consistently showed up before outcomes you didn't predict. Your losses that had a specific warning you repeatedly missed. Your winners that had continuation signals you repeatedly ignored. Once you find a repeating pattern in your own trade history, you have something actionable to practice in your next setup recognition session.
This is why keeping a trade journal and doing post-mortem replay are so tightly connected. The journal tells you which trades to go back and replay. The replay shows you what was actually there.
How to combine all three types
A well-structured practice week uses all three, not randomly but in a deliberate sequence.
Start each week with setup recognition drills. 20-30 minutes, one setup type, lots of reps. You're sharpening the pattern recognition edge before you practice deploying it.
Then move to full simulation sessions for the bulk of your practice time. Take actual trades with the setups you've been drilling. Log everything.
End the week with one post-mortem replay session. Go back through your simulation trades from the week, replay the ones that lost, look for the signals you missed. Then connect those signals back to your setup recognition drill criteria for next week.
That cycle of drill, simulate, review creates a feedback loop. Without the post-mortem, your practice is open-loop: you practice, you get outcomes, but you don't systematically extract what the outcomes are teaching you. The post-mortem closes the loop.
One drill most traders never try
Here's a specific exercise that sits between Type 1 and Type 2, useful for traders who find their recognition is good but their execution hesitates.
Set your replay to a slightly faster speed than comfortable. Identify setups in real time, but now add a rule: if you see a valid setup, you have to make an entry/no-entry decision within 5 seconds of identification. No deliberating. Call it, decide, either place the order or explicitly say "no trade, reason is X" and move on.
This drill specifically trains decision speed without losing selectivity. The trader who sees a setup and spends 45 seconds analyzing it is going to miss the entry in live markets. The trader who sees a setup and makes a snap decision without any real analysis is going to trade every impulse. The 5-second decision window trains the middle path: fast enough to execute in live markets, deliberate enough to have a real reason.
Start with one type
Open ChartMini TradeGame and decide before you click anything: which type of session is this? Setup recognition (I'm naming patterns before they resolve), full simulation (I'm trading as if live), or post-mortem (I'm reviewing a chart I've already seen)? Making that decision upfront changes how you approach everything that follows. Unstructured replay teaches unstructured thinking. Structured replay builds structured habits.
Common questions
How long should each type of session last? Setup recognition drills are effective in 20-30 minute blocks. Focus degrades fast with repetitive visual tasks. Full simulation sessions can run 60-90 minutes, though shorter is fine if you're logging every trade carefully. Post-mortem review is usually 30-45 minutes, enough to work through 3-5 trades in detail.
How often should I do each type? A rough guideline: setup recognition every practice day (it's quick and keeps the edge sharp), full simulation 3-4 times per week, post-mortem once per week or whenever you've accumulated 10+ trades to review.
Can I combine type 1 and type 2 in the same session? Yes, but keep them sequential. Run 20 minutes of setup recognition first without trading. Then switch modes and do full simulation. Mixing them simultaneously (identifying setups while also managing a position) is actually a good advanced drill once you're comfortable in both modes separately.
What if I don't have specific setups to drill yet? Start with the simplest pattern that relates to your intended trading style. For day traders, the opening range breakout is a good first setup recognition drill. For swing traders, daily close above a prior resistance level. Pick one, drill it to recognition fluency, then add the second.