Candle-Stepping vs Tick Replay: What Your Test Can't See

Two Different Experiences of the Same History

Candle-stepping is a slideshow: press next, a completed M15 bar appears — open, high, low, close, all at once. It’s what TradingView’s replay does, what most free tools do, and it’s genuinely useful: future data stays hidden, hindsight bias dies, and higher-timeframe decision practice works well.

Tick replay is a time machine: price ticks up, stalls, drops — the candle forms. You feel the stall before the breakout; you watch your stop survive a wick by half a pip, or not. Beyond realism, it carries information candle-stepping structurally lacks: the intrabar sequence and (when the data is bid/ask) the live spread at every moment.

What Each Test Can Honestly Conclude

Question your backtest asks	Candle-stepping	Tick replay
Did this daily-chart swing setup work?	✅ Yes	✅ Yes (overkill)
Was my 8-pip stop hit before the 12-pip target in that bar?	❌ A guess	✅ Known
Did my limit order actually fill?	❌ A guess (touch ≠ fill)	✅ Known
What did entering during news really cost?	❌ Invisible	✅ Visible (spread in data)
Does my candle-pattern entry survive watching it form?	❌ Untested	✅ Tested

The middle rows are the dangerous ones: candle-stepping doesn’t refuse to answer — it answers optimistically, and the test inherits the optimism (why live results then disappoint).

A Self-Test Worth Running Once

Take 20 trades from your strategy on candle-stepping replay, then replay the same dates at tick level and re-resolve every contested bar. If outcomes barely change, your strategy’s geometry is candle-safe — proceed with whichever workflow you prefer. If several trades flip, you’ve measured your own strategy’s granularity sensitivity, which is worth more than any general rule. (Free tick replay makes this experiment costless — StrategyTune streams bid/ask ticks in the browser; the tools comparison lists which tools offer which granularity, since several “replay backtesters” are candle-only.)

Speed Isn’t the Trade-off It Sounds Like

The intuition “ticks must be slower to replay” is backwards in practice: tick tools run at adjustable multipliers (up to 50,000× in StrategyTune’s case), so quiet stretches pass in seconds while the formation detail is available whenever you slow down. Candle-stepping’s real advantage is simplicity, not speed — and its real cost is every contested bar in your log.

Frequently Asked Questions

Is bar-by-bar replay useless, then?

Not at all — for swing strategies on H4-plus timeframes with wide stops, candle-stepping tests everything that matters, faster and simpler. It becomes inadequate exactly when trade geometry shrinks inside single bars: tight stops, intrabar limit fills, scalping, candle-formation entries.

What about tools that replay M1 candles to simulate intrabar movement?

A useful middle ground: stepping M1 candles inside an M15 trade resolves many — not all — contested fills, since ambiguity now only exists inside each minute. It's better than pure M15 stepping and worse than real ticks; for sub-5-pip stops the minute-level guesswork still bites.

Does tick replay show the spread too?

Only if the underlying data is bid/ask quotes rather than a single price series. Real quote ticks carry both sides, so the spread — and its widening at news and rollover — is visible and your simulated fills pay it. That's a data property, not a replay-mode property; check what the tool actually streams.

Which mode is better for building pattern-recognition skill?

Tick replay, clearly: patterns in textbooks are finished candles, but live trading shows you forming ones. Practicing on formation — watching a pin bar almost complete, then fail — trains the recognition you'll actually use. Candle-stepping trains recognition of completed structures only.