A strategy that reads the order book cannot be backtested the usual way, because historical bars do not contain the book. Market Replay is NinjaTrader 8's answer: it plays back recorded sessions with market depth included, as if they were happening live.

Why Market Replay and not the Strategy Analyzer

The Strategy Analyzer replays bars and, with Tick Replay, individual trades. It never replays resting orders. A Level 2 strategy such as Vantedge BookPilot therefore sees an empty book there and produces no trades. That is expected behaviour, not a fault. Market Replay data includes depth, so the strategy sees the same book updates it would see in real time.

Step 1: get the right data

In NinjaTrader 8, open Tools, then Historical Data, and use Load to download Market Replay data for the instrument and the days you want. The download includes depth. Pick several sessions, and include different kinds of day: a quiet rotational one, a trending one, one with a scheduled news release.

Keep a fixed set of test days. Replaying the same days after each change is what lets you compare settings fairly. BookPilot's signal engine runs on the timestamps of the market events, not the PC clock, so the same data produces the same signals each time.

Step 2: run it like a live session

Connect to the Playback connection, open an ES or NQ chart with Chart Trader, add BookPilot with your licence key and the playback simulation account, and arm it from the panel. If replay is paused, or no depth has arrived for 5 seconds, BookPilot refuses to arm and tells you why.

Let each session run continuously. Rewinding or jumping ahead resets the strategy's cooldowns and its hourly trade count, which makes the results look different from a real day. If you need to look at something again, finish the session first and replay it afterwards.

Step 3: keep a simple log

For each session, write down:

  • The number of trades, and how many were Pressure Break and how many Pulled Wall.
  • Win rate, average win and average loss.
  • How many entries were cancelled by the timeout or the run-away rule.
  • The reasons shown on the panel's Blocked line for signals that were not traded.

The Blocked line is the most useful diagnostic you have. It names the gate that stopped each signal: not armed, outside the trading window, against the bias, the hourly limit or Forecast Confirm.

Step 4: change one thing at a time

The BookPilot guide gives some patterns to look for:

  • Mostly blocked by Forecast Confirm or the bias: the filters are strict. Lower Confirm Threshold or raise Bias Threshold slightly.
  • Many trades losing straight away: tighten the scanners with a higher ratio or a longer Confirm.
  • Many cancelled entries: the limit offset is too deep or the Entry Timeout too short.
  • Many entries filled and then stopped quickly: the offset may be too deep, so you only get filled when wrong, or Confirm is too short.

Change a single setting, replay the full set of days, and compare with the log. Changing three things at once tells you nothing about which one mattered.

Step 5: respect the forecast file

The micro-forecast is a logistic model of the next 2-tick move. It starts with built-in weights, then fits its own weights per instrument once it has about 1,600 labelled samples, and refits every 200 new samples. The fitted model is saved to a file per instrument in the VantedgeAlpha folder under Documents\NinjaTrader 8 when the strategy stops.

That means early sessions behave differently from later ones. Give it a few sessions before judging Forecast Confirm. Keep the file when you only change trading settings. Delete it if you change the Flow Window, so the model starts again with the new input.

What replay cannot tell you

Replay fills are simulated. Your limit order's place in the queue, and whether it would really have filled, are estimates. Live latency differs too. Treat Market Replay as the place to understand the strategy's behaviour, then confirm it in real time on a simulation account before going further. The BookPilot user guide covers every setting, and the installation page covers setup and licence activation.

A good replay run is a reason to keep testing, not proof of future results; futures losses can be fast and large. Follow Market Replay with live SIM sessions before risking an evaluation, funded or live account.