A footprint strategy lives on bid/ask detail inside each bar. Run it in NinjaTrader 8's Strategy Analyzer without that detail and you get nothing, or worse, results that look plausible but are wrong. This guide covers how to set up a meaningful backtest with Tick Replay, using Vantedge FlowPilot as the example, and how to avoid fooling yourself with the results.

Why Tick Replay is non-negotiable

Historical bars normally store open, high, low, close and volume. A footprint needs much more: every trade, at its price, marked as a buy at the ask or a sell at the bid. Tick Replay makes NinjaTrader feed a strategy historical ticks one by one, as if they were arriving live.

FlowPilot sets Calculate to On each tick itself, because it checks for stacked imbalances on every trade, not just at the bar close. Without Tick Replay, the footprint stays empty and there are no trades. That is the most common reason for a blank backtest.

Setting up the test

  • If the option is hidden, enable Show Tick Replay under Tools > Options > Market Data.
  • Make sure you have tick-based historical data for the instrument, for example ES or NQ, over the whole test period.
  • In the Strategy Analyzer, choose the strategy, tick Tick Replay on the data series, and use the same bar type and size as your live chart.
  • Set realistic commission and slippage. Market entries, which FlowPilot uses, pay the spread and more in fast markets.
  • Start from the shipped defaults (the Balanced profile) before changing anything.

Tick Replay is heavy. Expect a multi-month test on NQ to take noticeably longer than a normal bar backtest.

What the backtest does not simulate

Knowing this prevents nasty surprises later.

  • Daily gates. FlowPilot's Daily Profit Goal, Daily Max Loss and Cooldown After Loss count real-time trades only and are not applied in backtests. Your backtest days have no lock, so check how those days would have looked with one.
  • The panel. Arm buttons, trend filters and mode switches you would click live do not exist in a backtest. Only the starting settings apply, so test in Auto, or account for the fact that Follow and Fade need manual arming.
  • Queue position and latency. Fills are modelled, not real.

An optimisation routine that holds up

One group at a time

Start with the signal: Imbalance Ratio and Stacked Requirement. Once those are settled, move to the regime settings, then the bracket. Changing everything at once makes it impossible to know what helped.

Enough data

Optimise over at least 2 to 3 months, then check the chosen settings on a later period they were not tuned on. That walk-forward check is the closest a backtest gets to telling you what happens next.

Plateaus, not peaks

If Imbalance Ratio 3.0 does well but 2.8 and 3.2 do badly, 3.0 is probably luck. Prefer a broad area of reasonable results.

The right metrics

Look at the number of trades, profit factor, maximum drawdown and average trade, not just net profit. A setting that only works with very few trades is usually overfitted.

Reading the trades, not just the summary

Open the trade list and pick a dozen entries. Load the Vantedge Footprint chart template on the same bars and check each one: was there a clear stack, did delta agree, and did the mode choose the direction you expected? This catches setup mistakes, such as the wrong bar type, that summary numbers hide.

After the backtest

A backtest is the first filter, not the last. Run the chosen settings on Sim101 or in Market Replay for several sessions to confirm the daily gates, cooldown and panel behave as you expect. The FlowPilot user guide covers every setting and the full optimisation checklist.

This is also one of the clearest differences between the two Vantedge strategies. FlowPilot can be backtested this way; BookPilot cannot, because historical bars carry no order book. The comparison page sets that out, and the FlowPilot product page lists what the plan includes.

Good backtest numbers are a reason to keep testing, not a forecast, and live futures trading can lose more than any test showed. Move from backtest to SIM before an evaluation, funded or live account.