A Replay Drill: 50 Setups in an Hour With Fixed Rules

[FXS Prop Desk]

Replay tools let you practise on historical data without risk. A structured drill gets more out of them than casual replay.

The drill:

1. Choose one setup with fixed rules.
2. Pick a random historical date.
3. Replay at speed, taking every valid setup.
4. Record entry, stop, target and result for each.
5. Aim for 50 setups in an hour.

What it trains:

• Recognising the setup quickly.
• Following rules without hesitation.
• Seeing the real win rate and distribution.

Tools include TradingView Bar Replay, MT5 visual mode and dedicated replay software.

After the drill, comparing the replay statistics with your live statistics for the same setup is revealing. Large differences usually point to execution or emotional factors rather than to the setup itself.

Do you use replay for practice, and how do you structure it?

[Fairman]

Pick a random historical date.

One thing that can quietly ruin the drill is seeing the chart before the replay starts. If you scroll to the date and the bigger picture is on screen, even for a second, the higher timeframe trend that followed is in your head, and the replay win rate comes out higher than live. Picking the date with a random number generator and jumping straight into replay mode with the right side hidden avoids most of it. It also helps to log setups you hesitated on or skipped, not just the ones taken. If the replay hit rate is good but there are lots of skipped valid setups, that's a recognition speed issue the drill can work on directly.

[Shadow Trader]

Blind date with an RNG is the part I stole immediately after Fairman's note. I used to scroll to a week I vaguely remembered as interesting, which quietly loaded the outcome into my head before replay started, and the hit rate looked better than live for months.

Logging skipped valids changed the drill more than logging the takes. When the replay card showed a solid win rate but twelve hesitated setups in an hour, the gap was recognition speed, not the rules themselves.

The replay versus live gap for me is almost always emotion. Same setup, same rules, yet live I cut winners earlier and move stops more, which the drill never trains unless I force a friction rule like writing the exit reason before I click.

I aim for fifty in an hour only on simple setups. Anything with a multi candle filter slows the pace, so I drop the target to thirty and keep the quality of the log higher.

Do you add any live like friction to the replay, such as a delay before entry is allowed, or keep it pure recognition practice? Pure recognition feels too clean compared with live.

[Shadow Trader]

FXS Prop Desk, one pattern showed up once I graded every replay setup for rule errors instead of only for result. Over ten drills of 50, the errors were not spread evenly through the hour, they piled up at the end.

Splitting each drill into five blocks of 10 setups, my average rule errors went 1, 1, 2, 3 and 4. That is 11 errors in 50, with 7 of them in the last 20 setups.

img_13633_r2.png (42.97 KiB) Viewed 6 times

The illustration shows those block averages. The last 20 minutes were training me to be sloppy at speed, which is the opposite of what the drill is for.

So I now run two blocks of 25 with a five minute break, and errors in the second block dropped to about 2. Fairman's skipped valids column fits here too, because my skips also cluster late.

Cost is the other thing bar replay leaves out. I subtract a fixed 0.8 pip per EURUSD trade from replay results, otherwise a scalping setup with a 5 pip target looks about 16% better than it can be live.

Does your 50 count include setups judged invalid, or only valid ones taken? That changes the pace a lot.