Model stats without distortion

4 min · 8 steps

1. Step 1 · Two buckets

   Put rule breaks in a separate bucket.

   A mistake trade isn't about the model — it's about discipline. Mix them and you won't learn about either.

   SOL 15m, Aug 23, 2025: an entry without MSS is a rule break. Its −1R doesn't go into the model stats.

   zonesweepentry without MSS

2. Step 2 · The whole sample

   The whole sample: the model is a loser.

   “Sweep + MSS + retest” by the book, no filters: 88 trades, total −13.3R, PF 0.79.

   trades

   88

   to target

   28%

   total

   −13.3R

   The sample: every setup picked by the model's rules. The stop sits right beyond the sweep, the target at the edge of the range. Your version of the model may differ, so track your own.

   cumulative R · 88 trades · 15m/1h/4h

3. Step 3 · The best streak

   The best 10 in a row don't count as statistics.

   The best streak of 10 trades in a row: +9.2R, PF 4.05. By this streak the model “works”, across the whole sample it doesn't.

   best streak · 10

   +9.2R

   PF 4.05

   to target 70%

   whole sample · 88

   −13.3R

   PF 0.79

   to target 28%

   best streak · 10 trades · 15m/1h/4h

4. Step 4 · Average R and PF

   Average R and PF tell you more about a model than winrate.

   The model's average win is +2.0R — which means you need at least 33% targets hit. It was 28%: an average R of −0.15R per trade.

   average win

   +2.0R

   targets needed

   33%

   average R

   −0.15R

5. Step 5 · Sample

   Draw a conclusion only from 30+ trades and no cherry-picked stretch.

   Count every model trade in sequence, across the whole history. A stretch that starts “from a good spot” is curve-fitting, not a result.

   Journal stats

   every trade in the journal

   the model only — no rule breaks

   the whole history, no cherry-picked stretch

   at least 30 trades

   conclusion

6. Step 6 · Find it yourself

   Model or rule break?

   SOL 15m, Aug 23, 2025: the trade closed on the stop. Tap the candle where it strayed from the rules.

   Practice on the chart

   This is a Find it yourself task: tap a candle or a level, and the lesson tells you right away if you got it.

   Create an accountThis step unlocks after sign-up. Your progress is saved.

   zonesweepentry without MSS

7. Step 7 · How it shows up for you

   Practice on the chart

   This is the How it shows up for you step: the lesson finds trades on this topic in your journal.

   Create an accountThis step unlocks after sign-up. Your progress is saved.

8. Step 8 · Recap

   Count the model and rule breaks separately.

   1

   model separate

   ›

   2

   whole sample

   ›

   3

   at least 30 trades

   Stats: setup slices

   conclusion

   I count the model separately from rule breaks, across the whole sample, with at least 30 trades.

   Stats slices calculate this from your own journal. You draw the conclusion — it isn't made automatically.