Decision quality is the score of the choice given what you knew when you clicked—not the rupee result five candles later.
Score the decision, not only the P&L. P&L is a lagging narrator; process is leading indicator for skill work.
Process rubric (0–2)
2 — Full adherence
- Setup grade met pre-rules
- Cap and window respected
- Urge logged if relevant
- Exit per plan
1 — Partial
- Minor deviation with awareness logged
- Or skip with weak reason but no entry harm
0 — Violation
- B-grade chase
- Cap break
- Peek / hindsight
- Revenge entry
- Unlogged impulse
Outcome does not upgrade 0 to 2.
Decision quality ledger
| Session | Decisions | Avg score | Cap OK? | Notes |
|---|---|---|---|---|
| Mon | 8 | 1.6 | Y | midday boredom |
| Tue | 6 | 2.0 | Y | — |
| Wed | 9 | 0.9 | N | tilt |
Review lowest-score sessions first—even if green P&L.
Why decision quality precedes expectancy claims
Indian retail discourse loves win rate. Win rate without grade logging is noise. Decision quality tracks whether you can execute a stated edge—edge itself unproven until live with costs.
No fake stats required: this is scoring discipline, not backtest marketing.
Failure modes
- Outcome grading
- Grade inflation on wins
- Ignoring skips in average
- Changing rubric weekly
- Chasing 2.0 average with timidity — never entering A-grade
NSE practice notes
Tag product (cash intraday vs F&O vs delivery). Decision quality for delivery includes hold-through rules unlike intraday square-off.
Liquidity: illiquid sim teaches different mistake profile—tag separately.
Replay Trader philosophy
Replay Trader targets 300 practice decisions in 30 days (₹999) to raise decision quality under randomised NSE history—not to promise profits.
Score honesty beats screenshot heroics. Educational simulation only—not investment advice.
Weekly lowest-score session replay
Re-run worst session mentally candle by candle—what single rule would have changed most decisions?
Decision quality without timidity
Average 2.0 with zero entries means you skipped everything—also failure mode. Entries must exist when A-grade appears.
Cross-session quality trends
Graph avg score by week roughly—hand drawn fine.
Quality vs taxes humility
High decision quality does not mean after-cost profitability.
Rubric versioning
If rubric changes, reset comparison windows—do not blend scores across versions.
Score distribution chart
Hand-draw histogram of 0/1/2 counts weekly.
Zero-score triage
All zeros reviewed before any twos celebrated.
Decision quality buddy
Swap anonymized score summaries.
Quality vs quantity month
Some weeks prioritize avg score over decision count—declare upfront.
Rubric poster
Laminate 0-2 definitions near screen.
Extended scored session example
| # | Action | Score | Why |
|---|---|---|---|
| 1 | Skip | 2 | No volume |
| 2 | Entry | 2 | A-grade |
| 3 | Exit stop | 2 | Valid stop |
| 4 | Skip | 2 | Post-loss pause |
| 5 | Entry | 0 | B-grade chase |
| 6 | Session stop | 2 | Cap after rule break review |
Session avg 1.67 with explicit 0 flagged for review—not hidden because win followed entry 5.
Quality-first month
Score every decision 0–2 for 30 days. Review zeros before celebrating twos. Decision quality work precedes any live expectancy conversation—always. Educational practice; not performance marketing.
Extended rubric examples (borderline cases)
Partial 1: entry A-grade but slightly late—aware, logged → 1 not 0.
Partial 1: skip with weak reason but no harm → 1.
Zero: B-grade known at click → 0 always.
Two: full adherence including exit.
Borderline cases debated in weekly review sharpen rubric faster than ignoring them.
Quality month beats win-rate month for readiness honesty.
Quality streak rule
Three sessions avg process ≥1.8 → maintain rules unchanged. Below 1.5 → simplify rules next week (fewer setups, tighter cap), not new indicators.
Quality streaks guide rule stability; win streaks mislead.
Zero-score triage order
Review all score-0 decisions before any score-2 celebration. Chronological order within zeros: cap breaks first, peek second, B-grade third. Triage order prevents lucky wins from hiding rule debt.
Quality metrics age better than win-rate screenshots in your journal archive.
Rubric poster update policy
Rubric changes allowed only on Sunday with version bump v1→v2 noted in journal. Midweek rubric edits are tilt tells. Poster on wall beats rubric in head during fast tape.
Weekly quality report template
Sessions: __ | Decisions: __ | Avg process: __ | Zeros: __ | Cap breaks: __ | Top contradiction: __
One page weekly. Report goes in binder front. Quality trend visible across months when win-rate screenshots fade.
Never average process scores on days you invalidated for peek. Invalid days exist to teach honesty; including them in averages rewards cheating statistically.
Zeros reviewed first, always.
Score 0 photo rule
If you take zero score session, photograph one-page contradiction note and file in binder before next session. Photo rule adds friction to hiding zeros—especially after green sim day tempts you to skip review.
Related reading
Replay Trader — randomised historical NSE practice, hidden dates, candle-by-candle reveal — is built for decision reps under uncertainty, not tip culture.
Practise today. Trade tomorrow.
Educational content only. Not investment advice. Not a recommendation to buy or sell any security. Practice and simulated results do not guarantee live results.