| Trend | What it likely means |
|---|---|
| Improving (mistake % down) | A smaller share of your trades involved a flagged mistake than last period — discipline is trending the right way. |
| Worsening (mistake % up) | A larger share of your trades involved a flagged mistake — worth a look at what changed (new setup, fatigue, revenge trading, overtrading a specific symbol, etc). Check the monthly table below for exactly when it started. |
| Flat | Not enough movement to draw a conclusion, especially on a small trade count. |
| Month | Trades | Mistakes | Mistake % |
|---|---|---|---|
| {{ row.label }} | {{ row.total }} | {{ row.mistakes }} | {% if row.pct is not None %}{{ row.pct|floatformat:1 }}%{% else %}—{% endif %} |
| No trades match the current filters. | |||
| Trend | What it likely means |
|---|---|
| Improving (cost smaller/less negative) | Mistakes cost you less this period — either you made fewer of them, or the ones you made were smaller (better damage control, e.g. cutting losses faster). |
| Worsening (cost bigger/more negative) | Mistakes cost you more this period — could be more frequent mistakes, or the same frequency with bigger losses each time. Sort the table below by Total to see which specific mistake is driving it. |
| Flat | Not enough movement to draw a conclusion, especially on a small trade count. |
A trade tagged with more than one mistake counts toward each one's total below, so the per-mistake rows can add up to more than the total above (which counts each mistake trade once).
RISING = happened more than once this period, also happened last period, and is happening more often now — a recurring pattern, not a one-off. BIGGEST IMPACT = one of the 3 costliest mistakes in the table below.
| Mistake | Trades | Total ({{ pnl_unit }}) | Avg ({{ pnl_unit }}) |
|---|---|---|---|
| {{ row.tag }} {% if row.rising %}RISING{% endif %} {% if row.high_impact %}BIGGEST IMPACT{% endif %} | {{ row.count }} | {% if row.total is not None %}{% if pnl_unit == "$" %}${{ row.total|floatformat:2 }}{% else %}{{ row.total|floatformat:2 }}R{% endif %}{% else %}—{% endif %} | {% if row.avg is not None %}{% if pnl_unit == "$" %}${{ row.avg|floatformat:2 }}{% else %}{{ row.avg|floatformat:2 }}R{% endif %}{% else %}—{% endif %} |
| No mistake-tagged trades match the current filters. | |||
Real = every trade. No mistakes = same trades restricted to no mistake tag (or only ones currently excluded from the mistake count, e.g. NONE) — "how it would have looked without the mistakes."
| Real | No mistakes | What it likely means |
|---|---|---|
| Worsening | Worsening | The decline isn't just about mistakes — your underlying setups/edge are losing too. Cutting mistakes alone won't fix this; the strategy itself needs a look. |
| Worsening | Improving or flat | Your clean trades are holding up fine or getting better, but mistakes are dragging the real numbers down. This is a discipline problem, not a strategy problem — fixing execution should show up directly in the real results. |
| Improving | Worsening | Overall results look better, but the mistake-free trades are actually getting worse — the improvement may be masking a fading edge (e.g. fewer mistakes this period, or a small sample skewing the "no mistakes" figure). Worth digging into before assuming things are genuinely better. |
| Improving | Improving | Genuine improvement — both discipline and edge are moving the right way. |
| Flat | anything | Not enough movement to draw a conclusion — likely normal noise, especially on a small trade count. |
Total R reflects whatever's currently selected via the Date filter above (all-time if none is set).
| Setup | Trades | Win rate | Total R | Avg R | Avg Max Move (R) | Avg Max Pullback (R) |
|---|---|---|---|---|---|---|
| {{ row.tag }} | {{ row.count }} | {% if row.win_rate is not None %}{{ row.win_rate|floatformat:1 }}%{% else %}—{% endif %} | {% if row.total_r is not None %}{{ row.total_r|floatformat:2 }}R{% else %}—{% endif %} | {% if row.avg_r is not None %}{{ row.avg_r|floatformat:2 }}R{% else %}—{% endif %} | {% if row.avg_max_move is not None %}{{ row.avg_max_move|floatformat:2 }}R{% else %}—{% endif %} | {% if row.avg_max_pullback is not None %}{{ row.avg_max_pullback|floatformat:2 }}R{% else %}—{% endif %} |
| No setup-tagged trades match the current filters. | ||||||
Trades reflect whatever's currently selected via the Date filter above (all-time if none is set). NONE (and any other tag currently excluded above) doesn't count as a mistake, and trades with no mistake tag at all are never counted.
| Situation | What it likely means |
|---|---|
| High mistake % + many trades | A real, actionable pattern — this setup consistently correlates with mistakes. Worth asking whether the setup itself invites bad behavior (overtrading it, chasing entries, ignoring rules because it "feels" high-probability) or whether execution specifically around this setup needs tightening. |
| High mistake % + few trades | Likely noise, not a pattern — a single mistake on a setup you've only taken once or twice will show 100%. Don't over-react until the trade count builds up. |
| Low mistake % + many trades | A setup you execute cleanly and consistently — whatever discipline you have here, it's working. |
| Low trade count overall | Not enough data yet to draw a conclusion either way. |
Priority to address: setups with both a high trade count and a high mistake % — those are costing the most in aggregate, not just individually.
| Setup | Trades | Mistakes | Mistake % | Cost R | Cost $ |
|---|---|---|---|---|---|
| {{ row.tag }} | {{ row.total }} | {{ row.mistake_count }} | {% if row.pct is not None %}{{ row.pct|floatformat:1 }}%{% else %}—{% endif %} | {% if row.mistake_total_r is not None %}{{ row.mistake_total_r|floatformat:2 }}R{% else %}—{% endif %} | {% if row.mistake_total_pnl is not None %}${{ row.mistake_total_pnl|floatformat:2 }}{% else %}—{% endif %} |
| No setup-tagged trades match the current filters. | |||||
"Hold time" is how long the trade was held (exit minus entry), not time of day. Buckets reflect whatever's currently selected via the filters above (all-time if no Date filter is set).
By default every row is a raw 5-minute bucket. Use Group nearby buckets above to merge adjacent buckets whose win rates are close together — a value around {{ default_duration_group_threshold }} percentage points is a reasonable starting point: tight enough to keep genuinely different durations apart, loose enough to smooth out noise from small buckets (and it can pull a too-small bucket over the {{ min_trades_for_optimal }}-trade "optimal" threshold by combining it with a similar neighbor). Too high a value risks blending durations that actually behave differently. Leave it blank to see every raw bucket.
Read each row as your own historical odds: if a bucket shows a 55% win rate over 200 trades, that's roughly what you can expect the next time you hold a trade that long — it's not a coincidence, it's your track record. If win rate (and Total R) climb the longer you hold, that's real evidence from your own data that staying in the trade works for you — a reason to trust the process in the moment, not second-guess it.
Success rate here is win rate (wins ÷ closed trades) for trades held that long. "Optimal" below is the bucket with the highest win rate among buckets with at least {{ min_trades_for_optimal }} trades — a bucket with 1-2 trades can show 100% purely by luck, so it's deliberately excluded from being called optimal until there's enough of a sample to trust.
| Situation | What it likely means |
|---|---|
| High win rate + low avg R | You're right often in this bucket but not making much when you are — small wins, or winners cut short. Win rate alone doesn't mean the duration is actually profitable; check Avg R alongside it. |
| Lower win rate + high avg R | Fewer winners, but they're big enough to matter — a duration you might be exiting too early on the losers rather than one to avoid entirely. |
| Very short holds (under 5 min) underperforming | Common pattern: fast in-and-out trades often correlate with impulsive entries rather than planned ones. Worth cross-checking against the Mistakes reports for the same period. |
| A bucket with very few trades | Not enough data to trust the number either way — that's why it's excluded from "optimal" below regardless of how good or bad it looks. Try raising "Group nearby buckets" above to combine it with a similar neighbor. |
No duration bucket has {{ min_trades_for_optimal }}+ trades yet, so no bucket is called out as optimal.
{% endif %}| Hold time | Trades | Win rate | Total R | Avg R |
|---|---|---|---|---|
| {{ row.label }} {% if row.is_optimal %}OPTIMAL{% endif %} {% if not row.eligible %}(too few trades){% endif %} | {{ row.count }} | {% if row.win_rate is not None %}{{ row.win_rate|floatformat:1 }}%{% else %}—{% endif %} | {% if row.total_r is not None %}{{ row.total_r|floatformat:2 }}R{% else %}—{% endif %} | {% if row.avg_r is not None %}{{ row.avg_r|floatformat:2 }}R{% else %}—{% endif %} |
| No closed trades match the current filters. | ||||
For each setup, the hold-time bucket with the best win rate among that setup's own trades (same {{ min_trades_for_optimal }}-trade minimum). Setups with too few trades in every bucket show "Not enough data" rather than a guess.
| Setup | Trades | Optimal hold time | Win rate | Trades (bucket) | Avg R | ||||
|---|---|---|---|---|---|---|---|---|---|
| {{ row.tag }} | {{ row.total_trades }} | {% if row.optimal %}{{ row.optimal.label }} | {{ row.optimal.win_rate|floatformat:1 }}% | {{ row.optimal.count }} | {% if row.optimal.avg_r is not None %}{{ row.optimal.avg_r|floatformat:2 }}R{% else %}—{% endif %} | {% else %}Not enough data | {% endif %}|||
| No setup-tagged trades match the current filters. | |||||||||
Win rate for every setup against every hold-time bucket at once. Always the raw 5-minute buckets, regardless of the "Group nearby buckets" control above — grouping is adaptive per setup, so different setups would merge different buckets and the columns wouldn't line up. Bold green is that row's best bucket among ones with {{ min_trades_for_optimal }}+ trades; a dash means no trades in that cell.
| Setup | {% for col in duration_matrix_columns %}{{ col }} | {% endfor %}
|---|---|
| {{ row.tag }} ({{ row.total_trades }}) | {% for cell in row.cells %}{% if cell and cell.win_rate is not None %} {{ cell.win_rate|floatformat:0 }}% ({{ cell.count }}) {% else %} — {% endif %} | {% endfor %}
| No setup-tagged trades match the current filters. |
When the trade was entered, in fixed 15-minute windows across the day (exchange local time) — not hold time, see the Hold time tab for that. Only windows that actually contain a trade are shown. Buckets reflect whatever's currently selected via the filters above (all-time if no Date filter is set).
Best win rate is the time window with the highest win rate among windows with at least {{ min_trades_for_optimal }} trades. Best result is the window with the highest total {{ pnl_unit }} among the same eligible windows — often the same window as best win rate, but not always: a window can win often in small increments (high win rate, modest total) while another wins less often but bigger (lower win rate, bigger total). A window can earn both badges at once.
| Time | Trades | Win rate | Total ({{ pnl_unit }}) | Avg ({{ pnl_unit }}) | ||
|---|---|---|---|---|---|---|
| {{ row.label }} {% if row.is_best_winrate %}BEST WIN RATE{% endif %} {% if row.is_best_result %}BEST RESULT{% endif %} {% if not row.eligible %}(too few trades){% endif %} | {{ row.count }} | {% if row.win_rate is not None %}{{ row.win_rate|floatformat:1 }}%{% else %}—{% endif %} | {% if pnl_unit == "$" %}{% if row.total_pnl is not None %}${{ row.total_pnl|floatformat:2 }}{% else %}—{% endif %} | {% if row.avg_pnl is not None %}${{ row.avg_pnl|floatformat:2 }}{% else %}—{% endif %} | {% else %}{% if row.total_r is not None %}{{ row.total_r|floatformat:2 }}R{% else %}—{% endif %} | {% if row.avg_r is not None %}{{ row.avg_r|floatformat:2 }}R{% else %}—{% endif %} | {% endif %}
| No trades match the current filters. | ||||||
For each setup, its own best-win-rate time window and best-result time window (same {{ min_trades_for_optimal }}-trade minimum) — often the same window, not always. Setups with too few trades in every window show "Not enough data".
| Setup | Trades | Best win-rate time | Win rate | Best result time | Total ({{ pnl_unit }}) | ||||
|---|---|---|---|---|---|---|---|---|---|
| {{ row.tag }} | {{ row.total_trades }} | {% if row.best_winrate %}{{ row.best_winrate.label }} ({{ row.best_winrate.count }}) | {{ row.best_winrate.win_rate|floatformat:1 }}% | {% else %}Not enough data | {% endif %} {% if row.best_result %}{{ row.best_result.label }} ({{ row.best_result.count }}) | {% if pnl_unit == "$" %}${{ row.best_result.result|floatformat:2 }}{% else %}{{ row.best_result.result|floatformat:2 }}R{% endif %} | {% else %}Not enough data | {% endif %}||
| No setup-tagged trades match the current filters. | |||||||||
⚠ Not financial advice. This report is statistical evidence from your own past trades only — not a guarantee or prediction of future results. Use it as one input alongside your own judgment, not a substitute for it.
"Capture ratio" is realized R divided by the trade's own best favorable move (in R) before exit — how much of the move that was actually on the table you kept. Trades with no favorable move at all (went against you from the first tick) aren't counted; there was nothing to capture. Trades with under $5 of computed risk are also excluded — same as the known-bad-stop cleanup elsewhere in this app, since a near-zero risk denominator blows the ratio up to a meaningless four/five-digit value. A handful of symbols also have cached daily-bar data wildly out of scale with the trade's actual entry price (likely reverse-split-unadjusted history from the data provider) — trades where that would produce a physically implausible max move (20x+ the position's own value) are excluded too.
A low capture ratio on a setup you otherwise trust is a signal to try holding longer or widening the target rather than partialing out early — the market was offering more than you took. A ratio near or above 100% means you're already capturing close to the full move; there's little room left to gain by holding longer there.
Coverage matters here. "Max move" only exists for a trade with cached intraday chart bars or an imported MFE value (TraderSync's own max-favorable-excursion column) — a trade with neither (a fetch that failed, or an older import outside the data provider's retention window) has no max move and is skipped by "Total available" and the capture ratio below. "Total realized" is unaffected — it's just pnl/risk, same as everywhere else in the app, so it always matches the period's real total. The coverage count shows how much of the period has a usable max move to actually measure capture ratio against.
{% if efficiency_summary and efficiency_summary.closed_trade_count %}(Realized R among just the {{ efficiency_summary.trade_count }} covered trades: {{ efficiency_summary.covered_realized_r|floatformat:1 }}R — this is what "Total available" and the capture ratio above are actually measured against, since max-move only exists for those trades. Avg per-trade capture: {% if efficiency_summary.avg_capture_pct is not None %}{{ efficiency_summary.avg_capture_pct|floatformat:1 }}%{% else %}—{% endif %} — this simple average is easily dominated by trades with a tiny favorable move and a big loss; the overall ratio above, which weights by size, is the more reliable headline number.)
{% else %}No closed trades with recorded max-move data match the current filters.
{% endif %}| Capture ratio | Trades | % of trades |
|---|---|---|
| {{ b.label }} | {{ b.count }} | {{ b.pct_of_total|floatformat:1 }}% |
| No closed trades match the current filters. | ||
Sorted so the setup with the most room to improve — the biggest gap between what the market offered and what got realized — shows up first.
| Setup | Trades | Avg Max-Move R | Avg Realized R | Capture ratio |
|---|---|---|---|---|
| {{ row.tag }} {% if not row.eligible %}(too few trades){% endif %} | {{ row.count }} | {{ row.avg_max_move_r|floatformat:2 }}R | {{ row.avg_realized_r|floatformat:2 }}R | {% if row.capture_pct is not None %}{{ row.capture_pct|floatformat:1 }}%{% else %}—{% endif %} |
| No setup-tagged trades match the current filters. | ||||
For each candidate target below, every trade whose own best favorable move reached that far is assumed to have exited exactly there (a resting order would have filled); a trade that never reached it keeps its real result unchanged. This is the same trustworthy-move assumption the capture ratio above already relies on. It answers "if I'd set a single fixed target across the board, would I have done at least as well as I actually did?"
{% if target_sim_rows %}| Target | Trades hitting it | Simulated total R | Change vs actual |
|---|---|---|---|
| {{ row.target|floatformat:2 }}R {% if target_sim_breakeven and row.target == target_sim_breakeven.target %}BREAK-EVEN OR BETTER{% endif %} {% if target_sim_best and row.target == target_sim_best.target %}(best){% endif %} | {{ row.hit_count }} | {{ row.total_r|floatformat:1 }}R | {% if row.delta_r > 0 %}+{% endif %}{{ row.delta_r|floatformat:1 }}R |
| No closed trades match the current filters. | |||
Always every month regardless of the Date filter above, so the trend stays visible. "Change" is versus the previous month shown. "Actual R" and "Trades" are every closed trade that month, matching the Trades page total — "Covered" is the chart-data subset the capture ratio is actually computed from; a low coverage count means the capture ratio that month is based on very little data.
| Month | Trades | Actual R | Covered | Realized R (covered) | Available R | Capture ratio | Change |
|---|---|---|---|---|---|---|---|
| {{ row.label }} | {{ row.closed_count }} | {{ row.true_realized_r|floatformat:1 }}R | {{ row.count }}{% if row.closed_count %} ({% widthratio row.count row.closed_count 100 %}%){% endif %} | {{ row.total_realized_r|floatformat:1 }}R | {{ row.total_max_move_r|floatformat:1 }}R | {% if row.capture_pct is not None %}{{ row.capture_pct|floatformat:1 }}%{% else %}—{% endif %} | {% if row.delta_vs_prev is not None %}{% if row.delta_vs_prev > 0 %}▲ +{% elif row.delta_vs_prev < 0 %}▼ {% endif %}{{ row.delta_vs_prev|floatformat:1 }}pp{% else %}—{% endif %} |
| No months with eligible trades yet. | |||||||
⚠ Not financial advice. This report is statistical evidence from your own past trades only — not a guarantee or prediction of future results. Use it as one input alongside your own judgment, not a substitute for it.
"Pullback" is the worst a trade ever moved against you before it resolved, in R (1.0R is your current stop distance, by definition). If a setup's average pullback is well under 1.0R, price rarely tests the full stop — that's room to tighten it. This uses the same trustworthy-data guards as Max-move efficiency (near-zero risk, cached bar data wildly out of scale with the entry price), so a corrupted trade can't skew the numbers.
The simulator below assumes a stop at a given fraction of your current distance would have closed the trade the moment price reached that far — a real historical event, not a guess, since the pullback is measured up to the trade's actual outcome. Tightening always saves on trades that were going to lose anyway, but it can also cut off a real winner that dipped hard before recovering — "Winners cut short" shows exactly how many of those there'd have been at each fraction, so a good choice balances total R against that cost. A trade priced above $10 never gets tightened below 4¢/share, even at fractions where the rest of its setup does — tighter than that sits inside ordinary spread and noise the simulation can't see, so it wouldn't survive real execution even if it wins on paper. That trade just keeps its real result at those fractions instead, so one hard-to-tighten trade can't block a fraction for the rest of its setup.
Sorted by average pullback — the setups at the top have the most historical room to tighten their stop. "Suggested stop" is this setup's own best-performing fraction from the same simulator, run against just its trades.
| Setup | Trades | Avg pullback | Median | P90 | Avg risk ($) | Suggested stop | Improvement |
|---|---|---|---|---|---|---|---|
| {{ row.tag }} {% if row.tight_candidate %}TIGHTEN?{% endif %} {% if not row.eligible %}(too few trades){% endif %} | {{ row.count }} | {{ row.avg_pullback_r|floatformat:2 }}R | {% if row.median_pullback_r is not None %}{{ row.median_pullback_r|floatformat:2 }}R{% else %}—{% endif %} | {% if row.p90_pullback_r is not None %}{{ row.p90_pullback_r|floatformat:2 }}R{% else %}—{% endif %} | ${{ row.avg_risk|floatformat:0 }} | {{ row.best_fraction|floatformat:1 }} × current | {% if row.best_delta_r > 0 %}+{% endif %}{{ row.best_delta_r|floatformat:1 }}R |
| No setup-tagged trades with recorded pullback data match the current filters. | |||||||
Applies to whatever's currently filtered above (all setups unless you've picked one via the Setup filter).
{% if stop_sweep_rows %}| Stop fraction | Trades affected | Winners cut short | Simulated total R | Change vs actual |
|---|---|---|---|---|
| {{ row.fraction|floatformat:1 }} × current {% if stop_sweep_best and row.fraction == stop_sweep_best.fraction %}BEST{% endif %} | {{ row.affected }} | {{ row.winners_cut_short }} | {{ row.total_r|floatformat:1 }}R | {% if row.delta_r > 0 %}+{% endif %}{{ row.delta_r|floatformat:1 }}R |
| No closed trades match the current filters. | ||||