A factual reconstruction of your real trading results — after every real cost.
Layers 3 & 4 · Error Memory + Personal Encyclopedia
The patterns in your trading
Each pattern below is detected from your own fills and rated by how much of your data supports it. Confirmed = 15+ trades and a strong effect.
PAYOFF-ASYMMETRYConfirmed
Disposition effect — small wins, big losses
Average win far smaller than average loss, so even a high win rate loses money.
At a 42% win rate you’d need to average ≥1.37 in win-per-loss to break even. You average 0.70.
CATASTROPHIC-TAILConfirmed
Catastrophic losses — no downside cap
A small number of trades allowed to fall past –30% do outsized damage.
257 trades worse than –30% lost –$41,609 — more than your entire net result.
HELD-TO-ZEROConfirmed
Names with zero wins
Assets traded repeatedly without a single profitable exit — held down rather than cut.
26 assets had 0 winning closes (e.g. W, SOMI, CAKE, LINK).
REPEAT-LOSSConfirmed
Returning to losing names
Names closed at a loss three or more times and still net-negative — re-entering what keeps costing you.
48 names lost ≥3 times and stayed negative. Worst: SLP –$10,818 across 172 trades (1% win rate).
CUT-WINNERS-EARLYConfirmed
Winners closed small
Winners taken at a small gain while losers run further than winners.
Median winner +10.8% vs median loser –12.4% — you let losers travel further than winners.
LOSS-CONCENTRATIONProbable
Loss concentrated in a few names
Most damage from a handful of assets, not broad bleed.
Top-5 losing names = –$28,251 (61% of net).
These entries are the seed of your personal trading encyclopedia. Anonymized, the same taxonomy feeds Isnadia’s collective Failure Encyclopedia — one spine, two failure layers.
Layer 2 · The Decision Gap
The price of one missing rule
Because price is continuous, a trade that closed at –50% did trade through –10% on the way down. So we can ask, soundly from your own data: what if a downside cap had been in place?
+$52,496
A simple –10% stop would have turned your –$46,550 into +$5,945. Same entries, same picks — one rule. That swing is the cost of decisions, not the market.
Caveats: winners are left untouched (no intrabar peak data, so no claim you “could have held longer”); stops are applied only where the trade actually closed below the level. This is a directional lower bound, not a wick-accurate backtest.
Reality Gap · Over time
Year by year — and the tell
Your one profitable year is the year you traded least. Activity and losses move together.
2024: 97 trades, the only green year. 2025: 944 trades, lowest win rate, your worst year by far. More clicks, more damage.