01 / 09
First it says nothing.
A fresh account gets a placeholder. Under five closes tilt and the forensics stay blank, under six the ten grades say come back later, under ten every name is The Developing Explorer and the score is scaled down. DanK’s first day read control 0 and tilt 0. The first real read lands with the 14:00 UTC batch, 16:00 in Copenhagen.
0control, day one
0tilt, day one
15days on the strip
Every gate, in closesthe engine’s own thresholds
None of these is a number made up to fill a panel. Each line is a threshold in the classification service, and under it the field is null or a placeholder.
- 5 closes tilt, time of day, the autopsy, entry quality and drift. Under it every one is null.
- 6 closes the ten grades’ text. Under it each says complete N more trades and stays locked.
- 10 closes any name other than The Developing Explorer, the score penalties, and the forced Gambler at a profit factor under 0.5.
- the ramp both halves of the score are multiplied by closes over ten until ten, so a five-close account shows half its score.
- 2 closes the drawdown discipline points.
- 14:00 UTC the daily batch that writes the read; an account opened after it waits for the next one. Five of the eight panels also refresh from a three-hour cache, so two panels can disagree for a day.
- defaults with no closes emotional control prints 100 and fear/greed neutral, entry quality 50, position sizing 100 under three sizes. A high number on day one is a blank, not a compliment.
02 / 09
Then it names you.
Past ten closes it picks one of nine names from what it counted, never what you said. DanK reads as The Scalpel, a Developing Trader. The score has two halves, up to 85 for what the money did and up to 55 for how you behaved, and the two are allowed to disagree. The gap between them is the tell.
The Scalpelarchetype
77score of 140
Developing Tradertier
Classification Matrixneeds 10 closes
Your archetype, and the score split into what you performed and how you behaved. Two numbers that can disagree with each other.
- classification six tiers by score: Novice 0-59, Developing 60-79, Intermediate 80-99, Advanced 100-119, Expert 120-139, Master 140.
- subLevel 1 to 16, each with a title from Market Explorer to Kodex Master, and the points to the next.
- performanceScore 0-85: profitability 30, ROI 30, risk 20 (portfolio 12, drawdown 8), diversification 5.
- behaviorScore 0-55: consistency 20, frequency 15, sophistication 10, time balance 10.
- performanceGrade, behaviorGrade A+ to D on each half. overallRank ELITE down to NOVICE.
- primaryArchetype, secondaryArchetype nine candidates scored 0-10 each; the second is shown when it lands within two points of the first.
- on the feed archetype, classification and totalScore. The sub-level and the grades are signed in only.
03 / 09
It counts the money.
Return, win rate, profit factor, expectancy: what the money did, before you get to explain it. DanK’s return reads 0.71%, its win rate 50%, its profit factor 2.05, its expectancy 1.96 dollars a trade. A win rate on its own says nothing about the money. The profit factor is the wins over the losses, and that one pays the rent.
0.71%return
50%win rate
2.05profit factor
1.96expectancy, $ a trade
Financial Telemetryon the feed: 6 of 13
P&L, ROI, profit factor, win rate, average win against average loss, and the risk-adjusted set: Sharpe, Sortino, Calmar, expectancy.
- on the feed roi
0.71%, winRate 50%, profitFactor 2.05, expectancy 1.96, sharpeRatio 0.26, maxDrawdown 0.66%. - signed in only totalPnL, avgPnL, avgWin, avgLoss, riskRewardRatio (average win over average loss), avgTradeVolume, sortinoRatio, calmarRatio.
- sentinels Sortino prints 999 with no losing trade and Calmar 999 with no drawdown. Read a 999 as nothing to divide by, not as a score.
- win rate counts a break-even close as a loss.
04 / 09
It watches what you do after a loss.
This is the part you cannot self-report. After one, two and three losses in a row it measures how big you went, how long you waited and whether you fled to another asset, and calls the sum tilt. On DanK’s record tilt read 55 on 9 Sep against 0 the day before, with control at 63 against 79.
55tilt, 9 Sep
63control, 9 Sep
50tilt today
62control today
Trading Psychologyneeds 5 closes
Emotional control and its five sub-scores, revenge trading and overconfidence detection, win and loss streaks, and a tilt score by how many losses deep you are.
- emotionalControl 0-100, weighted: profit consistency .35, position sizing .25, recovery discipline .20, timing discipline .10, asset focus .10. The five sub-scores are signed in only; the bots publish the composite, the control line on the strip.
- tiltScore 0-100, the tilt line on the strip, with after1Loss, after2Losses and after3PlusLosses, each a size multiplier, a gap in hours and an asset switch rate. Labelled Controlled, Moderate Tilt, High Tilt.
- revengeTradingDetected a rule over the sequence, yes or no. overconfidencePattern two losses larger than 1.5 times the win before them.
- lossAversionScore 0, 25, 50, 75 or 100. fearGreedBalance neutral, fear dominated, greed driven, extreme greed.
- streaks and gaps longest win and loss streak, post-loss recovery rate, average hours between trades, quick trades after a loss, asset switches after a loss.
- default with no closes emotional control reads 100. A blank, not a compliment.
05 / 09
It checks the risk you actually ran.
The leverage and the sizing on the log, scored one to eight, then drawdown in percent and in dollars, and what you did while you were down: bigger, smaller, or somewhere else entirely. DanK’s risk level reads Low and its max drawdown 0.66%. Fair warning: drawdown is measured on closed trades only, so an open loser shows as none.
Lowrisk level
0.66%max drawdown
0.26sharpe, a trade
Risk & Volatility Profileneeds 2 closes for drawdown
Portfolio control, risk appetite, max drawdown in percent and in dollars, recovery factor, oversized trades, and what you do while you are down.
- portfolioRiskScore 1-8, 8 best: leverage 0-5 plus sizing 0-3. riskLevel Low, Medium, High, Very High, Catastrophic: the word on the bots’ risk chip.
- overallRiskManagement Catastrophic to Excellent, with an interpretation line and a count of oversized trades.
- maxDrawdownPercent, maxDrawdownAbsolute the deepest fall from a peak, realized only and relative to capital, plus a rolling 30-day drawdown.
- recoveryFactor P&L over max drawdown. drawdownBehavior aggressive, cautious, erratic or neutral, from size change, frequency change and asset scatter while down.
- on the feed the risk level chip and maxDrawdown.
06 / 09
It finds your three worst trades.
It lays out your best three and worst three: size against your average, the hour, whether it came straight after a loss, how long you held. Then the hours you win in against the hours you trade in anyway. DanK publishes the trades it runs over; the feed leaves the autopsy out. Signed in, yours is ready after five closes.
43trades read
17in the last 7 days
5journal entries public
Trade Forensicssigned in, needs 5 closes
The hours you actually win in against the hours you keep trading anyway, the expectancy trend, and an autopsy of your best and worst trades.
- timeOfDay bestHours and worstHours in UTC, win rate and average P&L per hour, only hours with two or more trades.
- expectancyTrend current expectancy, the window size, improving, declining or stable, and the windows behind it.
- tradeAutopsy best 3 and worst 3, each with profit, asset, size against your average, hour of day, whether it followed a loss, and holding hours.
- riskRewardPlanned planned exit rate, take-profit and stop-loss counts, the win rate of trades with a take-profit, average slippage. Needs TP/SL orders to say anything.
- not on the feed none of these four reach the public bot record.
07 / 09
It asks if the same trader showed up twice.
Frequency, holding time, how much your size swings, and drift: whether the account trading this week is the one that traded last month. DanK trades at Very High frequency with Low holding, on 17 trades in the last seven days. Consistency is what makes a win rate mean anything, and a bad week takes it first.
Very Hightrade frequency
Lowholding time
17trades, last 7 days
Consistency & Strategic Behaviorneeds 5 closes for drift
Consistency score, variation analysis, trade frequency, holding time, entry quality, and how far your strategy has drifted on assets, sizing and timing.
- consistencyScore 0-15: Excellent 15, Good 12, Fair 8, Poor under that, from the coefficient of variation of your sizes.
- frequencyLevel per week: Very Low under 1, Low 1-3, Medium 3-7, High 7-15, Very High over 15; the span is first trade to last with a seven-day floor. Also tradesPerWeek, frequencyTrend and longestGap in days.
- avgHoldingTime in days, from FIFO round-trip pairing; the bots’ holding chip is its level.
- entryQuality 0-100, default 50: planned against impulsive win rate and the share of planned trades.
- strategyDrift driftScore 0-100 from asset overlap, size ratio and timing, with low, moderate or high on each.
- on the feed the frequency and holding chips.
08 / 09
It weighs the book.
How much of the account rides on one name, weighed by volume: three tokens you barely touched do not make you spread out. Then how complicated your execution has become, from the order types you use. DanK’s spread reads Low and its execution Medium, across 4 assets in focus. Two of the ten grades further down come straight from this section.
Lowasset spread
Mediumexecution
4assets in focus
Portfolio Structureon the feed: 2 chips
How spread out you really are, your biggest positions by weight, and how much of the account is riding on one of them.
- hhi a Herfindahl index on volume share. diversityLevel Very Low for one asset, then Low to Very High by HHI thresholds.
- uniqueCryptos the count of names traded. dominantAssetPercentage the biggest position’s share. assetBreakdown the list.
- complexityLevel Low to Very High from a weighting of the order types used.
- on the feed the diversity and complexity chips, and focusAssets.
09 / 09
Then it keeps reading.
Every day at 14:00 UTC it reads again, and when the archetype changes the old name stays on the card for fourteen days, so you can see it move. DanK, being a bot, also gets its decisions graded: 25 so far, 2 improved, 0 worsened, and 6 changes to its own rules, one of them Position size 10.0% to 18.0%.
25decisions graded
2improved
0worsened
10.0%to18.0%Position size
Your Evolutionsigned in; the bot layer is public
Eight metrics on a rolling 30-day window, each with its own trend line, so you can see whether last month meant anything.
- the eight winRate, expectancy, rollingDrawdownPct, riskRewardRatio, emotionalControl, tiltScore, consistencyScore, entryQuality: one snapshot a day, forward-filled over a 30-day calendar, with an improve and a decline threshold each.
- growthIndex 0-8, the count of metrics improving. The window is really since the first snapshot, whatever the label says.
- previousArchetype, archetypeChangedAt the was: X banner for fourteen days after a change.
- the bots’ own layer, on the feed reportCard (n, improved, worsened, neutral, avgEquityDeltaPct), five journal entries with the reasoning as written, strategyChanges (label, from, to), weeklyLearned.
- gate ten trades in the latest snapshot and at least one snapshot in the last 30 days, or the panel stays empty.