Muse Desk
What this is and how it is worked out
What this is: The bar at the top names the desk, shows that this is a paper account, counts the strategies, shows when the figures were made, and shows a row of small dots, one per strategy.
How it is worked out: The strategy count is the number of strategies in the desk’s strategy list. The snapshot time is the time the figures file was made, shown in New York time. Each dot is coloured by its strategy’s kind, using the colours from the strategy list, and a dot is filled only when that strategy currently holds at least one position.
What this is and how it is worked out
What this is: This page holds no numbers of its own. Every figure is loaded from the desk’s published data files, so what you see is the desk’s own record, not a page-side estimate.
How it is worked out: The desk rebuilds its figures file every 5 minutes. This page checks for a new file every 60 seconds, only while the page is visible, and redraws only when the file’s time stamp has changed. A rebuild is skipped when the records behind it have not changed. Before anything is published, the totals are checked against their own rows — a mismatch stops the publish — and the whole file passes one final check that removes internal codes and rewrites wording into plain words. The time stamp shown is moved forward to the newest item inside the file, so the stamp never claims to be older than its own contents.
Best setups
The desk's strongest current buy setups — names the desk has not entered yet, still intact and still in range, ranked by how many independent strategies agree on them, whether the desk's pattern engine confirms them, and how clean the entry looks. A name leaves this list the moment the desk enters it (it becomes a position, in the Position book below), its setup breaks, it runs too far, or its plan expires — and every departure is listed under the board with its reason. Every point of every score is explained on its card.
What this is and how it is worked out
What this is: This button turns on phone alerts for new desk entries, so a new position is announced instead of waiting to be noticed on the page.
How it is worked out: Alerts work only in the installed app, with notifications allowed. The page makes a 6-digit pairing code with a secure random number: 100,000 plus the remainder of the random number divided by 900,000. It then registers this device with the public key printed in the page code and sends the registration and the code to the relay address printed in the page code. On the desk side, an alert is created by comparing the earlier and new figures files: an alert means a name newly appears as entered, never just a name shown again.
What this is and how it is worked out
What this is: The desk’s strongest current buy setups, ranked. Each card shows the rank, the tier and score, whether the desk is watching or bought today, the last close, a small price chart, the entry, stop and target, the top reasons, and which strategies selected it.
How it is worked out: The score is out of 100, and every point must have a recorded reason or the scoring stops with an error. Agreement between different strategy groups: 1 group 6 points, 2 groups 16, 3 groups 24, 4 or more 30. Pattern engine: a confirmed swing pattern adds 14 when it is at most 3 trading days old, 4 when 4 to 10 days old, and 0 when older; pattern context only adds 7. Chart-review board: 10 points at a board score of 75 or more, 7 at 65 or more, 4 at 55 or more, 2 below that. Already entered by the desk: 12. An entry that was tried and turned away for a money reason only: 4. Fit with the market: on a fear or red day, a dip setup 8 and any other kind 3; in a rising market, momentum 6 and a dip 3; in a falling market, a dip 2 and any other kind minus 6; in a sideways market, 2. Reset at support: 8 when the close is within 1% either side of the 20-day average, 4 when the close is within one average daily range of that average or on the recent swing-low area. Stop distance, measured in 14-day average daily ranges: 1.5 to 2.5 ranges scores 8, 2.5 to 4 scores 5, 1 to 1.5 scores 2, under 1 range loses 4, over 4 ranges scores 1. Taken away: a run of 15% to 25% in 10 trading days loses 10; a close more than 2.5 ranges above the 20-day average loses 8; a close below the 20-day swing low loses 15. A name up 25% or more in 10 trading days has its finished score multiplied by 0.7. Tiers: 75 or more is Top tier, 55 or more is Strong, 35 or more is Watch, and below 35 a name does not appear at all. Before a name is shown, its plan is checked: the board and the plan must agree on where the setup came from; prices must run in order — stop below buy low, buy low at or below buy high, buy high below first target, first target below final target, all above zero; the reward-to-risk shown is recomputed from the plan and must match within 0.01; an option package’s most that can be lost or gained is recomputed from its parts; the plan must not have expired; and the headline reward divided by risk must be at least 1.0 or the name does not appear (option packages are governed by their spread checks instead). The small chart is the last up to 30 daily closes, with a gold dashed line at the entry price when that price falls inside the chart’s range; with no price series, no chart is drawn.
On the radar — no call yet
Names the desk is watching where the evidence does not add up to a call. One honest line each — no card until the desk can say what it expects.
What this is and how it is worked out
What this is: Board names whose full read reached no call. They scored enough to be watched, but the whole-picture read did not add up to a call, so they are listed here with one plain line each.
How it is worked out: A name is placed here when its full read carries the no-call flag. When the evidence behind a read is incomplete, the read is held back and only the held state and its plain reason are published. Names here are never dressed up as calls.
Call record
What this is and how it is worked out
What this is: One line counting the desk’s published calls: how many were made since the starting date shown, how many are still open, how many are closed, how many left the board, and the average result of the closed calls that carry a result.
How it is worked out: A call is a name the board published that is still open, or that later left the board because it was entered, expired, or was ended by its closing line. Names that were held back were never published and are never counted. The three parts must add up exactly to the number made. The average result, in R (one R is the amount risked on the trade), is printed only next to its count and the dates of the first and last calls, and the closed list is capped at 50.
Recently closed
Every name that leaves this list, with the reason and the outcome — kept here for at least five trading days, and permanently in the desk's record, so nothing quietly disappears.
What this is and how it is worked out
What this is: Every name that has left the Best Setups board, with the reason, the date, and the outcome when there is one.
How it is worked out: Each departure is added to a permanent record that is only ever added to. This list shows the last 5 trading days present in that record. For a name entered and closed the same day, the outcome in R is (exit price minus entry price) divided by (entry price minus stop). For a name held longer, the outcome is the amount-weighted average result of that name’s closed trades whose entry and exit dates span the departure. Otherwise the row simply says the position is open, or that it was closed.
On-demand reads
Full Desk reads the desk produced on request, for any stock — the same whole-picture read the daily setups get, written when asked and published here.
What this is and how it is worked out
What this is: Full desk reads made on request, for any stock, newest first. Each one is the same whole-picture read the daily setups get.
How it is worked out: The list is capped at 20 reads, newest first. Each read is built by the same builder that builds the daily full reads, and there is no separate scoring for a requested read.
Market and top-ranked stocks
Where the market stands today, and the stocks the desk's chart-review board ranks highest.
What this is and how it is worked out
What this is: Where the market stands: index charts for the S&P 500 and Nasdaq 100 funds, a label for the market’s direction, a fear-day label, the fear gauge, and for each index the close, the day’s move, and how far it sits from its 50-day average.
How it is worked out: For each index: the close is the last daily close; the day move is the last day’s percentage change; the distance from the 50-day average is (close divided by the 50-day average of closes, minus 1) times 100. Market direction is read from the Nasdaq 100 fund only: close above its 50-day average and a 20-day rise above +2% is a rising market; close below its average and a 20-day fall below −2% is a falling market; anything else is sideways. The 20-day move is the percentage change over 20 trading days. A fear day is any day when the S&P fund falls more than 1.5%, or the fear gauge rises more than 15% in a day, or the fear gauge closes at 25 or more, or either fund has a red day while the fear gauge is at 20 or more. The fear gauge is the VIX index, read from daily closes from a public data source because the broker does not carry it. The index charts are daily charts with the 20-, 50- and 200-day averages and the recent high and low drawn. A chart is held back when its own last close differs from the panel’s close by more than 0.01, so the panel never shows two disagreeing numbers.
Morgue Engine
Three measuring instruments the desk built for itself: a map of dealer positioning in the big index funds, a scored feed of the day’s options prints, and a ranked board of the strongest stock setups across 1,838 names. Every figure here was written into a permanent, add-only record before it was shown. The engine measures; it does not place trades.
The Morning Map — the positioning map and the qualified board as their own page → The Funnel Study — how the corpus delivers
Loading…
What this is and how it is worked out
What this is: The desk’s own positioning map, flow tape and setup board, rebuilt from measuring tools whose workings were published or recovered in full, and run on the desk’s own market data. Nothing here is bought from a vendor, and no figure is shown that was not recorded first.
How it is worked out: The Map adds up, for each index fund, the options dealers are likely holding: for every contract, its sensitivity to price moves times its open interest, with call contracts counted positive and put contracts negative. The total, in dollars per 1% move, says whether dealer hedging is likely to calm moves or amplify them; the walls are the strike prices with the largest call and put totals, and the flip is the price where the total changes sign. The short-sale ratio is the share of off-exchange volume sold short, from the free daily exchange file, a day behind. The Tape scores every options print of the day for the three funds: a heat score for size, urgency and unusualness, and a running total of premium, signed by whether each print lifted the offer or hit the bid. The Board scores all 1,838 names from eight groups of evidence — price action, relative strength, volume, trend, nearby support and resistance, sector strength, positioning and market conditions. A name qualifies only when at least five of the eight groups confirm, its score is at least 65 out of 100, at least 70% of the evidence could actually be measured, and it passes a liquidity and trend screen. A name with too little measurable evidence is left off the board rather than guessed at. The trigger prices come from an average-range ladder built on the previous close: the first trigger sits about a quarter of an average day’s range above it.
The full read
What this is and how it is worked out
What this is: The full read behind one setup: its charts, its plan, the desk’s whole-picture call, how it got here, the chart read, a lens-by-lens table, the wave count, the levels, what would end the setup, the strategies behind it, and two past-record lines.
How it is worked out: Reads are built over the stock’s last 320 daily bars, split-adjusted (about 6 months on the chart, plus the longer history the 200-day average needs). The call: experts for each family of evidence file a signed view — a direction times a weight. The balance is the sum of (direction times weight) divided by the sum of the weights. A balance of 0.45 or more reads strongly for, 0.18 or more for, −0.45 or less strongly against, −0.18 or less against, and anything in between is mixed. A mixed result is settled in a fixed order: price structure first, then momentum confirmed by volume, then the wider context. A failed ending line caps the call at against. The five strongest directional voices by weight decide; the rest are considered. A voice from a method that has not yet earned a proven record has its weight capped at 1.0 when the deciding voices are chosen. Incomplete evidence is refused: only the held state and its plain reason are published. Past-case line: when this shape has been measured over the desk’s 1,838-stock list, the read says ‘In past cases of this shape, the target was reached before the ending line in X% of N cases’; when the count is thin or unmeasured, a plain line says so, or nothing is printed. A single projected target also carries its measured rate of being reached first. Strategy record line: the plan strategy’s closed-trade record — win rate, number of trades, average result in R, and its dates — shown as evidence in the read only; it never adds to the score and never opens or closes a gate. Wave count: a turning point counts only at the bar that confirms it, 2 bars after the extreme; a swing must be at least 1.5 times the 14-day average daily range, with at least 3 bars between turning points; Fibonacci levels are allowed a 7% tolerance. The count comes out as active, unclear, no valid count, or ended, with a price that would end it; its fit score is a ranking index, never a chance, and the wave voice is one of the capped, not-yet-proven voices in the call. Levels shown: entry, stop and target from the plan, the last close, the 20-day average, and the recent 20-session swing low. Charts: the price pane draws the pattern boundaries, the levels, a Fibonacci retracement grid, the buy zone and the expected path; the volume pane, a 14-day relative strength pane (Wilder’s method), and a momentum pane are worked out over the same closes. A difference between price and an indicator is drawn only where the chart-read record carries one; a used-up difference is not drawn; and there is at most one such label per pane. Phones get a smaller, purpose-built picture of the same read.
How a setup earns its place
No single signal is trusted on its own. A stock becomes a setup here when several different ways of reading it point the same way at the same time. The desk reads the whole picture at once, never one signal at a time. Every setup is scored out of 100 across the lenses below, and every point is explained on its brief — nothing is a black box.
- Independent agreement. The desk runs dozens of strategies built on different ideas. When strategies that share nothing — different signals, different mathematics — select the same stock on the same day, that agreement is the strongest evidence the desk knows. Close cousins count once, so one family cannot stuff the ballot box.
- The pattern engine. A dedicated engine grades swing patterns: the shapes prices make as they base, pull back and turn. A fresh, confirmed pattern adds serious weight. No pattern, no points from this lens.
- The chart-review board. An independent nightly ranking of chart quality across a fixed board of stocks. It is background, never a trigger: it can support a setup the other lenses found, but it can never create one by itself.
- The market’s mood. The same setup means different things in different markets. Washed-out dips are at their best in fearful, falling markets; momentum needs a rising one. The desk reads the broad market first and scores the fit.
- A price sitting at a level. Setups score higher when the price has pulled back and reset onto something real — its recent average, or the zone of its recent lows — instead of dangling mid-range.
- The shape of the risk. A good setup knows where it is wrong. A defined exit level, far enough away that ordinary noise will not trigger it but close enough to cap the loss, earns points; a stop so tight that a normal day would shake it out loses them.
- The warning signs. Some things count against a setup no matter how many lenses agree: a price that has already spiked hard in a few sessions, one stretched far above its average, or one that has broken below its recent lows. Each of those deductions exists because that exact pattern has burned traders before.
The brief above walks the lenses for this one stock: the chart as it looked when the setup fired, the evidence lens by lens, and the list of what would end the setup. The list itself is a research starting point — the evidence is published with every setup so you can check the desk’s work and decide for yourself.
The whole-picture read
Every brief also carries a Desk read. It weighs everything the desk measures for the stock: the chart and its shape, volume, momentum, the wider market, and how fresh the data is. The few things that decide the call are named first. Where the evidence points both ways, the brief says so. Anything the desk does not measure for this stock is listed too. The read ends with what would change the call.
The Desk read is a weighed judgment, and it is deliberately kept out of the score: the number above never moves because of it. Every read is written into a permanent record before the outcome can be known, and the desk scores its past reads against what prices did over the following one, five and ten sessions. Only if that recorded track record earns it — by review and ruling, not by default — would the read ever be given weight in the ranking. Until then it is published so its judgment can be checked like everything else here.
The wider tape is measured the same way, every night, over the desk’s own scanned universe: how many stocks are rising and how many sit above their own averages, the market’s fear gauges combined into one mood reading, whether traders are reaching for risk or hiding from it across six benchmark pairings, which sectors and themes are leading and which are fading, and a five-check pulse on the two big index funds. Those readings sit in the market panel above and are weighed into every Desk read.
What this is and how it is worked out
What this is: The headline account numbers: the account value, the cash, and how many positions are open, with a note of how many strategies are holding.
How it is worked out: The account value, cash and buying power are the broker’s own figures for the paper account, printed as reported. The open position count is the number of position rows published below. The ‘strategies holding’ count is the number of strategies whose held value is above zero.
Exposure
The total value of everything the desk currently holds, and how much of the account that makes up.
Loading the latest figures…
What this is and how it is worked out
What this is: How much of the account is held in open positions: the gross held value, its percentage of the account with a meter, and a bar showing how the held value splits by strategy.
How it is worked out: Gross held value is the sum of the absolute market value of every published position row. A row with no market value adds nothing, and no second estimate (such as quantity times entry price) is ever used instead. The percentage is 100 times the gross value divided by the account value, rounded to one decimal place, using the account value — not a set budget. Each strategy’s value is the sum of the absolute values of its rows, and its share of the bar is that value divided by the total gross value.
Account
The paper account, as reported by the broker.
Loading the latest figures…
What this is and how it is worked out
What this is: The paper account in detail, as reported by the broker: its status, the cash held, the value held in positions, a bar showing cash against held value, the buying power, and the cash plus held equals account sum.
How it is worked out: Value held in positions is the signed sum of the published rows’ market values — gains add, short positions subtract. It is blank when rows exist but none carries a market value, and zero when the book is empty. A broker status of Active is shown as ‘Active — open and trading normally’. Buying power is the broker’s figure, printed as reported.
The desk's permanent record
Every action the desk takes is written into one record that is only ever added to, never rewritten. This is what is in it.
Loading the latest figures…
What this is and how it is worked out
What this is: A count of everything in the desk’s permanent record, split by kind of record, with each kind’s share.
How it is worked out: The counts are a straight count of every line in the desk’s add-only record. The total is the line count, and each kind is counted by its label: completed trades, journal entries, price records, nightly checks against the broker, trade reviews, and positions brought in at the start. A line that is not a proper record is skipped, never guessed at, and each share is that kind’s count divided by the total, times 100.
Position book
Each option contract covers 100 shares, so an option priced at $1.11 is worth $111 per contract. Rows grouped under one heading are parts of a single trade, with a combined total underneath. Words like call, put and strike are explained in the “Words we use” list on the Strategies tab.
Loading the latest figures…
What this is and how it is worked out
What this is: Every open position: the stock (options are written out in plain words), the strategy, the quantity, whether it is owned or sold first, the average entry, the current price, the market value, its share of the book, and its open profit or loss. Parts of one option trade are grouped under one heading with a combined total, and there is a total row for the whole book.
How it is worked out: A position is long (owned) when its quantity is above zero, and short (sold first) otherwise. Share of book is the row’s absolute market value divided by the sum of absolute market values of all rows, times 100. The profit-or-loss bar width is the row’s absolute open result divided by the largest absolute open result in the book, times 50% — half the bar on each side of zero. An option code is read as: the leading letters are the stock, the next six digits are the expiry date, C means call and P means put, and the last digits divided by 1,000 are the agreed price. Option parts are grouped by the same strategy, the same stock, and the same expiry date. One option contract covers 100 shares. Totals are sums of the rows shown, only.
Entered today — the plans these positions are managed by
These names were on the Best Setups list until the desk entered them. Entering a name moves it here: it is a position now, not a setup.
Closed today
These trades were entered and closed again today, so each one is finished. They are results, not current setups — they are listed here with the positions for the day's record.
What this is and how it is worked out
What this is: Two day lists inside the Position book. ‘Entered today’ shows the plans the positions the desk entered today are managed by. ‘Closed today’ shows today’s finished round trips: bought and sold again the same day.
How it is worked out: For an entered name, the stop shown is the plan’s trading stop, never the emergency backstop. The target comes from the plan’s exit target when the row itself carries none. The time bought and the quantity still held come from the broker’s position row — one source for both. When more than one strategy holds the same stock, a plain note on the row says so. An entered row with no target in any place carries an explicit note saying the target is missing, instead of a silent blank. A held read publishes only the held state and its plain reason. Closed rows reuse reads already made today — from today’s reads, then reads carried forward, then the earlier cycle’s — and no new read is ever built for a finished name.
Strategies
Every strategy the desk runs, and what each one is doing now. Select “How this strategy works” on any card for a plain-English explanation.
What this is and how it is worked out
What this is: One card per strategy: its plain name, a two-line summary, its kind, what it trades, whether it is holding positions now, and an expandable ‘How this strategy works’ note on the card itself.
How it is worked out: The card words — name, summary, longer explanation, what it trades, and kind — are written entries in the desk’s strategy list, one entry per strategy (63 entries when this note was written); they are words, not worked-out numbers. Whether a card says Holding or No positions comes from the live figures: holding means that strategy’s position count is above zero. The kinds are a fixed list of eight: Tensor strategies; Breakout and momentum; Pullbacks and mean reversion; Rule-based systems; Options — income; Options — directional and hedges; Systematic funds; and Crypto. Search matches only when every word typed appears somewhere in the name, summary, what it trades, longer explanation, kind, or the strategy’s name with its underscores read as spaces. The kind filter and the search apply together. Each strategy’s own trading rules are its own; the card’s expandable note is the plain-English place for them.
Loading the strategy list…
The mix, in pictures
What this is and how it is worked out
What this is: Pictures of the strategy mix: one dot per strategy grouped by kind, bars for how many strategies sit in each kind, bars for what the desk trades, and a grid crossing kind with what it trades.
How it is worked out: These are pure counts — no trading maths. A kind bar’s length is its count divided by the largest kind’s count, and the share printed is its count divided by the total number of strategies. The ‘what the desk trades’ bars count strategies by the what-it-trades entry on their cards. A grid cell is the number of strategies with that kind and that what-it-trades pair, and its shading deepens as the count grows. In the dot picture, a filled dot means that strategy’s position count is above zero.
Words we use
- Option: a contract giving the right to buy or sell something at an agreed price, by an agreed date.
- Call: an option that gains in value when the price rises.
- Put: an option that gains in value when the price falls.
- Strike: the agreed price at which an option lets you buy or sell.
- Premium: simply the price paid for an option — or the payment collected when one is sold.
- Spread: a paired option trade — one option bought and one sold on the same stock — so the most that can be made or lost is capped.
- Contract: one option contract covers 100 shares.
- Long: owned — the desk bought it and gains when the price rises.
- Short: sold first, as a bet the price will fall, and bought back later.
- Expiry: the date an option runs out.
- Straddle: buying a call and a put together on the same stock.
- Wheel: the repeating cycle of selling a put, possibly buying the shares, then selling calls against them.
- Ladder: options stacked at different prices and dates, like rungs of a ladder.
- Beta: a measure of how much a stock moves compared with the whole market; high-beta stocks move more.
- Bar: one period's price summary — where the price opened, its high, its low, and where it closed.
- ETF: a fund traded like a stock, holding a basket of shares or assets.
- Tensor: the desk's name for its pattern-scoring engine, which grades stocks for swing setups.
- VWAP: the day's average traded price, weighted by how much traded at each price.
What this is and how it is worked out
What this is: Plain-English definitions of the words used across this page, from Option and Strike to Tensor and the day’s average traded price.
How it is worked out: This list is fixed written text; nothing in it is worked out from live figures. Its contract definition — one option contract covers 100 shares — is the same rule the Position book uses.
Architecture
How this desk works, in plain English
Loading…
What this is and how it is worked out
What this is: How the desk is put together: a seven-step flow picture, a hub picture, numbered sections in plain English, a daily-rhythm timeline, and a highlighted safety section.
How it is worked out: The section words are fixed written notes, loaded with the page and set out by a small page-side formatter (bold, bullets and numbered lists). The flow picture is a fixed seven steps: Data, Scheduler, Strategies, Broker, Risk accounting, Record, Terminal. The timeline is display only: the page sorts the bold job names in the Automations section into four parts of the day — Before the open, During market hours, Evening and nightly, and Weekly — using a fixed word-to-time table. The timeline itself works out nothing about trading; the real job times live in the desk’s job list.
How it’s judged
How the whole desk works, and how its calls are made — the long version, in plain English
Loading…
What this is and how it is worked out
What this is: The long plain-English document behind this page: what the desk is, how it works end to end, how a Best Setup is made, every connection the desk uses, how to build your own, and a glossary — plus a link to the full architecture brief.
How it is worked out: This is explanatory writing, not a worked-out figure: six fixed sections and a 20-row list of every connection the desk uses, loaded with the page. The methods it describes are the ones set out in the Best Setups and Full read notes above.