A year of measurement, on one page, with the working attached.
Every figure here carries what it was measured on, how many events it came from, and what it was tested against.
Check any of it. Then go and try to do the same anywhere else in this category.
Three products agreed, and they were never asked to
Each one was measured separately, a full year each, and neither was allowed to borrow the other's answer.
All three came back at the same multiple. An amplified board travels about 1.8 times further in an hour than a compressed one.
This is what boards in each state have actually travelled. It is the number that sets your size and your stop before anything is on.
A symbol with no measured figure gets the words and no number. The field still tells you which state the board is in. How far is a question we answer only where we have earned an answer.
Five reads, and the record of each
These are SPY's figures. They are figures about SPY and nothing else, and Compass will not put one product's number under another product's name.
Forty logged live events is the bar before any of these is promoted. Until a read clears it the panel shows a count rather than a rate, because a percentage built on tens of events is a rumor whichever way it points.
THE FAST TAPE tells you when the tape is fast and going nowhere
That is where accounts give back a week, and it is close to invisible from the inside, because a fast tape feels like opportunity while it is happening.
It carries the largest sample of anything we measure. It exists to keep you out of it.
A position carried into the close draws down about half as much on a grind day
Measured across 64 of 234 sessions: a median 1.07 strikes against 1.86 on an amplified one.
It is a holding decision rather than an entry, so it lives in the field hover rather than on the strip. If you are already in and wondering whether to carry it, this is the number.
The shape that looks like a clean short, and is not one
An amplified board. Price under the Action Area. A put-weighted book, the heaviest node sitting just above price, and thin air behind it. Every part of that reads as a short.
We wrote it down before we measured it, the way a study is preregistered so nobody can move the target afterward.
Over the next half hour, price ran further the other way. Not one of the two thousand shuffled controls reached it.
The method, and the measured effect
- WINDOW
- the following 30 minutes
- SAMPLE
- 434 anchors across 82 sessions
- PRODUCT
- SPY
- CONTROLS
- 2,000 shuffled controls
- MEDIAN EFFECT
- 0.41 strikes the other way
So Compass draws that shape as friction and leaves the direction to you, and now it can say why with a number behind it.
Every number on this page had to survive the bench
Ideas get tested here whether or not we like where the answer points.
There is a bench inside the product, in the Engine Room, and it runs studies overnight. Members can open it and watch what is being tried. Under Test in the Logbook shows the reads that are still being measured, before they are allowed to say anything.
A read reaches the screen only after forty logged live events.
One read was retired in August, when three independent lines of evidence disagreed with the number we had published. It came off the screen the same week.
Every figure above is what is left after that. The failures are carried at the same volume as the findings, and that is the only arrangement under which the findings are worth anything.
Nobody can copy a year of outcomes
A screenshot is worth the same on day five hundred as on day one. A record is worth more every day.
One short paragraph closes each session, written from the session Compass recorded rather than from live state, so it still reads correctly on an app opened after the close.
The board opened amplified. A typical hour on an amplified SPY board has run 2.30 strikes. Today ran 0.67. Slower than the board suggested. The board read the day long.
One level was tested. It gave way. Price closed at 772.41, 3.2 strikes below where the recording started.
That second line is the whole point. The board is described the same way on the days it was long as on the days it was right, in the same voice, by the same paragraph.
One thing the close is not: a plan for tomorrow. By the closing bell on an expiry board the heaviest levels have collapsed onto price, because gamma concentrates at the money as expiry approaches. Writing down the closing King is nearly the same as writing down the closing price. The board has to be read live, on the day, which is the whole reason this product exists.
Ready to know which board you are on?
Join a limited group of traders helping shape TRT Compass before public release.
Join Private Beta