Five 09:30 cash-open sessions, each read at both durations Atlas IB offers, a 60-minute window and a 30-minute one. Ten readings, eight of them leans. Eight of the ten reached a rail, six of those on the favoured side and two on the other one, which is more windows settled than any week the record has published. Only Thursday ended with the session inside both targets, and it did so at both durations, off the same high.
Every window that appeared, in the order the indicator produced them.
| Session | IB range | Box tag | ▲ | ▼ | ∅ | n | State | Target (×0.5) | Outcome |
|---|---|---|---|---|---|---|---|---|---|
| 60-minute window · 09:30 to 10:30 | |||||||||
| Mon 14 Sep | 208.00 | ⚖️ Normal | 40% | 26% | 34% | 373 | abstain | ▲ 104.0 pt | ✅ 11:40 |
| Tue 15 Sep | 150.50 | ⚖️ Normal | 13% | 49% | 38% | 435 | 🧲 Atlas ▼ | ▼ 75.3 pt | 🧲✅ 10:50 |
| Wed 16 Sep | 109.75 | 💰 Accumulation | 44% | 20% | 36% | 466 | 🧲 Atlas ▲ | ▲ 54.9 pt | 🧲✅ 11:45 |
| Thu 17 Sep | 135.50 | ⚖️ Normal | 44% | 20% | 36% | 466 | 🧲 Atlas ▲ | ▲ 67.8 pt | ∅ neither |
| Fri 18 Sep | 151.25 | ⚖️ Normal | 18% | 41% | 41% | 410 | 🧲 Atlas ▼ | ▼ 75.6 pt | 🧲❌ 15:55 |
| 30-minute window · 09:30 to 10:00 | |||||||||
| Mon 14 Sep | 165.75 | ⚖️ Normal | 56% | 28% | 16% | 455 | 🧲 Atlas ▲ | ▲ 82.9 pt | 🧲✅ 11:30 |
| Tue 15 Sep | 81.00 | 💰 Accumulation | 56% | 28% | 16% | 455 | 🧲 Atlas ▲ | ▲ 40.5 pt | 🧲❌ 10:20 |
| Wed 16 Sep | 108.50 | ⚖️ Normal | 65% | 16% | 19% | 635 | 🧲 Atlas ▲ | ▲ 54.3 pt | 🧲✅ 11:05 |
| Thu 17 Sep | 132.00 | ⚖️ Normal | 50% | 36% | 14% | 365 | abstain | ▲ 66.0 pt | ∅ neither |
| Fri 18 Sep | 79.50 | 💰 Accumulation | 20% | 60% | 20% | 460 | 🧲 Atlas ▼ | ▼ 39.8 pt | 🧲✅ 10:00 |
▲ / ▼ / ∅ are the measured shares of comparable historical windows in which the upper target was reached first, the lower target was reached first, or neither was reached by the session close; n is how many such windows sit behind them. They are a historical null and not a prediction of the day being read. Every reading on this page was taken with the target multiple set to ×0.5, the setting the Atlas IB handbook recommends for NQ and the default in the current build, unchanged since w/c 17 August. These captures are 5-minute charts, so outcome times are the bar in which the rail was crossed and are good to the bar; every time on this page was read back off the 5-minute bar data rather than off the pictures. The session close is the 16:55 ET bar, and the two windows that reached neither rail were scored to it.
Two comparisons appear in every capture and they answer different questions. The box tag grades the window against its own last five sessions (💰 accumulation, ⚖️ normal, 🚀 expansion), while the status table's regime row grades it against the full seventeen-year history, and the two are allowed to disagree. They read the same way on seven of the ten windows this week. The three that split were both of Monday's windows, ⚖️ Normal on the box against expanded in the table, and Thursday's hour, ⚖️ Normal on the box against compressed. With the multiple fixed at ×0.5 the regime no longer moves the rails, so it is reported here but it is not doing anything to the targets. The Window strength row read 68% favoured-rail all week, #11 of 91 for the hour and #14 of 91 for the half hour; it describes the window rather than the day.
The red folder USD calendar carried one entry all week. The FOMC statement and projections landed at 14:00 ET on Wednesday and the press conference at 14:30; Monday, Tuesday, Thursday and Friday had no high-impact releases at all. Both Wednesday windows had been graded by 11:45, more than two hours before the statement, so no result on this page sits inside that block. Timing is all that is claimed for it.
The longer window waits for the whole first hour to complete before it commits. It gets a bigger, better-defined range to work from and, at a fixed multiple, a target that sits further away in points.
The first hour covered 208 points against a five-day average of 163.35, ⚖️ NORMAL on the box and expanded in the table. The low, 28,880.50, was the 09:30 bar itself and the high, 29,088.50, came at 10:10. The read was 40 / 26 / 34 from 373 sessions and the table's directional row printed 61%, one point under the 62% the Atlas mode asks for, so no magnet was placed and both rails printed grey.
Price fell to 28,955.75 by 10:40, turned, and rallied for the rest of the morning. The upper rail at 29,192.50 went on the 11:40 bar, which ran to 29,208.50. No lean had been offered, so the mark is a plain green tick recording that the larger of the two grey shares got there first. The move kept going to 29,305.75 at 13:55, 113.25 points beyond the rail.
Second draw of 40 / 26 / 34 on the record, after 9 September, and both abstained on the same 61%. That one ended inside both rails and this one cleared the upper. A threshold has to sit somewhere, and two draws either side of it say nothing about where.
150.5 points against a 196.05 average, ⚖️ NORMAL on both rulers. The high, 29,452.75, was set at 09:45 and was the highest price of the whole session. The low, 29,302.25, was set at 10:20, on the penultimate bar of the window, by the same fall that had just taken the half hour through its lower rail, against its long lean.
The read was 13 / 49 / 38 from 435 sessions, directional row 79%, the strongest of the five hours: magnet on the lower rail, target 75.3 points below the box. The window closed at 10:30 and the rail at 29,226.95 went on the 10:50 bar. Magnet and tick. The upper rail at 29,528.05 was never within 190 points of the day after the box closed.
Third draw of this cell on the record and its first payment: crossed against on 26 August, unreached on 31 August, paid today. The 13% on the upper side is a measured share of 435 comparable windows, not a claim about this one.
109.75 points against a 173.7 average, 💰 ACCUMULATION on the box and compressed in the table, the narrowest first hour of the week. Low 29,373.00 at 09:35, high 29,482.75 on the 10:00 bar, the first one after the half hour had closed. The read was 44 / 20 / 36 from 466 sessions, directional row 69%, magnet on the upper rail, target 54.9 points above the box.
The 11:05 bar reached 29,536.25, which cleared the half hour's rail by less than half a point and stopped 1.4 points under the hour's at 29,537.65. Price then backed off for half an hour before the 11:45 bar ran to 29,552.50 and settled it. Magnet and tick, forty minutes after the same move had already marked the shorter window.
The afternoon went the other way and kept going: 29,052.75 by 15:25, which took price through the unfavoured lower rail at 29,318.10 on the 14:35 bar, long after this window had been graded on the upper. Only the first rail counts, so that is context and not a result. The FOMC statement was at 14:00 and the press conference at 14:30; both this window and the half hour had resolved by 11:45, more than two hours earlier, and nothing on this morning is attributed to the release.
135.5 points against a 160.25 average, ⚖️ NORMAL on the box and compressed in the table, and the same cell as Wednesday: 44 / 20 / 36 from 466 sessions, directional row 69%, magnet on the upper rail. The box low, 29,605.75 at 09:35, was the session low as well, and the box high came on the 10:25 bar, so the window closed at its own extreme.
Target 67.8 points above the box put the rail at 29,809.05. Price dropped to 29,652.00 by 11:05, held between 29,652.00 and 29,767.00 until three o'clock, and then made the session high of 29,793.25 on the 15:50 bar, 15.8 points short. It fell back at once and the 16:55 bar closed near 29,720. The lower rail at 29,537.95 was never approached: the closest the whole session came to it was 67.8 points, at the box low itself.
Fourth draw of this cell: paid on 28 August, unreached on 10 September, paid on Wednesday, unreached today. Two of the four have now ended inside both rails, which is what a 36% ∅ share describes. The half hour sat 10.5 points from its own upper rail on the same bar, and Thursday is the only session of the five where neither duration resolved.
151.25 points against a 146.3 average, ⚖️ NORMAL on both rulers. The high, 29,845.25, came at 09:40 and the low, 29,694.00, at 10:15, the second of those set by the fall that had already paid the half hour. The read was 18 / 41 / 41 from 410 sessions, directional row 69%, magnet on the lower rail, target 75.6 points below the box at 29,618.40.
It got close. The session low of 29,659.25 at 12:25 finished 40.85 points above the rail. From there price climbed for the rest of the afternoon, and the 15:55 bar ran to 29,938.00 and through the upper rail at 29,920.85. Magnet and cross. The high of the day, 29,993.50, came at 16:10, six and a half points under thirty thousand.
This is the sixth draw of 18 / 41 / 41 and the first one marked against it: two paid and three unreached before today. It is also one of only four windows on the whole record to cross against an Atlas lean, and two of those four are on this page.
The shorter window commits half an hour earlier, from a smaller range, and its target is therefore usually nearer. It is a different measurement of the same morning, not a second opinion on it.
165.75 points in the first half hour, ⚖️ NORMAL against a 139.8 average and expanded in the table. Low 28,880.50 on the opening bar, high 29,046.25 at 09:45. The read was 56 / 28 / 16 from 455 sessions with a directional row of 67% and a decisive row of 84%, so the magnet went on the upper rail with the target 82.9 points above the box, at 29,129.15.
Price crossed it on the 11:30 bar, which ran to 29,164.50. Magnet and tick.
Read this mark and the hour's together. The two boxes share the 28,880.50 low, the second half hour added 42.25 points at the top and nothing at the bottom, and one rally cleared both rails ten minutes apart. They are separate windows and the scorer counts them separately, as it must, but a reader judging the week should count the morning once.
81 points against a 164.3 average, 💰 ACCUMULATION on the box and compressed in the table, the third narrowest half hour on the record. It drew the same cell as Monday, 56 / 28 / 16 from 455 sessions at 67% directional, so the same call was made from a box less than half the size: magnet on the upper rail, target 40.5 points above at 29,493.25.
The high after the box was 29,432.25 on the 10:00 bar, 61 points under that rail, and price then turned. The 10:20 bar fell to 29,302.25 and through the lower rail at 29,331.25. Magnet and cross, twenty minutes after the window closed at 10:00. The day's high had been set inside the box at 09:45 and was never revisited.
The same bar that marked this window against made the low of the hour's box, which then leaned short from it and was paid at 10:50. Tuesday is the only morning of the five on which the two durations favoured opposite rails, and both of them were graded within an hour of each other. Third draw of this cell: paid on 3 September, paid on Monday, against today.
108.5 points, ⚖️ NORMAL against a 134.1 average. 65 / 16 / 19 from 635 sessions with a directional row of 81% and a decisive row of 81% is the most one-sided reading of the week and the joint most one-sided the record has drawn since the multiple was fixed at ×0.5 on 17 August. Magnet on the upper rail, target 54.3 points above the box at 29,535.80.
The 11:05 bar topped out at 29,536.25, clearing the rail by 0.45 points. Magnet and tick. That is how the record grades it and it is worth saying plainly that half a point the other way is a blank instead. The hour's rail sat 1.85 points higher and was not reached until 11:45.
Fifth draw of this cell: paid on 26 August, against on 4 September, unreached on 9 and 11 September, paid today. Five draws of a cell measured over 635 sessions settle nothing about it in either direction. Both Wednesday windows were graded by 11:45, more than two hours before the 14:00 FOMC statement.
132 points against a 125.8 average, ⚖️ NORMAL on both rulers. The read was 50 / 36 / 14 from 365 sessions and the directional row printed 58%, four points under the test, so both rails stayed grey. The ∅ share of 14% is the lowest any window on this page carried, and ∅ is what happened.
The box ran 29,605.75 to 29,737.75, rails at 29,539.75 and 29,803.75. Price made 29,638.75 at 10:05, 99 points above the lower rail, and the session high of 29,793.25 on the 15:50 bar, 10.5 points under the upper. The same bar left the hour 15.8 points short of its own upper rail.
Second draw of 50 / 36 / 14, both of them abstains: crossed on 17 August, neither today. A window that abstains is still graded, on the larger of its two grey shares, and this one had 50% on the side that came closest. Ten and a half points is a near miss and it is recorded as a blank, because the record grades rails reached and not approaches.
79.5 points against a 123 average, 💰 ACCUMULATION on the box and compressed in the table, the second narrowest half hour on the record. The read was 20 / 60 / 20 from 460 sessions, directional row 75%, decisive row 80%: magnet on the lower rail, target 39.8 points below the box at 29,725.95.
The window closed on the 09:55 bar at its own low of 29,765.75. The 10:00 bar, the very next one, fell to 29,723.00 and through the rail. Magnet and tick, with no bar in between. That is the earliest any window on the record has been settled; the previous earliest was 10:17 on 6 August. The capture here was taken at the crossing plus ten minutes rather than plus five: at plus five the study still read IB building for the current window and was drawing the previous session's table. This picture is the aftermath rather than the moment.
The same fall carried on into the hour's window and made its box 71.75 points deeper at the bottom. Both durations then leaned short from the same morning; the nearer rail was reached in five minutes and the further one never was. First appearance of 20 / 60 / 20 anywhere on the record, and a single draw of a cell measured over 460 sessions is a single draw.
Eight of the ten windows on this page were settled by a rail, which is more than any week the record has published. Five of the eight leans were paid, more than any week has managed, and two of them were crossed the other way, which no week had done more than once before. A week that moves is not the same thing as a week that was right.
Two cells drew twice each this week with identical shares and bins both times, so the same call was made from two different mornings. 56 / 28 / 16 from 455 read Monday's and Tuesday's half hours, and the two boxes were 165.75 and 81 points wide, less than half the size on the second day and the same lean regardless. Monday paid, Tuesday was crossed. 44 / 20 / 36 from 466 read Wednesday's and Thursday's hours: Wednesday paid at 11:45, Thursday finished 15.8 points short. Four draws, two paid and two not, from readings that were identical before the session started. That is what a cell of 44% or 56% describes and it is the most useful thing on this page.
Tuesday is the only morning of the five where the two durations favoured opposite rails, and it is the clearest example on the record of why they are separate readings. The half hour leaned long off an 81-point box. Price turned and the 10:20 bar fell through its lower rail, which marked it against; that same bar set the low of the hour's box, which closed ten minutes later reading 13 / 49 / 38 and leaned short from it. The hour's rail went at 10:50. One morning, one direction of travel, one window paid and one crossed.
Friday produced the two ends of the range. The half hour was settled on the first bar after its box closed, at 10:00, which is the earliest resolution on the record. The hour, leaning the same way from a box the same fall had made deeper, was still 40.85 points from its rail at the session low and finished crossed against at 15:55. Same morning, same direction, opposite marks, and the difference between them was 35.8 points of target distance.
Across seven weeks the running count now stands at 70 windows: 27 reached the favoured or larger share first, 10 reached the other rail first, and 33 reached neither. Of the 44 that were in Atlas mode, 20 paid, 20 reached nothing and 4 were crossed against. Seventy windows cannot confirm or refute frequencies measured over hundreds of sessions, and a busy week is no more evidence than the flat one published a week earlier.
Five things, from the most active week the record has published.