Both charts fall in their final ten minutes. Before that interval began, IONQ was already below its hour-start price while PENN was substantially above it. The earlier path changes the comparison.
Two charts can end with a decline and still describe different sequences. IONQ at noon on May 12, 2026, and PENN at 11:04 a.m. Eastern on March 31, 2020, make that distinction visible.
PENN was original match 3 in our first frozen sample of 50 historical comparisons. This is a follow-up to that case, using the same evidence. We have not run another search, replaced a match, or opened a later price path.
Take the final decline out of the comparison
IONQ's displayed hour begins at 11:00. PENN's begins at 10:04. Both contain all 60 expected minute rows; both final ten-minute intervals are complete. Missing observations do not explain the difference between these two examples.
| Measurement before each cutoff | IONQ | PENN |
|---|---|---|
| First displayed open to final close | -4.57% | +3.47% |
| Final ten-minute interval, first open to last close | -1.41% | -3.32% |
| First displayed open to the open of that final interval | approximately -3.20% | approximately +7.03% |
The third row isolates where each chart stood before the matched ten-minute interval began. It is derived from the two published open-to-close measurements: (1 + full-hour return) / (1 + final-interval return) - 1. The inputs are the rounded values in the original evidence export, so this reconstruction is reported to two decimal places. It is not a return after the decision time.
This matters because the difference is already present before the final selloff. IONQ enters its last ten minutes below its hour-start price. PENN enters its last ten minutes above that reference, after an earlier advance. The larger final decline in PENN does not erase that earlier rise.
Read the sequence, then the last bars
In the original chart inspection, IONQ had declined, attempted a recovery, and turned lower again. PENN had made a strong advance, retreated and balanced, then declined late. Calling both simply a selloff loses the order of those moves.
The chart panels use different vertical scales. Their printed percentages matter: the final declines differ in size as well as in the path that preceded them. A visually similar slope is not a common amount of price movement.
The reference level here is the first open in the displayed hour, not yesterday's close, the regular-session open, or a cost basis for traders. We cannot infer who was trapped, which participants were selling, or why either move happened from these panels.
The finding
For these two examples, a matching late decline does not preserve the preceding direction. That is a concrete reason to inspect the lead-in before treating a short-window match as the same setup.
It is not evidence that one sequence subsequently performed better. The charts occur on different dates and at different times of day; a shared company event or market mechanism has not been established. The existing corpus also uses full-session eligibility, and its normalization postdates the IONQ seed. This remains a retrospective structural comparison.
A useful next question is whether a clearly specified recurring event produces comparable decline-and-recovery sequences. That question needs its own mechanism, counterexample, data checks and fresh confirmation sample before outcomes are inspected. These two charts help define what to distinguish; they do not supply a tested trading rule.
Read the original IONQ case and inspect all 50 matches. Download the original measurements.
Figures
Selection and evidence
Follow-up to CB-001. PENN remains original rank 3, the previously disclosed first match with a complete rising displayed hour. This article adds a decomposition of the already published pre-cutoff measurements. All 50 original members remain in the source case.
Source snapshot SHA-256: f6c1b05d94656906c23484c33be9d06e0093940756a8099fecebdd860b25178c. Original protocol and reconstruction receipt.
