Retrospective · No. 3
Retrospective: The Croissant Gap — August 30, 2026
A retrospective compiled now from aggregate records—not a forecast published in the past.
By Stock & Lattes Weekly Update
Published
Theo split a croissant and declared the empty middle a chart gap. Maya wanted to chase it; June quietly put the top half back and asked what would happen if the gap filled. The pastry disappeared before consensus arrived, which felt appropriately market-like.
387 model predictions entered the ledger; 0 currently have recorded outcomes and 387 remain pending.
Their scenarios below are fictional teaching examples, not observations about a current security, evidence of a placed trade, or promises of an outcome.
How the three see it
Maya: bullish
Educational scenario: an opening gap holds above the prior session high.
Only if: Consider strength only if the gap remains open through a retest on stable volume.
Then: A hold could favour continuation; a full gap fill invalidates the bullish scenario.
Theo: bearish
Educational scenario: an opening gap loses its first support shelf.
Only if: Consider weakness only if price accepts back below the prior session high.
Then: Acceptance below could favour a gap fill; a reclaim above the opening range cancels the scenario.
June: cautious
Educational scenario: price whipsaws around an unfilled opening gap.
Only if: Wait for both an opening range and a retest before assigning direction.
Then: Orderly confirmation may create a readable range; continued whipsaw leaves the scenario unresolved.
Counts, with caveats
- Predictions
- 65810
- Resolved
- 58091
- Pending
- 7719
These are aggregate records, not a performance or ranking claim. Legacy outcomes are unvalidated and unresolved records remain pending.
Methodology
- All figures are deterministic counts from TSX, US, and crypto prediction tables; no user, account, or personal data is included.
- Target outcomes are counted as correct; stops and expiries are shown separately; unresolved records remain pending.
- Yesterday separates the generated prediction cohort from outcomes actually resolved during that Toronto calendar day.
- Legacy outcomes have not been independently validated. No accuracy, best-buy, best-sell, or executed-trade claim is made.
- Earliest evidence capture is the first persisted reproducibility record, not the first calculation time. True legacy calculation time and raw historical math are not reconstructable.