Reading the Analysis Tab — Comparing Premarket vs Market-Open
The Analysis tab is the morning's reconciliation: which premarket trades held up after the opening drive, which got re-graded, which got killed, and which fresh setups appeared at the open. It runs around 10:15 AM ET, after both the premarket scan (8:30) and the market-open scan (9:45) have published. The LLM narrative reads the two side-by-side; the comparison table and the high-probability picks list are computed deterministically — no LLM judgment on the math.
The three sections, top to bottom
- LLM narrative. Written by the analysis routine around 10:15 AM ET. Reads the premarket and market-open scans side-by-side and explains what the open implied — confirmation, rotation, or a divergence from the premarket bias.
- High-probability picks. Deterministic. Filters to trades that satisfy all three conditions (in both scans · direction stable · grade ≥ A−). No LLM judgment — pure rule.
- Cross-comparison table. Every ticker from either scan, with lineage, premarket grade + direction, market-open grade + direction, grade Δ, direction Δ. The full audit trail.
Lineage legend
- Both (green) — appeared in premarket AND market-open. A continued plan. The most common lineage.
- New at open(sky-blue) — market-open introduced it. The premarket scan didn't flag this ticker; the opening drive surfaced a fresh setup. Often the highest-volatility opportunities but also the highest-uncertainty.
- Premarket only (amber) — premarket flagged it but market-open dropped it. Usually because the underlying moved away from the planned entry, news invalidated the thesis, or the opening drive made the setup stale.
Grade delta — what counts as a step?
The grade scale is ordered (A+ > A > A− > B+ > B > B− > C+ > C > C− > D+ > D > D− > F). Each step is one slot on that ladder. So premarket A− to market-open A is ▲ 1 step; premarket B+ to market-open A− is ▲ 2 steps. Step counts give you a quick sense of how dramatically the routine's opinion changed between scans.
What “Direction Δ” tells you
- same — strongest confirmation. Both scans agree on puts/calls/long/short. The trade is exactly as planned.
- flipped — opposite direction. Put → call or call → put. A strong reversal signal that something material happened at the open.
- → avoid — premarket flagged a live trade; market-open downgraded it to AVOID. Treat as a kill.
- from avoid— premarket said AVOID; market-open promoted it into a live trade. The opening drive must have created a setup the premarket didn't see.
How this fits with the other tabs
Pre-market is the original plan. Market-Open is the post-open re-grade. Analysis is the side-by-side comparison + LLM narrative explaining what changed and why. Trade Cards is the final merged plan (with post-close outcome stamps at end of day). Scorecard is the cross-day aggregate. Each tab is a different lens on the same daily flow — Analysis is specifically about the 8:30 → 9:45 transition.
Frequently asked questions
What does the Analysis tab show?
Three sections: (1) an LLM-written narrative at the top, comparing the premarket bias to what actually happened at the open; (2) a deterministic 'High-probability picks' list that filters to trades that survived both scans cleanly; (3) a full cross-comparison table showing every ticker that appeared in either scan, with grade and direction deltas.
What's the rule for 'High-probability picks'?
Three conditions must all hold: (1) the ticker appears in BOTH the premarket and market-open scans (no new-at-open, no premarket-only); (2) direction is stable — no flip from put to call, no shift to AVOID, no shift from AVOID; (3) the market-open grade is A− or better, OR the grade was upgraded into the A-tier at the open. The point is to surface setups where the morning re-grading confirmed (or strengthened) the premarket conviction.
What does the 'Lineage' column mean in the comparison table?
Where the trade came from. 'Both' = appeared in premarket AND market-open (a continued plan). 'New at open' = market-open introduced it (not in premarket — usually an opening-drive opportunity the premarket didn't anticipate). 'Premarket only' = premarket flagged it but market-open dropped it (often because the underlying moved away from the setup).
What do ▲ and ▼ symbols in 'Δ Grade' mean?
Grade movement from premarket to market-open. ▲ N = upgraded N grade steps (e.g. B+ to A is ▲ 2). ▼ N = downgraded N steps. A single dot (·) = same grade in both scans. A dash (—) = the ticker only appeared in one scan, so there's no delta to compute. Upgrades are green, downgrades are red.
What does 'Δ Dir' tell me?
How direction changed across scans. 'same' (green) = put stayed put, call stayed call — high conviction. 'flipped' (red) = put → call or call → put, a strong reversal signal. '→ avoid' (red) = market-open downgraded a live setup to AVOID. 'from avoid' = market-open promoted a premarket AVOID into a live trade. Each tells a different story about how the open changed the read.
Why is the narrative sometimes empty?
The narrative is written by the analysis routine that runs around 10:15 AM ET. If you load the page before 10:15 — or if the routine hasn't run yet for some reason — the narrative section will say 'Narrative is generated by the 10:00 ET analysis routine.' The deterministic table and high-probability picks are available the moment market-open publishes (~9:45) — they don't need the LLM.
How do I use this tab in practice?
Skim the narrative for the morning's character — was the open in line with the premarket bias, or did it diverge? Then scan the high-probability picks: these are the trades the two scans agreed on with A-tier conviction. Use the full comparison table to spot revisions you should be aware of — anything that flipped direction, anything that dropped to AVOID, anything that got upgraded into the A-tier. The TRADE CARDS tab shows the final merged plan; Analysis tells you HOW the plan evolved.
If a trade isn't a 'high-probability pick' should I avoid it?
Not necessarily. The filter is conservative on purpose — designed to surface the absolute clearest setups. Plenty of profitable trades won't qualify: a B+ in both scans is solid but doesn't meet the A− threshold; a single-scan opportunity ('New at open') might be the best trade of the day. Use the picks list as the highest-conviction shortlist; use the full comparison table to evaluate the rest.
What's the difference between Analysis and the merged Trade Cards?
Analysis is about HOW the plan evolved — the narrative and comparison emphasize the transition from premarket to market-open. TRADE CARDS shows the FINAL merged plan — strike, entry, targets, stop, rationale per ticker, with the post-close outcome stamped at the end of the day. Analysis is the 'change log'; Trade Cards is the 'current state'.
Why aren't 'avoid' direction trades in the high-probability picks?
An 'avoid' is the routine flagging the ticker as untradeable for the day — usually because of unfavorable IV, poor liquidity, an event in the way, or no clean setup. By definition there's no high-probability trade there. The cross-comparison table will still show avoids so you can audit the routine's reasoning, but they're excluded from the picks list since there's nothing to take.