📊

No data loaded

Upload a diagnostics_YYYYMMDD.json file or click “Load SAMPLE” to explore an example dataset with 31 run history rows across 10 tickers.

As of date
Last run snapshot
Settled ②
Resolved positions
Pending ②
Awaiting expiration
PUBLIC
Selection alpha (A − B)
Published vs full pool
Hypothetical EOD-mid, non-executable Settled & pending shown side-by-side Win rate never without avg return + worst case Premium mid beside net return
Cohort Comparison

A — Published: setups that reached the nightly Top 10 (our published ranking).  B — Full Pool: all setups that passed the liquidity gate (zero-skill baseline).  B′ — Naive Yield: top-10 sorted by raw annualised yield only (complexity baseline). A − B = selection alpha (the only outward headline). A − B′ = does the multi-factor model beat naive yield?

Headlines

② Settled and pending counts are shown side-by-side throughout.  ③ Win/loss frequency is always presented alongside avg return and worst case — never alone.

Score Calibration

Bucketed on the full pool (not published) so the low-score bucket exists. If high-score buckets do not beat low-score buckets, the weights are noise. Source: pool.

Avg return % by score bucket
Score band Avg return % Loss freq Worst case Settled Pending
Per-Ticker Results

Click a column header to sort. ③ Win rate is always shown alongside avg return and worst case on the same row — never alone.

Ticker Win rate % ③ Avg return % ③ Worst case % ③ Assign freq Settled Pending
Run History Archive

Append-only nightly snapshot. Pending rows (resolved = false) show “pending” in settlement columns. ④ Premium mid is shown alongside realised return.

Run date Ticker Expiration Strike Tier Rank Score Ann yield % Buffer % PoP % Snap price Premium mid ④ Realized % Outcome Assigned