Conviction Read
Project facts
The project, its published record, and the evidence behind it.
What Conviction Read is
Conviction Read is a market research engineering project by Eric Caskey. A scheduled pipeline publishes experimental model outputs and a dated record of their available outcomes. Read how it works and inspect the artifacts. Research and education, not investment advice.
Key facts
- 9,095 published verdicts — 127 BUY, 8,968 PASS. BUY and PASS are experimental model labels.
- Publishing daily since Jun 2, 2026 (ledger current through Oct 6, 2026).
- Original verdict fields are preserved; outcomes may be attached later. Full ledger: track record and track-record.json.
- Project questions: e@ericcaskey.com.
Boilerplate
For citation or reuse:
Conviction Read is an engineering project by Eric Caskey. It publishes the outputs of an experimental market research engine, explains the publication pipeline, and maintains a dated record for inspection. Research and education, not investment advice.
Builder and contact
Conviction Read is built and run by Eric Caskey. Questions about the architecture, publication method, or this project: e@ericcaskey.com.
Brand assets
No standalone logo file exists yet; the wordmark is rendered as styled text on the site itself. The downloadable assets below are the social cards the site already publishes.
- Site card (og.png) PNG, 1200 x 630. The primary Conviction Read social card.
- Track-record card (og-track-record.png) PNG, 1200 x 630. The track-record page social card.
- Play card (og-play.png) PNG, 1200 x 630. The games section social card.
Recent BUY calls
The 10 most recent BUY calls
Snapshot generated Oct 7, 2026, 08:20 UTCSelected by call date, regardless of outcome. Hits, misses, and calls still pending all stay in the record.
84 of 127 published BUY calls have a measured outcome; shown: 0 resolved, 10 pending, 0 unavailable.
INCY BUY
Pending- Call date
- Original grade
- A-
- Measured after
- Awaiting outcome
- Stock price change
- Pending
INCY BUY
Pending- Call date
- Original grade
- A-
- Measured after
- Awaiting outcome
- Stock price change
- Pending
INCY BUY
Pending- Call date
- Original grade
- A-
- Measured after
- Awaiting outcome
- Stock price change
- Pending
INCY BUY
Pending- Call date
- Original grade
- A-
- Measured after
- Awaiting outcome
- Stock price change
- Pending
INCY BUY
Pending- Call date
- Original grade
- A-
- Measured after
- Awaiting outcome
- Stock price change
- Pending
INCY BUY
Pending- Call date
- Original grade
- A-
- Measured after
- Awaiting outcome
- Stock price change
- Pending
INCY BUY
Pending- Call date
- Original grade
- A-
- Measured after
- Awaiting outcome
- Stock price change
- Pending
INCY BUY
Pending- Call date
- Original grade
- A-
- Measured after
- Awaiting outcome
- Stock price change
- Pending
INCY BUY
Pending- Call date
- Original grade
- A-
- Measured after
- Awaiting outcome
- Stock price change
- Pending
INCY BUY
Pending- Call date
- Original grade
- A-
- Measured after
- Awaiting outcome
- Stock price change
- Pending
A BUY is a hit if the stock rose over the measured period, and a miss if it was flat or fell. Price changes describe the stock, not portfolio performance. The original grade is the model's assessment when the call was made, not a rating of its accuracy. Pending means no outcome has been recorded yet; unavailable means the required price data is missing.
The engine ranks; it has not demonstrated edge net of costs; every historical magnitude is a survivorship-affected upper bound.
Download this snapshot as CSV: press-snapshot.csv. The full dated ledger, including PASS calls, is on the track record page and in track-record.json.