Week-ahead read: still choppy, coverage cooling
The model's week-ahead regime stays choppy. This window our analyzed cross-source coverage cooled to broadly cautious - the aggregate crypto-market read leaned bearish (16 mentions, down 8 versus the prior period) and the derivatives, CEX and DeFi sectors all skewed bearish, driven by BitMEX's announced permanent shutdown. Against that, Bitcoin still carries the deepest coverage in our corpus and holds a slim modeled upside lean (its standing weekly upside read sits near 37%, only a hair above its ~33% base rate). Alt breadth stayed thin - most large-caps are drifting sideways or lower alongside Bitcoin rather than leading - so there is no decisive market-wide direction and no broad rotation. The standing regime call (issued 2026-07-18) is left running unchanged.
What changed in the book
- New: a short-horizon (3-day) XRP upside lean at roughly 35%, a modest lift above its ~27% base rate. XRP's coverage this week skews positive and short-term momentum has firmed; the two align to the upside. This is part of our pre-registered short-horizon experiment and is excluded from our public accuracy record; confidence is low and the read is provisional.
- Left to run (unchanged): the standing 7-day upside leans on Ethereum (
42%), Bitcoin (37%), Zcash (57%), XRP (31%) and Solana (~37%). The engine re-scored each this run and none moved beyond our re-forecast threshold, so - because our records are immutable and never superseded - each is left to resolve as issued. - No new structural or relative calls. The fresh-Bitcoin-record read stayed near its ~9% recalibrated level; the altseason and major-drawdown reads produced no directional edge this run; and the Ethereum-, Solana- and XRP-versus-Bitcoin quarterly reads are already at their standing limit.
- No event calls. Our event inputs remain stale, so no resolvable macro or ETF event cleared the bar.
Receipts
Every call we make is scored by a deterministic resolver, and misses are reported alongside hits.
- Miss: our experimental 3-day Zcash upside lean (issued 2026-07-20, ~47%) resolved incorrect - Zcash closed down 2.38% over the window.
- Recent prior batch (resolved 2026-07-21): experimental 3-day Ethereum (+4.76%) and Bitcoin (+3.38%) upside leans both resolved correct; a 7-day XRP downside lean resolved incorrect (+4.63%).
Accuracy standing
Our engine was recalibrated on 2026-07-14, and its public skill claim is scored only on that current forecaster. That claim pool has not yet reached the 25-forecast minimum, so the headline Brier Skill Score is still gated (verdict: insufficient data) and we lead with reasoning rather than a headline accuracy number. For context only - and not a skill claim - across all forecasting epochs our resolved asset calls have run about a one-in-three hit rate and remain overconfident relative to outcomes; that gap is an engine-calibration matter, tracked openly, not something we hand-adjust in these write-ups.
