Regime
The standing regime read stays choppy (issued 16 October, runs to 23 October); nothing this run gives a reason to append a new one.
Receipts — resolved since the last run
All five are epoch-5 touch calls, graded on whether price reached the stated level at any point within the horizon.
| Call | Issued | Matured | Resolved | Outcome |
|---|---|---|---|---|
| Ethereum (ETH) — leans to reach −3.14% within the next 7 days | 11 Oct 14:29Z | 18 Oct 14:29Z | 18 Oct 17:39Z | Hit — level reached |
| Solana (SOL) — leans to reach −3.12% within the next 7 days | 11 Oct 14:29Z | 18 Oct 14:29Z | 18 Oct 17:39Z | Hit — level reached |
| Bitcoin (BTC) — leans to reach −2.31% within the next 7 days | 11 Oct 14:29Z | 18 Oct 14:29Z | 18 Oct 17:39Z | Hit — level reached |
| Ethereum (ETH) — leans to reach −1.43% within the next 3 days (experimental) | 15 Oct 15:11Z | 18 Oct 15:11Z | 18 Oct 17:39Z | Hit — level reached |
| Bitcoin (BTC) — leans to reach −1.00% within the next 3 days (experimental) | 15 Oct 15:11Z | 18 Oct 15:11Z | 18 Oct 17:39Z | Hit — level reached |
No misses resolved this run. These five are not five independent data points: all were short-side touches on co-moving majors over overlapping windows, so a single market-wide dip settled them together.
Accuracy — the honest headline
- Brier Skill Score vs the base rate: −0.0034 (epoch 5, n=10, 7-day claim calls). Our published probabilities are essentially level with simply quoting the historical base rate, and still fractionally worse. They have not beaten it yet.
- Murphy split: reliability 0.0375 (the over-confidence term, lower is better), resolution 0.045 (how much our probabilities discriminate). Resolution only just exceeds the reliability penalty, which is why the score sits near zero.
- The 1-day lane stands at a Brier Skill Score of −0.171 (n=23) — clearly worse than the base rate.
- The verdict remains insufficient data: ten calls is well below the 25 the public claim requires.
- Directional-accuracy note only (not a skill measure, blind to the probability we state): hit-rate minus base rate is −0.0113 on the same ten calls.
New calls this run
The conviction set is the same as yesterday's, plus Zcash: Ethereum, Solana, Zcash and Bitcoin all lean lower on every horizon. The lean comes mostly from price momentum. Coverage in this window was constructive on balance for the majors (one analyst, across two posts, framed Bitcoin, Ethereum and Solana as core holdings), and for Bitcoin our corpus input actually points the other way. Zcash fills the non-major slot, where an analyst read this window kept a bearish bias while price is held under its bands.
- Ethereum (ETH) — leans to reach −3.35% within the next 7 days — Price momentum on Ethereum leans lower, and its coverage in our corpus reads slightly less constructive than its own 90-day norm; the engine places roughly a 66% chance on a −3.35% touch within seven days, against a ~57% base rate for a move that size — a modest gap the engine has not yet shown it can earn.
- Solana (SOL) — leans to reach −2.46% within the next 7 days — Price momentum on Solana leans lower, and its coverage in our corpus reads slightly less constructive than its own 90-day norm; the engine places roughly a 76% chance on a −2.46% touch within seven days, against a ~68% base rate for a move that size — a modest gap the engine has not yet shown it can earn.
- Zcash (ZEC) — leans to reach −5.73% within the next 7 days — Price momentum on Zcash leans lower, and its coverage in our corpus is thin but reads slightly less constructive than its 90-day norm, including one analyst read this window keeping a bearish bias; the engine places roughly a 65% chance on a −5.73% touch within seven days, against a ~57% base rate for a move that size — a modest gap the engine has not yet shown it can earn.
- Bitcoin (BTC) — leans to reach −2.39% within the next 7 days — Price momentum on Bitcoin leans lower, and our corpus input on BTC leans mildly the other way, consistent with this window's largely constructive coverage, so the read rests on price momentum; the engine places roughly a 67% chance on a −2.39% touch within seven days, against a ~63% base rate for a move that size — a modest gap the engine has not yet shown it can earn.
- Ethereum (ETH) — leans to reach −2.20% within the next 3 days (experimental horizon) — Price momentum on Ethereum leans lower, and its coverage in our corpus reads slightly less constructive than its own 90-day norm; the engine places roughly a 67% chance on a −2.20% touch within three days, against a ~58% base rate for a move that size — a modest gap the engine has not yet shown it can earn.
- Zcash (ZEC) — leans to reach −3.75% within the next 3 days (experimental horizon) — Price momentum on Zcash leans lower, and its coverage in our corpus is thin but reads slightly less constructive than its 90-day norm, including one analyst read this window keeping a bearish bias; the engine places roughly a 67% chance on a −3.75% touch within three days, against a ~60% base rate for a move that size — a modest gap the engine has not yet shown it can earn.
- Bitcoin (BTC) — leans to reach −0.90% within the next day — Price momentum on Bitcoin leans lower, and our corpus input on BTC leans mildly the other way, consistent with this window's largely constructive coverage, so the read rests on price momentum; the engine places roughly a 62% chance on a −0.90% touch within a day, against a ~58% base rate for a move that size — a modest gap the engine has not yet shown it can earn.
Every probability is the engine's, uncalibrated and provisional. Touch probabilities sit near 60–75% because touch base rates are high at these levels. That is a property of the claim type, not evidence of skill. The complement of each call is "does not reach the level", not "rises".
Standing structural reads
No structural or relative row was appended. The Bitcoin record-high, altseason and majors-drawdown questions each already carry two standing reads. Today's re-scores of the Ethereum- and Solana-versus-Bitcoin questions are materially unchanged from the reads on the board.
