You asked
How do I play?
Abstained — no grounded answer; closest passages below counted as honest abstainAsked with: Casinobg-18121-en-1e
I looked through Casino's rulebook and couldn't find a solid answer to that one — and I'd rather tell you so than guess at your table. Try asking it another way, or ask me about setup, turns, or how the game ends: those I know cold.
The closest rulebook passages are below — check them before ruling.
whom it would be dealt may reject it, and it must then be placed in the middle of the stock, and he must be given the top card. If a card is exposed in the last round, the dealer must take it, and must allow the player to whom it would have been dealt to draw a card from the dealer's hand, face down
Casino — Foster's Complete Hoyle (1914) 5 opponent’s build may be increased by cards from the hand only. In the four-handed game, partners may take in one another’s builds, or may make builds which can be won by the card declared in the partner’s hand. For instance; One player builds an 8, and his p
The simple rule to be remembered is that no combination of cards once announced, and left on the table, can be changed, except by the addition of a card from the hand of some player. Taking In. Any player who has made a build is obliged either to win it, when it is next his turn to play, or to win s
The last trick is usually made by the dealer, who always keeps back a court card if he has one, to pair one already on the table. Irregularities in Play. If any person plays out of his proper turn, the card so played is laid aside as exposed, until it comes to his turn, when it is simply placed on t
If points are played for, the lower score is usually deducted from the higher, and the difference is the value of the game. It is sometimes agreed that if the winner has twice as many points as his adversary, he shall be paid for a double game. Dealing. After the cards have been properly shuffled th
🔧 debug — where this ruling came from
- outcome
- ABSTAINED — Abstained — no grounded answer; closest passages below
- warm abstain
-
- this ruling — W1 — generic warm abstain · receipt: none (generic)
- corpus profile
- live,vps
- answerer pin (deploy-frozen, attribution only)
- gemma4:26b think:false 5571076f3d70
- timings
- search 206 ms · ask 1332 ms · total 1540 ms
- single-flight (PR-5)
- queued 0 ms · generation 1540 ms
- retrieval refs (raw) — what the PASSAGE SEARCH returned, not what the answerer read
- [{"page": 3, "score": 0.048652, "snippet": "whom it would be dealt may reject it, and it must then be placed in the middle of the stock, and he must be given the top card. If a card is exposed in the last round, the dealer must take it, and must allow the player to whom it would have been dealt to draw a card from the dealer's hand, face down", "headingPath": "CASSINO.", "sharpsignalDocId": 226}, {"page": 5, "score": 0.047131, "snippet": "Casino — Foster's Complete Hoyle (1914) 5 opponent’s build may be increased by cards from the hand only. In the four-handed game, partners may take in one another’s builds, or may make builds which can be won by the card declared in the partner’s hand. For instance; One player builds an 8, and his p", "headingPath": "", "sharpsignalDocId": 226}, {"page": 4, "score": 0.046486, "snippet": "The simple rule to be remembered is that no combination of cards once announced, and left on the table, can be changed, except by the addition of a card from the hand of some player. Taking In. Any player who has made a build is obliged either to win it, when it is next his turn to play, or to win s", "headingPath": "", "sharpsignalDocId": 226}, {"page": 5, "score": 0.04598, "snippet": "The last trick is usually made by the dealer, who always keeps back a court card if he has one, to pair one already on the table. Irregularities in Play. If any person plays out of his proper turn, the card so played is laid aside as exposed, until it comes to his turn, when it is simply placed on t", "headingPath": "", "sharpsignalDocId": 226}, {"page": 2, "score": 0.045262, "snippet": "If points are played for, the lower score is usually deducted from the higher, and the difference is the value of the game. It is sometimes agreed that if the winner has twice as many points as his adversary, he shall be paid for a double game. Dealing. After the cards have been properly shuffled th", "headingPath": "CASSINO.", "sharpsignalDocId": 226}, {"page": 3, "score": 0.045228, "snippet": "For instance: He holds an 8, and there are upon the table four cards only:--5 3 6 and 2. By combining the 6 and 2, and the 5 and 3, two Eights will be formed, and the sweep is made. Sweeps are usually marked by leaving the cards with which they are made face upward at the bottom of the tricks taken ", "headingPath": "CASSINO.", "sharpsignalDocId": 226}, {"page": 7, "score": 0.044583, "snippet": "The game is scored on a cribbage board, every point being pegged immediately; that is, every spade, every Ace, the Cassinos and the sweeps. There is nothing to count at the end of the hand but the cards. Sixty-one points is game, once round the board and into the game hole. In this variation, no mor", "headingPath": "SPADE CASSINO. · DRAW CASSINO.", "sharpsignalDocId": 226}, {"page": 6, "score": 0.044563, "snippet": "This game is usually marked with counters, or pegged on a cribbage board. Nothing is scored until the end of the hand, when each side reckons and claims its points. In order to avoid disputes there should be a previous understanding as to what points go out first in a close game. In the absence of a", "headingPath": "TWENTY-ONE POINT CASSINO.", "sharpsignalDocId": 226}]
- citations (raw) — what the answer cited, chosen from a retrieval this page cannot show
- []
- boost provenance — errata + base rulebook (E2/E3)
-
no registered errata card matched this query; base rulebook doc 226 boosted; rulesage.errata.boost-enabled=true, rulesage.errata.supersedence-enabled=true
base rulebook doc 226 boosted
- answered-before recall
-
no past answered ruling for this game matched closely enough
similarity threshold in force: 0.55 (rulesage.answered-before.similarity-threshold; the exact-question tier is not gated by it)
- debug row recorded
- Sep 6, 2026 08:03 UTC
retrieval — what the passage search returned
what this block IS: the PASSAGE SEARCH's own results — the same search that fills this page's closest-passages list — and NOT the passages the answerer read. Answering runs its OWN retrieval: a far wider candidate pool, re-ranked down to a final handful, plus floors this search never applies — the base rulebook, surviving errata corrections, and one slot per book on a multi-book read. Each of those armed only if THIS ask sent what it needs: the boost block names the ids it forwarded, and the scope line names the books it read. A document the answer CITES that no row here names is therefore EXPECTED, not a defect: two searches, two candidate pools, two orders. The answerer's own final rows are shown in their own block below, beside these and never mixed into them.
score = the fused retrieval score (reciprocal-rank fusion over the search arms) — NOT a cosine similarity. This ruling ran the THREE-ARM fuse (full-text + vector + the doc-priority boost arm, which carried the base rulebook and/or the errata corrections this query matched), whose arithmetic ceiling is 0.04918 — all three arms ranking the same chunk #1 (ceiling = arms/(k+1) at k=60). The measured healthy band 0.016–0.033 was measured on TWO-arm asks and does not bound this one: its high end is the two-arm ceiling, and a third arm has room above it. A chunk only one arm ranked highly lands near 0.016 (that one arm's #1 vote), so 0.03 here is an excellent hit, not a 3% match.
per-arm ranks (the rank each search arm gave a chunk): NOT DISCLOSED by the gateway. Fusion happens gateway-side and only the FUSED score crosses the wire, so rulesage cannot show them without a gateway change. The rank column below is the FUSED rank (position in the returned list), which is real.
token count per chunk: NOT DISCLOSED — chunk token counts are not reported on the ask path. The character counts below are of the bounded snippet/quote rulesage actually received — a real measurement of what we hold, never an estimate of the chunk.
| fused rank | document | heading | page | fused RRF score | snippet chars |
|---|---|---|---|---|---|
| #1 | RULEBOOK Casino — Rulebook (corpus doc 226) | CASSINO. | p.3 | 0.048652 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | 300 |
| #2 | RULEBOOK Casino — Rulebook (corpus doc 226) | p.5 | 0.047131 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | 300 | |
| #3 | RULEBOOK Casino — Rulebook (corpus doc 226) | p.4 | 0.046486 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | 300 | |
| #4 | RULEBOOK Casino — Rulebook (corpus doc 226) | p.5 | 0.04598 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | 300 | |
| #5 | RULEBOOK Casino — Rulebook (corpus doc 226) | CASSINO. | p.2 | 0.045262 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | 300 |
| #6 | RULEBOOK Casino — Rulebook (corpus doc 226) | CASSINO. | p.3 | 0.045228 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | 300 |
| #7 | RULEBOOK Casino — Rulebook (corpus doc 226) | SPADE CASSINO. · DRAW CASSINO. | p.7 | 0.044583 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | 300 |
| #8 | RULEBOOK Casino — Rulebook (corpus doc 226) | TWENTY-ONE POINT CASSINO. | p.6 | 0.044563 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | 300 |
what the answerer read — the passages it actually had in front of it
what this block IS: the passages the ANSWERING read put in front of the answerer, in the order it read them — the other half of the two reads described above, and the one that decides what the answer could possibly say. Each row names the book it came from and where in that book it sits, what it scored when the search arms were fused, which arm carried it there, and whether the re-ranker chose it or a floor guaranteed it a slot. A row here that no passage-search row names is EXPECTED: a far wider pool, re-ranked, with floors the passage search never applies.
candidate pool: the answering read scored 19 passages, out of a ceiling of 80 set gateway-side. Both figures are reported by the gateway for THIS ask — measurements of what actually happened, not constants read out of anyone's source and printed here on faith.
floors armed for this ask: the base rulebook — armed · surviving errata corrections — not armed · one slot per book on a multi-book read — not armed. Every floor is stated on every ask, armed or not, so a floor that did not apply here is never mistaken for one this app does not know about.
arms fused for this ask: vector + full-text + the doc-priority boost arm. Stated separately from the rows because an arm that ran and carried nothing is a different fact from an arm that never ran.
| read at | document | heading | page | fused rank | fused RRF score | carried by | how it got its slot |
|---|---|---|---|---|---|---|---|
| #1 | RULEBOOK Casino — Rulebook (corpus doc 226) | CASSINO. | p.3 | #9 | 0.029762 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | vector + full-text | the re-ranker chose it |
| #2 | RULEBOOK Casino — Rulebook (corpus doc 226) | SPADE CASSINO. · DRAW CASSINO. | p.7 | #7 | 0.044583 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | vector + full-text + the doc-priority boost arm | the re-ranker chose it |
| #3 | RULEBOOK Casino — Rulebook (corpus doc 226) | p.5 | #4 | 0.04598 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | vector + full-text + the doc-priority boost arm | the re-ranker chose it | |
| #4 | RULEBOOK Casino — Rulebook (corpus doc 226) | CASSINO. | p.3 | #6 | 0.045228 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | vector + full-text + the doc-priority boost arm | the re-ranker chose it |
| #5 | RULEBOOK Casino — Rulebook (corpus doc 226) | p.4 | #16 | 0.026876 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | vector + full-text | the re-ranker chose it | |
| #6 | RULEBOOK Casino — Rulebook (corpus doc 226) | TWENTY-ONE POINT CASSINO. | p.6 | #8 | 0.044563 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | vector + full-text + the doc-priority boost arm | the re-ranker chose it |
| #7 | RULEBOOK Casino — Rulebook (corpus doc 226) | CASSINO. | p.3 | #14 | 0.027598 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | vector + full-text | the re-ranker chose it |
| #8 | RULEBOOK Casino — Rulebook (corpus doc 226) | p.4 | #3 | 0.046486 (healthy, below this ruling's 3-arm ceiling 0.04918 (the measured band 0.016–0.033 was taken on two-arm asks and does not bound a 3-arm one)) | vector + full-text + the doc-priority boost arm | the re-ranker chose it |
chosen chunks — what the answer actually cited
token count per chunk: NOT DISCLOSED — chunk token counts are not reported on the ask path. The character counts below are of the bounded snippet/quote rulesage actually received — a real measurement of what we hold, never an estimate of the chunk.
| # | document | heading | page | quote chars |
|---|
generation — timing and the path taken
- timing split
- retrieval 206 ms · first delta 1426 ms · total 1540 ms (retrieval and first-delta both measured from the START of gateway work, so first-delta INCLUDES retrieval — do not subtract)
- full breakdown
- queued 0 ms · retrieval 206 ms · ask 1332 ms · gateway work 1540 ms · total 1540 ms
- tokens in (the prompt the answerer read)
- 3890
- tokens out (what the answerer generated)
- 9
- tok/sec
- 180.0 tok/sec — 9 output tokens in 0.050 s of generation
- live seat (read from the running child)
- answering — gemma4:26b think:false · num_ctx 32768 (live-verified 1s ago)
- answerer host (measured by rulesage)
- reachable (for 22h, checked 1s ago)
- answerer node (the seat proxy's own header — this probe, not this ask)
- node1
- stream path
- streamed — the gateway pushed at least one accumulated draft
- watchdog
- first-token watchdog: 30s per arming window (rulesage.table.first-token-watchdog-seconds), re-armed once by the passages frame — so the worst case before a reader falls back to polling is 60s. Whether it actually fired is a BROWSER-side decision that never reaches the server, so it is not recorded here. The stream path above is the server's half of that story: whether a delta was emitted at all.
- num_ctx · think posture
- both are the CHILD's settings, not rulesage's, and no ask reply carries them — an answer never states which context window or think posture produced it. The live-seat line above reads them from the running child instead, where the gateway can answer for itself; it describes the child NOW, not this ask.
disposition
- outcome
- ABSTAINED — Abstained — no grounded answer; closest passages below
- abstain rationale
-
I don't have enough information.
BOILERPLATE, not reasoning: this is the answerer's fixed abstention phrase, the same sentence it emits whenever it declines. It records THAT the answerer declined and never why — no per-ruling reason was captured, because none was sent. Hostile corpus/LLM text, rendered escaped
setup-shape — why the checklist band did or didn't render
- decision
- NOT_SETUP_SHAPED — no setup-shape pattern matched this question
errata registry — what the gate above had to match against
this game registers NO errata cards (edition 152), so "no registered errata card matched" above could never have matched one
Errata scope: the gate above matched this ruling's own edition only. When an ask also reads an expansion, that expansion's own registered errata cards stay outside the match — a stated phase-1 limitation, not a failed lookup.