Draw and discard — the turn-window math

A turn on a rummy table has three windows. The first window is the draw: the seat takes a card from the closed deck or the open discard pile. The second window is the discard: the seat places one card face-up on the open pile. The third window is the show (only on the final turn): the seat places the 14th card face-down and announces a valid show. The chapter walks through the priority order the desk has read on published flows.

ScopeDraw · Discard · Show window
Last reviewed06 Aug 2026 · Edition 342
Editorial distancePriority order, not a winning system

The draw, the discard and the turn window

A draw-and-discard turn-flow diagram on a print-journal editorial rail

A turn on a rummy table has three windows. The first window is the draw: the seat takes a card from the closed deck or the open discard pile. The second window is the discard: the seat places one card face-up on the open pile. The third window is the show (only on the final turn): the seat places the 14th card face-down and announces a valid show.

The draw priority is the closed deck first. The published research on rummy turn policy consistently shows that a closed-deck draw is the most common draw on a Points table; the published rate the desk has read is one closed-deck draw for every four discards at the lower stake band. The closed-deck draw is the seat’s protection against picking a card the table has read.

The open discard is the seat’s read of the table. A discard pick is a read of the card the previous seat has decided not to keep. A discard pick is also a read of the card the previous seat wants the table to pick. The desk flags the open-discard pick as a high-information move; the seat that picks the open discard is also the seat that is most likely to be read by the next seat.

The discard priority, in order

The first discard priority is the deadwood card that does not fit any meld. A high-point deadwood card that cannot be grouped is the seat’s first discard. A high-point deadwood card that could be grouped with a future draw is the seat’s second discard. A low-point deadwood card that could be grouped with a future draw is the seat’s third discard.

The second discard priority is the joker that cannot anchor a meld. A joker with no base card is a wild card; the joker without a base card carries zero meld value and is the seat’s second discard priority. The joker with a base card is part of a meld and is not a discard priority.

The third discard priority is the high-point card that breaks a pure sequence. A high-point card that breaks a pure sequence is the seat’s third discard priority; the seat that discards the high-point card is opening a pure-sequence slot. The pure-sequence slot is filled by a future draw or a discard pick.

The fourth discard priority is the middle-of-sequence card. A middle-of-sequence card that breaks a long sequence is the seat’s fourth discard priority. The seat that discards the middle-of-sequence card is closing one sequence and opening two shorter ones; the future draw determines which of the two shorter sequences is built out.

The observation read of the discard stream

The discard stream is a public record. Every card the table has discarded is face-up on the open pile, and the seat that picks from the open pile is a read of the table. A discard pick is a read of the card the previous seat has decided not to keep. A discard pick is also a read of the card the previous seat wants the table to pick. The desk flags the open-discard pick as a high-information move; the seat that picks the open discard is also the seat that is most likely to be read by the next seat.

The show-cycle read is the count of discards since the last show. A show-cycle read of zero means the table is on the first deal of a new round. A show-cycle read of one means the table is on the second deal. The desk has read published research that shows a seat that tracks the show cycle is materially more likely to make a correct drop decision than a seat that does not track it.

The seat-position read is the relative position of the seat to the dealer. A seat to the dealer’s left is the first to act on every deal. A seat to the dealer’s right is the last to act before the dealer. The first-to-act seat has the most information at the draw window and the least information at the show window. The last-to-act seat has the least information at the draw window and the most information at the show window.

A discard-priority read is the read of the discard stream against the seat’s own hand. The seat that has built a clear pure-sequence path by turn three will pick the open discard only if the discard extends the path. The seat that has not built a clear pure-sequence path by turn three will pick the open discard only if the discard opens a new path. The desk flags the misread that a discard pick is independent of the seat’s own hand; the discard priority is route-relative to the hand, not the stream.

A seat-rotation read is the read of the seat position against the discard stream. A first-to-act seat on a fresh deal has a closed-deck draw; the seat cannot read the discard stream because the open pile is empty. A first-to-act seat on a mid-deal has an open discard pile; the seat can read the discard stream from the previous seat’s discard. The desk flags the mid-deal open-pile read as a high-information move; the seat that reads the open pile is also the seat that is most likely to make a correct drop decision.

Operational reference for this page

The pages on boltfantasy.com are written by the editorial desk. Each page walks through the published material on a single topic; the operator-specific UI is the source of truth and must be verified on the operator’s own published material before the reader commits a rupee. The pages are not legal or financial advice; the pages are editorial guidance on the published rules, the published drop ladder, the published chip math, the published safety surface and the published comparison dimensions. The reader verifies the operator’s current page before relying on any figure listed here.

A published number is a number the operator has published on the operator’s own page. The desk has read the published numbers on a small set of operators; the desk does not synthesise or extrapolate. Where the operator publishes a number, a window or a document list, the desk records it; where the operator does not, the desk records the gap. The reader uses the desk’s record to make their own comparison; the desk does not publish a winner.

A published window is a period the operator has published on the operator’s own page. The desk has read the published windows on a small set of operators; the desk does not synthesise or extrapolate. The reader verifies the operator’s own page before relying on any window listed in the editorial chapter. The comparison ledger records the windows the desk has read against a published primary source.

A published document list is a list the operator has published on the operator’s own page. The desk has read the published document lists on a small set of operators; the desk does not synthesise or extrapolate. The reader verifies the operator’s own page before relying on any document list listed in the editorial chapter. The comparison ledger records the document lists the desk has read against a published primary source.

A restricted-state list is a list the desk has read against a published primary source. The restricted states for real-money rummy, as of the desk’s last review (Edition 342, 06 Aug 2026), include Telangana, Andhra Pradesh, Assam, Odisha, Sikkim, Nagaland, Meghalaya and Tamil Nadu. The desk’s list is a desk record, not a legal interpretation. The reader verifies the local state gaming act before playing. A state change to the restricted list is recorded in the news register with the source date and the desk’s verification status.

A support line is a contact the desk has read against a published primary source. The desk does not operate a support line; the desk links to the published support lines operated by recognised public-health bodies. The reader is encouraged to contact a support line if the play has stopped being a leisure activity. The chapter’s last review date is recorded on the page footer.

A review cadence is the desk’s published review window. The desk reviews each chapter on a quarterly cadence. The chapter’s last review date is recorded on the page footer; the reader verifies the local rule before relying on the chapter’s published figures. A chapter that has not been reviewed in the last quarter is flagged in the chapter’s footer; the reader verifies the operator’s current page before relying on the chapter’s published figures.

A red flag is a published signal the desk has read against a primary source. The desk flags any operator that does not publish a deposit cap, a session timeout or a self-exclusion switch. The desk flags any operator that publishes a dispute window shorter than the desk’s published baseline. The desk flags any sign-in surface that does not match the operator’s published domain. The comparison ledger records the red flags the desk has read against a published primary source.

A primary source is a document the operator or regulator has published on its own page. The desk reads the primary source and records the desk’s own test ticket where the operator publishes a verifiable flow. The desk does not synthesise or extrapolate. The reader verifies the operator’s own page before relying on any figure listed in the editorial chapter.

A chapter cross-reference is the link between chapters on a single topic. The format-split chapter is referenced from the games hub, the strategy hub and the reviews hub. The drop-vs-show chapter is referenced from the strategy hub and the games hub. The dispute-handling chapter is referenced from the reviews hub and the safety hub. The reader uses the cross-references to navigate to the chapter that matches the decision they are facing. The cross-references are not commercial links; the cross-references are the desk’s own editorial navigation.

A reading order is the desk’s recommended sequence for a reader who is new to the chapter. The reader starts with the chapter’s first section, walks through the published material, and finishes with the operational reference. The reading order is not a personal recommendation; the reading order is the desk’s own editorial sequence. The reader may skip sections they are already familiar with; the reader verifies the operator’s current page before relying on any figure listed in the chapter.

A glossary entry is the desk’s published definition of a term used in the chapter. The glossary entries are linked from the chapter’s operational reference; the reader uses the glossary to look up an unfamiliar term without leaving the chapter. The glossary is not exhaustive; the glossary covers the terms the chapter uses most often. The reader uses the desk’s broader editorial chapters to look up terms the chapter does not define.

Editorial pickBolt first-party route
Open Editorial Pick