A discard is a choice. The next card is not.

Define the boundary in three beats

Beat one: you may choose which card leaves. Beat two: you may choose a first drop or a middle drop if the window is open. Beat three: you may choose to wait for a pure sequence before a show. Stop there. The named wild, the opponent’s hold and the next stock card do not take instructions.

A worked discard, then an uncertain twin

You hold Q of spades with no matching jack or king and no second queen. Returning the queen is a legal reduction of deadwood. The alternative is keeping the queen because someone might discard a matching queen later. That second path is legal and uncertain. This desk will not dress the second path up as a tip.

Deadwood points are the only honest scoreboard mid-hand

If the opponent showed now, what would you pay? That number, not a story about momentum, is the mid-hand fact. High unmelded face cards are expensive on Points and they are fuse-length on Pool. Count them in pips. If you cannot count them, you are not choosing. You are hoping.

Middle-card holds have a cost

A 7 that could join 5-6 or 8-9 looks useful. It also sits as 7 deadwood if the table ends now. Holding it is legal. Calling it correct requires knowledge of the closed deck that nobody at the table has. Mark it as a hold with a price, not as a principle.

Drop timing is a limit, not a personality test

First drop exists because some deals are already expensive before you take a closed card. Middle drop exists because new information can still be bad. Waiting for a miracle third life is not a strategy. It is the absence of a pre-set stop. Put the stop on the limit panel before the deal, not after the third face card.

What this column refuses to do

It will not publish captain-style pick lists. It will not rank Luckywin tables. It will not claim a discard sequence that beats other players. Those claims need data this desk does not have.

How this differs from how to play

How to play asks whether a finished spread is legal. This column asks which unfinished choices are still yours. A player can know the pure-sequence rule and still treat every hold as destiny. That is the friction this desk removes.

After you accept the boundary

If you wanted format math, open the format desk. If you wanted a sample show, open how to play. If you wanted a promise, stop. There is not one on this desk.

Price the hold in pips

A connector that might become a run still counts as its rank if the opponent shows now. Write that rank. If the number is larger than a middle drop, you are paying for a story about the closed deck.

The legal column is the discard you can explain in one sentence. The unknown column is everything that sentence cannot reach. Keep them on opposite sides of the paper divide.

Stop rules are strategy

A pre-set loss number is a discard of the next buy-in. It is as much a choice as returning a queen. People treat limits as mood. This desk treats them as legal choices made early.

Define the boundary in three beats

Legal choice: which card leaves your hand. Unknown outcome: which card arrives next. Visible cost: deadwood points if a show happens now. Strategy on this desk stays inside that boundary. It does not sell a guaranteed winning line.

A worked discard names why a card is safer to release than its twin. An uncertain twin stays uncertain. Middle-card holds have a cost when they block a pure sequence. Drop timing is a limit, not a personality test.

Deadwood points are the only honest scoreboard mid-hand. Pretty shapes that still owe a pure run are not safer than an ugly hand that can show. Open how to play if the pure-sequence rule is still fuzzy.

What this column refuses to do

It will not promise a hit rate. It will not rank opponents from a blank file. It will not turn format math into a lucky charm. Points, Deals and Pool change how expensive a wrong show is; they do not change the pure-sequence rule itself. Format cost lives on the format desk.

How this differs from how-to-play: teaching explains validity. Strategy assumes validity and talks about choices under uncertainty. Mixing those jobs produces confident language on hands that cannot show.

After you accept the boundary, set a stop that matches the clock on the limit panel. A clean discard plan with no loss cap is still an unfinished session plan.

Carry the boundary back to the lobby

If the lobby hides invalid-show cost, strategy cannot price a failed claim. Fix the lobby rows first via the usability checklist and the format ledgers. Do not use strategy prose to paper over missing product facts.

Short sessions favour fewer speculative holds. Long Pool fuses punish failed shows harder. Choose the clock on the compare, then return here for discard boundaries inside that clock.

This column ends when the next card is unknown. That sentence is the product. Anything that pretends the next card is known has left the desk.

Deadwood as a mid-hand scoreboard

Count unmelded points before you admire a shape. A hand that looks one card from beauty can still sit on a high deadwood number if the pure sequence is missing. The scoreboard does not grade beauty. It grades leftover pips when someone shows.

When two discards look equal, prefer the one that lowers deadwood without breaking your only pure-run path. When you cannot tell, you are inside the unknown half of the boundary. Do not narrate certainty you do not have.

Format multiplies the cost of a wrong show. It does not change the boundary. Price the show on the ledger for your clock, then discard inside the boundary you can defend.

Questions this desk can answer

Can a discard note promise a win?

No. A note can name a legal choice. It cannot name the next closed card or an opponent’s hidden set.

Is holding a middle card always correct?

No. Holding a connector can reduce deadwood or it can trap high pips. The outcome stays unknown.

Re-check a sample show, then set a loss limit before another buy-in.