Practice Case
Your current direction still has one anchor, but the backup just gained a second connected clue. What is the best table response?
Learn
When to stop chasing a weak plan and move to a better one.
A switch should have a reason you can explain. Useful reasons include a backup direction gaining support, the current direction losing its anchors, or public table information making the current path harder.
One disappointing draw is not usually enough. It becomes more meaningful when it joins other clues: your current anchors stop helping, the backup gains connected support, or shared support can move with you.
Keep the exact annual-card target private. In public-safe language, say that the current direction has weakened, the backup has gained evidence, or the rack still needs one more clue before moving.
Use a three-question comparison: What still supports the current direction? What actually supports the backup? Which tiles help both? If the backup has only one interesting tile, hold. If the current direction has lost anchors and the backup has connected clues, prepare to move.
Do not treat switching as a sign that the earlier decision was wrong. A good switch means the information changed. The skill is noticing that change before the old plan becomes a trap.
Use this six-step switch-or-hold routine: 1. Name the current anchor. 2. Name what weakened. 3. Name the backup evidence. 4. Save shared support if possible. 5. Move unsupported tiles into release. 6. Choose either hold for one more clue or switch now with a reason.
Run the routine after a meaningful draw, after a visible exposure changes table information, before discarding a tile that supports the backup, and after several turns without improvement. If the answer is still muddy, make the smallest release that keeps both reasonable options alive.
At a social table, use process language: I am keeping the shared support, or my backup finally has more evidence. Do not quote official-card hand text, values, notation, or layout details.
Practice Case
Your current direction still has one anchor, but the backup just gained a second connected clue. What is the best table response?
A player draws one tile that does not help the current direction and immediately wants to abandon the plan.
What should they check first?
Answer: Check whether the current anchors and shared support are still real before switching.
One bad draw is information, but it is not the whole rack. The player needs evidence from the full rack and table.
The current direction feels slow, but the backup idea is based on one isolated tile and no shared support.
Should the player switch?
Answer: Not yet. Hold or make a small release while waiting for the backup to earn real evidence.
Switching into an unsupported idea replaces one weak plan with another. A backup needs anchors or connected support.
Another player exposes a generic group, and the table information makes your current path feel harder. Your backup already has a cluster and shared support.
What does the visible information do?
Answer: It becomes one more clue in the comparison. If the current direction weakens and the backup is supported, switching may now be reasonable.
Visible information can change direction strength, but it should be combined with rack evidence.
Practice Case
You draw one tile that does not help your current direction. What is the best beginner response?
Practice Case
Which clue most supports switching direction?
Practice Case
Which tile group should you try hardest to preserve during a switch?
Practice Case
Which public-safe explanation is best when teaching switching?
A player has protected the same direction for several turns, but their best pair is gone, useful tiles have not appeared, and a backup cluster has grown.
What should they do next?
Answer: Compare the two directions and switch if the backup now has stronger rack evidence.
The decision is not emotional; several clues show the original plan weakening while another plan improves.
A player knows the current direction is fading, but they have spent several turns protecting it and feel reluctant to let it go.
What should guide the decision now?
Answer: The current rack and table evidence should guide the decision, not the effort already spent.
Past effort is not rack support. If the current direction has lost evidence and the backup has gained it, holding only because of sunk cost is risky.
A weak tile does not support the current direction, but it might support a backup. The player is about to discard without comparing both plans.
What repair habit helps?
Answer: Pause before discarding, compare current and backup support, then decide whether the tile is release, shared support, or a reason to switch.
A discard can close a door. The switch-or-hold routine helps the player release weak tiles without accidentally dropping backup evidence.