Refine the list and choose careful crafts

Choose one repeated problem

Use your playtest log to name the issue. “I often miss an early answer” gives you a job to fix. “I lost my last game” gives you too little information.

Start with a flex slot when possible. Before cutting a card, check its job, its support links, and any tutor that can find it. A weak-looking piece may be keeping an important package working.

Write the cost of a swap

An extra early answer can help you survive. It also takes space from a threat, draw tool, or late card. Say what the new card adds and what the old card provided.

For example: “I will replace one slow extra threat with an early answer. I expect safer starts, but fewer threats in long games.” This is a test, not a guaranteed gain.

Save the old version. Change a small number of cards so you can understand the result. Test one change at a time when the deck's main plan already works.

Use narrow answers for a reason

A tech card answers a certain problem. It may do little in other games. Record how often the problem appears and whether the tech card is useful when you draw it.

Facing one unusual deck once is weak evidence for changing your whole list. Review several games and the mix of opponents you actually meet.

Decide whether to craft

Separate needed cards from optional upgrades. Ask whether the card completes the plan, whether you will use the deck, and whether an owned card can fill the same job.

Waiting is a valid choice for an untested idea. A theorycraft may change after its first games. Use the collection guide before spending dust or giving up cards you may want later.

Keep or undo the change

Compare the new version with the question you wrote. Review games where the swap helped and games where its cost mattered. A short win streak alone cannot settle the choice.

Keep changes with a clear purpose and useful evidence. Undo a swap when it damages the main plan more often than it fixes the target problem. Then choose the next question.