Minimal-Change Team Repair

Diagnosis comes before repair

A repair needs a target. If you cannot say what the change is supposed to solve, you are experimenting. Experimentation is fine, but it should not be confused with a controlled fix.

Good repair targets are concrete: an explicit coverage gap, a priority-access issue, a missed speed floor, or another supported condition. The narrower the target, the easier it is to judge whether the edit succeeded.

Small changes preserve cause and effect

If one move changes and the rest of the team stays the same, the before-and-after comparison is easy to understand. The same is true for a limited EV adjustment or another bounded field. When four slots change at once, even a better result does not tell you which edit actually mattered.

There is also a practical benefit: you keep the sequencing, roles, and matchup knowledge you already developed with the team. Constant total rebuilds make every new problem feel unfamiliar.

There is another advantage to small edits: they are easier to undo. If a new move solves the target but makes two common positions worse, you can return to the previous set without reconstructing the whole team. That reversibility is useful when you are testing ideas rather than pretending every candidate is a permanent upgrade.

A candidate list is not a ranking

PokeTeamOn can generate deterministic candidates for supported repair targets. The list is not a secret competitive leaderboard. The first item is not guaranteed to be the best strategic choice, and candidate order should not be treated as a recommendation score.

Two legal moves can solve the same coverage target while having very different consequences for the set. One may preserve utility; another may remove something the team depends on. The tool can show the proposal and re-evaluate the target. You still choose.

Preview before you touch the working team

The repair flow is preview-first for a reason. Compare the old value with the proposed one, check the target being addressed, and look at the evidence after re-evaluation. If the change is a move, ask what the old move was doing. If it is an EV adjustment, inspect the stats that went down as well as the stat that went up.

A valid candidate can still be strategically wrong. Rejecting it is part of the workflow, not a failure of the tool.

Re-test the same problem

After a preview, rerun the exact condition that motivated the repair. If the target was a speed floor, keep the same floor. If it was type pressure, keep the same type. If it was coverage, keep the same target combination.

This gives the comparison meaning. Changing both the team and the test at the same time makes it impossible to tell whether the edit actually solved the original issue.

Set a change budget before you start

It helps to decide how much of the team you are willing to disturb before looking at candidates. If the team is already performing well, the budget might be one move or a small EV change. If the core idea is failing, one member replacement may be reasonable. Setting the boundary first prevents a narrow repair session from quietly turning into a complete rebuild.

A change budget is not a hard competitive rule. It is a debugging tool. You can always widen the scope later if the evidence says the smaller fix is not enough.

Check for collateral damage before accepting the idea

One improved signal can hide a new problem. A coverage move may remove recovery. A faster spread may make a defensive slot too fragile. A new Tera type may solve a weakness but compete with the team’s primary Tera plan.

Run the analyzer again, revisit the member’s role, and ask whether the cost is acceptable. Sometimes the best repair is no repair at all because the weakness is manageable and every available fix damages the team more than it helps.