POKEMAS LAB GUIDE

How to use the firepower team optimizer

Written by SciPocket · Updated: October 20, 2026

Choose the pairs you want to use, then find a team and action order for your battle. The current setup has five steps: Battle → Pairs → Training → Actions → Review. Blank slots can be filled automatically, so you can start with just one pair.

Open Team Firepower Explorer ↗

Find your first team

1. In Battle, choose a format and evaluation period. If unsure, start with Optimize by sync move count through the sixth sync move.
2. In Pairs, select one pair you want to use. You can leave the other two slots blank.
3. In Training, match move levels, EX settings and the grid energy limit to your game. You can start with automatic grid selection.
4. In Actions, you can leave sync move users automatic and continue with Review these settings.
5. In Review, check your conditions and press Find a team of three. Inspect the selected grids and action order in the results too.

1. Choose the battle and evaluation period

Choose Damage Challenge, Gym battle, Tower, or Optimize by sync move count in the first step. Even the same team can produce different results when the time limit or opponent conditions change.

Gym battle ticket settings

One ticket sets 90 seconds and starting sync buff 0; two set 180 seconds and buff 2; three set 270 seconds and buff 5. One and two tickets are for Extra Battles. Check cheer, special rules and opponent passives separately. Choosing Normal allows only Normal-type damage at neutral effectiveness.

Time-based evaluation runs until the time limit; sync-count evaluation runs through the selected sync move. Both include the first sync move and attacks during setup. Time-based evaluation does not stop at six sync moves and includes attacks after the last sync move until the deadline.

Single-target versus total damage

Single-target damage compares damage to the central opponent; total damage sums damage to all opponents. Selecting both runs separate searches. Total-damage searches also consider single-target moves, so a team with a strong single-target move can rank higher.

2. Choose pairs and a search mode

Use Find a team including these pairs (optional) to select one or two pairs and fill the rest automatically. If you already have three, choose Calculate only the three selected pairs. To compare alternatives, choose Compare the selected trio with other teams.

Physical/special filters apply to the main attacker’s moves. They do not require supporting pairs to have the Support role. To evaluate only physical or special damage across the entire team, also check physical/special damage immunity in the opponent settings.

Exclude unowned pairs in the roster selector. All pairs are eligible initially. Rarity/acquisition and optional type filters are combined with your roster. Free-pair filters include past distributions and do not guarantee current availability.

You can also describe your conditions in a sentence. Check the interpreted pairs and battle settings, then correct them in the setup steps if needed. A text request does not guarantee that every condition was interpreted as intended.

3. Match training and grids

Check the default move level, ★6 EX/EX role and Superawakening. Selected pairs can have individual move levels; automatic slots use the default. Superawakening is evaluated under Move Lv. 5 conditions.

The current default is automatic grid selection with an energy limit of 70. If you have not expanded energy in the game, choose your available limit from 60 to 70. The limit is a maximum, not a requirement to spend all energy.

Automatic prioritizes the team, damage-focused prioritizes the pair itself, and support-focused prioritizes contributions to the other two pairs. Open Review or edit grid to inspect the board. To lock only that pair’s grid, set its individual grid policy to Manual; the others can stay automatic.

Automatic grids compare candidates with checked unlock conditions and connection paths. This does not exhaustively optimize every panel combination. Use View selected grid in the results to inspect the panels actually used.

4. Set sync users and actions

You can choose the first, second and later sync move users. Leave them blank to search automatically. For a Support A → Support B → attacker sequence, enter each pair in the corresponding field.

The action search checks regular/S-move requirements and remaining uses. Review the available additional settings when an S-move requires HP, received hits or other conditions. Unspecified requirements are not all assumed to be satisfied.

Timing and field durations are estimates

A cycle contains three ally actions and one opponent action, with 19 seconds as the default. Each action takes one quarter of that estimate. Sync moves are separate from the three ordinary slots and may activate at countdown 0 when the action-order constraints permit. Consecutive actions by the same pair are normally blocked, including across opponent actions and sync moves. In time mode, only the last cycle containing an ally action permits immediate post-sync follow-ups, consecutive actions and multiple ordinary slots for the same pair. The timer is approximated as paused during sync animations; this is not an exact reproduction of in-game timing.

Weather, terrain, zone and circle expiry are checked at each attack. Field effects and type rebuffs are not applied retroactively to earlier attacks. When changing timing settings, also inspect remaining field durations and the action order.

5. Review conditions and search

The review screen summarizes the battle, pairs, training, grids and sync users. Your inputs remain when returning to earlier steps. Start with the standard search depth and use the wider search when you want more candidates.

Press Find a team of three to start. If it takes too long, select one or two pairs or narrow your roster. You can stop a running search. Run a new search when you change conditions and want to compare other teams too.

6. Read scores, grids and action order

After searching, only the top three teams receive an additional comparison of physical/special boost timing. Training settings and sync grids stay unchanged, and a new action order is adopted only if its damage index improves. This does not guarantee the best possible action order.

Check the per-sync interval score, cumulative score, estimated elapsed time and sync user, rather than just the ranking. The B1 interval includes attacks before the first sync move and that sync move itself. In time-based results, the final interval covers attacks after the last sync move until the deadline.

Inspect individual actions, countdown and sync buffs in the action details, and activation/expiry times in the field and rebuff history. Remaining-use changes show consumption of limited-use moves. Read the unevaluated effects and assumed conditions too.

The base comparison value is a damage index, not actual HP damage or DPS. Damage Challenge also uses damage estimates assuming Defense 83, Sp. Def. 83 and stat-drop mitigation 5. Do not treat these as interchangeable with general indices; compare results under the same displayed evaluation conditions.

7. Recalculate and share an article’s team

Open the Team Builder from an article’s result card to load the shared three pairs, grids, calculation conditions and action order. Check the original conditions first. After changing move levels, use Recalculate this trio. Automatic grids are selected again; check unlock conditions for manual grids.

The trio in a shared link is separate from your own roster. To search your other pairs, use Return to my roster and run a new search. Share this result creates a link containing the team, grids, conditions and result, without your entire roster.

Shared results replay the action order with the current calculation code. Updates or conditions can change values, and older links may retain earlier evaluation settings. Check the article date and result assumptions.

8. Scope and limitations

The model handles move power, offensive stats, stat ranks, sync buffs, supported passives/grids, boost consumption, S-move unlocking/resealing, use limits, countdown acceleration, fields and per-target rebuffs. Defense/Sp. Def. drops are reflected when supported evaluation settings can track them deterministically; drops from an attack take effect after that attack.

The search limits candidates, grids and action orders; it does not guarantee a global optimum. Gauge waiting, HP changes, survival, every opponent action, probabilistic recovery and some unique multipliers are not fully reproduced. Even a high-scoring team needs a separate check that its action order can be executed in battle.

Gym opponent passives are separate from weakness and special rules. R3 is treated as fixed conditions including R1/R2, without simulating knockouts or round transitions. Compare like with like: one versus three opponents, physical versus special, and sync-count versus timed evaluation.

Open Team Firepower Explorer ↗

Understanding damage indices and comparison settings (SciPocket) ↗