0. Five-second answer
Keep the original 40. Keep attacking on 4 PP and 6 PP. Just do not dump your entire hand into one board wipe.
Against Pirate Royal, the clean plan is to stop the followers that create Flags, then send the rest of your damage to the enemy leader.
The key correction is simple:
“Do not develop on 4/6 PP” is wrong.
Develop a first wave while keeping a second wave in hand.
If everyone goes home because a board wipe might happen, your company collapses before the pirates do.
1. Plain-language glossary
- PP: the points you spend to play cards that turn.
- Board: the followers currently in play.
- Go face: attack the enemy leader instead of a follower.
- Board wipe / AoE: one effect that damages several followers at once.
- Storm: a follower can attack the enemy leader immediately on the turn it is played.
- Lethal: you can win this turn with the damage you already have.
- Mulligan: your opening-hand redraw.
- Second wave: cards you deliberately keep in hand so you can rebuild after a wipe.
You are not taking a vocabulary exam. The actual exam is “make the other leader reach zero.”
2. What was tested
The local study compared 96 Aggro Nightmare candidate lists and four play plans against 18 fixed public Pirate Royal lists from the official Deck Portal corpus.
There were 5,904 engine games in total. The holdout stage contained 1,944 games; the final baseline itself played 648 of those.
| Metric | Final 40 + normal plan |
|---|---|
| Win rate over 648 holdout games | 91.4% |
| Worst Pirate list | 69.4% |
| Bottom-five CVaR | 83.3% |
| Going first | 92.6% |
| Going second | 90.1% |
| First/second gap | 2.5 pts |
| Average finish | Turn 7.36 |
| Robust score | 0.8489 |
Do not read 91.4% as a real ladder win rate. It is a local deterministic human-prior-v1 comparison. Its useful message is that the unchanged list and normal policy beat the tested alternatives under the same rules.
3. Final 40: zero changes
- 3× 渇仰の悪魔
- 3× ゴーストエスケーパー
- 3× デーモンドラム・ラズ
- 3× タイトロープキャット
- 3× 夜の唱のライブ
- 3× 傍死のアナテマ・徒姫
- 3× キュートな悪魔・リリム
- 3× 強襲の特攻隊長
- 3× 鉄鎖の繋がり
- 3× 激烈の副総長
- 3× ガロダートVSゼット
- 2× 魅惑のサキュバス・リリム
- 2× 誠実なる呪い屋・スージー
- 2× 幽冥の中尉
- 1× ダークナイト・マーシャ
Decision: KEEP_BASELINE. Nothing added, nothing removed. The matchup-specific optimization ended by recommending no matchup-specific deck change. Very efficient meeting: ninety-six candidates later, “leave it alone.”
4. Mulligan
Usually keep
魅惑のサキュバス・リリム, 渇仰の悪魔, ゴーストエスケーパー, デーモンドラム・ラズ, キュートな悪魔・リリム, 強襲の特攻隊長, 幽冥の中尉.
Conditional keep
鉄鎖の繋がり if you already have an early play and want an immediate answer to an important Pirate follower.
Usually send back
傍死のアナテマ・徒姫, 激烈の副総長, and duplicated expensive cards.
A hand full of late cards is a pirate ship still stuck in the parking lot.
5. The big correction: you still develop on 4 PP and 6 PP
アージュドール makes one Flag and deals 2 to all opposing followers at 4 PP. At 6 PP, it makes two Flags and deals 4 instead.
That does not mean “play nothing.” It means:
- Build a board and keep pressure.
- Often aim for roughly two useful bodies, adjusted for health and your hand.
- Do not empty your hand.
- Prefer cards that do something immediately, direct damage, or immediate leader attacks.
- If the opponent spends 4 or 6 PP wiping, rebuild with the cards you saved.
The rule is not “do not eat the snacks.” It is “do not eat the entire field-trip bag three minutes after the bus leaves.”
6. Turns 1–3: board plus leader damage
Use cheap cards to build your board and chip the enemy leader. You do not need to clean every enemy follower. Remove the pieces that seriously disrupt you or generate more Pirate value, then send spare damage to the leader.
Ask in this order every turn: Can I win now? If not, which enemy follower actually matters? Then how much damage can still go to the leader?
7. Turns 4–6: first wave plus second wave
This is not the “stop attacking” phase. It is the “attack without donating your whole hand” phase.
Respect the 2-damage wipe around 4 PP and the 4-damage wipe around 6 PP. Keep some refill, immediate damage, or Storm in hand. If the opponent pays a large chunk of PP to clear wave one, wave two comes down next.
One lifeboat is clever. Putting the entire crew into the same sinking boat is not.
8. What to remove: stop the Flag factory
The simulation play guide prioritizes:
- 外洋の斥候
- 渦潮の砲手
- 荒波の副船長
- バルバロス
The Flag itself is not something you solve by attacking it. It starts at Countdown 7, advances when Pirate Royal plays spells, and deals 2 to your leader when it expires.
So yes, the Flag is annoying. No, punching the Flag is not a menu option. Stop the people printing Flags.
9. 7 PP: Barbaros is the deadline
逆行の咎人・バルバロス costs 7 PP, creates another Flag, advances all of Pirate Royal's Flags by 5, and can attack immediately.
Before that turn, count Flag damage, remaining spell acceleration, Barbaros's attack, and your own damage from hand. The simulator's average finish was turn 7.36, so the preferred mindset is to end the game rather than plan a long vacation after Barbaros arrives.
10. Cards to hold
When possible, keep one copy of:
- 夜の唱のライブ
- ガロダートVSゼット
Spend them if they produce guaranteed lethal now or if holding them would let the next Pirate board kill you.
Before self-damage takes your leader to 12 or below, check whether ガロダートVSゼット can be used this turn or next. While in hand, it gets cheaper at the end of your turn when you are at 12 or less; it can become an immediate finisher or a large defensive reset.
11. How you actually finish
Do not demand all 20 damage from a permanent board. Early damage comes from the board; late damage increasingly comes from your hand.
Every turn add up current attacks, direct damage, immediate attacks, 夜の唱のライブ, and ガロダートVSゼット. A wiped board is not the end of your offense. It is the cue for the second wave.
12. Why Marsha → Baal was rejected
The last surviving swap was 1× ダークナイト・マーシャ → 1× 元素の共鳴・バアル.
On unseen seeds, baseline was 91.4%. The Baal tempo plan fell to 88.0% (-3.4 pts; 95% paired interval about -6.3 to -0.8). The race plan fell to 87.2% (-4.2 pts; about -7.6 to -1.2).
The worst single matchup improved a little, but average performance, bottom-five performance, and first/second balance got worse. It was basically installing a better fire alarm and discovering the front door no longer opens.
13. The six-Legendary wallet problem
This 40 contains six confirmed Legendary copies:
- 3× 傍死のアナテマ・徒姫
- 3× ガロダートVSゼット
At 3,500 Red Ether per normal Legendary, crafting all six from zero costs 21,000 Red Ether. Pirate Royal may be the opponent, but your wallet is the one getting raided.
Important: the 96-list experiment optimized matchup performance, not crafting cost. It did not impose “maximum three Legendaries” or “under 10,000 extra Ether.” A budget version therefore needs a separate cost-constrained simulation. This article does not pretend that an untested cheap cut is proven.
14. One-screen cheat sheet
Count lethal first.
Turns 1–3: cheap board + leader damage.
4 PP: respect 2 AoE, but keep attacking.
6 PP: respect 4 AoE, but keep attacking; save wave two.
Ignore the Flag itself; remove the Flag makers.
Try to keep one 夜の唱のライブ and one ガロダートVSゼット.
Finish around the 7 PP Barbaros deadline with damage from hand.
One line: Stop the Flag factory, not the Flag. Attack on 4/6, keep a refill, and end it by 7.
15. Scope and sources
Local run: set09-aggro-nightmare-vs-pirate-royal-v1; 9 public Aggro Nightmare submissions as candidate input, 18 Pirate Royal public lists, 96 candidate decks, four play plans, 5,904 games total, 100% rule coverage, zero unsupported cards, zero rule gaps, zero hidden-information violations, and 23 SHA-256-tracked files.
Current card-effect checks used Cygames' Set 9 material plus Game8/GameWith pages for アージュドール, バルバロス, ガロダートVSゼット, 傍死のアナテマ・徒姫, and current Red Ether crafting rates.
