Short answer: Set 9 Aggro Nightmare (Abysscraft) is not a deck that keeps flooding the board from turn 1 to the end. It wins best in three gears: borrow the board early, stop fighting for it in the middle, then finish straight from your hand.
Early: Last Words followers stick around and hit face → Mid: spells go to the enemy leader's face → Late: 2-cost Storm, 4-cost Storm, and Garodeth VS Zeta close it out.
Winning the board back is not the goal. The moment the enemy leader dies, their whole board is gone anyway.
1. Bottom line: a burn and Storm deck that only rents the board early
Putting together recent games, public lists updated on August 27, 2026, and earlier local tests, the core looks like this:
| Job | Card | Current take |
|---|---|---|
| Sticks around early | Nether Lieutenant | lean 3 |
| Cheap instant damage | Wicked Lesser Mummy | lean 3 |
| Instant damage past Ward | Ansurge | lean 3 |
| Buffs your Storm finishers | Susie | lean 2 |
| Finisher | Garodeth VS Zeta | 2–3, depending on your collection |
| Mostly board clearing | Ghost Escaper | lean 0 |
| Mid-game board recovery | Tightrope Cat | lean 0 |
GameWith's August 27 list runs Lieutenant 3, Mummy 3, Ansurge 3, Susie 2, VS Zeta 3. Game8's list from the same time runs Mummy 3, Ansurge 3, Susie 2, VS Zeta 2. The 40 cards have not fully settled yet.
So don't hunt for one "perfect" 40. Look at the direction: cut board recovery, add instant damage.
2. Why Witch (Runecraft) suddenly got hard: Drain pushes your lethal back a turn
Experiment Witch's Ecstatic Talent summons Immersed Experiment, and when its Super-Evolve condition is met it can give that Experiment Drain (it heals the leader for the damage it deals). Immersed Experiment has Rush and can grow to 5/5.
For the aggro player, that's the worst case.
If you casually drop small followers on turns 5–6, Witch says:
"Thanks. I'll Rush into that. And heal while I'm at it."
You are basically delivering their healing punching bags to the door.
You lose differently to Witch now not just because your board gets cleared. Your board turns into their healing.
3. Gear 1: you still have to play real aggro early
Dropping the board later doesn't mean you can just throw spells from the start. On turns 1–3, followers have to do the first chunk of damage. If you can't chip the opponent enough here, spells and Storm alone won't reach later.
What matters is less how many followers you put down and more whether some damage is still left after they get hit by Rush or evolves.
Nether Lieutenant is a 2-cost 1/1, but its Last Words summons a new Lieutenant, gives it +1/+0 and Rush. Even if the first one is removed, your board isn't fully empty.
You keep the board not to control it, but so something is still there to hit face next turn. That's why 3 Lieutenants feel better than 2.
4. Gear 2: don't make "take the board back" your goal
Playing Ghost Escaper (a 3/2 with Rush) or Tightrope Cat, which leans on getting the board back mid-game, raises one question:
You took the board back. Then what?
This deck isn't trying to protect 20 life. Taking some hits actually moves you closer to VS Zeta's discount.
In an earlier test tuned only for the Pirate Royal (Swordcraft) matchup, with VS Zeta locked at 1 copy and no extra legendaries, all 3 Ghost Escapers and all 3 Tightrope Cats were cut. Lesser Mummy, Ansurge, Lieutenant and direct-damage cards went in.
That result was for Pirate Royal only, not a rule for every matchup. Still, it points the same way as these games.
In short: spend PP on the enemy leader, not on their board. The one exception: stop any damage that would surely kill you next turn.
5. Gear 3: from mid-game, spells to the face and Storm all at once
After the early chip damage, it's stronger to save instant damage in your hand and unload it in the same turn than to add one small follower every turn.
Wicked Lesser Mummy becomes a 2-cost Storm follower with Necromancy 4. Ansurge is a 4-cost Storm follower that can attack past Ward.
So late in the game you get: Opponent: "I played a Ward." Ansurge: "Don't care."
When Susie evolves, she gives a follower in your hand +3/+0. That makes Ansurge 5 Storm damage and VS Zeta 11.
Instead of running 3 Susies for a long board fight, about 2 copies, as a booster for your Storm cards, fit the way this deck wins now.
6. Against Witch: on turns 5–6, "don't play it" is an attack
This was one of the biggest takeaways from these games.
Against Witch on turns 5–6, try not to play followers you don't need.
Even if their Experiment has Rush and Drain, it can't turn Rush into healing if you have nothing on board to hit.
Meanwhile, use spells that hit face. Keep the 2-cost Storm, 4-cost Storm and VS Zeta in hand. Then dump them all next turn.
So: don't give them a board to heal off. Don't give them a safe turn to land a big stabilizing card. Don't drip damage out; stack it into one turn.
There was one big Witch answer card I kept calling "the lion" because of its art. I couldn't match that name to a card in the main public lists I checked, so this article doesn't name it officially.
The rule of thumb is still clear: if Witch gets one turn to stabilize, it suddenly gets rough. If that card doesn't show up, you can usually push through.
So on turns 5–6 there are spots where it's better not to play something just because you can, and save everything for one big lethal turn.
7. VS Zeta: 12 life is the gas pedal, 2 life is "you're not coming in"
While Garodeth VS Zeta is in your hand, its cost goes down by 1 at the end of your turn if your life is 12 or less.
20–13: taking some hits is fine
12: the discount switch turns on
12–3: the usable range for Storm mode
2 or less: Storm mode is dangerous
Storm mode deals 2 damage to your own leader.
At 2 life, it goes: "VS Zeta, 8 Storm damage, I win!" → you take 2 → you die before it even attacks.
VS Zeta wants you to take damage so it gets cheaper, then refuses to show up for work if you take too much.
So it's not "anything goes once I'm at 12 or less." Once the discount is on, keep at least 3 life if you plan to use Storm mode.
8. How many copies: Lieutenant 3, 2-cost Storm 3, 4-cost Storm 3, Susie 2
Nether Lieutenant: lean 3. You want to draw it early. Even if it's removed, a new one stays. The point isn't defense; it's leaving your next hit to the face on board.
Wicked Lesser Mummy: lean 3. A 2-cost Storm lets you stack damage late while using all your PP, like spell + Mummy or Mummy + Ansurge.
Ansurge: lean 3. 4-cost Storm that ignores Ward. It fits the "drop the board" plan almost too well.
Susie: lean 2. Strong, but she costs 4. Some earlier simulations kept 3, but I never ran a direct test locking her at 1, 2 and 3 copies. Current public lists also include 2-copy builds.
Garodeth VS Zeta: 2–3. Crafting one costs 3500 Red Ether. Going from 1 to 3 copies costs 7000. If your collection is thin, finish the cheap Storm cards first, get the 2nd VS Zeta next, and decide on the 3rd based on your Ether and the meta.
9. How this lines up with earlier simulations: read the direction, not a win rate
On day one of Set 9, a local comparison ran 2,324 games across 32 supported decks. Aggro Nightmare came out on top with an 81.0% average and 63.9% in its worst matchup.
But that model had limits, like a max of 8 turns. It does not mean you'll win 81 out of 100 ranked games.
A Pirate Royal-only study ran 5,904 games, plus 1,944 final check games on unseen random seeds, and the base 40-card list scored 91.4%. That number only covers 18 Pirate Royal decks, not real win rates across the whole meta.
What matters is what kept showing up under different conditions: deal damage early. Don't lean too hard on board control. Add more Storm and direct damage. It works even without every pricey legendary.
These games explained a lot about the reasons behind those numbers.
10. The Red Ether problem: your wallet dies before your opponent
As of August 2026, crafting costs Bronze 50, Silver 90, Gold 750, Legendary 3500.
Dismantling a normal legendary gives 1200, and a premium one gives 2500. On top of that, you can mostly only dismantle extra copies, so the old trick of "dust every unused legendary for a new deck" doesn't really work anymore.
So the deck's 41st card is Red Ether. Even though only 40 fit.
If you're short on Ether, finishing the cheap Storm and early-game cards first makes the deck more consistent than crafting a 3rd copy of an expensive card.
11. Fast games are great, but "losing doesn't cost rating" depends on your rank
This build plays fast games. If your early damage doesn't land, you see the loss coming early. If it lands, from about turns 5–7 you can focus on counting lethal.
According to the official help page, losing a normal Ranked Match does not lower your BP or MP. In Grand Master, though, your per-class rating (CR) goes up and down with wins and losses.
So: while you're climbing on BP/MP, you can grind lots of short games. Once you're chasing CR in Grand Master, losses hit your rating like normal.
12. Final takeaway: the board is a rental for the early game, not the goal
The biggest lesson is this:
Owning the board is worth nothing on its own. It's only worth as much damage as it does to the enemy leader.
Early, use Last Words cards like Lieutenant to leave damage that's hard to get rid of. But don't keep doing the same thing into turns 5–6.
Against Witch especially, don't feed Drain targets. Save your spells and Storm cards and throw them together. Turn on VS Zeta at 12 life, and finish before you drop to 2.
Early: stick around with Last Words and hit face
Mid: drop the board, spells to the face
Late: 2-cost Storm + 4-cost Storm + VS Zeta
Against Witch: on turns 5–6, not playing a follower is also an attack
Life: turn it on at 12, keep 3 or more
Deckbuilding: instant damage over board recovery
You don't need to control the board to win.
Killing the enemy leader is the strongest board clear in the game.
