Dhelmise deck: the core, the flex slots, and what beats it

What Dhelmise looks like when you open the box

Across the 35 lists we counted, this deck runs 21 to 28 Pokémon, 25 to 32 Trainers and 7 to 8 Energy. Its attacks are paid for in Psychic.

On results pages and on stream you will see it written « dhelmise-pbl ». Every card named in that label turns up in the tables below, which is what makes it a fair name for the deck.

The bit they tack on the end, « PBL », is a set code: Pitch Black. They spell it out because several cards share that name, and this deck plays that particular printing.

Look at the Energy: 7 to 8 out of 60, when half the 97 decks we track run 10 or more. This one is lean. It lives on cheap attacks and leans on Trainers to find the right Energy at the right moment rather than stockpiling it.

It sets up fast. Banette is a Stage 1 Pokémon: a Basic, one evolution on top, and you are attacking. Two cards instead of three. You lose fewer turns building, and you rebuild much more easily when an early knockout lands on you.

Look at the Trainers: 25 to 32 cards out of 60, against a median of 31 across the 97 decks we track. Inside that, 10 to 15 Supporters, and the rule there is strict — one a turn, even if you are holding three. The 11 to 16 Items you can play all of in the same turn. It also brings up to 4 Stadiums, and 57 of the 97 decks we track bring that many or more. Only one Stadium sits on the table at a time: the one you play sends the one already there to the discard pile.

What Tools and Stadiums do — the two Trainer families we skipped here — and why the Energy count says so much: how to read a deck list.

How Dhelmise actually wins

Banette attacks with Puppet Pull for 80. It needs 1 Psychic Energy attached, and it has 80 HP. And the number is not everything the card says. There is one more line printed under the attack, and it is the one everybody skips: "You may search your deck for a card and put it into your hand. Then, shuffle your deck."

One Energy to attack is enormous, and it means the attack cost will never be what slows you down. What slows you down here is getting Banette up. Once the line is standing you swing every turn with nothing to recharge, and you can start again immediately when your attacker falls.

Good news on mobility: pulling Banette back to the Bench costs you 1 Energy attached to it, and that is all. Retreating does not cost you your turn, so you step back and swing anyway. That matters enormously the day your opponent lands something nasty on it.

Lillie's Clefairy ex carries the "ex" suffix, and your opponent takes 2 Prize cards instead of one when Lillie's Clefairy ex falls. They need 6 to win: 3 well-placed knockouts on that card beat you, where it would take 6 anywhere else on your Bench.

The attackers in the deck's core, with their best attack, its cost and their Weakness

Pokémon	HP	Attack	Damage	Cost	Weakness	Retreat
Close BanetteStage 1	80	Puppet Pull	80	Psychic	Darkness ×2	1
Close DhelmiseBasic	140	Vengeful Anchor	30+	Psychic	Darkness ×2	3
Close Lillie's Clefairy exBasic	190	Full Moon Rondo	20+	Psychic + Colorless	Metal ×2	1
Close ShuppetBasic	50	Hang Down	10	Psychic	Darkness ×2	1

The core: the 15 cards every list runs

These 15 cards appear in all 35 lists, without a single exception. That is 51 of the 60 slots. The other 9 are decided player by player.

Look at how those 51 slots break down: 24 Trainers, 21 Pokémon, 6 Energy. Nothing stands out in that split. What sets this deck apart is how thick its core is: 51 of the 60 slots are the same for everybody, when half the 97 decks we track fix 34 or fewer. Fewer choices left to you than usual, and fewer ways to get them wrong.

15 of those 15 cards do not sit at the same count from list to list: Banette (2 to 4), Dhelmise (3 to 4) and Poltchageist (2 to 4), plus 12 more. Add those spreads up and you get 26 copies the 35 lists do not agree on, inside the core itself.

What "3 to 4" means in the copies column, what the "TWM 130" code under each name is for, and why we count dozens of lists instead of one: how to read a deck list.

The cards present in every list counted, with the number of copies

Copies	Card	Printing	Kind	Rotation
2–4	Close Banette	PBL 34	Pokémon	stays
3–4	Close Dhelmise	PBL 39	Pokémon	stays
2–4	Close Poltchageist	PBL 5	Pokémon	stays
3–4	Close Shuppet	PBL 33	Pokémon	stays
1–4	Close Sinistcha	PBL 6	Pokémon	stays
1–2	Close Lillie's Clefairy ex	JTG 56	Pokémon	stays
2–4	Close Gwynn	PBL 78	Trainer	stays
2–4	Close Lillie's Determination	MEG 119	Trainer	stays
3–4	Close Poké Pad	POR 81	Trainer	stays
3–4	Close Ultra Ball	MEG 131	Trainer	stays
2–3	Close Boss's Orders	MEG 114	Trainer	stays
1–4	Close Prism Tower	CRI 80	Trainer	stays
2–4	Close Night Stretcher	ASC 196	Trainer	leaves
3–4	Close Telepathic Psychic Energy	POR 88	Energy	stays
1–4	Psychic Energy	MEE 5	Energy	stays

The 41 cards that change from list to list

41 cards fight over the 9 slots that are left: 0 above three lists in four, 29 below one in four, and the percentage on each row says how many players took it.

Not one of these cards reaches three quarters of the lists. The 9 free slots in this deck are genuinely open, and two players can build it very differently without either being wrong.

The dividing line runs through Bloodmoon Ursaluna ex, at 51 %: 18 of 35 players took it, 17 put something else there. The ones who run it bring 1 copy. Right at the bottom of the table, Pokégear 3.0 (8 lists out of 35) and Air Balloon (8 lists out of 35).

Why a card at 45 % is not a bad card, and what order to buy all this in: how to read a deck list.

The cards some lists run and others don't, most common first

Lists running it	Card	Copies	Printing	Kind
74 %26/35	Close Shaymin	1	DRI 10	Pokémon
71 %25/35	Close Special Red Card	1–2	CRI 82	Trainer
69 %24/35	Close Dunsparce	1–3	JTG 120	Pokémon
69 %24/35	Close Dudunsparce	1–3	TEF 129	Pokémon
69 %24/35	Close Patrat	1	CRI 70	Pokémon
69 %24/35	Close Legacy Energy	1	TWM 167	Energy
54 %19/35	Close Buddy-Buddy Poffin	1–2	TEF 144	Trainer
51 %18/35	Close Bloodmoon Ursaluna ex	1	TWM 141	Pokémon
46 %16/35	Close Spiritomb	1	PBL 35	Pokémon
34 %12/35	Close Lana's Aid	1	TWM 155	Trainer
31 %11/35	Close Switch	1	MEG 130	Trainer
29 %10/35	Close Latias ex	1	SSP 76	Pokémon
23 %8/35	Close Pokégear 3.0	1–3	SVI 186	Trainer
23 %8/35	Close Air Balloon	1–2	ASC 181	Trainer
20 %7/35	Close Fezandipiti ex	1	ASC 142	Pokémon
20 %7/35	Close Meowth ex	1	POR 62	Pokémon
20 %7/35	Close Explorer's Guidance	1–4	TEF 147	Trainer
20 %7/35	Close Brilliant Blender	1	SSP 164	Trainer
14 %5/35	Close Flutter Mane	1–2	TEF 78	Pokémon
14 %5/35	Close Psyduck	1	ASC 39	Pokémon
11 %4/35	Close Dudunsparce ex	1	JTG 121	Pokémon
11 %4/35	Close Secret Box	1	TWM 163	Trainer
9 %3/35	Close Lucky Helmet	2–3	TWM 158	Trainer
9 %3/35	Close Team Rocket's Watchtower	2	DRI 180	Trainer
9 %3/35	Close Surfer	1	SSP 187	Trainer
6 %2/35	Close Moltres	1	PFL 14	Pokémon
6 %2/35	Close Morty's Conviction	2	TEF 155	Trainer
6 %2/35	Close Team Rocket's Petrel	1–2	DRI 176	Trainer
6 %2/35	Close Handheld Fan	1	TWM 150	Trainer
6 %2/35	Close Kieran	1	TWM 154	Trainer
6 %2/35	Close Ruffian	1	JTG 157	Trainer
6 %2/35	Close Prism Energy	1	ASC 216	Energy
3 %1/35	Close Drakloak	3	TWM 129	Pokémon
3 %1/35	Close Dreepy	3	TWM 128	Pokémon
3 %1/35	Close Dragapult ex	1	TWM 130	Pokémon
3 %1/35	Close Dunsparce	1	TEF 128	Pokémon
3 %1/35	Close Tatsugiri	1	TWM 131	Pokémon
3 %1/35	Close Naveen	2	POR 79	Trainer
3 %1/35	Close Black Belt's Training	1	JTG 143	Trainer
3 %1/35	Close Jamming Tower	1	TWM 153	Trainer
3 %1/35	Close Payapa Berry	1	SCR 141	Trainer

The most common Dhelmise list, all 60 cards

Here they are in full, ready to build. 51 slots come from the core, the cards all 35 lists run. The remaining 9 come from the table just above, working down it: the cards the most players chose, each at its most common copy count, until the total lands on 60.

This list is nobody's list in particular. It takes, slot by slot, what the most people played, under the guardrails a judge would hold you to: never more than four copies of a card, and one ACE SPEC. Behind the core it took 6 more cards to reach 60. The whole thing comes to 28 Pokémon, 25 Trainers and 7 Energy.

One thing to know before you build it: the 35 lists we counted are all from players who came in the top twenty of their tournament. Nobody publishes the lists of the players who lost. So what you have in front of you is what the players finishing near the top bring most often.

The text under the table is the one the game expects: copy it, paste it into the deck manager, and the list builds itself.

There is no price total on this page. The readings we have are Cardmarket's, in euros, and the shop link further down sends you to TCGplayer, which quotes in dollars. Putting a euro figure next to a dollar shop would be one currency too many. The French and German pages carry the total.

Why we only read lists from players who came in the top twenty of their tournament, and what that changes about every percentage on this page: how to read a deck list.

Pokémon: 28
4 Banette PBL 34
4 Dhelmise PBL 39
4 Poltchageist PBL 5
4 Shuppet PBL 33
4 Sinistcha PBL 6
1 Lillie's Clefairy ex JTG 56
1 Shaymin DRI 10
3 Dunsparce JTG 120
2 Dudunsparce TEF 129
1 Patrat CRI 70

Trainer: 25
4 Gwynn PBL 78
4 Lillie's Determination MEG 119
4 Poké Pad POR 81
4 Ultra Ball MEG 131
3 Boss's Orders MEG 114
3 Prism Tower CRI 80
2 Night Stretcher ASC 196
1 Special Red Card CRI 82

Energy: 7
4 Telepathic Psychic Energy POR 88
2 Psychic Energy MEE 5
1 Legacy Energy TWM 167

Total Cards: 60

Which decks beat Dhelmise at tournaments

Here is how Dhelmise did against every other deck at tournaments. Real games, real players sitting across the table from each other, and the platform those tournaments run on counts them one by one. All of them were played after Pitch Black became legal for tournaments, and we read the table on 5 September 2026.

The format matters more than people think. When the next set lands, decks get rebuilt around the new cards, and a matchup that leaned one way can swing the other. What you are reading holds for today's format.

We start with the opponent that gives you the most trouble, because that is where the work is before your next tournament. The further down the list you go, the better it goes for you.

Always look at the number of games next to the score. Six wins out of ten happens on its own, even with a middling deck: it is the same as flipping six heads in ten tries. Six hundred out of a thousand, and luck has nothing to do with it. Below 100 games we show nothing at all, because at that count half the matchups look like they lean one way when they lean nowhere.

Against Marnie's Grimmsnarl ex / Froslass: over 278 games, 80 won and 194 lost. 4 ended with no winner.

Against N's Zoroark ex: over 356 games, 131 won and 220 lost. 5 ended with no winner.

Against Latias ex / Mega Kangaskhan ex: over 131 games, 48 won and 82 lost. One ended with no winner.

Against Festival Lead: over 432 games, 183 won and 240 lost. 9 ended with no winner.

Against Cynthia's Garchomp ex: over 110 games, 42 won and 67 lost. One ended with no winner.

Against Mega Excadrill ex: over 576 games, 249 won and 321 lost. 6 ended with no winner.

Against Dragapult ex / Blaziken ex: over 442 games, 190 won and 243 lost. 9 ended with no winner.

Against Dragapult ex: over 552 games, 241 won and 301 lost. 10 ended with no winner.

Against Teal Mask Ogerpon ex / Meganium / Hydrapple ex: over 114 games, 48 won and 65 lost. One ended with no winner.

Against Raging Bolt ex: over 127 games, 58 won and 69 lost.

Against Toucannon: over 220 games, 110 won and 108 lost. 2 ended with no winner.

Against Dragapult ex / Dusknoir: over 408 games, 233 won and 171 lost. 4 ended with no winner.

Against Slowking: over 402 games, 240 won and 151 lost. 11 ended with no winner.

Against Alakazam / Dudunsparce: over 471 games, 288 won and 176 lost. 7 ended with no winner.

Against Mega Lucario ex: over 133 games, 100 won and 32 lost. One ended with no winner.

Against Mega Lucario ex / Hariyama: over 131 games, 106 won and 23 lost. 2 ended with no winner.

What Dhelmise costs to build, and how long it will last

You need 56 different cards to cover every version of this deck, 15 of which nobody argues about. We do not have card prices, so we are not going to put a number on it.

All 35 lists bring an ACE SPEC, and that is where they part ways: 24 took Legacy Energy, 7 took Brilliant Blender and 4 took Secret Box. You are allowed ONE ACE SPEC card in your whole deck, so a single copy covers it whatever it costs. New to this deck? Take Legacy Energy and watch what the others do with theirs before you switch.

Then there is time. 1 of the 15 core cards carries the letter H, the one that goes at the next rotation: Night Stretcher. The core table marks it. In copies, that is 2 of the 51 slots the core fills. Which day it lands, we do not know yet.

What Dhelmise struggles against

The attackers in this deck are Weak to Darkness and Metal, so they take DOUBLE damage from those types. 22 of the 97 decks we track attack in exactly those types.

Concretely, here is what is waiting for you: Mega Excadrill ex, Slowking (Metagross) and N's Zoroark ex, plus 19 more decks. The table below takes them one at a time, and those are the lists you practise against first before an event.

This table counts one thing, Weakness, and that comes to 22 decks out of the 97 we track. The other 75 do not attack into yours, and this table says nothing about them. You will still run into them.

Going the other way, your attackers are Psychic, and 6 of the decks we track have an attacker Weak to one of those: Mega Lucario ex, Mega Lucario ex / Hariyama and Greninja ex, among others. That is you doubling, for once. Second half of the table.

When an attack does nothing at all

There are two ways to hurt the Pokémon across the table, and the game does not treat them the same. An attack with a number printed next to it deals damage. An attack with no number, whose text just says "place 2 damage counters", does what the rules call an effect. That matters, because a handful of Pokémon have an Ability that prevents the effects of your opponent's attacks, and their card spells it out in brackets: damage is not an effect.

This deck has an answer to that. Banette, Poltchageist, Shuppet and Sinistcha all have Hide 'n' Sneak, and 35 lists out of 35 play them. Against an attack that only places counters, those Pokémon take nothing at all. Not one counter. An attack with a number printed next to it still puts its damage on them, mind you. Careful, though: the Ability protects the card that has it, and only that card. The rest of your bench takes everything.

Of the decks we track right now, 4 play an attack like that. They are in the table below, with the Pokémon carrying it. When you run into one, put one of those Pokémon in the Active Spot: the marker attack can no longer touch it. Your opponent will have to reach for printed-damage attacks — most of these decks carry some too. The wall buys you turns, it does not win the game for you.

The other way round, this deck attacks with counters too: Poltchageist with Furtive Drop and Sinistcha with Matcha Spin. Those attacks have no damage number printed next to them at all. In front of a Pokémon that shuts off effects, they place nothing. Zero.

Of the decks we track, 6 play a Pokémon like that, and they are in the table below. The day you see one come across the table, stop counting on your counters. You need an attack with a real number written on it. You have one: Puppet Pull, on Banette, for 80.

Questions people ask

What is the Dhelmise deck list?

There isn't one, so we built one for you: all 60 cards are written out further up this page. Across 35 tournament lists harvested between 11 August 2026 and 4 September 2026, 15 cards turn up in every single one, 51 slots out of 60. The remaining 9 change from player to player, with 41 different cards competing for them. Both tables are above.

How many lists is this counted on, and where do they come from?

35 lists, played across 33 different tournaments, harvested between 11 August 2026 and 4 September 2026 from play.limitlesstcg.com. All 35 were read card by card, and every card recognised.

Where do I start if I want to build Dhelmise?

With the whole core: 15 cards, 51 slots. Inside it, buy the 24 Trainer slots and the 6 Energy first: they carry over into your next deck. Leave the 21 Pokémon slots for last. Only then work down the flex table and add the cards the most lists agree on.

What beats Dhelmise?

Its attackers are Weak to Darkness and Metal, so they take double damage from an attacker of that type. 22 of the 97 decks we track play an attacker of that type: Mega Excadrill ex, Slowking, N's Zoroark ex and Marnie's Grimmsnarl ex / Froslass, among others.

Will Dhelmise survive the next rotation?

Not intact: 1 core cards out of 15 carry the letter H, starting with Night Stretcher.

Are a lot of people playing Dhelmise?

We have read 35 lists of it, which puts it 20th of the 97 decks we track by lists read. The platform those tournaments run on counts 1,610 entries for it, 3.95 % of the field, across all its players. That figure is theirs.

These pages are about the same corner of the game. If you came away from this one still wondering about something, start with one of them.