Common Offensive Cores

Built for the aggressive-Offense player, and built from the data. Every core on this page was chosen the same way: from the Teammates %s in the May-2026 gen3ou (1760) usage report — the mons that actually appear together on winning offensive teams — not from “these two have nice synergy on paper.” Where a popular-sounding pairing isn’t in the data, it’s called out in What the data does not support instead of being dressed up as a core.

A core in ADV is not a list of mons that “go well together” — it is a machine for removing the one thing that stops you from winning. The defensive spine of the tier (Skarmory + Blissey + a bulky Water + Snorlax) recovers forever, so you cannot brute-force it. You win by deleting the specific wall that blocks your win condition, then sweeping into the hole. Read this alongside Building Aggressive Offense, Sample Teams (these cores realized as full builds), and the Top Threats dossiers (bold names link out). Verify any matchup math in the Mechanics & Calc Reference.

The two rules that drive everything here: (1) Physical/special is by TYPE, not by move — so Crunch and Shadow Ball are special, and Steel resists both (Dark→Steel and Ghost→Steel are each 0.5×). (2) There is no Choice Scarf, so speed comes from Dragon Dance / Agility / Salac Berry / paralysis, never from an item. Internalize both before you read a single set.

How this page is built (data-driven)

The Smogon moveset report lists, for each Pokemon, the teammates it appears with more often than chance — the conditional usage minus the baseline. That number is directional: “A wants B” (A→B%) can be large while “B wants A” (B→A%) is small, usually because B is a hub that everything wants.

• A mutual pair is two mons that each list the other. Its strength is the smaller of the two directional %s (min(A→B, B→A)) — a pair is only as real as its weaker half.
• A fully-mutual trio is three mons all pairwise mutual; its strength is the weakest of the six edges.
• The top-10 cap caveat: each mon lists only its top-10 teammates. A low-usage trapper like Magneton (8.9% usage) shows a one-sided link to Metagross (57.4%) because Metagross’s own top-10 is saturated by higher-usage partners — that’s a strong pairing the report can’t show as mutual, not a weak one. So this page reads both the mutual table and the strong one-directional links.

The cores split into three engines:

• Trap (★2 DragMag, ★3 ZapDug) — physically remove the wall with Magnet Pull or Arena Trap, then sweep into the gap.
• Overload (★1 sand spine) — present two or three simultaneous threats whose walls don’t overlap, so the opponent’s defensive allocation breaks.
• Out-chip (★4 Spikes special) — stack Spikes + Sand + phazing and grind the recovery-spamming backbone into KO range.

★5 (Aerodactyl) is a speed-control + revenge shell that bolts onto any of the three.

The through-line: “no shared answer.” A core wins when the wall that stops member A is the very thing member B trains to remove — or when A and B share no common defensive answer at all. The data simply tells us which of those reciprocities the playerbase has actually found.

The cores at a glance

Everything sits on one hub: Tyranitar is the #1 teammate of nearly every offensive mon in the tier (Aerodactyl→TTar 76.3, Starmie→TTar 68.2, Salamence→TTar 61.9, Zapdos→TTar 61.6, Gengar→TTar 56.9, Metagross→TTar 52.8, Jolteon→TTar 59.1). “ADV aggressive offense,” in the data, is “Tyranitar + sand + a rotating offensive cast.” Sand chip is the background hum behind every core below.

★1 — The sand spine: Tyranitar + Metagross + Salamence

Members: Tyranitar + Metagross + Salamence.

This is the single most data-supported aggressive structure in the tier, so it leads the page. The three are a fully-mutual trio with a weakest edge of 35.5%, and the strongest all-offense mutual pair in the entire tier is its Metagross⇄Salamence link (46.9 / 51.5). The other edges: Tyranitar⇄Salamence 37.9 / 61.9, Tyranitar⇄Metagross 35.5 / 52.8. Three physical breakers, sand chipping behind them, that the two great physical walls of the tier cannot all be kept alive against.

Why it works

The whole tier’s physical defense routes through two bodies — Skarmory and Swampert — and each of them walls two of the three while folding to the third’s coverage. That’s the overload:

• Skarmory resists Metagross’s Meteor Mash (Steel→Steel/Flying = 0.5×), is immune to Earthquake (Flying), and walls DD Salamence’s Hidden Power Flying (Flying→Steel/Flying = 0.5×) and Rock Slide (Rock→Steel/Flying = 1× but eaten by its Defense). But it takes 2× from a Fire move (Fire→Steel/Flying = 2×), so MixMence’s Fire Blast and Magneton’s HP Fire break it, and CB Metagross’s Explosion blows through its physical bulk (Explosion halves the target’s Defense in Gen 3).
• Swampert resists Tyranitar’s Rock Slide (Rock→Water/Ground = 0.5×), tanks Earthquake (Ground = 1×), and walls CB Metagross’s neutral hits (Meteor Mash = 0.5×) — but it is 4× weak to Grass, so MixMence’s Hidden Power Grass 2HKOs it (Grass→Water/Ground = 4×).
• Tyranitar itself is the glue: Sand Stream chips every non-Rock/Ground/Steel switch-in 1/16 a turn, accelerating every revenge race, and DD Tyranitar is widely regarded as the best sweeper in the format once its checks are gone.

So the opponent cannot keep both walls: commit Skarmory and MixMence’s HP Grass deletes the Swampert behind it; commit Swampert and Fire Blast / Explosion deletes the bird. Whoever they spend their switch on, the other breaker walks in.

This core also contains the classic double-Dragon-Dance overload as a sub-mode: run DD Tyranitar and DD Salamence, hold both, and force the opponent to answer one with a switch — then set up the other on the predicted move. Their revenge answers don’t overlap (see Watch out).

Key sets

Lum Berry on DD Tyranitar absorbs the Thunder Wave / Will-O-Wisp that would otherwise stop the boost cold. The MixMence variant is what makes the Swampert overload real — DD Salamence alone can’t break Swampert (Earthquake is neutral, Rock Slide resisted), so if your plan leans on cracking bulky Ground/Water, run the mixed set.

How to play it

Lead Tyranitar to set sand and threaten immediate pressure. Scout which physical wall the opponent leans on from their early switch-ins — what comes in on Tyranitar’s Rock Slide / Earthquake tells you whether the answer is Skarmory or Swampert — then bait it and break it with the member that beats it: Explosion/Fire on Skarmory, HP Grass on Swampert. Explosion is a one-time resource: spend it on the specific wall standing between you and a sweep, then clean with whichever DD’er the opponent can no longer answer.

Watch out

• The two Flying sweepers have different revenge weaknesses, by design. Salamence is Ice 4× (Ice→Dragon/Flying); a bulky-Water Ice Beam revenges it. Tyranitar is Fighting 4× and is raced through its Ground weakness by Earthquake from CB Aerodactyl and Dugtrio (Ground→Rock/Dark = 2×) once chipped. No single revenge move covers both — that’s the overload working — but it means you set up only after the specific revenge threat to the sweeper you’re committing is gone.
• Phazing undoes a boosted sweep. Skarmory Roar and Suicune/Milotic Roar blow a +1 sweeper out. This is exactly why the spine so often runs a Magneton (★2) bolted on to pre-remove the bird.
• Choice-locking is exploitable. A locked CB Metagross Meteor Mash invites Swampert or a bulky Water in for free; read the wall before you commit.

★2 — DragMag: Magneton + the physical spine

Members: Magneton + a physical breaker it clears the path for — Metagross, Salamence, or Gyarados.

Magneton’s whole correlation profile is “the trapper bolted onto ★1.” Its top teammates are Metagross 57.4, Snorlax 57.4, Salamence 52.4, Celebi 43.0, then Aerodactyl 30.3 and Tyranitar 29.5 — and it lists Gyarados at 19.7 in a clean mutual pair (Gyarados→Magneton 31.8). Because Magneton sits at 8.9% usage, those links are one-directional (the partners’ own top-10s are full), but the top-10 cap caveat means they read as strong, not weak. Magneton exists to trap the Steel (Skarmory, Registeel, Forretress) that walls those physical sweepers.

Why it works

Magnet Pull traps Steel-types on the field; bring Magneton in on the Skarmory switch and remove it before it can Spike or phaze:

• Thunderbolt hits Skarmory 2× (Electric→Steel/Flying = 1×·2× = 2×).
• Hidden Power Fire hits Skarmory 2× and OHKOs Forretress 4× (Fire→Bug/Steel = 2×·2× = 4×).
• The HP Fire vs HP Grass slot is the real decision: HP Fire reliably 2HKOs Skarmory and kills Forretress; HP Grass trades that for 4× on Swampert (Grass→Water/Ground). For a pure DragMag whose only job is to kill the bird, run HP Fire.

With the Steel gone, the breaker has no switch-in:

• DD Salamence at +1 outruns the unboosted field; its STAB Hidden Power Flying is physical (Flying is a physical type), so it fires off its enormous Attack.
• CB Metagross — the strongest link (57.4) — no longer faces the one body that resists Meteor Mash and shrugs Earthquake; Meteor Mash (physical Steel) snowballs with its own Attack-boost chance.
• DD Gyarados adds Intimidate to ease setup; after +1 it outpaces base 115s.

Magneton does double duty: trapping the opposing Spiker also denies the opponent Spikes and removes the Steel that would switch into your Tyranitar.

Key sets

Pair Magneton with the CB Metagross and DD Salamence sets from ★1. Substitute on Magneton lets you scout the switch and block status while you line up the trap.

How to play it

Sequence the trap before the sweep, never after. Pivot Magneton in on a predicted Skarmory switch, remove it, then bring the sweeper in on a forced switch and boost. Don’t reveal the sweeper early — keep the opponent unsure whether Magneton is even paired with a setup mon until the bird is already dead.

Watch out

• You spend a turn (and often Magneton’s life) on the trap. If the opponent never sends the Steel in, Magneton is a slow base-70 liability.
• Dugtrio revenge-kills Magneton — Arena Trap pins it and Earthquake is 4× (Ground→Electric/Steel = 2×·2× = 4×).
• The sweeper still answers to revenge killers. CB Aerodactyl Rock Slide is 2× into Salamence and Gyarados; bulky-Water Ice Beam is 4× into Salamence. Keep a scout for the mirror and don’t set up into a healthy Suicune.

★3 — ZapDug: Dugtrio + a special wincon (Zapdos / Suicune)

Members: Dugtrio + a special win condition it frees — Zapdos or Suicune.

Special offense in ADV runs into exactly one wall: Blissey eats every special hit and Soft-Boils it back. Dugtrio’s job is to trap and delete Blissey, then let an unchecked special attacker run. The data backs the trap targets, not a forced “dual-trapper” pairing: Dugtrio’s mutual partners are bulky bodies it picks off — Suicune 37.9/50.4, Claydol 46.3, Blissey 29.6/46.6, Snorlax 27.9/44.2 — plus its two offense-relevant special partners, Zapdos (25.2/32.3) and Gengar (18.6/29.8). The canonical aggressive build is ZapDug.

Why it works

Blissey is grounded and not a Steel, so Arena Trap holds it:

• Beat Up abuses Blissey’s poor Defense — each strike lands off a healthy teammate’s base Attack, so a full team trades Beat Up + Earthquake to remove the universal special wall. (Beat Up uses party Attack, so Choice Band is irrelevant to it; it’s the intended Blissey-chip tool.)
• Dugtrio also chips the secondary special checks — Tyranitar (Earthquake = 2×, Ground→Rock/Dark) and a worn Snorlax — that special attackers otherwise grind against.

With Blissey gone, the partner sweeps:

• CM Suicune (Dugtrio’s strongest link, 50.4) becomes self-sufficient: behind Calm Mind it overwhelms Zapdos, Celebi, and Snorlax; Surf is 2× on uninvested Tyranitar (Water→Rock/Dark); with Blissey removed only Milotic still answers it.
• Zapdos breaks the Skarmory + Blissey sandwich with Thunderbolt (2× Skarmory) / Hidden Power Ice (4× Salamence and Flygon), and its Ground immunity gives it free switch-ins.

Key sets

How to play it

Bait Blissey in, then trap it. Beat Up math depends on your team being healthy — every fainted or chipped teammate weakens the hit — so trap Blissey earlier rather than at the very end of a long game. Then bring the wincon in on a forced switch and set up.

Watch out

• Gengar lures and OHKOs Dugtrio — Levitate dodges Earthquake (Ground → 0×) and Ice Punch is 2× (Ice→Ground). The single biggest threat to the trap; don’t auto-pivot Dugtrio into a Ghost.
• Dugtrio’s correlation skews bulky. Its top partners (Suicune, Claydol, Blissey, Snorlax) say it lives as comfortably on balance as on hyper-offense — ZapDug is the aggressive slice of a broader Dugtrio pattern, not proof Dugtrio “wants” two trapper slots (see What the data does not support).
• Suicune is still stopped by Milotic and Toxic (Toxic out-paces Calm Mind). Frail trappers and Jolteon die to anything they can’t outspeed once sand chips them.

★4 — Special Spikes: Skarmory + Gengar + Starmie / Jolteon

Members: a Spiker (Skarmory) + Gengar (spinblock) + fast special sweepers (Starmie, Jolteon, Zapdos), with phazing to force chip.

This is the densest cluster in the offense data. Skarmory⇄Gengar 36.4/49.7, Skarmory⇄Starmie 36.6/61.2, Gengar⇄Jolteon 33.1/72.5 (Jolteon→Gengar 72.5 is the single highest offense→offense link in the tier), Gengar⇄Starmie 27.0/33.0, Starmie⇄Jolteon 22.2/39.6 — and it forms real trios: Skarmory+Gengar+Starmie 27.0 and Gengar+Starmie+Jolteon 22.2. Where the trap cores delete Blissey, this core out-chips it.

Why it works

The pieces interlock around hazard pressure:

• Skarmory lays up to three Spikes layers (1/8 → 1/6 → 1/4 per switch) and phazes with Roar to rack chip on the body it forces in. (Spikes is the only entry hazard in Gen 3 — no Stealth Rock, no Toxic Spikes — so layers + spin-denial is the chip engine.)
• Gengar is the linchpin: Levitate + Ghost typing blocks Rapid Spin (a Normal move that simply fails while a Ghost is in), so your layers stay down. It adds Will-O-Wisp to cripple physical switch-ins and Explosion as a momentum trade.
• Fast special sweepers force the walls in and out. Jolteon and Zapdos Roar the Blissey switch back into the hazards and spread Thunder Wave as de-facto speed control; Starmie keeps your hazards live by spinning the opponent’s away while pressuring with Bolt-Beam.

Over a few cycles, Spikes + Sand + phazing pushes special hits into KO range even through Blissey, which can’t out-recover layered chip while it’s forced in and out.

Key sets

Skarmory’s Spikes set uses Roar, not Whirlwind — Whirlwind + Drill Peck is an illegal egg-move combination in Gen 3.

How to play it

Set Spikes early and protect Gengar at all costs — it’s the keystone that keeps the layers down. Then play for forced switches: every time Blissey comes in to wall a special attacker, Roar it back out and let the hazards bite. The win is cumulative chip, not a single hit — prioritize tempo and switch-pressure over greedy KOs.

Watch out

• If the Spikes get spun, the engine stalls. Opposing Claydol, Starmie, and Forretress all spin; keep Gengar alive, and beware Tyranitar’s Pursuit (Dark→Ghost/Poison = 2×, special-typed but still super-effective) plus Sand chewing Gengar’s 60/60/75.
• Gengar’s own Ghost STAB is resisted by Steel. Shadow Ball is Ghost→Steel 0.5× into Skarmory, and only neutral into Metagross (Ghost→Steel/Psychic = 0.5×·2× = 1×) — so Gengar breaks specially via Thunderbolt/Ice Punch, not with its STAB.
• Blissey is never fully removed, only worn down. A fresh full-HP Blissey can still reset a sweep — keep chip flowing and don’t let Aromatherapy + Soft-Boiled undo a game’s work.

★5 — Aerodactyl pressure: Aerodactyl + Celebi / Zapdos

Members: Aerodactyl + a special partner — Celebi or Zapdos — on the Tyranitar hub.

The data surfaces Aerodactyl as a genuine offensive engine, not just a lead: Aero⇄Celebi 38.8/42.6 is the strongest mutual pair involving a pure-offense fast mon outside the ★1 spine, with Aero⇄Zapdos 30.5/31.4, Aero⇄Tyranitar 28.2/76.3, and Aero⇄Skarmory 31.4/52.1 behind it. It anchors real trios — Tyranitar+Aerodactyl+Zapdos 22.1, Tyranitar+Aerodactyl+Celebi 20.8, Metagross+Aerodactyl+Celebi 21.6. CB Aerodactyl is base-130 immediate pressure; the special partner gives it the wallbreaking it lacks.

Why it works

CB Aerodactyl is the premier no-setup revenge piece. Rock Slide is 2× into Salamence, Zapdos, and Gyarados and 4× into Charizard and Moltres (Rock→Fire/Flying); Earthquake handles Steels and Electrics; Double-Edge is the hard neutral STAB; and it’s immune to Ground (Flying), so it pivots into Earthquakes for free. The partner covers what Aerodactyl can’t:

• Celebi is the defensive complement: it resists the exact moves that kill Aerodactyl — Thunderbolt (Electric→Grass/Psychic = 0.5×) and Surf (Water = 0.5×), the bulky-Water/Electric attacks that hit Aero’s 2× Water/Electric — while Calm-Minding or Leech-Seeding up as a second special breaker. In return Aerodactyl revenges the fast Flying/Rock/Fire threats (opposing Salamence, Zapdos, Charizard, the Aerodactyl mirror) that hit Celebi for 2×–4×. That reciprocity is why it’s the strongest of Aero’s pairs.
• Zapdos is the more aggressive partner: shared Ground immunity, Thunderbolt + HP Ice to break Skarmory and the dragons. Note the honest cost — Aero and Zapdos share a 2× Rock weakness, so one fast Rock Slide (CB Aerodactyl, CB Tyranitar) threatens both (see Anti-synergy). Celebi is the cleaner defensive fit; Zapdos is the higher-pressure one.

Key sets

Pair with the Zapdos set from ★3/★4 when you want the aggressive variant. (Celebi’s listed ability is cosmetic — it has only Natural Cure in Gen 3.)

How to play it

Lead or pivot Aerodactyl to apply immediate Rock Slide pressure and revenge the tier’s fast Flyers; bring the special partner in behind Tyranitar’s sand to break the fat core. Aerodactyl is a hit-and-run piece — its 2× Water/Electric/Ice/Rock/Steel list means it folds to most bulky-Water and Electric attacks, so don’t leave it in.

Watch out

• Aerodactyl is frail and broadly checked. Bulky Waters Surf it (2×), Suicune/Starmie/Jolteon Thunderbolt it (2×), and any faster Rock or a Steel revenge it. It pressures; it doesn’t tank.
• Celebi is 4× weak to Bug — Megahorn from Heracross deletes it, and it takes 2× from Fire/Ice/Flying/Ghost/Dark. It’s the glue, not a wall; keep it off the fast Flyers Aerodactyl is there to revenge.

The raw backbone — top mutual offense pairs

The offense-relevant mutual pairs, straight from the report (min = the weaker direction; both directions shown). These are the bricks every core above is built from.

Strong one-directional links (top-10 cap; the partner’s own list was full). These are the DragMag backbone and the niche breakers that want the spine but are too rare to be listed back:

Fully-mutual offense trios

The all-offense (or offense-led) trios where every edge is mutual, ranked by weakest edge. Most are just ★1, ★4, or ★5 with a hub piece attached — which is the point: these cores genuinely co-occur as units.

Heracross & the secondary breakers

Heracross is a top-20 aggressive breaker, but its correlation profile is honest about where it lives: its strongest mutual partners are Snorlax (24.8/54.5) and Suicune (22.3/41.2) — bulky bodies — with a genuinely aggressive slice in Jolteon (29.7/30.1) and the Jolteon+Heracross+Camerupt trio (28.3). The data says Heracross is more of a balance-leaning breaker than a hyper-offense staple; the popular idea of welding it to a DragMag trapper isn’t in the report (next section).

What it does: SubSalac or SD Heracross is the answer to the fat Normal/Steel backbone. Megahorn is 4× on Celebi (Bug→Grass/Psychic) and 2× on Starmie; Focus Punch is 4× on Tyranitar (Fighting→Rock/Dark), 2× on Snorlax/Blissey/Registeel. It is hard-walled by anything Flying (4× weak, e.g. Skarmory Drill Peck) and by Gengar (immune to Fighting via Ghost, resists Megahorn 0.25×), so it pairs with a bulky Normal partner that grinds the walls Heracross can’t, while Heracross removes the Snorlax/TTar that wall it.

What the data does not support

The most important section, and the one the rest of the page exists to earn. These pairings sound like cores and get repeated, but the May-2026 report shows no above-chance co-occurrence — so they’re not on this page as cores.

• Magneton + Dugtrio “dual trappers.” Neither lists the other in its top-10 (Magneton’s: Metagross, Snorlax, Salamence, Celebi, Aerodactyl, Tyranitar, Suicune, Swampert, Gyarados, Medicham; Dugtrio’s: Suicune, Blissey, Claydol, Snorlax, Zapdos, Gengar, Metagross, Swampert, Celebi, Skarmory). They live on different archetypes — Magneton on the physical spine, Dugtrio on bulkier balance. Stacking two slow trappers purely to “remove the whole backbone” costs two non-attacking slots and isn’t what winning teams actually do. Run one trapper that fits your plan.
• Magneton + Heracross. A staple-sounding “trap Skarmory for SD Heracross” pairing — but Heracross is absent from Magneton’s list and Magneton from Heracross’s. Heracross’s trapper, if any, isn’t Magneton; its data partners are bulky (Snorlax/Suicune) and fast-offense (Jolteon).
• Dugtrio + Starmie / Jolteon as a trap core. Dugtrio is mutual with Zapdos and Suicune, not with Starmie or Jolteon — those two are members of the ★4 Spikes wing (they pair with Gengar/Skarmory), not beneficiaries of a Dugtrio trap. Build them as out-chip pieces, not as “free the sweeper after Dugtrio traps Blissey.”

The lesson: a mechanism can be plausible and still not be a core. Reciprocity has to show up in what people actually win with.

The defensive spine you’re racing

You’re building to beat these — and the data shows them as the tier’s tightest defensive clusters, which is exactly why brute force fails:

• Tyranitar + Swampert + Skarmory (trio, weakest edge 36.3) — the sand-balance physical backbone. Tyranitar⇄Swampert is the single strongest pair in the whole tier (55.5/83.8).
• Snorlax + Suicune + Claydol (trio, 36.3) — the bulky-offense/CuneDol spine.
• Skarmory + Blissey (39.7/55.0) — the canonical SkarmBliss that every core above is engineered to delete or out-chip.

Your cores don’t out-stat these; they remove a load-bearing piece (trap, Explode, or out-chip) and pour through the hole.

Anti-synergy & redundancy

• Two slow trappers (Magneton + Dugtrio) — covered above; two non-attacking slots is too passive for aggressive offense. Pick the one that removes your wall.
• Aerodactyl + Zapdos share a 2× Rock weakness — a single fast Rock Slide (CB Aerodactyl, CB Tyranitar) revenges both. Pairing Aerodactyl with Celebi instead removes the shared answer.
• Double Steel/Psychic (Metagross + Jirachi) — same typing, same checks (Fire/Ground), redundant rather than overloading. Two breakers should split the walls, not share them.
• Salamence + Gyarados together is not redundant on revenge — Salamence is 4× Ice, Gyarados is 4× Electric — but both are Flying sweepers that fold to a fast Rock Slide and both want the same Skarmory removed first. One DD Flyer plus a non-Flying breaker spreads the answers better.

Legality & mechanics traps

Every set above was validated against the Gen-3 (gen3ou) legality rules and the type-based physical/special split:

• Physical/special is by TYPE. Salamence’s STAB Hidden Power Flying is physical (Flying); Metagross’s Meteor Mash is physical (Steel); but Crunch, Pursuit (Dark) and Shadow Ball (Ghost) are special, and Steel resists both at 0.5×. A Choice Band boosts only physical Attack, so a special-typed move on a CB set (Fire Punch, HP Grass) does nothing — those are never on the CB sets here. The lone CB exception is Beat Up on Dugtrio: each hit uses a party member’s base Attack, so the Band is irrelevant to it.
• No Choice Scarf in Gen 3. Speed comes from Dragon Dance / Agility / Salac / paralysis — never an item.
• Skarmory cannot run Whirlwind + Drill Peck (incompatible egg moves); it phazes with Roar.

Building from a core

These engines are seeds, not teams. Most aggressive builds stack two of them: the ★1 sand spine almost always runs inside a ★2 DragMag shell so the phazer gets removed before the sweep; ★4 Spikes special layers a ★3 Dugtrio on top to delete Blissey instead of merely chipping it; ★5 Aerodactyl bolts onto any of them for speed control without a Scarf. The recurring support shell is one trapper (Magneton or Dugtrio — not both) + a spinblocker (Gengar) when you run Spikes + a fast revenge piece (Aerodactyl, Jolteon).

To turn any of these into a full six, work through Building Aggressive Offense, then study the same cores as complete, vetted builds in Sample Teams. Cross-reference every member against its Top Threats dossier, and check any damage claim in the Mechanics & Calc Reference.

The one-line summary: ADV aggressive offense doesn’t out-muscle the SkarmBliss backbone — it removes a load-bearing piece of it (trap, Explode, or out-chip) and sweeps into the gap. Every core above is a different tool for doing exactly that, and every one of them is a pairing the winning teams in the data actually run.