Arkham Horror LCG Deckbuilding Guide 2026: Roles, Upgrades, Tommy, Marie & Campaign Teams

Reliability begins with five layers. First, the deck has a job: discover clues, control enemies, or flex between both. Second, it has repeatable tools for that job rather than one lucky event. Third, it has enough resources and card draw to deploy those tools before the agenda accelerates. Fourth, it compresses actions—more clues, damage, movement, cards, or resources per action. Fifth, it carries survival and emergency answers without burying the engine.

Write the job at the top of the deck list. For every card, ask which layer it serves and when you are happy to draw it. If the answer is “someday, in a cool situation,” it is a flex slot, not a foundation.

Start with the stat line and health/sanity, but do not stop there. Read the investigator ability, elder-sign effect, traits, deckbuilding options, deck size, signature card, signature weakness, and any additional requirements on the reverse. Those lines define both the legal card pool and the behavior the deck is rewarded for repeating.

A four in a primary stat is a beginning, not an engine. Ask what assets raise or leverage that stat, what events convert it into progress, and what icons rescue the most important tests. Then read the weakness as a design brief: it tells you which resource, action, location, or timing problem the deck must survive.

There is no universal count that survives every investigator and card pool. Use functional coverage instead. A specialist needs multiple repeatable ways to perform the primary job, several cheap ways to find or pay for them, and events or icons that cover the opening turns before the engine arrives. A flex deck needs fewer peak tools but more overlap: a clue card that also moves, an enemy answer that also protects, or an asset that supports multiple test types.

Goldfish the opening five cards ten times. Record whether the deck can contribute by turn two, whether it can afford its assets, and which card is repeatedly stranded. That evidence is more valuable than copying a ratio from a collection twice your size.

Build the table’s coverage before polishing individual decks. True solo needs one investigator who can discover clues, handle or escape an enemy, move, and survive. Two-handed solo is cleanest with a clue lead and an enemy lead. Two human players can use the same split or two complementary flex investigators. Four players can support true specialization: dedicated clue and enemy leads plus two flex/support decks.

Every deck still needs a weak-case plan. The clue lead needs an evade, movement, or emergency enemy answer when isolated. The fighter needs a way to collect a last clue or support a test. Support must accelerate scenario progress, not merely make the table comfortable while doom advances.

Yes, especially for a multiplayer table that wants a Guardian with a real economic identity. The Tommy Muldoon Investigator Deck is a current ready-to-play product with an upgrade package. Fantasy Flight lists it at $18.99 MSRP and requires a Core Set.

Tommy rewards deliberate damage and horror placement on assets that can protect the team and return value when defeated. The trap is playing every protective asset merely because it can soak. The deck still needs tempo: a weapon or enemy-control plan, a way to pay for the board, and enough positioning to protect someone who matters. In a two-player pair, match him with a decisive clue investigator rather than another character waiting to be rescued.

Yes, for players who enjoy flexible spell sequencing and card selection. The Marie Lambeau Investigator Deck is a current ready-to-play Mystic deck with campaign upgrades and the same Core requirement.

Her 2026 design rewards managing damage and selecting from additional cards rather than simply stacking one large stat. The attraction is flexibility: Mystic tools can cover clues, enemies, protection, and scenario tests. The cost is sequencing. You must know which spell is the engine, which effect is the emergency answer, and how the deck pays for both. In true solo, include movement and a plan for an early enemy before the spell suite is established.

Spend the first experience on consistency. Upgrade or replace the card you are unhappy to see, improve the primary tool’s reliability, or add a stronger search/economy piece. Next, compress actions: upgraded cards that discover additional clues, deal more damage, move while doing useful work, or draw without surrendering tempo.

After the engine works, repair the weakness the campaign exposed—horror, economy, mobility, boss damage, or encounter protection. Luxury cards come last. Do not hoard experience for a dramatic fifteen-XP future while the present deck fails to start.

Respect card levels and investigator access. Experience cost is not always the large printed number you remember from a deck list; check the actual card version and current rules before recording the purchase.

Choose by the missing role at your table. Tommy is the clean buy for protection and enemy control. Marie is the spell-forward flexible choice. Carolyn Fern fits a larger group that can convert healing and support into tempo. Andre Patel is the clue-forward Seeker. Miguel de la Cruz is the resilient Survivor flex.

Buy one, play it unchanged, then customize from evidence. Five boxes at once create five unopened syllabi, not mastery.

ArkhamDB is the established browser deckbuilder and public deck library. arkham.build offers another modern deck-construction surface. Arkham Cards combines deck access, campaign tracking, chaos bag management, and scenario bookkeeping across iOS and Android.

Use public lists as annotated evidence, not scripture. Read why a card is present, compare the author’s collection and player count with yours, then test the opening. For video discussion, PlayingBoardGames’ Gathering guide is a useful example of focused investigator analysis. Avoid copying a deck built for taboo rules, four-player support, or a full card pool into a different environment without reading the assumptions.

Test the deck against the next scenario, then record three things: which card was dead, which action was repeatedly inefficient, and which problem required another investigator to abandon a job. Those notes determine the next upgrade more honestly than a tier list.

When the team is ready for a mechanically dense campaign, continue to Volume III: The Circle Undone Hosting and Community Toolkit. If the round itself is still uncertain, return to Volume I: The First-Night Field Manual.