Build decks

What you’ll accomplish

By the end of this workflow, you will have a target deck in Pegasus that’s tied to your real inventory. You’ll read its three readiness layers, understand why a list can be legal but unbuildable, and know when to archive instead of delete. The next workflow (Analyze decks and scores) takes it from there.

Before you start

• At least one card in your inventory — see Organize inventory if you haven’t done the import yet.
• A rough idea of the format you’re building toward. Goat, Edison, HAT, current TCG/OCG — whichever it is, set it deliberately. Format drives almost every downstream evaluation.

Steps

1. Create the target deck

   Open Deck Lab and click New deck. Pegasus needs two things to start: a name and a format. The format field is the single most consequential choice on this form — it determines which banlist applies, how many copies of each card are legal, and which meta the analysis layer compares your list against. Setting it later than now means re-running every evaluation.

   Screenshot pending /img/docs/build-decks/build-newdeck-form.png

   Format is the most load-bearing pick on this form. Don't leave it on default and 'fix it later.'

2. Read the readiness layers

   A deck carries three independent badges and each one answers a different question:

   • Legal — Does the active banlist allow this list as-is?
   • Buildable — Do you own (or can you reasonably acquire) every card in it?
   • Cost — How much would the missing cards cost at current prices?

   A deck can be legal but unbuildable (you don’t own three Pot of Greed), buildable but priced past your budget, or fully buildable but structurally fragile. There is no single “this deck is fine” badge. Read all three together.

   Screenshot pending /img/docs/build-decks/build-readiness-layers.png

   When the layers disagree, that's the signal — not the noise.

3. Edit main, extra, and side

   Drag cards into their zones from the search panel. Pegasus analyzes main and extra together — combo pairs frequently cross the boundary (engine cards in main pulling extra-deck win conditions), and splitting the analysis loses that signal entirely.

   The side deck is evaluated separately because it never participates in game-one combos. It runs against the format meta to tell you if you’re covered for the matchups you’ll actually face.

   Screenshot pending /img/docs/build-decks/build-edit-grid.png

   The main/extra split is structural, not analytical. They're one deck for scoring.

4. Archive when done

   The temptation is to save every fresh brew as a live target. Resist it. Five active targets are useful — forty-five make readiness noisy and turn the acquisition planner into a wish list. Archive a deck the moment you decide you won’t realistically build or test it in the next few weeks. Archive isn’t delete: version history stays, and you can resurrect a list any time the meta shifts.

   Screenshot pending /img/docs/build-decks/build-archive-modal.png

   Archived decks vanish from active scoring but remain reachable under History.

What success looks like

Your Deck Lab grid shows the new deck with three live badges colored green, amber, or red. Clicking through opens the deck’s analysis page where coverage, simulation metrics, and a per-card shortage list become available — that’s the next workflow.

If your three badges all read green and there’s still a “Review” callout, that’s not a bug. It’s the structural-soundness check telling you the list is buildable but the combo distribution is off.

Common pitfalls

• Wrong format set on creation. Cascades into false-positive banlist warnings on cards that are legal in the format you actually play. Fix it via deck settings — don’t recreate the deck.
• Analyzing main and extra separately. Some users do this from habit (the YDK file ships them in separate sections). Don’t. You’ll lose all combo-pair signal that crosses the boundary, and the readiness score will read higher than it should.
• Hoarding stale targets. The acquisition planner ranks by aggregate readiness gain across all active targets. One stale Goat brew you’ll never build can shift the priorities of a deck you genuinely want to play this weekend.

Going deeper

The three readiness layers are derived from the scoring model — coverage, consistency, and risk weights tuned per format. The “buildable” check is a join between the deck list and your inventory, respecting location and availability flags. Format selection feeds into the data sources lookup for prices and meta evaluation.

If you want to understand why a deck scored the way it did, Analyze decks and scores breaks it down to per-card contribution.

For the field-by-field reference of the deck-building surface — new-deck form, editor, archive flow — see Deck Lab (reference).

Next steps

Analyze decks and scores Coverage, consistency, risk — explained per card, not just totaled.

Readiness and acquisition Turn shortages into a prioritized buy list.

Manage deck versions Branch, swap, and roll back when a card change underperforms.