Quick Analysis

What this surface is

Quick Analysis is the deck-evaluation surface. It combines a structural read (coverage on 5 dimensions), an empirical read (simulation across ~10,000 opening hands), a risks list (structural warnings outside the scoring weights), and a findings panel (named patterns Pegasus detected). It does NOT include card-by-card breakdown — that lives inside the coverage drilldown.

Where to find it in the app

From any deck in Deck Lab, open the Analyze tab — URL pattern /decks/<id>/analyze. Re-run by clicking “Run analysis” in the top toolbar.

Screenshot pending /img/docs/reference/quick-analysis-overview.png

A typical analysis layout for an evaluated deck.

Layout

Region	Purpose
Top toolbar	Run analysis, version selector, export, share
Bottleneck callout	One-line summary of the weakest dimension + suggested direction
Coverage radar	5-axis spider chart
Simulation panel	Observed rates from sampled opening hands
Risks list	Structural warnings (not part of score)
Findings panel	Named patterns Pegasus detected (engine identification, combo lines)
Deck-style inference	Inferred archetype/style — affects how scores are graded

Coverage dimensions catalog (5)

Dimension	What it measures	Scale	High-score example	Low-score example
Access	Reliability of reaching the engine card	0-4	Searcher-heavy combo decks	Hand-trap-only stun
Conversion	Reliability of access → payoff	0-4	Linear combo with closer	Brick-heavy decks
Protection	Resilience against interaction	0-4	Negate-stack control	Going-second blowouts
Answer density	Density of cards that disrupt opponent	0-4	Hand-trap-heavy lists	Pure combo with no traps
Recovery	Ability to keep playing after the first exchange	0-4	Grind-game control	Single-line combo

A score is directional — meant to guide decisions, not replace testing. Different deck styles have different expected shapes (see deck-style inference below).

Simulation metrics catalog

Sample size defaults to 10,000 opening hands. Hand size and mulligan rules follow format conventions.

Metric	Definition	Healthy range (varies by style)
Starter access rate	% of opening hands containing at least one starter card	Combo: 70%+ · Control: 50%+
Extender access rate	% of hands containing engine extenders	Combo: 60%+ · Control: 30%+
Interaction access rate	% of hands with disruption cards	Control: 70%+ · Combo: 30%+
Playable hand rate	% of hands that have a non-trivial line	90%+ for all decks
Dead hand rate	% of hands that brick or do nothing	≤10% for all decks
Starter + extender rate	% of hands with both	Combo: 40%+
Engine online rate	% of hands resolving the deck’s primary plan	Combo: 50%+ · Control: n/a

Structural risks taxonomy

Risks are warning flags, not part of the coverage score. A deck can score 4/4/3/3/3 and still carry a severe risk.

Risk	Trigger	Severity inputs
Brick risk	High dead-hand rate in simulation	Dead hand rate, starter density
Single point of failure	Game plan hinges on one resolving card	Card-criticality + counter density in meta
Normal summon dependence	Plan collapses to any normal-summon negate	Normal-summon-tagged starter ratio
Extra deck dependence	Folds to extra-deck disruption	Extra-deck reliance + Nibiru/Forbidden Droplet exposure
Graveyard dependence	Folds to graveyard hate	Graveyard-recursion count
Going-second fragility	Needs to go first to function	Lack of board-breakers in main
Low follow-up risk	Runs out of gas in grind games	Recovery score + searcher count in grind

Severity is rendered as a chip color: gray (info) / amber (consider) / red (matters in many metas).

Findings panel

Findings are named patterns Pegasus detected — engine identifications, combo lines, package compositions. Each finding is a one-liner with evidence cards listed underneath.

Finding kind	Example	Evidence
Engine identification	”Snake-Eye engine”	Lists the named cards triggering the match
Combo line	”1-card combo line — Diabellstar → Princess”	Step-by-step card sequence
Package match	”Adventurer Token package”	Lists the package cards present
Tech choice	”Maxx C tech inclusion”	Lists the unusual card

Deck-style inference

Pegasus infers a deck style from card composition. The inferred style determines which coverage shape is “good”:

Style	Expected coverage shape
Combo	High access, high conversion, lower answer density
Control	High protection + answer density + recovery, lower conversion
Blind second	High answer density + payoff conversion
Stun	Moderate access + very high protection
Hybrid	Square radar — moderate everywhere

Override the inference via deck settings if it mislabels — analysis grading depends on it.

Interactions

• Click a coverage axis → drill into that dimension’s evidence (per-card contribution)
• Hover a simulation metric → see the calculation breakdown
• Click a risk chip → expand the explanation + mitigation hints
• Click a finding → highlight the evidence cards in the deck editor
• “Re-run analysis” → triggers a fresh simulation; uses the current cached structural read unless deck changed

Edge cases

Cuidado

• Coverage and simulation disagree. Simulation is empirical and usually wins — the structural model has context boundaries. A 4/4 coverage with 28% starter access in simulation means the model misread the engine.
• Deck style inferred wrong. Common with hybrid builds or experimental engines. Override in deck settings — otherwise the scoring is graded against the wrong rubric.
• Stale analysis badge appears when the deck list changed after the last run. Re-run before drawing conclusions.
• Sample-size noise. Below ~5,000 hands, simulation metrics carry ≥2% margin. Don’t make changes based on sub-percent deltas.

Related workflows

Analyze decks and scores The end-to-end workflow for reading an analysis.

Compare and improve decks Turn an analysis weakness into a candidate swap.

Related references

• Scoring model — coverage weights, simulation methodology, calibration corpus
• Compare — the surface that consumes two analyses
• Combos — combo-line detection details