Opponent Decks

Clusters opponents whose observed non-land cards overlap heavily, then reconstructs what they are likely playing. Similarity is the overlap coefficient: shared cards divided by the smaller of the two observed sets, so a game that revealed fewer cards is not penalised. Pick a format to avoid mixing Standard, Brawl, Explorer, drafts and tutorials.

Opponent appearances 64

Clusters found 0

Similarity threshold 75%

Minimum shared cards 2

Format: Overlap threshold: Minimum appearances: Minimum observed cards: Minimum shared cards:

No clusters at 75% — closest pairs

Opponents	Format	Overlap	Shared cards	Shared non-land cards
Ninjaman vs manabu94	Standard	60%	3	Ajani's Pridemate, Haliya, Guided by Light, Hinterland Sanctifier
KonGik vs Atlas Mario	Standard	57%	4	Ajani's Pridemate, Hinterland Sanctifier, Leonin Vanguard, Sheltered by Ghosts
mike vs KonGik	Standard	50%	4	Ajani's Pridemate, Healer's Hawk, Leonin Vanguard, The Theorist, Jace Beleren
Player 2 vs Dragon_Axolotl	Standard	50%	3	Erode, No More Lies, Quick Study
Ninjaman vs KonGik	Standard	43%	3	Ajani's Pridemate, Hinterland Sanctifier, Leonin Vanguard

Lower the overlap threshold or the minimum shared cards to form a cluster from these pairs.

Counted: visible hand cards, draws, permanents they played, cards returned to their hand, and every card that hit their graveyard or exile — including resolved spells, discard, surveil, mill and discover/exile effects. Moves carry their owner, so a spell cast in response on either player's turn is attributed to the player who cast it. Tokens are ignored entirely. Hidden information is never guessed; when an older row has no owner information, non-spell cards known to belong to you are filtered out. Quantities are lower bounds, and lands are ignored for similarity and omitted from the inferred list.