← Game Corner x Constructed x Saturday

Fire Umbra GuardianTier 1

at Game Corner x Constructed x Saturday

decks
2 (25.0% of field)
matches (non-mirror)
7 5-1-1
overall winrate
78.6%
ELO-adjusted winrate
69.4% avg ELO +9.7 per match
expected value
58.7% share-weighted (raw)
EV (ELO-adj)
59.5% share-weighted (skill-controlled)
Pilot Skill ↑ top
mean ELO 1374 · median 1374 · range 1340–1408 (2 pilots with an ELO rating)
1029 Q1 1281 Q3 1404 1408
50% of pilots are in the field's top quartile, 0% in the bottom. This deck is favored by stronger players — weight the headline winrates accordingly (the ELO-adjusted column already controls for this at the match level).
Head-to-Head Matchups

Winrates vs every other archetype, sorted by the opponent's field share. Mirrors excluded. Shrunk WR uses a Beta(2, 2) prior so a 1-0 cell doesn't read as 100%; it feeds the EV calc above. Adj WR is the skill-controlled view: derived from per-match ELO deltas (0.5 + avg ELO Δ / K=32, same shrinkage), so it tells you what would have happened on equal-rating play. A faded Raw WR means under 10 matches — lean on the Shrunk WR column there.

Opponent Their share Matches W-L-D Raw WR Shrunk WR Adj WR ELO Δ/match
Wind Shadowstrike 25.0% 2 1-0-1 75.0% 58.3% 61.3% +10.9
Fire Aetherwing 12.5% 0
Water Umbra Guardian 12.5% 1 1-0-0 100.0% 60.0% 54.0% +6.3
Wind Astra Cleric 12.5% 2 2-0-0 100.0% 66.7% 71.0% +20.2
Wind Wakeup Combo 12.5% 2 1-1-0 50.0% 50.0% 49.9% -0.1
Decklists · 2 from this event

The 2 public Fire Umbra Guardian decklists from this event, ordered by final placement — click a pilot for the full list.

Mirror Analysis

In 1 mirror match, here are the card-count differences that most often separated the winning deck from the losing one. Staples (cards both decks run at the same count) are filtered out — only flex slots show up below. Cards whose count only differed in fewer than 3 mirrors are also held back as too sparse to call a pattern (still visible in the raw table further down). Small sample warning: 1 mirror is borderline; treat these as suggestive, not conclusive.

No card differed in at least 3 of the 1 mirror we have. The raw table below has the sparse signals that didn't clear the threshold — treat them as noise unless the sample grows.

Show all 25 cards with non-zero signal (raw numbers, including 25 held back from the grid above)
Card Edge In winner In loser Same Direction
Resolute Stand -1.00 0 1 0
Reduce to Ash -1.00 0 1 0
Cheshire Cat, Impish Grin +1.00 1 0 0
Staggering Strike -1.00 0 1 0
Palvor Sword +1.00 1 0 0
Grande Aiguille -1.00 0 1 0
Feu Awakening -1.00 0 1 0
Flamme Sorcel +1.00 1 0 0
Heavenly Guide +1.00 1 0 0
Rictus Tiding -1.00 0 1 0
Three of Hearts +1.00 1 0 0
Gentle Respite +1.00 1 0 0
Lost Providence -1.00 0 1 0
Inert Sword +1.00 1 0 0
Nullifying Lantern -1.00 0 1 0
Fast Cure -1.00 0 1 0
Turbo Charge -1.00 0 1 0
Infernal Vessel +1.00 1 0 0
Crystallized Destiny +1.00 1 0 0
Safeguard Amulet +1.00 1 0 0
The Looking Glass +1.00 1 0 0
Grim Pastiche -1.00 0 1 0
Tariff Ring -1.00 0 1 0
Carter, Synthetic Reaper -1.00 0 1 0
Regal Inquisition +1.00 1 0 0

JSON: /api/events/47865/archetypes/Fire Umbra Guardian