About the Deck Lab · Current experiment

Blind Deck Optimizer · Unlimited (2ED)

Can a deck learn to win?

Blind EC 93/94 Lab V2 · Unlimited

Start with the cards, not a winning deck list. The optimizer builds legal Magic decks, plays them against each other with Forge AI, then changes the promising ones and tests again. Follow the decks that emerge—and see whether they get better against a fixed set of opponents.

Experiment complete Last activity

Live progress

8,092 training games played

Connecting to live updates…
Experiment complete

The final population and diagnostics are frozen.

80 epochs completed Open-ended experiment · no fixed game limit
Epoch Complete 92 / 100 games 92.0% Results build toward the next population update
Experiment work in flight — games Checking running and prefetched work… Checking shared runner fleet…
Latest report Waiting for a report Completed games and failed attempts appear here automatically See recent games
01 / BUILD

No borrowed deck lists

“Blind” means no human card rankings or established strategies. Card text, mana costs, and legality are the starting clues.

02 / EVOLVE

Play, change, repeat

Forge AI plays both sides. Results help select survivors and propose changed decks for the next epoch.

03 / CHECK

Measure real progress

Separate games against fixed reference decks test improvement. Those results never guide which decks survive.

Watching the experiment requires no account. Open a deck to inspect its cards. These results describe Forge AI play within this experiment; they do not establish a human tournament ranking.

The current leaders

Leading decks

Top 12 by field score · select a deck to inspect it
Rank Deck Deck family Learned strength Field score Wins–losses–draws
1 —
Blind 93/94 E027-04 #2965 · born epoch 27
Black Vise / Mesa Pegasus / Ankh of Mishra +4.767 ± 0.522 87.4% 100% elite chance 13–10–0 23 games · 57%
2 —
Blind 93/94 E057-06 #3207 · born epoch 57
Black Vise / Mesa Pegasus / Ankh of Mishra +4.728 ± 0.577 87.0% 100% elite chance 3–1–0 4 games · 75%
3 ↑ 3
Blind 93/94 E067-04 #3285 · born epoch 67
Black Vise / Mesa Pegasus / Ankh of Mishra +4.694 ± 0.577 86.7% 100% elite chance 6–5–0 11 games · 55%
4 ↑ 4
Blind 93/94 E025-06 #2951 · born epoch 25
Black Vise / Mesa Pegasus / Ankh of Mishra +4.605 ± 0.534 85.8% 100% elite chance 15–12–0 27 games · 56%
5 —
Blind 93/94 E074-05 #3342 · born epoch 74
Black Vise / Mesa Pegasus / Ankh of Mishra +4.488 ± 0.579 84.5% 100% elite chance 1–3–0 4 games · 25%
6 ↑ 1
Blind 93/94 E071-06 #3319 · born epoch 71
Black Vise / Mesa Pegasus / Ankh of Mishra +4.477 ± 0.583 84.4% 100% elite chance 3–1–0 4 games · 75%
7 ↑ 6
Blind 93/94 E039-07 #3064 · born epoch 39
Black Vise / Mesa Pegasus / Ankh of Mishra +4.415 ± 0.551 83.7% 100% elite chance 8–8–0 16 games · 50%
8 ↓ 4
Blind 93/94 E030-05 #2990 · born epoch 30
Black Vise / Mesa Pegasus / Ankh of Mishra +4.408 ± 0.562 83.6% 100% elite chance 9–3–0 12 games · 75%
9 ↑ 7
Blind 93/94 E033-07 #3016 · born epoch 33
Black Vise / Mesa Pegasus / Ankh of Mishra +4.305 ± 0.545 82.3% 100% elite chance 11–10–0 21 games · 52%
10 ↓ 1
Blind 93/94 E079-03 #3380 · born epoch 79
Black Vise / Mesa Pegasus / Ankh of Mishra +4.256 ± 0.588 81.7% 100% elite chance 2–1–0 3 games · 67%
11 ↓ 1
Blind 93/94 E050-06 #3151 · born epoch 50
Black Vise / Mesa Pegasus / Ankh of Mishra +4.161 ± 0.563 80.5% 100% elite chance 8–9–0 17 games · 47%
12 New
Blind 93/94 E080-03 #3388 · born epoch 80
Black Vise / Mesa Pegasus / Ankh of Mishra +4.073 ± 0.590 79.3% 100% elite chance 2–0–0 2 games · 100%

Held-out validation

Is the population actually getting stronger?

Excluded from training and ratings

We test decks against the same fixed opponents throughout the experiment, so strength can be compared across epochs. Higher Battle Rating (BR) means stronger decks.

Current check E76 32 / 32 games complete
32 of 32
Latest complete E76 8 production references · 14 diagnostics · suite reference-league-v3
Frontier change since E16 0.0 pts 25.0% → 25.0% at E76 All-check trend: -1.4 pts per 10 epochs
Upper quartile change since E16 0.0 pts 25.0% → 25.0% at E76 All-check trend: +1.9 pts per 10 epochs
Median change since E16 +12.5 pts 12.5% → 25.0% at E76 All-check trend: +1.8 pts per 10 epochs
Lower quartile change since E16 0.0 pts 12.5% → 12.5% at E76 All-check trend: +0.7 pts per 10 epochs
Evidence included 17 complete checks 528 held-out games Partial checks stay on the chart but are excluded here.

Change compares the first and latest complete checks. “All-check trend” fits every complete check, so read it as direction—not certainty—until the intervals narrow.

Strength against a fixed scale

Each point is an independent held-out checkpoint at the labeled epoch; whiskers show an approximate 95% score interval.

Held-out population strength from epoch 1 through epoch 76 Frontier, upper-quartile, median, and lower-quartile active-deck win plus half-draw rates against the fixed reference league. Exact values follow the chart. E01 E16 frontier: 25.0%, 2–0–6 W-D-L E16 upper quartile: 25.0%, 2–0–6 W-D-L E16 typical deck: 12.5%, 1–0–7 W-D-L E16 lower quartile: 12.5%, 1–0–7 W-D-L E16E19 E21 frontier: 50.0%, 4–0–4 W-D-L E21 upper quartile: 25.0%, 2–0–6 W-D-L E21 typical deck: 25.0%, 2–0–6 W-D-L E21 lower quartile: 12.5%, 1–0–7 W-D-L E21 E26 frontier: 12.5%, 1–0–7 W-D-L E26 upper quartile: 0.0%, 0–0–8 W-D-L E26 typical deck: 12.5%, 1–0–7 W-D-L E26 lower quartile: 25.0%, 2–0–6 W-D-L E26 E31 frontier: 25.0%, 2–0–6 W-D-L E31 upper quartile: 12.5%, 1–0–7 W-D-L E31 typical deck: 12.5%, 1–0–7 W-D-L E31 lower quartile: 25.0%, 2–0–6 W-D-L E31 E36 frontier: 25.0%, 2–0–6 W-D-L E36 upper quartile: 25.0%, 2–0–6 W-D-L E36 typical deck: 0.0%, 0–0–8 W-D-L E36 lower quartile: 12.5%, 1–0–7 W-D-L E36E37 E41 frontier: 50.0%, 4–0–4 W-D-L E41 upper quartile: 12.5%, 1–0–7 W-D-L E41 typical deck: 12.5%, 1–0–7 W-D-L E41 lower quartile: 37.5%, 3–0–5 W-D-L E41 E46 frontier: 25.0%, 2–0–6 W-D-L E46 upper quartile: 25.0%, 2–0–6 W-D-L E46 typical deck: 25.0%, 2–0–6 W-D-L E46 lower quartile: 0.0%, 0–0–8 W-D-L E46 E51 frontier: 12.5%, 1–0–7 W-D-L E51 upper quartile: 37.5%, 3–0–5 W-D-L E51 typical deck: 12.5%, 1–0–7 W-D-L E51 lower quartile: 12.5%, 1–0–7 W-D-L E51E55 E56 frontier: 37.5%, 3–0–5 W-D-L E56 upper quartile: 37.5%, 3–0–5 W-D-L E56 typical deck: 12.5%, 1–0–7 W-D-L E56 lower quartile: 25.0%, 2–0–6 W-D-L E56 E61 frontier: 12.5%, 1–0–7 W-D-L E61 upper quartile: 25.0%, 2–0–6 W-D-L E61 typical deck: 12.5%, 1–0–7 W-D-L E61 lower quartile: 12.5%, 1–0–7 W-D-L E61 E66 frontier: 25.0%, 2–0–6 W-D-L E66 upper quartile: 37.5%, 3–0–5 W-D-L E66 typical deck: 12.5%, 1–0–7 W-D-L E66 lower quartile: 50.0%, 4–0–4 W-D-L E66 E71 frontier: 25.0%, 2–0–6 W-D-L E71 upper quartile: 12.5%, 1–0–7 W-D-L E71 typical deck: 37.5%, 3–0–5 W-D-L E71 lower quartile: 12.5%, 1–0–7 W-D-L E71E73 E76 frontier: 25.0%, 2–0–6 W-D-L E76 upper quartile: 25.0%, 2–0–6 W-D-L E76 typical deck: 25.0%, 2–0–6 W-D-L E76 lower quartile: 12.5%, 1–0–7 W-D-L E76
No held-out scores were recorded before E16. Rates count wins plus half of draws.

Basic land → Mox checks

Paired, held-out games compare otherwise identical fixed decks using the same opponent, orientation, and planned seed schedule. Positive values favor the Mox version. Individual rows test legal one-card swaps; the all-five-Mox row deliberately amplifies the effect as a high-signal simulator and analysis regression alarm, not as a legal deck estimate.

Distribution-free 95% bounds
Paired held-out estimates for replacing a basic land with each Mox
CardMox advantagePaired comparisonsConclusion
Mox Emerald +12.5% -84% to +100% 8 Inconclusive
Mox Jet 0.0% -96% to +96% 8 Inconclusive
Mox Pearl 0.0% -96% to +96% 8 Inconclusive
Mox Ruby 0.0% -96% to +96% 8 Inconclusive
Mox Sapphire 0.0% -96% to +96% 8 Inconclusive

Standard-candle ledger

Results from the latest complete check, from each candidate's point of view.

Record order: W–D–L
Held-out win, draw, and loss records against each standard candle
Fixed referenceFrontierUpper quartileMedianLower quartile
01 40 Mountain / 20 Mons's Goblin Raiders 0–0–0 0 / 0 played — 0–0–0 0 / 0 played —
02 40 Forest / 20 Grizzly Bears 0–0–0 0 / 0 played — 0–0–0 0 / 0 played —
03 40 Plains / 20 Savannah Lions 0–0–0 0 / 0 played — 0–0–0 0 / 0 played —
04 40 Mountain / 20 Lightning Bolt 0–0–0 0 / 0 played — 0–0–0 0 / 0 played —
05 40 Plains / 20 Savannah Lions 0–0–0 0 / 0 played — 0–0–0 0 / 0 played —
06 39 Plains / 1 Mox Pearl / 20 Savannah Lions 0–0–0 0 / 0 played — 0–0–0 0 / 0 played —
07 40 Island / 20 Merfolk of the Pearl Trident 0–0–0 0 / 0 played — 0–0–0 0 / 0 played —
08 39 Island / 1 Mox Sapphire / 20 Merfolk of the Pearl Trident 0–0–0 0 / 0 played — 0–0–0 0 / 0 played —
09 40 Swamp / 20 Scathe Zombies 0–0–0 0 / 0 played — 0–0–0 0 / 0 played —
10 39 Swamp / 1 Mox Jet / 20 Scathe Zombies 0–0–0 0 / 0 played — 0–0–0 0 / 0 played —
11 40 Mountain / 20 Mons's Goblin Raiders 0–0–0 0 / 0 played — 0–0–0 0 / 0 played —
12 39 Mountain / 1 Mox Ruby / 20 Mons's Goblin Raiders 0–0–0 0 / 0 played — 0–0–0 0 / 0 played —
13 40 Forest / 20 Grizzly Bears 0–0–0 0 / 0 played — 0–0–0 0 / 0 played —
14 39 Forest / 1 Mox Emerald / 20 Grizzly Bears 0–0–0 0 / 0 played — 0–0–0 0 / 0 played —
15 Crap Wall Deck · Floor strength anchor · production BR 917 1–0–0 1 / 1 played 1–0–0 1 / 1 played 1–0–0 1 / 1 played 1–0–0 1 / 1 played
16 Vintage Green Stompy · Lower strength anchor · production BR 1515 0–0–1 1 / 1 played 1–0–0 1 / 1 played 0–0–1 1 / 1 played 0–0–1 1 / 1 played
17 Vintage Goblins · Middle strength anchor · production BR 1592 0–0–1 1 / 1 played 0–0–1 1 / 1 played 0–0–1 1 / 1 played 0–0–1 1 / 1 played
18 Vintage Mono Red · Middle strength anchor · production BR 1601 0–0–1 1 / 1 played 0–0–1 1 / 1 played 1–0–0 1 / 1 played 0–0–1 1 / 1 played
19 Vintage Red Green Ramp · UpperMiddle strength anchor · production BR 1639 0–0–1 1 / 1 played 0–0–1 1 / 1 played 0–0–1 1 / 1 played 0–0–1 1 / 1 played
20 Vintage Crusaders · UpperMiddle strength anchor · production BR 1670 1–0–0 1 / 1 played 0–0–1 1 / 1 played 0–0–1 1 / 1 played 0–0–1 1 / 1 played
21 Vintage Bad Moon · Strong strength anchor · production BR 1743 0–0–1 1 / 1 played 0–0–1 1 / 1 played 0–0–1 1 / 1 played 0–0–1 1 / 1 played
22 Vintage Mono Blue · Strong strength anchor · production BR 1800 0–0–1 1 / 1 played 0–0–1 1 / 1 played 0–0–1 1 / 1 played 0–0–1 1 / 1 played
View exact trajectory data
EpochLearning gamesValidationFrontierUpper quartileMedianLower quartile
E16 1,600 32 / 32 25.0% (7%–59%) · 2–0–6 25.0% (7%–59%) · 2–0–6 12.5% (2%–47%) · 1–0–7 12.5% (2%–47%) · 1–0–7
E19 1,900 28 / 28 — · 0–0–0 — · 0–0–0 — · 0–0–0 — · 0–0–0
E21 2,100 32 / 32 50.0% (22%–78%) · 4–0–4 25.0% (7%–59%) · 2–0–6 25.0% (7%–59%) · 2–0–6 12.5% (2%–47%) · 1–0–7
E26 2,600 32 / 32 12.5% (2%–47%) · 1–0–7 0.0% (0%–32%) · 0–0–8 12.5% (2%–47%) · 1–0–7 25.0% (7%–59%) · 2–0–6
E31 3,100 32 / 32 25.0% (7%–59%) · 2–0–6 12.5% (2%–47%) · 1–0–7 12.5% (2%–47%) · 1–0–7 25.0% (7%–59%) · 2–0–6
E36 3,600 32 / 32 25.0% (7%–59%) · 2–0–6 25.0% (7%–59%) · 2–0–6 0.0% (0%–32%) · 0–0–8 12.5% (2%–47%) · 1–0–7
E37 3,700 28 / 28 — · 0–0–0 — · 0–0–0 — · 0–0–0 — · 0–0–0
E41 4,100 32 / 32 50.0% (22%–78%) · 4–0–4 12.5% (2%–47%) · 1–0–7 12.5% (2%–47%) · 1–0–7 37.5% (14%–69%) · 3–0–5
E46 4,600 32 / 32 25.0% (7%–59%) · 2–0–6 25.0% (7%–59%) · 2–0–6 25.0% (7%–59%) · 2–0–6 0.0% (0%–32%) · 0–0–8
E51 5,100 32 / 32 12.5% (2%–47%) · 1–0–7 37.5% (14%–69%) · 3–0–5 12.5% (2%–47%) · 1–0–7 12.5% (2%–47%) · 1–0–7
E55 5,500 28 / 28 — · 0–0–0 — · 0–0–0 — · 0–0–0 — · 0–0–0
E56 5,600 32 / 32 37.5% (14%–69%) · 3–0–5 37.5% (14%–69%) · 3–0–5 12.5% (2%–47%) · 1–0–7 25.0% (7%–59%) · 2–0–6
E61 6,100 32 / 32 12.5% (2%–47%) · 1–0–7 25.0% (7%–59%) · 2–0–6 12.5% (2%–47%) · 1–0–7 12.5% (2%–47%) · 1–0–7
E66 6,600 32 / 32 25.0% (7%–59%) · 2–0–6 37.5% (14%–69%) · 3–0–5 12.5% (2%–47%) · 1–0–7 50.0% (22%–78%) · 4–0–4
E71 7,100 32 / 32 25.0% (7%–59%) · 2–0–6 12.5% (2%–47%) · 1–0–7 37.5% (14%–69%) · 3–0–5 12.5% (2%–47%) · 1–0–7
E73 7,300 28 / 28 — · 0–0–0 — · 0–0–0 — · 0–0–0 — · 0–0–0
E76 7,600 32 / 32 25.0% (7%–59%) · 2–0–6 25.0% (7%–59%) · 2–0–6 25.0% (7%–59%) · 2–0–6 12.5% (2%–47%) · 1–0–7

Live result feed

What just played—and what it changes

Newest first · records update now · rankings follow the live model
Current ranking Live · E80 128 decks ordered by field score
New evidence 92 games since the last model refresh
Next ranking refresh E81 8 more games

A result immediately adds a win, draw, or loss to both decks. The shared model then refits at the epoch boundary, which can move every deck because evidence carries across related cards and strategies.

Finished Result Where they rank now Ranking impact
Win
Blind 93/94 E053-04 defeated Blind 93/94 E080-06
#58 vs #78 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E056-02 defeated Blind 93/94 E080-08
#86 vs #128 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E047-03 defeated Blind 93/94 E080-08
#85 vs #128 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E080-06 defeated Blind 93/94 E072-07
#78 vs #110 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E073-07 defeated Blind 93/94 E080-06
#69 vs #78 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E063-04 defeated Blind 93/94 E080-08
#102 vs #128 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E062-02 defeated Blind 93/94 E080-08
#105 vs #128 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E045-06 defeated Blind 93/94 E080-08
#112 vs #128 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E052-03 defeated Blind 93/94 E080-08
#92 vs #128 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E079-02 defeated Blind 93/94 E080-08
#95 vs #128 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E060-06 defeated Blind 93/94 E080-08
#91 vs #128 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E077-06 defeated Blind 93/94 E048-03
#81 vs #96 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E077-05 defeated Blind 93/94 E080-08
#106 vs #128 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E041-02 defeated Blind 93/94 E080-06
#66 vs #78 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E080-01 defeated Blind 93/94 E080-05
#88 vs #115 Feeds E81 +1 win / +1 loss now · model update pending
Win
Blind 93/94 E046-03 defeated Blind 93/94 E080-08
#94 vs #128 Feeds E81 +1 win / +1 loss now · model update pending

Complete field

Every deck currently in rotation

128 of 128 decks

This is the complete 128-deck field currently eligible for self-play—incumbents plus any challengers—ranked by predicted score against the field. Movement compares this live order with E79.

Rank Move Deck Deck family Field score Learned strength Wins–losses–draws
1 of 128 —
Blind 93/94 E027-04 #2965 · born E27 · Mana and copies
Black Vise / Mesa Pegasus / Ankh of Mishra 87.4% 100% elite chance +4.767 ± 0.522 13–10–0 23 games · 57%
2 of 128 —
Blind 93/94 E057-06 #3207 · born E57 · Text package
Black Vise / Mesa Pegasus / Ankh of Mishra 87.0% 100% elite chance +4.728 ± 0.577 3–1–0 4 games · 75%
3 of 128 ↑ 3
Blind 93/94 E067-04 #3285 · born E67 · Mana and copies
Black Vise / Mesa Pegasus / Ankh of Mishra 86.7% 100% elite chance +4.694 ± 0.577 6–5–0 11 games · 55%
4 of 128 ↑ 4
Blind 93/94 E025-06 #2951 · born E25 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 85.8% 100% elite chance +4.605 ± 0.534 15–12–0 27 games · 56%
5 of 128 —
Blind 93/94 E074-05 #3342 · born E74 · Mana and copies
Black Vise / Mesa Pegasus / Ankh of Mishra 84.5% 100% elite chance +4.488 ± 0.579 1–3–0 4 games · 25%
6 of 128 ↑ 1
Blind 93/94 E071-06 #3319 · born E71 · Text package
Black Vise / Mesa Pegasus / Ankh of Mishra 84.4% 100% elite chance +4.477 ± 0.583 3–1–0 4 games · 75%
7 of 128 ↑ 6
Blind 93/94 E039-07 #3064 · born E39 · Mana and copies
Black Vise / Mesa Pegasus / Ankh of Mishra 83.7% 100% elite chance +4.415 ± 0.551 8–8–0 16 games · 50%
8 of 128 ↓ 4
Blind 93/94 E030-05 #2990 · born E30 · Text package
Black Vise / Mesa Pegasus / Ankh of Mishra 83.6% 100% elite chance +4.408 ± 0.562 9–3–0 12 games · 75%
9 of 128 ↑ 7
Blind 93/94 E033-07 #3016 · born E33 · Mana and copies
Black Vise / Mesa Pegasus / Ankh of Mishra 82.3% 100% elite chance +4.305 ± 0.545 11–10–0 21 games · 52%
10 of 128 ↓ 1
Blind 93/94 E079-03 #3380 · born E79 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 81.7% 100% elite chance +4.256 ± 0.588 2–1–0 3 games · 67%
11 of 128 ↓ 1
Blind 93/94 E050-06 #3151 · born E50 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 80.5% 100% elite chance +4.161 ± 0.563 8–9–0 17 games · 47%
12 of 128 New
Blind 93/94 E080-03 #3388 · born E80 · Text package
Black Vise / Mesa Pegasus / Ankh of Mishra 79.3% 100% elite chance +4.073 ± 0.590 2–0–0 2 games · 100%
13 of 128 ↓ 1
Blind 93/94 E075-03 #3348 · born E75 · Text package
Black Vise / Mesa Pegasus / Ankh of Mishra 77.9% 100% elite chance +3.971 ± 0.577 3–0–0 3 games · 100%
14 of 128 ↓ 3
Blind 93/94 E046-05 #3118 · born E46 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 77.1% 100% elite chance +3.911 ± 0.574 4–3–0 7 games · 57%
15 of 128 ↓ 12
Blind 93/94 E016-05 #2878 · born E16 · Mana and copies
Black Vise / Mesa Pegasus / Ankh of Mishra 77.0% 100% elite chance +3.906 ± 0.520 12–8–0 20 games · 60%
16 of 128 ↑ 2
Blind 93/94 E028-04 #2973 · born E28 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 76.6% 100% elite chance +3.878 ± 0.561 10–5–0 15 games · 67%
17 of 128 ↑ 37
Blind 93/94 E070-04 #3309 · born E70 · Local substitution
Roc of Kher Ridges / Regrowth / Balance 73.9% 100% elite chance +3.696 ± 0.522 13–8–0 21 games · 62%
18 of 128 ↑ 5
Blind 93/94 E019-07 #2904 · born E19 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 73.0% 100% elite chance +3.641 ± 0.550 11–11–0 22 games · 50%
19 of 128 ↑ 25
Blind 93/94 E055-07 #3192 · born E55 · Text package
Black Vise / Mesa Pegasus / Ankh of Mishra 72.7% 100% elite chance +3.624 ± 0.560 7–5–0 12 games · 58%
20 of 128 ↑ 70
Blind 93/94 E062-04 #3245 · born E62 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 72.1% 100% elite chance +3.584 ± 0.518 12–12–0 24 games · 50%
21 of 128 ↓ 1
Blind 93/94 0732 #2535 · born E00 · Balanced random
Black Vise / Mesa Pegasus / Ankh of Mishra 72.0% 100% elite chance +3.577 ± 0.547 9–9–0 18 games · 50%
22 of 128 ↑ 9
Blind 93/94 E026-05 #2958 · born E26 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 71.5% 100% elite chance +3.546 ± 0.544 11–12–0 23 games · 48%
23 of 128 ↑ 48
Blind 93/94 E063-06 #3255 · born E63 · Macro
Holy Strength / Righteousness / Black Ward 71.4% 100% elite chance +3.538 ± 0.497 17–10–0 27 games · 63%
24 of 128 ↑ 13
Blind 93/94 E059-05 #3222 · born E59 · Local substitution
Holy Strength / Righteousness / Black Ward 70.8% 100% elite chance +3.505 ± 0.550 4–7–0 11 games · 36%
25 of 128 ↑ 31
Blind 93/94 E021-02 #2915 · born E21 · Text package
Roc of Kher Ridges / Regrowth / Balance 70.7% 100% elite chance +3.498 ± 0.500 26–28–0 54 games · 48%
26 of 128 ↑ 17
Blind 93/94 E062-05 #3246 · born E62 · Text package
Holy Strength / Righteousness / Black Ward 70.5% 100% elite chance +3.487 ± 0.541 7–7–0 14 games · 50%
27 of 128 ↑ 14
Blind 93/94 E066-05 #3278 · born E66 · Mana and copies
Holy Strength / Righteousness / Black Ward 70.5% 100% elite chance +3.485 ± 0.533 7–7–0 14 games · 50%
28 of 128 ↓ 13
Blind 93/94 E020-06 #2911 · born E20 · Mana and copies
Black Vise / Mesa Pegasus / Ankh of Mishra 68.6% 100% elite chance +3.371 ± 0.547 4–9–0 13 games · 31%
29 of 128 ↑ 52
Blind 93/94 E048-02 #3131 · born E48 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 67.7% 100% elite chance +3.317 ± 0.543 9–11–0 20 games · 45%
30 of 128 ↑ 15
Blind 93/94 E065-04 #3269 · born E65 · Local substitution
Holy Strength / Righteousness / Black Ward 67.4% 100% elite chance +3.295 ± 0.545 5–4–0 9 games · 56%
31 of 128 ↑ 55
Blind 93/94 E067-06 #3287 · born E67 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 67.3% 100% elite chance +3.290 ± 0.486 19–19–0 38 games · 50%
32 of 128 ↑ 50
Blind 93/94 E068-07 #3296 · born E68 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 67.0% 100% elite chance +3.275 ± 0.494 15–17–0 32 games · 47%
33 of 128 ↑ 47
Blind 93/94 E049-07 #3144 · born E49 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 66.2% 100% elite chance +3.228 ± 0.489 20–18–0 38 games · 53%
34 of 128 ↓ 20
Blind 93/94 E072-06 #3327 · born E72 · Mana and copies
Thoughtlace / Library of Leng / Orcish Oriflamme 66.0% 100% elite chance +3.217 ± 0.569 5–2–0 7 games · 71%
35 of 128 ↑ 61
Blind 93/94 E041-04 #3077 · born E41 · Macro
Mox Ruby / Ironclaw Orcs / Kudzu 66.0% 100% elite chance +3.214 ± 0.527 16–14–0 30 games · 53%
36 of 128 ↑ 28
Blind 93/94 E025-05 #2950 · born E25 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 65.6% 100% elite chance +3.191 ± 0.507 25–32–0 57 games · 44%
37 of 128 ↑ 31
Blind 93/94 E026-04 #2957 · born E26 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 65.5% 100% elite chance +3.189 ± 0.495 20–24–0 44 games · 45%
38 of 128 ↓ 2
Blind 93/94 E029-05 #2982 · born E29 · Text package
Black Vise / Mesa Pegasus / Ankh of Mishra 65.4% 100% elite chance +3.183 ± 0.533 9–11–0 20 games · 45%
39 of 128 ↑ 30
Blind 93/94 E020-03 #2908 · born E20 · Random restart
Black Vise / Mesa Pegasus / Ankh of Mishra 65.3% 100% elite chance +3.173 ± 0.482 36–52–0 88 games · 41%
40 of 128 ↑ 52
Blind 93/94 E042-04 #3085 · born E42 · Local substitution
Mox Ruby / Ironclaw Orcs / Kudzu 64.0% 100% elite chance +3.101 ± 0.535 11–11–0 22 games · 50%
41 of 128 ↑ 21
Blind 93/94 E060-04 #3229 · born E60 · Local substitution
Holy Strength / Righteousness / Black Ward 64.0% 100% elite chance +3.100 ± 0.540 6–10–0 16 games · 38%
42 of 128 ↑ 62
Blind 93/94 E075-04 #3349 · born E75 · Mana and copies
Holy Strength / Righteousness / Black Ward 63.1% 100% elite chance +3.052 ± 0.496 14–14–0 28 games · 50%
43 of 128 ↓ 24
Blind 93/94 E070-06 #3311 · born E70 · Local substitution
Thoughtlace / Library of Leng / Orcish Oriflamme 63.0% 99% elite chance +3.042 ± 0.583 4–1–0 5 games · 80%
44 of 128 ↑ 8
Blind 93/94 E058-06 #3215 · born E58 · Local substitution
Holy Strength / Righteousness / Black Ward 62.8% 99% elite chance +3.032 ± 0.551 3–3–0 6 games · 50%
45 of 128 ↑ 28
Blind 93/94 E028-03 #2972 · born E28 · Macro
Holy Strength / Righteousness / Black Ward 62.3% 99% elite chance +3.002 ± 0.535 11–12–0 23 games · 48%
46 of 128 ↓ 20
Blind 93/94 E075-07 #3352 · born E75 · Local substitution
Thoughtlace / Library of Leng / Orcish Oriflamme 62.0% 99% elite chance +2.988 ± 0.572 3–2–0 5 games · 60%
47 of 128 ↑ 25
Blind 93/94 E035-06 #3031 · born E35 · Local substitution
Holy Strength / Righteousness / Black Ward 61.9% 99% elite chance +2.979 ± 0.512 13–13–0 26 games · 50%
48 of 128 ↑ 29
Blind 93/94 E040-07 #3072 · born E40 · Macro
Holy Strength / Righteousness / Black Ward 60.6% 99% elite chance +2.908 ± 0.517 15–17–0 32 games · 47%
49 of 128 ↑ 18
Blind 93/94 E039-02 #3059 · born E39 · Local substitution
Holy Strength / Righteousness / Black Ward 60.1% 99% elite chance +2.880 ± 0.532 11–12–0 23 games · 48%
50 of 128 ↑ 38
Blind 93/94 E053-06 #3175 · born E53 · Local substitution
Black Vise / Mesa Pegasus / Ankh of Mishra 59.6% 99% elite chance +2.852 ± 0.513 8–22–0 30 games · 27%
51 of 128 ↓ 12
Blind 93/94 E054-04 #3181 · born E54 · Mana and copies
Thoughtlace / Library of Leng / Orcish Oriflamme 59.5% 98% elite chance +2.848 ± 0.550 7–10–0 17 games · 41%
52 of 128 ↑ 33
Blind 93/94 E042-07 #3088 · born E42 · Local substitution
Holy Strength / Righteousness / Black Ward 59.4% 99% elite chance +2.844 ± 0.501 16–20–0 36 games · 44%
53 of 128 ↓ 7
Blind 93/94 E070-07 #3312 · born E70 · Local substitution
Thoughtlace / Library of Leng / Orcish Oriflamme 59.1% 98% elite chance +2.823 ± 0.547 7–10–0 17 games · 41%
54 of 128 ↑ 29
Blind 93/94 E038-06 #3055 · born E38 · Local substitution
Holy Strength / Righteousness / Black Ward 58.7% 98% elite chance +2.806 ± 0.522 14–19–0 33 games · 42%
55 of 128 ↑ 52
Blind 93/94 E054-06 #3183 · born E54 · Mana and copies
Mox Ruby / Ironclaw Orcs / Kudzu 57.6% 98% elite chance +2.744 ± 0.514 11–22–0 33 games · 33%
56 of 128 ↑ 20
Blind 93/94 0252 #2055 · born E00 · Balanced random
Roc of Kher Ridges / Regrowth / Balance 57.3% 97% elite chance +2.725 ± 0.527 27–34–0 61 games · 44%
57 of 128 ↑ 32
Blind 93/94 E057-04 #3205 · born E57 · Local substitution
Holy Strength / Righteousness / Black Ward 57.2% 98% elite chance +2.721 ± 0.507 16–12–0 28 games · 57%
58 of 128 ↑ 33
Blind 93/94 E053-04 #3173 · born E53 · Mana and copies
Black Vise / Mesa Pegasus / Ankh of Mishra 57.1% 98% elite chance +2.713 ± 0.519 21–18–0 39 games · 54%
59 of 128 ↓ 6
Blind 93/94 E015-05 #2870 · born E15 · Text package
Thoughtlace / Library of Leng / Orcish Oriflamme 56.9% 98% elite chance +2.704 ± 0.501 27–45–0 72 games · 38%
60 of 128 ↑ 15
Blind 93/94 E068-05 #3294 · born E68 · Local substitution
Holy Strength / Righteousness / Black Ward 56.3% 96% elite chance +2.672 ± 0.549 3–7–0 10 games · 30%
61 of 128 ↓ 13
Blind 93/94 E079-05 #3382 · born E79 · Local substitution
Thoughtlace / Library of Leng / Orcish Oriflamme 56.3% 95% elite chance +2.670 ± 0.577 1–3–0 4 games · 25%
62 of 128 ↑ 35
Blind 93/94 E064-05 #3262 · born E64 · Text package
Black Vise / Mesa Pegasus / Ankh of Mishra 55.8% 98% elite chance +2.643 ± 0.481 12–24–0 36 games · 33%
63 of 128 ↓ 31
Blind 93/94 E071-05 #3318 · born E71 · Local substitution
Thoughtlace / Library of Leng / Orcish Oriflamme 55.3% 95% elite chance +2.617 ± 0.575 2–2–0 4 games · 50%
64 of 128 ↑ 36
Blind 93/94 E051-05 #3158 · born E51 · Text package
Holy Strength / Righteousness / Black Ward 55.1% 97% elite chance +2.604 ± 0.490 18–24–0 42 games · 43%
65 of 128 ↑ 34
Blind 93/94 E033-03 #3012 · born E33 · Local substitution
Holy Strength / Righteousness / Black Ward 54.2% 96% elite chance +2.555 ± 0.506 16–27–0 43 games · 37%
66 of 128 ↑ 32
Blind 93/94 E041-02 #3075 · born E41 · Text package
Black Vise / Mesa Pegasus / Ankh of Mishra 53.7% 95% elite chance +2.530 ± 0.522 20–22–0 42 games · 48%
67 of 128 ↑ 45
Blind 93/94 E022-01 #2922 · born E22 · Text package
Black Vise / Mesa Pegasus / Ankh of Mishra 53.3% 95% elite chance +2.508 ± 0.494 41–24–0 65 games · 63%
68 of 128 ↓ 2
Blind 93/94 0350 #2153 · born E00 · Balanced random
Thoughtlace / Library of Leng / Orcish Oriflamme 52.5% 94% elite chance +2.464 ± 0.501 39–55–0 94 games · 41%
69 of 128 ↑ 10
Blind 93/94 E073-07 #3336 · born E73 · Local substitution
Holy Strength / Righteousness / Black Ward 52.3% 92% elite chance +2.450 ± 0.531 5–7–0 12 games · 42%
70 of 128 ↑ 24
Blind 93/94 E032-07 #3008 · born E32 · Mana and copies
Roc of Kher Ridges / Regrowth / Balance 52.0% 93% elite chance +2.434 ± 0.512 19–35–0 54 games · 35%
71 of 128 ↑ 31
Blind 93/94 E063-07 #3256 · born E63 · Text package
Black Vise / Mesa Pegasus / Ankh of Mishra 51.9% 92% elite chance +2.427 ± 0.529 8–20–0 28 games · 29%
72 of 128 ↑ 37
Blind 93/94 E043-05 #3094 · born E43 · Local substitution
Mox Ruby / Ironclaw Orcs / Kudzu 50.8% 90% elite chance +2.368 ± 0.515 20–25–0 45 games · 44%
73 of 128 ↑ 20
Blind 93/94 0085 #1888 · born E00 · Balanced random
Black Vise / Mesa Pegasus / Ankh of Mishra 47.8% 85% elite chance +2.204 ± 0.500 35–31–0 66 games · 53%
74 of 128 ↑ 27
Blind 93/94 E056-06 #3199 · born E56 · Text package
Holy Strength / Righteousness / Black Ward 47.8% 85% elite chance +2.203 ± 0.493 7–25–0 32 games · 22%
75 of 128 ↓ 14
Blind 93/94 E034-01 #3018 · born E34 · Macro
Granite Gargoyle / Black Lotus / Pearled Unicorn 44.3% 73% elite chance +2.007 ± 0.499 24–9–0 33 games · 73%
76 of 128 ↑ 27
Blind 93/94 E027-02 #2963 · born E27 · Text package
Roc of Kher Ridges / Regrowth / Balance 44.0% 73% elite chance +1.995 ± 0.479 30–36–0 66 games · 45%
77 of 128 ↓ 42
Blind 93/94 E015-06 #2871 · born E15 · Local substitution
Roc of Kher Ridges / Regrowth / Balance 39.6% 54% elite chance +1.744 ± 0.505 15–11–0 26 games · 58%
78 of 128 New
Blind 93/94 E080-06 #3391 · born E80 · Local substitution
Roc of Kher Ridges / Regrowth / Balance 39.3% 52% elite chance +1.722 ± 0.602 7–17–0 24 games · 29%
79 of 128 ↑ 32
Blind 93/94 E061-07 #3240 · born E61 · Mana and copies
Mox Ruby / Ironclaw Orcs / Kudzu 38.8% 50% elite chance +1.695 ± 0.474 33–20–0 53 games · 62%
80 of 128 ↑ 28
Blind 93/94 E030-03 #2988 · born E30 · Text package
Roc of Kher Ridges / Regrowth / Balance 38.6% 49% elite chance +1.682 ± 0.479 24–42–0 66 games · 36%
81 of 128 ↓ 48
Blind 93/94 E077-06 #3367 · born E77 · Local substitution
Roc of Kher Ridges / Regrowth / Balance 38.6% 49% elite chance +1.682 ± 0.563 5–6–0 11 games · 45%
82 of 128 ↓ 33
Blind 93/94 E031-03 #2996 · born E31 · Mana and copies
Roc of Kher Ridges / Regrowth / Balance 37.9% 46% elite chance +1.642 ± 0.497 16–13–0 29 games · 55%
83 of 128 ↑ 27
Blind 93/94 E030-04 #2989 · born E30 · Text package
Holy Strength / Righteousness / Black Ward 37.1% 42% elite chance +1.593 ± 0.485 17–57–0 74 games · 23%
84 of 128 ↓ 55
Blind 93/94 E045-07 #3112 · born E45 · Text package
Roc of Kher Ridges / Regrowth / Balance 36.9% 42% elite chance +1.582 ± 0.536 6–9–0 15 games · 40%
85 of 128 ↓ 34
Blind 93/94 E047-03 #3124 · born E47 · Local substitution
Granite Gargoyle / Black Lotus / Pearled Unicorn 36.4% 39% elite chance +1.552 ± 0.514 18–5–0 23 games · 78%
86 of 128 ↓ 28
Blind 93/94 E056-02 #3195 · born E56 · Macro
Roc of Kher Ridges / Regrowth / Balance 36.0% 37% elite chance +1.526 ± 0.521 16–5–0 21 games · 76%
87 of 128 ↑ 26
Blind 93/94 E046-06 #3119 · born E46 · Macro
Mox Ruby / Ironclaw Orcs / Kudzu 35.7% 36% elite chance +1.513 ± 0.493 33–11–0 44 games · 75%
88 of 128 New
Blind 93/94 E080-01 #3386 · born E80 · Macro
Granite Gargoyle / Black Lotus / Pearled Unicorn 35.2% 36% elite chance +1.482 ± 0.586 6–2–0 8 games · 75%
89 of 128 ↓ 39
Blind 93/94 E065-06 #3271 · born E65 · Local substitution
Roc of Kher Ridges / Regrowth / Balance 35.1% 33% elite chance +1.474 ± 0.518 9–10–0 19 games · 47%
90 of 128 ↑ 24
Blind 93/94 E078-05 #3374 · born E78 · Mana and copies
Mox Ruby / Ironclaw Orcs / Kudzu 34.8% 33% elite chance +1.457 ± 0.527 10–7–0 17 games · 59%
91 of 128 ↓ 67
Blind 93/94 E060-06 #3231 · born E60 · Local substitution
Granite Gargoyle / Black Lotus / Pearled Unicorn 34.3% 32% elite chance +1.427 ± 0.587 9–1–0 10 games · 90%
92 of 128 ↓ 33
Blind 93/94 E052-03 #3164 · born E52 · Local substitution
Granite Gargoyle / Black Lotus / Pearled Unicorn 33.6% 28% elite chance +1.384 ± 0.524 13–5–0 18 games · 72%
93 of 128 ↓ 23
Blind 93/94 E035-07 #3032 · born E35 · Macro
Roc of Kher Ridges / Regrowth / Balance 33.4% 25% elite chance +1.374 ± 0.481 22–21–0 43 games · 51%
94 of 128 ↓ 73
Blind 93/94 E046-03 #3116 · born E46 · Local substitution
Granite Gargoyle / Black Lotus / Pearled Unicorn 33.2% 27% elite chance +1.358 ± 0.550 15–4–0 19 games · 79%
95 of 128 ↑ 21
Blind 93/94 E079-02 #3379 · born E79 · Mana and copies
Ley Druid / Firebreathing / Ironroot Treefolk 33.0% 27% elite chance +1.348 ± 0.571 4–8–0 12 games · 33%
96 of 128 ↑ 19
Blind 93/94 E048-03 #3132 · born E48 · Text package
Mox Ruby / Ironclaw Orcs / Kudzu 32.6% 20% elite chance +1.320 ± 0.452 43–10–0 53 games · 81%
97 of 128 ↑ 23
Blind 93/94 E076-08 #3361 · born E76 · Random restart
Merfolk of the Pearl Trident / Clone / Counterspell 32.0% 24% elite chance +1.280 ± 0.586 22–8–0 30 games · 73%
98 of 128 ↓ 64
Blind 93/94 E036-01 #3034 · born E36 · Local substitution
Granite Gargoyle / Black Lotus / Pearled Unicorn 30.7% 18% elite chance +1.202 ± 0.546 14–10–0 24 games · 58%
99 of 128 ↓ 74
Blind 93/94 E048-04 #3133 · born E48 · Mana and copies
Granite Gargoyle / Black Lotus / Pearled Unicorn 30.6% 16% elite chance +1.192 ± 0.499 32–7–0 39 games · 82%
100 of 128 ↓ 83
Blind 93/94 E062-03 #3244 · born E62 · Local substitution
Mox Ruby / Ironclaw Orcs / Kudzu 29.7% 15% elite chance +1.133 ± 0.537 16–6–0 22 games · 73%
101 of 128 ↓ 74
Blind 93/94 E073-04 #3333 · born E73 · Text package
Roc of Kher Ridges / Regrowth / Balance 29.7% 16% elite chance +1.133 ± 0.570 4–3–0 7 games · 57%
102 of 128 ↓ 47
Blind 93/94 E063-04 #3253 · born E63 · Text package
Roc of Kher Ridges / Regrowth / Balance 29.2% 13% elite chance +1.103 ± 0.520 9–13–0 22 games · 41%
103 of 128 ↑ 18
Blind 93/94 E079-06 #3383 · born E79 · Local substitution
Merfolk of the Pearl Trident / Clone / Counterspell 29.2% 18% elite chance +1.100 ± 0.642 8–4–0 12 games · 67%
104 of 128 ↑ 13
Blind 93/94 E057-05 #3206 · born E57 · Mana and copies
Ley Druid / Firebreathing / Ironroot Treefolk 29.0% 11% elite chance +1.086 ± 0.496 13–45–1 59 games · 23%
105 of 128 —
Blind 93/94 E062-02 #3243 · born E62 · Text package
Roc of Kher Ridges / Regrowth / Balance 28.7% 10% elite chance +1.070 ± 0.477 21–14–0 35 games · 60%
106 of 128 ↓ 28
Blind 93/94 E077-05 #3366 · born E77 · Text package
Roc of Kher Ridges / Regrowth / Balance 28.4% 13% elite chance +1.050 ± 0.566 4–4–0 8 games · 50%
107 of 128 ↓ 33
Blind 93/94 0273 #2076 · born E00 · Balanced random
Roc of Kher Ridges / Regrowth / Balance 27.9% 7% elite chance +1.011 ± 0.464 33–31–0 64 games · 52%
108 of 128 ↓ 66
Blind 93/94 E079-07 #3384 · born E79 · Macro
Mox Ruby / Ironclaw Orcs / Kudzu 27.8% 12% elite chance +1.005 ± 0.596 4–2–0 6 games · 67%
109 of 128 ↓ 46
Blind 93/94 E059-07 #3224 · born E59 · Mana and copies
Roc of Kher Ridges / Regrowth / Balance 27.6% 9% elite chance +0.992 ± 0.513 13–13–0 26 games · 50%
110 of 128 ↓ 26
Blind 93/94 E072-07 #3328 · born E72 · Text package
Mox Ruby / Ironclaw Orcs / Kudzu 27.3% 8% elite chance +0.974 ± 0.508 25–9–0 34 games · 74%
111 of 128 New
Blind 93/94 E080-02 #3387 · born E80 · Local substitution
Ley Druid / Firebreathing / Ironroot Treefolk 27.1% 11% elite chance +0.958 ± 0.608 3–3–0 6 games · 50%
112 of 128 ↑ 6
Blind 93/94 E045-06 #3111 · born E45 · Local substitution
Mox Ruby / Ironclaw Orcs / Kudzu 26.7% 5% elite chance +0.934 ± 0.475 43–16–0 59 games · 73%
113 of 128 ↓ 91
Blind 93/94 E060-02 #3227 · born E60 · Text package
Mox Ruby / Ironclaw Orcs / Kudzu 24.3% 6% elite chance +0.761 ± 0.585 5–7–0 12 games · 42%
114 of 128 ↓ 84
Blind 93/94 0814 #2617 · born E00 · Balanced random
Granite Gargoyle / Black Lotus / Pearled Unicorn 23.7% 2% elite chance +0.716 ± 0.478 39–17–0 56 games · 70%
115 of 128 New
Blind 93/94 E080-05 #3390 · born E80 · Local substitution
Merfolk of the Pearl Trident / Clone / Counterspell 21.8% 4% elite chance +0.565 ± 0.648 2–3–0 5 games · 40%
116 of 128 ↓ 78
Blind 93/94 E070-05 #3310 · born E70 · Text package
Roc of Kher Ridges / Regrowth / Balance 21.2% 2% elite chance +0.520 ± 0.582 2–5–0 7 games · 29%
117 of 128 ↓ 77
Blind 93/94 E052-07 #3168 · born E52 · Text package
Roc of Kher Ridges / Regrowth / Balance 19.8% 1% elite chance +0.405 ± 0.534 9–8–0 17 games · 53%
118 of 128 ↓ 12
Blind 93/94 E078-06 #3375 · born E78 · Text package
Holy Strength / Righteousness / Black Ward 18.4% 0% elite chance +0.284 ± 0.508 22–20–0 42 games · 52%
119 of 128 ↑ 3
Blind 93/94 E078-04 #3373 · born E78 · Random restart
Black Vise / Mesa Pegasus / Ankh of Mishra 18.0% 1% elite chance +0.243 ± 0.569 20–17–0 37 games · 54%
120 of 128 ↓ 92
Blind 93/94 E051-07 #3160 · born E51 · Local substitution
Roc of Kher Ridges / Regrowth / Balance 17.7% 0% elite chance +0.217 ± 0.556 13–9–0 22 games · 59%
121 of 128 ↓ 34
Blind 93/94 E069-06 #3303 · born E69 · Local substitution
Roc of Kher Ridges / Regrowth / Balance 14.4% 0% elite chance -0.105 ± 0.560 10–10–0 20 games · 50%
122 of 128 ↑ 3
Blind 93/94 E079-01 #3378 · born E79 · Local substitution
Aspect of Wolf / Scavenging Ghoul / Word of Command 8.8% 0% elite chance -0.824 ± 0.627 1–7–0 8 games · 12%
123 of 128 New
Blind 93/94 E080-04 #3389 · born E80 · Text package
Aspect of Wolf / Scavenging Ghoul / Word of Command 7.6% 0% elite chance -1.024 ± 0.641 1–2–0 3 games · 33%
124 of 128 ↑ 2
Blind 93/94 E079-08 #3385 · born E79 · Random restart
Ley Druid / Firebreathing / Ironroot Treefolk 7.3% 0% elite chance -1.070 ± 0.588 7–11–0 18 games · 39%
125 of 128 ↓ 2
Blind 93/94 E078-03 #3372 · born E78 · Local substitution
Time Walk / Lifetap / Forcefield 7.2% 0% elite chance -1.093 ± 0.605 9–1–0 10 games · 90%
126 of 128 ↑ 1
Blind 93/94 E079-04 #3381 · born E79 · Local substitution
Giant Spider / Rod of Ruin / Wall of Brambles 7.2% 0% elite chance -1.097 ± 0.586 4–0–0 4 games · 100%
127 of 128 New
Blind 93/94 E080-07 #3392 · born E80 · Local substitution
Giant Spider / Rod of Ruin / Wall of Brambles 7.0% 0% elite chance -1.120 ± 0.602 0–2–0 2 games · 0%
128 of 128 New
Blind 93/94 E080-08 #3393 · born E80 · Random restart
Personal Incarnation / Drain Power / Gloom 2.0% 0% elite chance -2.565 ± 0.758 8–29–0 37 games · 22%

Niches without labels

Discovered deck families

Clusters formed from what the decks contain

Cross-learning

What the model is learning about cards

Shared coefficients across the entire population View all 289 cards

These are associations inside the decks Forge has played—not isolated card win rates and not proof of causation. Nonland effects are separated from mana-base and color signals so strong land coefficients do not hide the spells. Uncertainty and appearance counts show how much evidence sits behind each signal.

Positive nonland signals

  1. Juggernaut2,521 games · 196 decks
    +1.121± 0.085
  2. Obsianus Golem1,716 games · 169 decks
    +0.979± 0.132
  3. Cockatrice750 games · 85 decks
    +0.925± 0.152
  4. Royal Assassin1,611 games · 103 decks
    +0.813± 0.155
  5. Hypnotic Specter882 games · 72 decks
    +0.802± 0.179
  6. Elvish Archers345 games · 60 decks
    +0.773± 0.233
  7. Counterspell501 games · 80 decks
    +0.767± 0.157
  8. Mesa Pegasus1,492 games · 110 decks
    +0.730± 0.124
  9. Fire Elemental839 games · 63 decks
    +0.727± 0.149
  10. Sengir Vampire576 games · 75 decks
    +0.711± 0.226

Negative nonland signals

  1. Basalt Monolith1,104 games · 172 decks
    -1.163± 0.139
  2. Circle of Protection: Green674 games · 66 decks
    -0.949± 0.175
  3. Drain Power162 games · 54 decks
    -0.947± 0.305
  4. Circle of Protection: Red1,022 games · 92 decks
    -0.925± 0.149
  5. Tranquility476 games · 61 decks
    -0.831± 0.184
  6. Demonic Hordes1,016 games · 82 decks
    -0.799± 0.157
  7. False Orders1,136 games · 84 decks
    -0.723± 0.198
  8. Word of Command504 games · 60 decks
    -0.721± 0.171
  9. Power Surge550 games · 68 decks
    -0.680± 0.200
  10. Prodigal Sorcerer799 games · 89 decks
    -0.670± 0.199

Mana-base / color signals

Land coefficients often describe which colors and mana bases are succeeding, rather than the isolated power of a land.

  1. Tropical Island nonbasic land · 3,329 games · 277 decks
    +0.820± 0.146
  2. Tundra nonbasic land · 3,326 games · 279 decks
    +0.636± 0.163
  3. Badlands nonbasic land · 3,814 games · 296 decks
    +0.551± 0.146
  4. Scrubland nonbasic land · 4,343 games · 315 decks
    +0.311± 0.116
  5. Island basic land · 7,137 games · 785 decks
    -0.257± 0.252
  6. Bayou nonbasic land · 3,469 games · 311 decks
    +0.251± 0.148
  7. Underground Sea nonbasic land · 2,773 games · 261 decks
    -0.216± 0.156
  8. Savannah nonbasic land · 3,835 games · 288 decks
    +0.196± 0.142
  9. Swamp basic land · 6,318 games · 704 decks
    -0.189± 0.258
  10. Plateau nonbasic land · 2,830 games · 246 decks
    -0.133± 0.170
  11. Volcanic Island nonbasic land · 3,812 games · 291 decks
    -0.008± 0.151
  12. Taiga nonbasic land · 3,403 games · 252 decks
    +0.006± 0.148

Evolution log

Recent epochs

Newest first
  1. E80
    Blind 93/94 E043-06 leader after 8,000 total games
  2. E79
    Blind 93/94 E027-04 leader after 7,900 total games
  3. E78
    Blind 93/94 E048-04 leader after 7,800 total games
  4. E77
    Blind 93/94 E048-04 leader after 7,700 total games
  5. E76
    Blind 93/94 E027-04 leader after 7,600 total games
  6. E75
    Blind 93/94 E043-06 leader after 7,500 total games
  7. E74
    Blind 93/94 E043-04 leader after 7,400 total games
  8. E73
    Blind 93/94 E043-04 leader after 7,300 total games
Experiment details Population, selection decisions, and runtime versions
Population 1,568 1,024 initial candidates
In rotation 128 target 128 active decks
Bootstrapping 0 initial field calibrated
Living species 12 discovered, not predefined
Births per epoch 8 legal, seeded mutations
Hall of fame 8 historical self-play leaders