Miracle Luffy EB05 Guide

• Bonney is the engine: 5 DON!! for a 7000 that draws 2 and plays two 2000-or-less Characters from your hand. In players' Miracle games she played Hiyori 519 times, Nami 458 and Lucky Roux 262. In our Pro AI's close calls, Hiyori beat Nami as the pick 43 times in 59.
• Hiyori turns into a 5-cost: trash her to play a Land of Wano Character costing 5 or less from your hand, then draw 1. Players played Kouzuki Oden with her 578 times, Kawamatsu 47: Oden is the target (it draws 1, and one of their Characters can't attack or block until the end of their next turn). Hiyori into Kawamatsu is the fallback when there's no Oden: players who did it won 38% of those 42 games.
• Law plays Bonney: 6 DON!! for a 6000 that returns one of your Characters to your hand and plays a 5-cost-or-less from your hand, rested. Bonney's On Play still happens. Players did it 359 times.
• The rest is digging and resting: Garp, Otama and Emporio Ivankov dig on your 1st turn; Nami rests 2 of theirs; Vander Decken IX K.O.s one that's rested; Electrical Luna keeps them down; one Shanks rests them all.
• The Leader still defends: with a DON!! on it and 5 or fewer active DON!!, each DON!! you rest on their turn is +2000 for that battle. Nami is the only Straw Hat Crew Character, so it's your Leader or Nami.

The list

The Perfect 10 Luffy: this exact list is the one our Pro AI plays in the Perfect 10 challenge.

GARP + BONNEY = KEEP. ONE OF THEM, WITH 1-COSTS = KEEP. NEITHER = MULLIGAN.

• Why Garp and Bonney: Garp digs 4 deep for a card that costs 3 or more (Bonney, Hiyori, Law, Oden are all in reach). Bonney is the chain itself.
• Keep: Monkey D. Garp, Emporio Ivankov, Kawamatsu, Jewelry Bonney, Kouzuki Hiyori (Garp and Bonney). Keep: You Can Be My Samurai!!!, Emporio Ivankov, Jewelry Bonney, Kouzuki Oden, Trafalgar Law (Bonney with a 1-cost). Mulligan: Lucky Roux, Kouzuki Oden, Nami, Jewelry Bonney, Electrical Luna (a lone Bonney with late cards). Mulligan: Kouzuki Hiyori, Nami, Kawamatsu, Trafalgar Law, Lucky Roux (no Garp, no Bonney).
• Late cards pull a hand toward a mulligan: Electrical Luna, Nami and Shanks don't do anything early, and You Can Be My Samurai!!! needs two Characters down first; Otama and Ivankov pull it toward a keep.
• These are players' own choices, not the same hands played both ways: read them as what works for players, not proof.

• Our Pro AI chooses: the side it picks in that matchup. The last column is how its own games went each way, the same shuffles played going 1st and going 2nd (the mirror only: outside it, our Pro AI has fewer than 15 games against any one deck, too few to quote).
• Small samples: under 15 games a side, read the players' columns as a lean, not a rule.

• Your Leader swings from your 2nd turn on: our Pro AI's Leader attacked on 114 of 116 of its 2nd turns.
• Going 2nd, cast You Can Be My Samurai!!! on your 2nd turn: rest the 1-costs that already dug (our Pro AI rested Otama and Garp most) and draw 2. Our Pro AI cast it on 37% (22 of 59) of those turns, players on 25%.
• On 7 DON!!, Law beats Bonney: Law (6) puts Bonney out for free, so it's the same chain plus a 6000. In our Pro AI's close calls, Law was the better play 27 times in 42. Return a spent Garp or Otama to Law: it digs again when you replay it. Bonney comes in rested, but her -1000 still works.
• Vander Decken IX's best target is their rested Shanks: trash Kawamatsu (a Fish-Man) and K.O. it for 2 DON!!. That was our Pro AI's most common Vander Decken K.O. (31 of its 69 Vander Decken K.O.s).
• Shanks on 10 DON!!: our Pro AI's came down on its 5th turn going 2nd (23 of the 33 games it came down), later going 1st.

• Rest exactly enough: (their power - yours) ÷ 2000, rounded down, + 1. The attacker wins ties.
• Players used the Leader's defense on 28% of the attacks on their Leader: 23% at 4 Life, 30% at 2, 49% at 0. They used a counter card on 10% at 4 Life and on 33% at 3.
• Our Pro AI at 4 Life almost never used a counter card (3% of 142 attacks), and rested DON!! for its Leader on 18%. At 3 and 2 Life it countered 35-41% of attacks; at 0 Life, 71% (105).
• It rested DON!! for its Leader on 25% of the 861 attacks on it, a bit fewer than players' 28%: after a Bonney or Law turn there's usually nothing active to rest.
• Your counters: Otama and Kawamatsu are your 2000s; Garp, Ivankov, Vander Decken IX, Lucky Roux, Hiyori, Nami and Kouzuki Oden are 1000s. Bonney, Law and Shanks have none.

• Players' numbers are this week's EB05 ladder games with Miracle lists (16+ Land of Wano cards), duplicate uploads once, farmed games removed (the same two accounts playing each other 8+ times, one side winning 85%+). Our Pro AI's are its own games with this exact list against each deck's most played list, the same shuffles played going first and going second.
• The chain decides the hard ones. In players' games, when Bonney's chain came by their 4th turn they won 57% against Xebec (27% without it), 50% against Kaido (10% without) and 86% against Buggy (43% without).
• Our Pro AI's column is the mirror only: outside it, our Pro AI has fewer than 15 games against any one deck, too few to quote.
• Seat shares count every Miracle list (8 or more Land of Wano cards); the win rates use the 16+ lists.

1. Going 1st in the mirror. Players who chose went 1st 61% of the time (33 choices), but players' mirrors went 62% going 2nd (32 games) and 35% going 1st (34). Our Pro AI goes 2nd there. Its own mirrors split about equally (54% going 1st, 52% going 2nd, 51 games), so this one rests on players' games.
2. Spending counter cards at 4 Life. Players used one on 10% of the attacks on their Leader at 4 Life (657 attacks). Our Pro AI: 3% (142 attacks). Take the hit at 4 and keep the card. At 0 Life it's the other way: our Pro AI countered 71% of attacks (105), players 47%.
3. Trashing a Bonney when your board is full. When a 6th Character comes in, our Pro AI replaces a spent Garp, Otama or Emporio Ivankov, never a Bonney: her -1000 works every turn, even while she's rested. Replacing Garp instead of a Bonney was the better play 54 times in 57; Ivankov instead, 38 in 39; Otama, 19 in 19.

• The key pick: Bonney's first play. From that spot, Hiyori as the pick won 74% of the time, Nami 50%, Emporio Ivankov 36%.
• Their answer was the same chain a turn later: in the mirror, both sides have it, so the counter war and Shanks decide it.

• The decks that beat Miracle most this week (players' Miracle games): yellow Enel (Miracle won 27%), Kaido (32%) and Xebec (40%).
• Its Oden locks your best Character for a turn: in our Pro AI's games it locked a Jewelry Bonney most often (78 of 253 locks), then a Kouzuki Oden.
• Its Nami rests your Blockers most often (Nami and Lucky Roux in the mirror), then your Bonney; Vander Decken IX knocks out what's rested, your Shanks above all.
• After your own Shanks swings, it's rested: against Miracle, that's a 10-cost Vander Decken IX can knock out for 2 DON!!.
• Its chain on time is the danger: our Pro AI won 79% of the games where Bonney's chain came by its 3rd turn, 50% when it never came.

Play Miracle Luffy against our Pro AI, free in your browser at optcgwebsim.com. Perfect 10 is open until November 15: go 8-2 or better against our Pro AI and you get paid, up to $1,000 for a perfect 10-0.

Players: this week's EB05 ladder games with Miracle lists (16+ Land of Wano cards), 403 with a winner, duplicate uploads once, farmed games removed. Our Pro AI: 116 of its own games with this exact list (51 of them mirrors), the same shuffles played going 1st and going 2nd; its keep and turn-order picks are the ones it plays with on the site.

More guides: Xebec, Buggy, the Straw Hat Luffy, yellow Enel, Kaido, Sabo, purple Enel, Law, Luffy ST10, Nico Robin, Luffy & Ace, Sugar, Rosinante, Shirahoshi, Moria, Hancock.