The Next Six Cards

I’ve been thinking about mulligans a lot recently.

I’d assert that the Paris mulligan is the most revolutionary thing to happen in Magic since the game’s invention if I had any real proof to back that up, but even without that kind of certainty it’s still pretty darn important.

Magic started without any formal rules regarding opening hand composition, so local and house rules were the norm. When the DCI started they used the most common mulligan rules, which said that if you had Magic’s equivalent of euchre’s “farmer’s hand”—that is, you had all lands or none—you could reveal your hand to your opponent, shuffle in, and draw seven. Upon your reveal, your opponent could, regardless of his hand’s contents, could “ride the mulligan” and do the same. And once you mulliganed, that’s it, you were done. If you still had all lands or none, you had to deal with it.

(Mana screw is part of the game, of course, and we’ve all experienced it at some level and been royally cheesed off by it. When you’re playing a game based on probability, though, you either learn to build better mana bases or learn to expect it. I recommend the former.)

Anyway, a new system was developed by tournament organizer Matt Hyra, who used the system we know today as the Paris mulligan in local unsanctioned tournaments. The new rules added a new strategic element to the mulligan, but raised the question: “Is it fair in constructed?” Constructed decks could use combos (such as they were back in 1997), and some cards—specifically Balance—get much better with fewer cards in hand.

In 1997, just before Pro Tour Paris actually, the first official Paris mulligans happened at PT Boston and PT Los Angeles as a test in the limited format. Unfortunately, those tournaments were smaller and not a great test for the new rules. PT Paris came around and ProsBloom was the deck du jour, so the DCI could finally see how the new mulligan worked in an environment with actual combos to mulligan towards. (Actually, PT Paris was supposed to have used the original “farmers hand” rules, but the event staff forgot to change the players information packets from PTs LA and Boston. A fortunate mistake indeed.)

In the end, things worked out just fine, and the Paris mulligan rules were adopted permanently by the DCI and were eventually added to the formal rules of the game.

Anyway, I built ChannelBall in the Microprose Magic game. It’s a lot of fun to play since every time I combo out, I get to shout “Boom, bitch!” at my computer:

4x Channel
4x Fireball
4x Disintegrate
4x Black Lotus
4x Lightning Bolt
4x Regrowth
4x Wheel of Fortune
4x Howling Mine
3x Mana Flare
2x Bazaar of Baghdad
4x Mox Ruby
4x Mox Emerald
4x Taiga
5x Forest
6x Mountain

I’m sure you’re all aware how the deck works—Channel your life into a huge red burn spell as soon as possible. Howling Mine is a really good draw engine in an environment where your deck and your playskills are better than your opponents’. Mana Flare is my backup win condition for when I’ve lost too much life to Channel or just can’t find one. Regrowth works as free mana with Lotus and keeps me alive with Lightning Bolt.

And Wheel of Fortune is an in-game mulligan.

To combo, I need to find four mana, at least two of which is green, and at least one of which is red. The deck’s best opening hand is Channel, Black Lotus, an X-spell, and a red mana source as it means an undisruptable (no Force of Will) first turn win. So it’s entirely possible to play the “lay-down loner” and win without seeing the opponent’s first land. If you go second, there are few cards that can stop you: Lightning Bolt, Force Spike, and Healing Salve come to mind, but two of those mean we draw and the third just means you need to draw a Lightning Bolt.

According to the Deck-u-Lator, an old probability calculator from the turn of the century, my ChannelBall deck has a 6.6% chance of drawing the Channel, Lotus, fireball, red mana Hand of First-turn Destruction on the play and almost 10% on the draw. Of course, other hands (like Mox, Mox, Mox, land, Fireball, Channel, random or Lotus, Regrowth, Forest, Channel, Fireball, random, random) win first turn too.

Usually the deck wins easily by turn three.

The point I’m slowly approaching, though, is that the Microprose game still uses mulligan rules from 1995. If ChannelBall could Paris mulligan, it would be so much better. Opening hands with one land would be a thing of the past, likewise hands with five lands and a Regrowth.

Likewise hands like: Fireball, Fireball, Disintegrate, Forest, Lightning Bolt, Forest, Mox Emerald.

Or: Channel, Channel, Regrowth, Mana Flare, Forest, Mountain, Regrowth.

Or, I swear: Regrowth, Regrowth, Regrowth, Channel, Regrowth, Taiga, Black Lotus.

The deck still has a 3.3% chance to win first turn, even with only six cards. So I guess my point is that even though this deck is horrendously broken (and probably not even as broken as it could be), it would be even more reliable if it could Paris mulligan.

A similar issue exists in Vintage, playing Belcher.

4x Goblin Charbelcher
4x Empty the Warrens

4x Simian Spirit Guide
4x Elvish Spirit Guide
4x Rite of Flame
4x Tinder Wall
4x Land Grant
3x Seething Song
3x Chromatic Star

4x Goblin Welder
4x Pyroblast
2x Living Wish

1x Wheel of Fortune
1x Memory Jar
1x Lotus Petal
1x Lion's Eye Diamond
1x Mana Vault
1x Sol Ring
1x Mana Crypt
1x Channel
1x Chrome Mox
1x Black Lotus
5x Moxen
1x Taiga

Sideboard
3x Xantid Swarm
1x Bazaar of Baghdad
1x Mishra's Workshop
1x Ancient Tomb
1x Tolarian Academy
1x Tin Street Hooligan
1x Uktabi Orangutan
2x Ancient Grudge
2x Shattering Spree
1x Gorilla Shaman
1x Taiga

The important cards to consider are Empty the Warrens, Goblin Charbelcher, Wheel of Fortune, and Tinder Wall. If any of those shows up in my opening hand, I’m more than likely able to call it successful. Aside from four Pyroblasts, four Goblin Welders, and two Living Wishes (which sort of count anyway), the rest of the deck is mana so I’m likely to have the ability to cast one of those.

When I put the numbers into the Deck-u-Lator, I set it up like this:

Wins 10:1
Red Mana 12:1
Accel 16:2
Other 22:0

I figured that having red mana (Black Lotus, Lotus Petal, Mox Ruby, Land Grant, Taiga, Simian Spirit Guide) in the opening hand is important to get things started, as are two pieces of acceleration (off Moxen and colorless artifact mana, Elvish Spirit Guide, Tinder Wall, and Rite of Flame but not Seething Song).

This configuration gives me a rough estimate of 30% success on the play and 42% on the draw—not too bad. Mulliganing gives me another 19% success rate at six cards and 10% at five.

Of course, these numbers aren’t quite right, as most of the red mana also counts as mana acceleration, having Belcher doesn’t necessarily require red mana, Seething Song and Channel are both in the deck, etc. Plus, knowing that most of the deck (36 cards) is mana producers or mana multipliers, lets you find a win condition on the assumption you’ll find the means to use it.

Honestly, it’s more solid than it looks, and to me it already looks pretty solid. I’ve played well more than a hundred goldfishes and I have yet to feel bad about mulling once or twice if a hand is weak. I have never yet mulled past four, nor had to. Maybe it’s me being lucky, but I’m confident that the results would hold true over the longer course of Belcher’s career.

Here are the three “rules” I keep in my head for mullagining with Belcher:

1. Be aggressive: Your best bet is to put something together before your opponent has time to put a defense together, so point yourself at that goal.

2. Find a winner: Have a possibly game-winning card in your opening hand. If you can play and fire it without too much trouble, Belcher is the best one, especially if you can do both on one turn. Empty the Warrens is next as it only costs four mana; it’s not hard to have 8 or 10 goblins spill out, but that’s still at least a two turn clock. Wheel and Jar are about equal, and each has is positives. Jar gets bonuses for Welder being in hand, but it costs two more mana—meh, whatever.

3. Don’t be stupid or unreasonable, but protection is good: Belcher’s strategy is fragile, no doubt. Discard, counters, Null Rod, Pithing Needle, whatever, you won’t be able to plan for everything, but don’t play into it. For example, if your hand is all artifacts, that’s awesome because it’s mostly permanent sources of mana, but it also means you die to Null Rod. If you’re playing against Fish, you might consider shipping it. At the same time, if that hand can play through a Force and win, go for it!

I kept a hand in that January tournament that included two Xantid Swarms, a Red Elemental Blast, and no threat. Bad idea. I was definitely afraid of counters, when I should have been trying to throw threats out there to counter. Don’t be that guy. Don’t be me; be bold.

Actually, now that I look at it, those are pretty good general rules for mulliganning. Not only do you get to set up your first few turns and ensure you get some mana, you can adjust your entire strategy. Find some threats and protection or disruption and go for the glory. Mulligans are your friend. Mulligans are your friend! Mulligans only sometimes a free Duress for your opponent.

I’ll go to six.