Deep Analysis — Public Service Announcement

Read Richard Feldman every Thursday... at StarCityGames.com!
Richard is on a mission. For the past few weeks, Mike Flores has been lending his support to the “four and one” school of deckbuilding: this theory explains why having any number of a particular card in your deck other than “four” and “one” is inherently flawed. Richard disagrees. In today’s Deep Analysis, he attempts to set the record straight with what we in England call “fightin’ talk.” Let battle commence!

Okay, enough is enough. At first I thought this whole “four and one” thing was harmless, but now that it seems to be doing some actual damage to the Magic community, I have to speak up.

Awhile ago, skilled penman Mike Flores wrote that “Magic is a game where the following things are the most important: 1) cost, 2) power, 3) speed, and 4) consistency.” I agree with this. Next, he wrote that “Cost is the single most important factor… the best cards in the game tend to find an intersection at [cost and power], and the best decks combine [cost, power, and speed], if not [those three things and consistency].” I was still with him there, and continued to agree when he made the very good point that, “Strategically the goal in Magic, and in any strategy game, is to make every game the same.”

Mike concluded from this that the way to get every game of Magic to be as close to the same as possible is to “stuff your deck with fours.” That’s pretty much true, but I would argue that a key modification of his goal is important before we start chasing after it. I would say that, “Strategically the goal in Magic, and in any strategy game, is to make every game the same — with you winning it.

Without that last bit, I would be applauded for presenting the most consistent deck in Constructed Magic:

60 Forest

You thought fours were good? Wait ‘til you see sixties.

Every game with this deck will play out exactly the same way, unless you change the way you play it. That would be awesome, except that you’ll be losing all of them. Obviously the consistent thing we want to be doing is winning, as there is no inherent value in a deck that does the same thing every game.

The problem with the “fours and ones are the best” dogma is that it ignores the possibility that a deck can be full of twos and threes — and therefore, I agree, less consistent than a four-and-one setup would have been — can still be better than the four-and-one alternatives in the format.

Pretend, for a second, that the following deck is Constructed-legal.

30 Black Lotus
15 Channel
15 Fireball

This is a more reasonable extrapolation of Mike’s logic. Every game you want to draw two copies of Lotus, one of Channel, and one of Fireball. Thus, this list is one part Channel, one part Fireball, and two parts Lotus. As far as making every game go the same goes, this list is awesome.

Joey Bloggs brings this deck to a tournament and gets paired against me. He casts Lotus, Lotus, Channel, Fireball for lots. I die on the first turn. The second game goes the same way, and Joey goes on to win the tournament without dropping a game. Man! What an insane deck.

The next week, there is another tournament. Joey brings the same deck, and is once again paired against me. On his first turn, his deck does what it does once again — consistent beast that it is — and he hits me with Lotus, Lotus, Channel, Fireball. This time, however, I cast Misdirection on the Fireball, send it back to Joey, and win the game. Game two, he does the same thing, and I do the same thing. Thus ends the match in my favor.

Convinced that it was a fluke, Joey presses on to the next round. Unfortunately, someone else got the memo on his deck, and Joey is once again crisped by his own Fireball in both games.

The next week, Joey comes back with his Channel/Fireball deck and goes home with nightmares of Misdirection floating through his brain.

Frustrated, he decides to alter his deck. Knowing he will face Misdirection, he moves some cards around to make room for Defense Grid. His new list looks like this:

20 Black Lotus
14 Channel
14 Fireball
12 Defense Grid

Joey’s deck is now less consistent.

But it is better.

Joey has to take a couple more mulligans the next time he plays the deck, but he goes on to win the tournament anyway through several Misdirection-wielding opponents.

In order to combat Misdirection, Joey has had to sacrifice some of the consistency he had before. Whenever he sees an opening hand of Lotus, Lotus, Lotus, Channel, Channel, Defense Grid, Defense Grid, he has to throw it back. In his original setup, there’s a good chance one of those Grids would have been the Fireball he desired; such is the price he has paid for the ability to beat Misdirection.

Back when I last played Psychatog, I had a post-board gameplan against Ichorid that involved having two copies of Meloku in my deck. Had I played four copies, I would have drawn him quite often in my opening hand — a harmful blank against Ichorid’s blistering early game — so I played two instead, and used cantrips and Fact or Fictions to increase the odds that I would see one as the game progressed to its later stages. Playing zero copies would have meant that I would have had to win all my long games on Psychatog alone, a strategy Ichorid can blunt with Stinkweed Imp or ignore by racing with flying 3/1s.

Meloku was the perfect finisher once I hit the five mana mark, but an awful, awful card to see early. I won many games on the back of drawing into Meloku for the win, and also took a few extra mulligans because he was in my opening hand instead of a defensive spell. True enough, the presence of the two-of meant my deck was less consistent…

… but it was better.

When you can successfully play a deck like Joey’s original 30/15/15 configuration of ChannelBall, that’s awesome. You lucked out. When Joey played his deck at the first tournament, he was fortunate to have stumbled upon a configuration that was both maximally consistent and correctly balanced. However, in the second tournament, the maximally consistent version of his deck was no longer optimal. The best deck for the Misdirection-heavy format actually turned out to be ChannelBall with Defense Grid, even though that was a strictly less consistent build than the original ChannelBall.

Bear in mind that, all along, Mike has been advocating decks that are predominantly fours and ones. He allows for certain “exceptions,” but expects that all good decks will be almost exclusively fours and ones.

For example, Adrian Sullivan recently stated that his Pro Tour deck configuration from back in the day was correct to play 2 Lay Waste and 3 Avalanche Riders. He said that his testing showed he wanted at least ten creatures, five Land Destruction spells, and the maximum number of Lay Wastes because they seemed more powerful than Avalanche Riders.

Mike called this an error in the forums:

“The reason that non-4/non-1 is ‘bad’ is not because you are a bad person or anything, it’s because you don’t know what card is best.”

According to Mike’s logic, three and two was obviously terrible because Adrian himself had determined that Lay Waste was more powerful than Avalanche Riders; the more powerful card obviously should have been a four-of. The Riders should have occupied the fifth Land Destruction slot as “Lay Waste Number Five” (you know, like how Jade Leech is Blastoderm number five), because, you know, that’s just how you do it. “Magic is simple, QED,” or whatever that line is that he gets away with.

A reader with a keen memory may recall that Mike was not specifically addressing Adrian’s list when he made that comment. Perhaps he meant it in the general case, and did not mean to imply that people who had actually figured out their ratios did not know what card was best.

That is exactly what’s wrong with it.

If the rule only diagnoses you when you’re already making a different mistake — that is, improperly valuing your cards — then it is worthless at best and misleading at worst. If you see someone’s decklist, and they should clearly be playing four copies of a powerful card that will not hurt them when drawn early and/or in multiples, it’s ridiculous to tell them that “The gods of 4/1 say you should play four of this card, my son.” It makes much more sense to instead tell them, “This card is powerful, it does not hurt you when drawn early and/or in multiples, so you should cut this less-powerful card and play four.” I mean, since that’s actually their problem and all.

Now, there is something to be said for searching for a deck that naturally works out to be all fours in the first place. If you design a deck that only plays cards that are good when you draw them early, and also when you draw them in multiples, guess what? You have a consistent deck! You will indeed suffer from fewer awkward draws, and that factor will work to your advantage throughout the tournament. However, what’s more important than that is making sure your deck is playing the cards that are best for it, and in the correct ratios.

For example, if you are working on a control deck where you are certain the best finishers are Akroma, Angel of Wrath and Teferi, Mage of Zhalfir, there’s a pretty low chance that playing four of those two Legends is going to be correct. Rather than cramming all four copies of both in there and suffering from clunky early draws, or cutting down to one copy of each that you won’t successfully dig to in a lot of games, the correct solution — if you insist on settling for nothing less than the consistency of a four-and-one list — is to just play a different deck.

If two Legends are your best finishers, and they don’t fit as four-ofs or as one-ofs, then your deck is not set up to be a four-and-one deck. Sorry. Abandon this deck idea and continue searching for one that fits the mold. Do not try and contort the existing strategy into a four-and-one configuration, because you will end up doing more damage than good. If four-and-one consistency is truly what you desire, you are better off starting over from scratch than mangling your deck’s ratios in order to fit with the “in” crowd.

In other words, it’s great when you find the optimal deck for a format and its ratios work out to be all fours. That means that not only is the deck strong, it is also consistent. However, when the list that yields the best results in testing cannot work out as a sequence of mostly fours and ones, don’t panic.

If you panic, you might find yourself in a position like Mike did last week. Mike’s B/W deck has these conspicuously high counts all over the place — four Isamaru, when none of the established Zoo decks play more than three, four Umezawa’s Jitte as if the Legend Rule didn’t exist, four Mortify to go along with the four Smothers — as if everyone’s got so many targets that you should unquestionably max out on both – and four Ghost Councils like the first one’s going to die.

Everyone I know of who saw that B/W deck immediately thought “Legend Rule much?” yet Mike didn’t address the elephant in the room anywhere in the article. I’ll be honest, I read that list full of awkward fours and I saw a deck designer who was torn between the desire to balance his deck properly and the fear of becoming a laughing stock if he posted a list with three Umezawa’s Jitte and three Isamaru after carrying on for two weeks about how “Almost always when you find a deck that is not predominantly fours and ones, it is in some ways terrible.”

This is why I had to speak up.

If you think that you can find an optimal deck for every tournament that will work out to be predominantly ones and fours, feel free to chase after it. But if you try to contort your existing deck to be something it’s not, you will increase not only your deck’s consistency, but also its suckage.

The lesson to be learned from the “fours and ones” theory is not that you should aspire to play decks that are all fours and ones… but that you should consider not only power and speed, but also consistency in the cards you play. Consider that if a card fits into your deck as a four-of, then it is probably a card you are fine with drawing at any point in the game. If a card is simply not correct as a four-of, that implies that there are some fairly common circumstances under which you are unhappy to see it. That doesn’t mean you should cut it straightaway, or bump it up to a four-of in defiance of what your testing has told you; no, it should merely send you looking for more consistent alternatives. If you can’t find any that compare with its power level, then keep it as a two or three.

Take Umezawa’s Jitte for example. If I’m playing three Umezawa’s Jitte because it stinks against TEPS, is too slow against Tron, and drawing multiples in game 1 hurts me against most other decks, I should start thinking if I could find a card that could fill the same role without being clunky for me in so many circumstances. Let’s say I examine Sword of Fire and Ice, Sword of Light and Shadow, and Armadillo Cloak. I find all of them lacking. Does this mean that I should cut Jitte? Max out on Jitte? No — both of those would reduce the power level of my deck. It’s better to have three Jittes and enjoy the power level that those three afford than to play four and introduce clunky draws (the opposite of consistency) or to cut them entirely and smash the power level of my deck in the kneecaps for no reason other than trying to fit a dogma.

Ideally, Jitte would be more like Remand. Remand is great early, and fine late; if it works well in your deck, and if you’ve got room for it, you can play it as a four-of and expect that it will not lead to clunky draws. Rarely will you hear someone complain that they drew too many Remands early. If you fill your deck with cards that fit this description, you will minimize your clunky draws. However, if the best deck, in your estimation — even factoring in consistency — is all twos and threes, all Jitte-like cards, let’s say — not only can it still be correct, it can still be the best deck.

Don’t let Mike tell you otherwise.

Next week, more on Extended. For real this time!

See you then,

Richard Feldman
Team :S
[email protected]