"I Came to Play (Second)"
Zvi Mowshowitz


Before I begin, I want to update everyone on the status of my interview with Kyle Murray. So that he could make sure his answers were correct, I gave him the opportunity to respond in writing, but unfortunately I have yet to receive his reply. I will post it as soon as it is available, but there are currently delays on the other end. Now on to the bulk of the article, which concerns the decision on whether to play first in limited. I don't choose to do so anymore, and I'm here to give an explanation.

Wizards first introduced the play/draw rule for the first Pro Tour. It was a great innovation, changing going first from a gigantic advantage to a moderate advantage in constructed and an open question in limited. At first, I was choosing to draw first in constructed. What was the chance to play a land compared with an extra card? Later on as decks got more efficient it became clear that it was better to play first in most matchups. Over time, it became important enough that winning when going second is sometimes called "breaking serve." No one uses it in real life because it's more than a little silly, but every time there's commentary on a top eight and airtime to fill the question gets dragged back out. Randy Buehler and his partner in commentary will then take three games and draw large conclusions about the importance of going first in a matchup.

My editor and teammate Scott Johns is always there in my playtesting to bring up the issue of who is going first. Going second is always nerve-wracking to him. How do you deal with the onslaught that's bound to come out with a perfect sligh curve? He should know because we're testing on Apprentice and therefore I will always come out with that perfect sligh curve. (For those not aware, I have a near-magical ability to draw well with Apprentice) At least, that's the way it seems to him. Where things get strange is that he will talk about a matchup being "shaky" when going second even in limited, whereas I will almost never think that way. In fact, in limited right now I'm choosing to draw first more often than not, and this article is here to tell you why. Before I start the explanation, a truism: In any given situation where we reveal all information, if one player wants to draw first then so does the other.

The first principle is that if playing first would result in either player being mana screwed or color screwed enough to seriously mess up their game, both players probably want to draw first. This is true whether or not drawing first instead would result in their problem being solved. If it gets rid of the problem completely, excellent. If it doesn't, the problem will be there for one less turn. Either way it's a one turn improvement. I would estimate that if you add together all mana and color screws and extreme mana floods, which for this purpose are the same thing most of the time, I would estimate 40% of limited games fall into this category. My quick survey from a Magic Online tournament revealed a large quantity of bad draws and frustrated players. If your shuffling is better than Magic Online's, that means you're doing better than a true randomizer - in other words, you'd be a savage cheater.

The reason this has occurred to me is that I've been playing on Magic Online for real prizes. There's nothing like putting a little money on something to make people shut up and pay attention, and this is one of the things it has drawn my attention to. For a while I just couldn't seem to catch a normal match. Things have gotten better, but I've been paying attention to how many games don't involve one or both players suffering enough that drawing first is superior on this basis rather than the comparison between card advantage and tempo that is normally the trade off. If anything, the estimate has been low. Very often I thank my lucky stars I'm drawing first, and very rarely do I think the opposite.

Now let's move on to the games where both players have playable draws. These are the more interesting questions. How often is the tempo gained worth the card lost? That brings up the question of how often tempo is gained. The problem is that often the extra card drawn by not playing first will result in that player filling a hole in his mana curve. How often have you played 'the amazing second turn Basking Rootwala' or other such gem? This applies both to not having the right color of mana and to not having the right casting cost spell in hand to use your mana. If playing first results in you missing the third or fourth turn drop, or if it causes your opponent to make his where he would have missed, then playing first is about even in terms of tempo and the extra card draw carries the day. We can throw in a good chunk of the remaining matches by adding this category.

In the remaining games, the tempo gain is not an illusion. The question is, is it worth a card? That depends on how the game develops in both situations. Sometimes the fact that creatures just cast remain untapped ends up minimizing the difference between going first and second. Other times, one player is attacking on the ground while the other plays flyers that can't afford to block. In a third situation, one player is desperately trying to get his defenses up. When someone has a bad mana curve to begin with, he will often need to take the first turn in order to make up for lost time, but he risks even greater disaster if mana issues come up.

Many matches will stabilize at some point, with both players essentially living off the top of their decks while perhaps one or two creatures come in for some damage each turn. When the game ends up in one of these situations, the extra card normally ends up being balanced against a few life points. How many life points it ends up as determines whether it is worth it. Most of the time, situations where the table stabalizes favor the player with the extra card.

The two situations where playing first is better are where one player is trying for a large tempo advantage or where the game is a race. Much of the reason most players play first is because the 'normal' game of Magic in most matchups will fall under one of these two categories. One player will have flyers, or one player will play a tempo deck, or often both. When these effects are extreme, playing first is a considerable advantage in 'normal' games. Every now and then drawing first will only draw you one more card you can't cast in time for it to make a difference.

But in general, I've reversed the usual thinking. Normally players will play first unless they have a good reason to go second. At this point, I will go second until I have a damn good reason to go first.

At this point, I will go second until I have a damn good reason to go first...

If you've never experienced it, it is very strange what the die roll is like once you've made this change. It has happened before, especially in sealed deck. In those formats, many of the top players chose to draw first rather than play first. The odd part about it was that they knew that most opponents they didn't recognize would choose to go first if they won the die roll, and they also knew that if they chose to draw first then their opponents would reconsider that decision later in the match. In short, they wanted to lose the die roll. The difference now is that I don't find my opponents reconsidering their decisions. If I had, I probably wouldn't have written this article, because I would be more interested in preserving this advantage. As it is, given that this is on the premium side, I feel safe enough talking about it.

The last question that has to be asked is whether this carries over to constructed. My experience is that I'm often tempted to draw first when I feel I have a large advantage, but I almost always manage to think better of it in time. A much larger percentage of the bad draws in constructed involve draws that are slow rather than missing lands or spells, and these decks are far more consistent. If they stall on land, it is much more likely to stop them from casting two spells a turn or one or two spells at the high end of their curve than it is to keep them out of the game entierly. And in most matchups, if you think about it long enough you can find a key strategic reason to want to take the first turn.

The exceptions remain matchups where neither player is threatening to break through in any meaningful way. This works the same way that it does in limited. In traditional control matchups neither player wants to cast an early spell, so there's no reason to play first. However, modern control matchups seem to be different because there are permanents worth casting. Casting a second turn Nightscape Familiar or a third turn Psychatog is not only worthwhile, it can be both risk free for you and problematic for your opponent to try and react to. The 'threat' at this point in the game is Fact or Fiction, Probe or other card advantage. You can therefore cast your spells without giving your opponent a chance to do anything damaging. In fact, there's not much the player going second can do for a while because drawing cards would just cause him to discard.

In other long games, the first few turns are important for similar reasons. Resolving a second turn Merfolk Looter is well worth the card. These days, there always seems to be a reason. Even in matchups that are all about land, like the monoblack mirror matchup, going first is important because it lets you cast Rancid Earth and Braids first.

Since I wrote my initial explanation, I've played a lot more matches, and I am now even more comfortable with my decision. If you haven't thought about this decision recently, I recommend revisiting it.

- Zvi Mowshowitz
zvimowshowitz@yahoo.com