Most Magic players operate with the knowledge that speed kills. The decks that are generally successful have this undeniable advantage of speed. Tournament-worthy cards are generally evaluated on a speed basis where "faster is better". The true difference between the different formats, Type I, Type II, and even Sealed Deck, to some extent can be described as a difference in speed. In Type I, it is not unusual to have three to four sources of permanent mana, along with a 3/4 flyer (Serendib Efreet) on turn one. In Type II, it is not unusual to have one source of permanent mana, and a 2/2 flyer (Hypnotic Spectre, by way of a Dark Ritual) on turn one. In Sealed Deck, it would be almost inconceivable for any of the above.
When it comes to the constructed deck formats, most good players build decks with a decided emphasis on speed. This is particularly true of the intermediate players who have gone through the beginner's stage of thinking that "bigger is always better" and stuff their decks with every large creature they could lay their hands on. Once they realize that a Scryb Sprite is often better than a Shivan Dragon in tournaments, they begin to build fast decks. As a result, a large number of intermediate-level decks focus much of its resources on generating speed, whether they do so with fast mana (green, black, artifacts w/ Mana Vault, etc.), or with cheap creatures and spells (R/G weenie/blaster is a good example).
And yet, in higher levels of play, decks which are "slower" win consistently. One of the strongest deck types to have emerged in pre-1997 Type II format is a control deck based around Winter Orb and Armageddon, the so-called Prison decks. Prior to, and even during, the Black Summer of Necro of 1996, the U/W sceptre-lock deck (often called the Weissman deck) was one of the strongest contenders in any tournament. Even with the dramatic changes to Type II created by the 1997 bans on restricted cards, two of the most effective deck types in existence -- the U/W CounterPost and the U/R CounterBurn -- are not blazing fast speed-demons. Rather, they are slow, methodical, and consistent game winners.
These facts lead one to ask, "What does 'speed' really mean?"
It is impossible to play Magic at the highest levels, or to build decks capable of winning consistently, without some understanding of 'speed' and its true meaning. Those players who are successful do have a grasp of the meaning of speed, even if they have never really consciously thought about it. I suppose that is natural talent. Most of us, however, are not so gifted and conscious reflection and study of the concept of 'speed' is undeniably beneficial.
Speed can mean several different things in Magic. Conceptually, we could break it down into four basic components: offensive speed,defensive speed, resource speed, and strategic speed. These four concepts are related, which makes it difficult to understand which aspect of deck speed you are dealing with. Nonetheless, once each component concept is understood, it is possible to gain better understanding of deck speed and how deck design relates to speed.
One popular method of measuring deck speed is the "goldfish test", in which you assume that the opponent is completely helpless. Then, starting with your initial draw, you attempt to kill your opponent in as few turns as possible. Some Type I decks average five turns on this test.
Offensive speed is the measure which most people use to evaluate a particular deck's speed. However, it is incomplete as a measure of deck speed, as a number of the fastest decks never win tournaments. For example, a R/G weenie horde/blaster deck with Giant Growths and Bloodlusts is one of the fastest designs in terms of offensive speed. However, few of them ever win major tournaments. Some would say that the deck has no "staying power" and believe that there is a tradeoff between deck speed and staying power. As I will explain below, I belive this dichotomy is unnecessary.
Basically, defense is the ability to disrupt the opponent's game. Certainly, staying alive despite opponent efforts can be described as disrupting the opponent's game. Preserving your assets, your game resources, is also disruption. If your opponent Lightning Bolts your Wall of Wood so as to be able to attack you with his Black Knight, and you Healing Salve the Wall of Wood, you have disrupted his game plan at least for the time being. Casting Armageddon when neither of you has any creatures in play, but you are holding land, could be defensive since it is bound to disrupt your opponent's game plan while your own game plan might be built around just such a move.
A deck with defensive speed,then, will quickly disrupt the opponent's strategy and preserve its assets. Most players think of blue counterspells and white anti-permanent spells when they think of defense, but this is needlessly narrow. The Necrodeck of 1996 was one of the most defensive decks around with incredible defensive speed coming from 4 Hymn to Tourach and 4 Strip Mines. Nonetheless, for starters, it might be best to think of the U/W deck as the archetype of speed defense since it is capable of dealing with most creatures with a Swords to Plowshares and Wrath of God, most permanents with Disenchant, and all spells with Counterspell. The U/W deck can defend starting from the first turn with both Swords to Plowshares and Force of Will. That this defensive speed disrupts the opponent's strategy is not seriously contested, as most players would think twice about doing anything against a good U/W deck, counting the mana on the table, the number of cards in hand, the expression on the opponent's face, etc. before making a move.
"Resource" in tournament Magic comes in three forms: life points, time, and cards. If you run out of life resources, you lose. If you run out of time resources, you either lose or draw, but cannot continue to play. And of course, that cards are a resource is hardly contested -- the idea of "card advantage" is well developed by now, thanks to the insight of Brian Weissman and others.
Refining the concept further, however, we note that cards are further broken down into different classes: permanents, mana, and "escrow". Permanents are obvious -- creatures, enchantments, artifacts, and often special lands which put a strategy into effect. Mana is also obvious. Given the structure of Magic, without adequate mana, it is impossible to play. Escrow, however, is a difficult concept to capture. At its most basic, escrow refers to cards in hand -- cards which in and of themselves do very little, but by virtue of being able to do something impact the game dramatically. The same can be said of cards in the library and in some circumstances, of cards in the graveyard. The point is that escrow resources are resources which could be used but have not yet been used for one reason or another. For example, if I have four cards in hand, a creature, a land, a Lightning Bolt, and a Shatter, none of those four impact the game simply by being in my hand. They will impact the game when they are played -- their playability depends upon the mana resources in play among other things, but the value of the four cards as a resource depends also on their being available to me. The same can be said of cards in my library -- if I have used three of the four Disenchants in my deck, then I only have one escrow resource left to deal with the Stasis that my opponent has put in play.
Understood this way, resource speed depends on manipulating these three types of card resources. There are ways to manipulate both life and time resources, but all of those ways (short of cheating or delaying) depend on cards. Now, speed creatures or inexpensive artifacts (such as Telim Tor's Darts) might grant permanent resource speed. Dark Ritual, mana creatures, and artifact mana gives mana resource speed. Jayemdae Tome, Soldevi Excavations, and cantrips give escrow speed either through card drawing or library manipulation.
Now then, we might begin to describe resource speed. Playing a first turn Bird of Paradise does not in and of itself damage the opponent, or set up defenses. It may make both of those possible, but taken in its individual capacity, all that a first turn Bird of Paradise does is provide mana resources quickly. A Jayemdae Tome does nothing in and of itself either in defensive or offensive terms. What it does is provide escrow resource speed by moving a card from one state of escrow (the lower level of being unknown in the library) to another (the higher level of being known in the hand). It is possible to evaluate relative resource speed by comparing a Bird of Paradise to a Fellwar Stone to a Mana Prism, or a Sylvan Library to a Jalum Tome to a Jayemdae Tome. In Type I play, the difference between Ancestral Recall and Braingeyser is that of resource speed, for example.
We might judge a deck to have resource speed if it is able to deploy its card resources quickly -- permanents and mana -- while manipulating its escrow resources.
The answer to the question depends on a number of factors as different decks have different strategies, and the better decks have multiple strategies. One might get a sense of strategic speed by comparing a superior deck with an inferior one or by comparing different strategies.
For example, take two R/G creature/burn decks. One uses Grizzly Bears and Nettletooth Djinns along with Bolts and Incinerates. The other uses Jolrael's Centaurs and Jungle Wurms along with Hammer of Bogardan and Kaervek's Torch. Obviously, the first deck would be able to put its creature strategy in place much faster than the second, and will also be able to put its burn strategy into place faster. The second deck may be more effective over the long haul, since a Centaur has abilities that a Bear doesn't, and the Jungle Wurm is bigger than the Nettletooth. The Hammer is also a great card over the long-haul, and a Torch is capable of doing more damage than an Incinerate. We cannot say offhand which deck is better. All that is clear is that the first deck is strategically faster.
Or, for two different strategies, take a R/G weenie horde/blaster deck and a U/W CounterPost deck. The first can put out a land and a Llanowar Elf on turn one. The second can put out a land and Swords the Elf on turn one. On the second turn, we might see an Incinerate from the R/g deck, but nothing from the U/W deck. It would be erroneous, however, to believe that the U/W deck is necessarily slower as a result. The U/W player could be holding 2 Arcane Denials and a Counterspell with mana to power them. Its defensive and disruptive strategy is fully in place with two mana, just as the offensive strategy of the R/G deck is fully in place with two mana on turn two. The R/G weenie might kill an opponent in five turns, while the U/W deck might kill an opponent in 30 turns. Nonetheless, both decks can be said to have strategic speed.
How do these concepts blend together? The key is strategic speed. A deck is fast if it is able to unfold its plan of action quickly. I feel that in Type I, "quickly" means within the first two turns as going two turns without being able to put your strategy (or at least one of them) into action will usually spell defeat. Of course, one must take into account a number of interactive factors as well, such as what the opponent is doing. However, barring a truly unfortunate draw, or heavy interference, if a deck cannot put its strategy into play within two turns, it should be considered a slow deck. In Type II, "quickly" means within the first four turns for the same reasons, but lacking the speed cards like Moxen or Ancestral Recall, it is normally acceptable to put strategy into action in four turns. These generalizations must be qualified.
Strategic speed depends crucially upon resource speed, which is the reason why most of the dominant decks in Type I require at least on-color Moxen, if not then entire set plus the Black Lotus. In Type II, smaller creatures always have a speed advantage over larger creatures, even though bigger is better in the long run.
Strategic speed always consists of defensive speed (in terms of disruption) and offensive speed as a secondary matter. This is a crucial point -- that offensive speed is not strategic speed. In both Type I and Type II, a deck must be able to conceivably put up a defense on turn one, even if it is something as simple as bolting a Hypnotic Spectre that the opponent put out with a Dark Ritual. Even in decks which specialize on offensive speed, that offense must be so fast as to disrupt the opponent's game plan, or there is little hope of success.
Let us examine this principle from the perspective of two extremely effective and extremely fast decks of totally different character: a U/W CounterPost deck and a R/B Greed deck. Both are products of the post 1997 environment and both demonstrate the principle of strategic speed.
The U/W ConterPost, so named because of its many counterspells and its use of the Kjeldoran Outpost as the primary, if not the only, source of offense, utilizes an almost entirely defensive/disruptive strategy. It seeks to prevent the opponent from playing his game until card advantage can be established through the use of the Outpost, of Wrath of God, Nev's Disk, Disrupting Sceptres, or Jayemdae Tomes. It uses the full panoply of white anti-permanent spells -- Swords to Plowshares, which can deal with just about any creature in Type II, with one white mana; Wrath of God, which can deal with almost all the creatures that StP cannot; Disenchant, which eliminates 95% of the non-creature permanents which pose a threat. In addition, it uses 8-10 counterspells to prevent the opponent from playing out his strategy. Although it can bring out offense on turn three with an Outpost, the CounterPost does not concentrate on offense so early. Instead, it will bid its time and deploy its full defensive scheme (multiple countermagic, removal of anything on the field, card drawing, etc.) until the sheer force of multiple Kjeldoran Outposts can overrun the opponent's defenses. In other words, it is capable of playing its game, of deploying its strategy, from turn one, and controlling the game by turn five or so. It requires resource speed in terms of mana and escrow, as cards in hand often directly affect defensive strategy: if no StP in hand, the CounterPost will counter a Hypnotic Spectre, but otherwise, let it through.
The R/B Greed deck, similar in format to the R/B Necrodeck which was popular prior to the 1997 bans, plays with a decided focus on offense. Its early defense is its offense and the speed with which it sets up its offensive resources, forcing the opponent to become reactive thereby disrupting his game. It uses 8 creatures with casting cost 2 or lower, 4 with 3 casting cost, all of which can be brought out on turn one with a Dark Ritual. In the early-mid game, the deck might use other means of disruption -- discard with Stupor, or land destruction with Choking Sands (which also does damage). By the late game, the deck focuses entirely on finishing off the opponent with direct damage, fueled by Greed or other card-drawing resources. Its defense is pretty much limited to direct damage to kill off dangerous creatures -- such as opponent Hypnotic Spectres -- or to allow its own creature through. The mindset of the Greed player is to reduce the opponent's life as quickly as possible, relying on Drain Life to keep himself alive in the meantime. Again, it's possible to see that the R/B Greed deck also plays its game from turn one and could control the game by turn three or four with Stupors and Choking Sands or with overwhelming damage. One of the most important reasons why this deck can win consistently is the Dark Ritual, which provides a mana resource boost unparalleled in current Type II play. The possibility of putting out a 2/2 creature with Protection from White on turn one pushes the deck from being merely fast to being fast enough to disrupt the opponent's game.
What is important to note in comparing these two very different, yet very successful, decks is that the main focus of strategic speed is on defense/disruption. Even in the R/B Greed deck, given a choice between Stupor and a Choking Sands on a non-basic land, the Choking Sands is preferable in most situations because it does damage. The opponent has to start getting worried about taking damage quickly. Otherwise, winning the game becomes too difficult if the opponent can take the damage, but play his own game to the fullest.
There are other insights about deck speed in the Magic world, which can be addressed by examing a specific defensive/offensive/resource speed. For example, in defense, is mana denial faster for disruption than permission? What about card discard? These are questions that each player must answer for himself.