As a software architect, one tool that has been very useful over the years is the identification and use of patterns. The concept of design patterns originated in traditional building architecture where similar problems were frequently encountered. Over time, standard solutions to these problems were documented and reused by other architects, thus allowing a buildup of professional knowledge and a common language for discussing the various solutions. An example of such a pattern in traditional building architecture is the arch. Patterns also creep up in software design, and they can apply to cube design as well.
A pattern is a recurring solution to a common problem. For example, in software engineering, the Singleton Pattern is a solution to the requirement that there be only a single copy of a particular object at run-time. Cubes also traditionally follow a singleton pattern. But why?
In cube design, the singleton pattern refers to something slightly different: the cube contains only a single copy of each card. This is not strictly required; if you've ever played in a booster draft then you've probably seen more than one copy of a single card, especially among the commons. In fact, contemporary booster packs sometimes contain premium foils that can mean a single pack could contain two copies of the same card (one foil, one non-foil). On the other hand, booster packs otherwise do have a kind of contract with players: with the exception of premium foil cards, booster packs will not contain any duplicate cards. This helps to smooth out the drafting and avoids clumping of cards. It is made possible by the fact that booster packs are assembled in a controlled printing and packaging process that ensures the contract is met.
When you're building booster packs yourself everything is done manually. It is already tedious enough shuffling up several hundred cards; I don't also want to sort cards to ensure that multiple copies of the same card don't end up in the same pack. If there is only one copy of each card, then it is not necessary to do any additional sorting, as there will never be duplicates.
A happy by-product is that the singleton pattern provides a great deal of variety during a draft. The result is much more akin to the feel of a draft environment that includes all three sets of a block, rather than the early triple-large set drafts.
When multiple cards that provide a similar effect do appear in the same pack, this can actually create some interesting decision points for drafters. While you may always choose Lightning Bolt over pretty much any other burn spell, you might find the choice between Firebolt and Searing Spear more interesting.
Including only one copy of each card also creates some interesting decisions for the design of the cube. If you were selecting only a single burn spell in Red, for example, you might choose something like Lightning Bolt for your set. When a new card comes along like Searing Spear, you can instantly see that it's not quite as good as Lightning Bolt so you might decide to just leave it be. However, in a singleton cube, you often want to accumulate a critcal mass of a similar type of effect. This redundancy basically acts as a bit of a counter-measure to the singleton pattern and may result in you including both Lightning Bolt and Searing Spear in your cube.
I follow the singleton pattern strictly for my traditional "best of" cube. For the most part, the themed cube that I am currently designing will also follow this pattern - with one exception. For a number of reasons, I have some issues in my White section which can be helped greatly if I include multiple copies of Squadron Hawk. Squadron Hawk, of course, is a very average card if there is only one copy available, and as a result it is almost universally shunned in cubes. However, if I were to include, say, four copies of Squadron Hawk, then the card becomes significantly stronger. In fact, it can create some interesting draft decisions in and of itself. It is very much a gamble to take the first copy of Squadron Hawk since it is not very good on its own. Collecting additional copies, though, can be very advantageous and even very powerful depending upon the draft environment that you're creating. In M11 draft, it was a very strong card indeed.
But what about that issue of having multiple copies in the same pack? Actually, I think that in this case it could once again make for some interesting draft decisions. If it's early in a pack and you're looking at two copies of Squadron Hawk, do you take one? If you do, what are the chances that someone else will take the other one, thus reducing the value of the first? What if you pass both? You can then no longer get both, meaning that you're passing on them entirely. Since these decisions will only come up periodically and since I'm planning to restrict it to the one card, I think that this sort of thing is fine. I just wouldn't want it to happen with a lot of cards; it makes a better exception than a rule.
So, the singleton pattern solves the problem of duplicates and simultaneously adds to the overall experience. As a result, most cube designers choose to follow this pattern by default. As I continue to explore cube design in more detail, I'll be on the lookout for more patterns used by other cube designers to solve common problems and apply them to my own cube designs wherever they seem appropriate.