The core tension
Every deck list is a compromise between two goals. Power is how strong your cards are when they work. Consistency is how often your deck actually delivers a working game. A list full of powerful cards that demand specific conditions may produce spectacular wins and frequent stumbles. A list of simpler cards may rarely win spectacularly but seldom misfire.
Probability turns this from a matter of taste into something you can reason about. The question is not “which cards are best?” but “which set of cards produces the best results across many games, given how often each one is available when I need it?”
Power is an average, consistency is a spread
If you play many games, your deck has an average win rate. That average is made of a mix of games: ones where everything lined up, ones where it did not, and ones in the middle. Two decks can share the same average and have very different shapes.
- Deck A wins strongly when its key combination appears and loses badly otherwise.
- Deck B wins a little more than half of its games in a steady way.
Over a long session of play, you may find the two decks perform similarly. In a short event, Deck A produces more extreme results either way, which is the same difference between low variance and high variance designs that appears throughout game design. Which you prefer depends on whether you need to avoid bad outcomes or to reach a high ceiling.
Putting numbers on a risky card
Suppose your deck depends on drawing a particular card, and you run three copies in a 60-card deck. The chance of at least one in an opening hand of seven is about 31.5 percent. With four copies it is about 40 percent. With two copies it is about 22 percent.
Now consider what that means for a deck whose whole plan needs that card. In roughly six out of ten games with four copies, you start without it. You may find it later, since each extra draw helps, but meanwhile you are playing a different game. A deck designed around a single dependency commonly adds redundancy: other cards that fill a similar role, or draw effects that dig for the key card.
Consistency, in this sense, is often bought by replacing one very strong card with several slightly weaker ones that do roughly the same job. You give up a little peak strength, and you get more games where the plan comes together.
The cost of a consistent card
Consistency has a price. Spots in the list used for card draw, searching or duplicate effects are spots not used for the strongest threats. Some questions to ask when comparing the two options:
- How often does the powerful card do nothing because its conditions are unmet?
- What does the card do when drawn late, or drawn alongside other situational cards?
- How much worse is the reliable alternative when it is drawn at the right time?
- How many slots does the consistency package consume?
A simple estimate: if a strong card is playable in about half the games in which you draw it, its effective contribution is roughly half its stated strength, and the unplayable copies are dead cards in hand. This is a rough guide, not a rule, and the exact share depends heavily on the deck.
Redundancy and the diminishing returns problem
Adding copies helps less each time. Going from three to four copies of a card increases the chance of seeing it in seven cards by only about 8 percentage points (from about 31.5 to about 40). A fifth effectively similar card does help, but at some point each added slot buys very little extra reliability while still using up deck space that could carry something else.
This is why many builders aim for a sensible level of reliability rather than certainty. No realistic configuration is guaranteed; the hypergeometric distribution shows that even strong draw odds leave a real chance of missing, and that chance is worth planning around rather than hoping to eliminate.
Resources are part of consistency too
Many games require resource cards or a particular colour or type of resource. A deck with demanding requirements has an extra source of inconsistency: the cards themselves may be fine, but you cannot afford them. This is where the resource ratio matters, and it is why greedy lists that spread across several resource types are considered risky. The more specific the requirement, the more the deck relies on drawing the right supporting cards.
Opening hands: where consistency shows first
The first few cards are where failures are most visible. A deck that can reliably produce a functioning opening hand lets you stop worrying about mulligans and plan from the start. Judging whether to keep or redraw a starting hand is easier when most of your hands are fine. Decks built for power often need hands containing very specific pieces, which makes the decision harder and the redraws more frequent.
Testing rather than guessing
Opinions about consistency tend to be coloured by memory. Players remember the games where the key card never appeared and tend to blame the list. This is a form of confirmation bias, since the misses are vivid and the many ordinary games are forgotten. A few suggestions:
- Record results over a reasonably large number of games before changing a list.
- Use a game simulation, even a basic one in a spreadsheet, to measure how often you hold a working hand by a given turn.
- Compare the calculated frequency with what you actually see, keeping in mind that short-run results routinely differ from the averages.
Practical guidelines
- If your plan depends on one card, count your copies and your substitutes, and calculate the chance you have access to one by the turn you need it.
- Prefer cards that are useful in most situations when you are unsure; reserve conditional cards for the number of slots you can afford to lose.
- Consider the format. When games are short, one stumble can decide the match; when games are long, you have time to recover.
- Do not assume the answer is always “more consistent”. A reliable deck that is too weak to beat anything in its environment is also a poor choice.
There is no single correct balance. A useful way to frame the decision is to decide how much average strength you are willing to give up for each reduction in the chance of a dead game, and to be honest that those chances are estimates.