Two different skills
In constructed play you choose from your whole collection, or from every legal card, and bring a prepared list. In a draft you start with nothing. Each player opens a pack, takes one card, passes the rest along, and repeats until the packs are gone. The deck is then built from only what you picked, and it is usually smaller than a constructed deck.
This changes what is being tested. Constructed play rewards preparation and knowledge of the field. Limited play rewards adaptability, evaluation of cards you have not seen much of, and sound deck construction under constraint. Many players find that skill in one does not carry over automatically to the other.
Why picks are not just “take the best card”
The card you want most is not always the card you should take, for a few reasons.
The pool is shared. What you pass becomes part of your neighbour’s deck, and what you receive tells you something about theirs. A strong card in a colour or category that no one else is taking can be a signal that it is underdrafted.
Decks have shape. A limited deck needs a handful of different things: cheap plays, a few strong top-end cards, removal, and enough creatures or units. An excellent card that does not fit your colours or plan is worth less than an average one that does.
Consistency counts for more. Decks are small and cannot repeat their best cards, so averages matter. The ideas in deck building principles apply here, but with fewer options and more variance in what you have.
Early, middle and late picks
A rough pattern is useful, though it is a tendency and not a rule.
- Early picks favour raw power and flexibility. You do not yet know what you will play, so take the strongest card that keeps options open.
- Middle picks begin to follow what is flowing. If a particular style keeps coming back to you, that can mean the players passing to you are not taking it.
- Late picks are about filling the shape of the deck: the cheap creature you are short of, the removal spell you need, a card that completes the plan.
Many drafters talk about “staying open”, meaning they avoid committing too early. That is sensible because early commitment ignores information. But staying open forever has a cost too, as a deck split over too many styles tends to be inconsistent.
Signals and what they can and cannot tell you
If a powerful card shows up late in a pack, it suggests the players before you did not take it. This is a signal, but a noisy one. People value cards differently, they sometimes make mistakes, and some of what you see is just what the packs happened to contain. It resembles reading a table where hidden information is partly revealed by what others choose to do.
A reasonable approach is to treat a signal as a mild nudge. One late strong card is weak evidence; a pattern across several picks is stronger. Rely on the pattern rather than the single example, a point with obvious parallels to judging any result by sample size.
Evaluating cards without a tier list
In constructed play, the community usually settles on reference lists fairly quickly. In limited play, card rankings exist but depend on the set and on how people draft it. Be careful with such rankings: how tier lists are made and misread applies here, since a ranking reflects a certain environment and a certain level of play.
When you have no reference, a few questions help:
- Does the card affect the board when it is played, or does it need setup?
- Does it trade evenly with an opponent’s card, or better than evenly?
- Is it playable on its own, or does it depend on specific partners?
Cards that answer the first two well and the third with “yes, on its own” tend to be reliable picks. Cards that need partners are better later in the draft, once you know whether those partners are present.
Building the deck
At the end of the draft, you assemble a deck from the pool. Typical limited decks run a fixed number of lands or resource cards, around seventeen in a forty-card deck in many card games. That is only a starting point, since the number should follow from the cost of your cards. If your average card is cheap, you can run fewer; if it is expensive, more.
Check the curve: if you have too many expensive cards, you may stall early. Concerns about mana curve and tempo are as important here as in constructed. For a seventeen-land, forty-card deck, the chance of a seven-card hand with no lands is roughly 1.3 percent, and with exactly one land roughly 9 percent, which is part of why people decide carefully about keeping hands. The details of those choices are covered in mulligan decisions.
What to practice
- Pick quickly early, slowly late. Early choices have fewer consequences than they feel like they do.
- Keep notes on what you took and why. Looking back at picks that did not make the deck shows where your evaluation was off.
- Count playables. If you reach the last few picks short of playable cards, adjust earlier next time.
- Review games, not just results. A loss might reflect variance in the draw rather than bad drafting, so judge the deck over several matches.
Drafting is a skill that improves with repeated exposure. Even strong drafters will sometimes end up with a deck that does not come together, because the packs you open and the choices of the players near you are outside your control.