UTILS.
100% in-browser
🃏

Hypergeometric Draw Calculator (MTG / TCG)

Hypergeometric probability calculator for MTG and card games: the odds of drawing your card, exactly, at least, at most k, and at least one.

Opening hand = 7 cards. On the play by turn t, set cards drawn to 7 + t.

About this tool

The Hypergeometric Draw Calculator tells you how likely you are to draw the cards you need from a deck, the exact maths behind every Magic: The Gathering or trading-card-game consistency question. Everything runs in your browser — no data leaves your device. Enter the deck size, how many copies of the card it contains, how many cards you will draw, and how many you want.

The probability of drawing exactly k copies is P(X = k) = C(K, k)·C(N − K, n − k) / C(N, n), the hypergeometric distribution, where N is the deck, K the copies, and n the cards drawn. To stay accurate for large decks it evaluates every binomial coefficient with log-factorials — lnC(n, k) = lnΓ(n+1) − lnΓ(k+1) − lnΓ(n−k+1) — then exponentiates, so nothing overflows. From the same terms it derives P(at least k), P(at most k), P(at least one) = 1 − P(0), and the expected number of copies drawn, n·K/N.

For an opening hand draw 7 cards; on the play by turn t you will have seen 7 + t cards, so set the drawn field accordingly. The copies wanted cannot exceed the copies in the deck, and you cannot draw more cards than the deck holds — those cases are flagged rather than guessed.

Frequently asked questions

When should I use the hypergeometric distribution?
Whenever you draw without replacement from a fixed deck — the classic "what are my odds of drawing at least one of my four copies?" question. Because each card drawn changes the deck, this is exact where a simple percentage-per-card estimate is not.
How do I model 'by turn 3 on the play'?
Set the cards drawn to 7 plus the number of turns. On the play you draw no card turn one, so by your turn 3 you have seen 7 + 2 = 9 cards; on the draw add one more for the extra draw step.
What is the expected copies value?
It is n·K/N, the average number of copies you will draw across many games. It can be a fraction — 0.7, say — since it is a long-run mean, not a count you will see in any single game.
Why do I get an invalid result?
The copies wanted cannot exceed the copies in the deck or the cards drawn, and the deck size must be at least the copies and the cards drawn. Fix any of these and the probabilities appear immediately.

More tools