Probability finding a card in $n$ draws?

581 Views Asked by At

I remember seeing this as a trick before. I tried to recreate it but it's proving to be difficult.

I have a deck of $13$ cards, consisting of ace through king. I have a card number in my head (ace, $2$-$10$, jack, queen, king). What is the probability of finding it in $n$ draws? After each draw I put the card back and shuffle the deck.

1

There are 1 best solutions below

0
On BEST ANSWER

As it is essentially drawing with replacement, we notice that any particular draw will be the card we are looking for with probability $\frac{1}{13}$ and will not be with probability $\frac{12}{13}$.

As each draw is independent of one another, the probability of a specific sequence of draws will be the product of the probabilities of each draw individually. E.g. Ace-Ace-Nonace in that order will occur with probability $\frac{1}{13}\cdot\frac{1}{13}\cdot\frac{12}{13}$.

There are two possible interpretations of the question I can think of.

What is the probability that the first ace is drawn on the $n$'th draw?

This follows the geometric distribution with $p=\frac{1}{13}$. This corresponds to getting $n-1$ non-ace cards in a row followed by an ace. This occurs then with probability $\frac{12}{13}\cdot\frac{12}{13}\cdots\frac{12}{13}\cdot\frac{1}{13}=(\frac{12}{13})^{n-1}\cdot\frac{1}{13}$

What is the probability that within the first $n$ draws, at least one ace is drawn? (but it doesn't necessarily need to be the $n$'th draw where the ace occurs)

Getting at least one ace within the first $n$ draws means that the first ace we see occurs within the first $n$ draws, so we may simply sum over the geometric distribution up to $n$. $\sum\limits_{k=1}^n(\frac{12}{13})^{k-1}\frac{1}{13}$ which may be simplified using what you know about geometric series.

An alternate method to calculate this is by recognizing that getting at least one ace within the first $n$ draws is the opposite event of getting no aces within the first $n$ draws. Getting no aces corresponds to not getting an ace for the first draw followed by not getting an ace for the second draw followed by...etc...

Getting no aces in $n$ draws has probability then $\frac{12}{13}\cdot\frac{12}{13}\cdots\frac{12}{13}=(\frac{12}{13})^n$ so getting at least one ace in $n$ draws has probability $1-(\frac{12}{13})^n$. Had you simplified your geometric sum above, it would have arrived at this answer as well.


A word of warning, I mentioned that $Pr(A\cap B)=Pr(A)\cdot Pr(B)$, that is the probability event $A$ and event $B$ simultaneously occur is the probability of the product of their respective probabilities. This is true only for independent events, and is in general not true for arbitrary events.