Introduction
Before starting, I would like to specify a few things:
- The deck has 52 cards. From 1 (Ace) to 10, J, Q, K, all in four different suits: spades, diamonds, club and hearts. I don't add jokers because that is far beyond my possibilities of calculations.
- For better precision, I won't use chance, instead I would use combinations or frequency / total possible combinations.
- This is based on poker, but it isn't exactly poker. I mean, for example, the frequency on a pair in 5-cards poker is $2860={13 \choose 1}{4 \choose 2}{12 \choose 3}{4 \choose 1}^3$, this formula that can be found on Wikipedia already excludes the chance of having not two but three or four cards with the same rank (three or four of a kind) and also excludes the chance of a full house (three of a kind + pair). What I want to discover, is not the frequency of a pair, instead, is the frequency of having at least two cards of the same rank. I'm not using the rules of poker, so I don't want to remove from the frequency the combinations that are "better". For example, the hand Q Q Q J J, which normally would be a full house, for this calculation I want it to be included in the "pair" calculation, because if you have three Qs, that means that you have at least a pair of Q, same with the Js.
What do I want?
What I want to calculate is the frequency of drawing:
- At least two cards of the same rank (a pair).
- At least two cards of the same suit (a flush of 2).
- At least two consecutive cards (a straight of 2).
- At least two consecutive cards of the same suit (a straight of 2 flushed).
All the previous frequencies I wanted to calculate them with:
- A 2-card hand.
- A 3-card hand.
- A 4-card hand.
- A 5-card hand.
- A 6-card hand.
- A 7-card hand.
My Calculations
Before starting to show my math I will show you how I got it. At wikipedia, the chance of a four of a kind if a 5-card deck is $642={13 \choose 1}{4 \choose 4}{12 \choose 1}{4 \choose 1}$. I understand this as choose one of 13 ranks, then choose all four suits of that rank, for the last card choose one other rank and then one suit for that card. Then I tried to recreate the formula with different reasoning. $642={13 \choose 1}{4 \choose 4}{52-4 \choose 1}$ which I understand as choose one of 13 ranks, then choose all four suits of that rank, finally choose one card other than the four previously used.
Below I will show you a table with my math.
Hand | 2 cards | 3 cards | 4 cards | 5 cards | 6 cards | 7 cards |
---|---|---|---|---|---|---|
Total combinations | $1.326={52 \choose 2}$ | $22.100={52 \choose 3}$ | $270.725={52 \choose 4}$ | $2.598.960={52 \choose 5}$ | $20.358.520={52 \choose 6}$ | $133.784.560={52 \choose 7}$ |
Pair | $78={13 \choose 1}{4 \choose 2}$ | $3.900={13 \choose 1}{4 \choose 2}{52-2 \choose 1}$ | $95.550={13 \choose 1}{4 \choose 2}{52-2 \choose 2}$ | $1.528.800={13 \choose 1}{4 \choose 2}{52-2 \choose 3}$ | $17.963.400={13 \choose 1}{4 \choose 2}{52-2 \choose 4}$ | $165.263.280={13 \choose 1}{4 \choose 2}{52-2 \choose 5}$ |
Flush of 2 | $312={13 \choose 2}{4 \choose 1}$ | $15.600={13 \choose 2}{4 \choose 1}{52-2 \choose 1}$ | $382.200={13 \choose 2}{4 \choose 1}{52-2 \choose 2}$ | $6.115.200={13 \choose 2}{4 \choose 1}{52-2 \choose 3}$ | $71.856.600={13 \choose 2}{4 \choose 1}{52-2 \choose 4}$ | $661.053.120={13 \choose 2}{4 \choose 1}{52-2 \choose 5}$ |
Straight of 2 | $208={13 \choose 1}{4 \choose 1}^2$ | $10.400={13 \choose 1}{4 \choose 1}^2{52-2 \choose 1}$ | $254.800={13 \choose 1}{4 \choose 1}^2{52-2 \choose 2}$ | $4.076.800={13 \choose 1}{4 \choose 1}^2{52-2 \choose 3}$ | $47.902.400={13 \choose 1}{4 \choose 1}^2{52-2 \choose 4}$ | $440.702.080={13 \choose 1}{4 \choose 1}^2{52-2 \choose 5}$ |
Straight of 2 flushed* | $52={13 \choose 1}{4 \choose 1}$ | $2.600={13 \choose 1}{4 \choose 1}{52-2 \choose 1}$ | $63.700={13 \choose 1}{4 \choose 1}{52-2 \choose 2}$ | $1.019.200={13 \choose 1}{4 \choose 1}{52-2 \choose 3}$ | $11.975.600={13 \choose 1}{4 \choose 1}{52-2 \choose 4}$ | $110.175.520={13 \choose 1}{4 \choose 1}{52-2 \choose 5}$ |
*At 12 cards $27.917.689.800={13 \choose 1}{4 \choose 1}{52-2 \choose 10}$ while the total number is $15.820.024.220={52 \choose 12}$
My problem is that as you can see, when you draw multiple hands, the combinations of for example a pair becomes higher than the total possible combination, a thing that is absurd and can be easily demonstrated with an example hand: 1, 2, 3, 4, 5, 6, 7, there are 7 different cards, none is a pair. Of course, a flush of 2 cards is allowed to get 100% if you draw 5 cards (it is impossible to get 5 different suits because there only exist 4 different); however, it should not reach 100% in a 4-card hand (an easy example would be drawing a spade, a club, a diamond and a heart, none is repeated).
Question: What am I missing with my calculations?
As my method of calculation of the extra cards after forming what I want (for example, the three cards remaining after having a pair using 5 cards) I used ${52-2 \choose 3}$ instead of what appears in Wikipedia, I made an experiment trying to get the same results of Wikipedia with my method.
The frequency of three of a kind in Wikipedia is $54.912={13 \choose 1}{4 \choose 3}{12 \choose 2}{4 \choose 1}^2$ which I understand as choose one card of the 13 ranks, then choose three of the four suits for that card, finally choose 2 cards that aren't from the previous rank, these cards each can be of any suit.
If I use ${13 \choose 1}{4 \choose 3}{52-4 \choose 2}=58.656$ I have to eliminate the frequency in which that two extra cards are a pair, because that would fall into Full House from normal poker (following Wikipedia), so $58.656-3.744=54.912$ which is fine. Now I go one step further.
If I use ${13 \choose 1}{4 \choose 3}{52-3 \choose 2}=61.152$ I have to eliminate the frequency in which that two extra cards are a pair (because of full house), and the frequency in that the remaining card of the suit appears and transform my three of a kind into a four of a kind $61.152-3.744-624=56.784$ which isn't 54.912. I discovered that if I multiply (in the previous formula) the frequency of a full house by 1.5, or I multiply the frequency of a four of a kind by 4 I get exactly 54.912, but I'm not sure why, and that might be what makes my math surpass the 100% (when comparing the frequency/total possible combinations).
Do you have any idea what I'm doing wrong?
What would be considered an answer?
An explanation of which mathematical expression I am understanding wrong, a correct expression to replace that part, an explanation of what means that suggested new expression and one example using any of the combinations that appear in my table above that surpass the 100% using that new suggested expression.
For example: "When you say this expression ... you are wrong because that means instead ... What you have to use is ... because that means ... For example, if we use this formula to get the frequency of a pair with a 7-cards hand (in which your math surpassed 100%), we get ... which is correct because ..."
Note: I'm not a math teacher, and I'm not studying for a math career, so please try to not be too much complicated in the explanation. I just know combinations because at university we had 3 classes of hypergeometric distributions (I'm studying business administration). This is not for the university, work or gambling, actually, it is for Dungeons & Dragons 5e.