Poll
| 25 votes (49.01%) | ||
| 16 votes (31.37%) | ||
| 7 votes (13.72%) | ||
| 4 votes (7.84%) | ||
| 12 votes (23.52%) | ||
| 3 votes (5.88%) | ||
| 6 votes (11.76%) | ||
| 5 votes (9.8%) | ||
| 12 votes (23.52%) | ||
| 10 votes (19.6%) |
51 members have voted
Applying my conjecture that the probability of reaching a point at any time during a session is double the probability of ending the session at/beyond that point , we need the z-score corresponding to the probability of 1/3 * 1/2, which is 0.967. Then take 1200^.5 * 1.1547 * 0.967 to get the answer of 38.7 units * $500 = $19,350.Quote: Ace2You go to Vegas over a long weekend to play a single-deck blackjack game with a 0% edge ($500 minimum). You decide to play 1,200 hands but you will quit if you bust or double your initial bankroll before reaching 1200 hands.
What size bankroll should you bring to give yourself a 1/3 chance of busting, 1/3 chance of doubling and 1/3 chance of finishing the 1200 hands without busting or doubling?
Assume a standard deviation of 1.1547, flat betting $500 one hand at a time and perfect basic strategy to realize the 0% edge
link to original post
Verification: Knowing the standard deviation and edge, you can easily calculate that this game is statistically equivalent to a bet with a 3/7 probability of winning 7 for 3. Markoving 1200 bets shows that a bankroll of 38 units * $500 = $19,000 gives bust/double/finish probabilities of 33.4%/33.4%/33.2%. I believe this is the closest you can get to 1/3.
So I'd bring $20,000
Gin and Win
The object of trying to put as many cards as you can in sets, with the goal of minimizing dead wood, is also the goal of Gin and Win. It should be emphasized that pairs and flushes of at least three cards count as sets in Gin and Win. The rules start as follows:
1. Cards are ranked as in poker, except aces are low only.
2. To begin, each player must post an Ante bet.
3. The player and dealer are each dealt seven cards, face down, from a normal 52 card deck.
4. The player looks at his hand and removes any combinations of pairs, trips, and quads (rank melds), or any flushes of three cards or more cards (suit melds). One card cannot be part of more than one set.
5. The remaining cards are known as "Dead Wood" and are set apart from the cards that belong to a set.
Math puzzle:
With what probability will a player be dealt a hand with "no sets" and thus have seven Dead Wood cards?
With what probability will a player be dealt a hand with exactly five Dead Wood cards?
The second part is that to get no flushes the suits have to be divided 2-2-2-1. There are four ways to pick the singleton suit, so let's assume it's Clubs, thus the doubletons are Spades, Hearts and Diamonds. There are seven different ranks, so the ways to assign the Spades to them is 7*6/2=21. This leaves five non-Spades. The ways to assign Hearts is 5*4/2=10. Similarly Diamonds = 3*2/2=3. Leaving the Club assignment. The total number of ways for the suits is 4 * 21 * 10 * 3 * 1 = 2520.
Of the 1716 ways for seven different ranks, only 393 have no possible 3-card straights (I could only do that using a spreadsheet).
So there are 393*2520 hands = 990,360.
The total number of ways to have 7 cards from 52 is 133,784,560.
So the chances of such a hands = 990,360 / 133,784,560 which is about 0.7% (or one hand in 135).
13 * combin(12,5) * (12*10*3) / combin(52,7)
Quote: Ace25 deads
13 * combin(12,5) * 16 / combin(52,7)
link to original post
In order to have exactly 5 dead cards, two of your cards must be a pair, and the other 5 cards must have no pairs and your seven cards must have no more that 2 cards in any given suit.
Quote: Ace2^^^That’s exactly what I calculated, though I did edit my answer
link to original post
Well you got the denominator right, lol. I am struggling with your numerator.
Try answering the question about 7 deads, which may illuminate the path to wisdom on 5 Deads.
Quote: gordonm888This easy math puzzle is based on the rules of Gin and Win.
Gin and Win
The object of trying to put as many cards as you can in sets, with the goal of minimizing dead wood, is also the goal of Gin and Win. It should be emphasized that pairs and flushes of at least three cards count as sets in Gin and Win. The rules start as follows:
1. Cards are ranked as in poker, except aces are low only.
2. To begin, each player must post an Ante bet.
3. The player and dealer are each dealt seven cards, face down, from a normal 52 card deck.
4. The player looks at his hand and removes any combinations of pairs, trips, and quads (rank melds), or any flushes of three cards or more cards (suit melds). One card cannot be part of more than one set.
5. The remaining cards are known as "Dead Wood" and are set apart from the cards that belong to a set.
Math puzzle:
With what probability will a player be dealt a hand with "no sets" and thus have seven Dead Wood cards?
With what probability will a player be dealt a hand with exactly five Dead Wood cards?
link to original post
gordonm888,
I see no mention of straights in item 4. Do they not count in Gin and Win?
Dog Hand
Quote: DogHandI see no mention of straights in item 4. Do they not count in Gin and Win?
link to original post
No, they do not. See part 5 of this Nevada Gaming Commission document about the game.
Quote: ThatDonGuyQuote: DogHandI see no mention of straights in item 4. Do they not count in Gin and Win?
link to original post
No, they do not. See part 5 of this Nevada Gaming Commission document about the game.
link to original post
That is correct.
Dog Hand
13 *6 * combin(12,5) * (4*3*10*3) / combin(52,7)
13 * 6 ways to choose the pair
C(12,5) ways to choose ranks of the five
4*3 ways to choose suits of the five (221)
10*3 ways to assign the suits to the five
Caveat: combinatorics aren’t my strongest skill
Quote: charliepatrickThe first part is to get no pairs, all the ranks of the cards have to be different. At a later stage let's look at which of those could make a 3-card straight.
The second part is that to get no flushes the suits have to be divided 2-2-2-1. There are four ways to pick the singleton suit, so let's assume it's Clubs, thus the doubletons are Spades, Hearts and Diamonds. There are seven different ranks, so the ways to assign the Spades to them is 7*6/2=21. This leaves five non-Spades. The ways to assign Hearts is 5*4/2=10. Similarly Diamonds = 3*2/2=3. Leaving the Club assignment. The total number of ways for the suits is 4 * 21 * 10 * 3 * 1 = 2520.
Of the 1716 ways for seven different ranks, only 393 have no possible 3-card straights (I could only do that using a spreadsheet).
So there are 393*2520 hands = 990,360.
The total number of ways to have 7 cards from 52 is 133,784,560.
So the chances of such a hands = 990,360 / 133,784,560 which is about 0.7% (or one hand in 135).
link to original post
To confirm the comments of others, in this puzzle (and in Gin and Go) one cannot meld straights.
Quote: charliepatrick^ I erroneously had assumed it was like regular gin. It does make the maths easier as you don’t have to check the unequal ranks.
So instead of 393*2520 it’s 1716*2520 giving 4,324,320 or about 3.23%.
link to original post
I get a slightly higher answer. I am having difficulty critiquing your answer because you have used numbers rather than formulas such as c(n,m) for combinations.
The second number is how many ways the suits of seven different cards can happen so the suit distribution is 2-2-2-1.
The first part is that there can be four suits for the singleton, and once that has been determined then the other suits are fixed.
The next part is how those suits can be distributed onto the ranks.
The singleton can be put in one of 7 positions.
The highest of the other three suits can be put in 6*5/2 of the remaining 6 positions.
Similar the next suit can be put in 4*3/2 of the remaining positions.
The last suit is now set.
So you have 4 * 7 * 6*5/2 * 4*3/2 * 1 = 2520.
2520 * 1716 = 4324320.
7 | 4 | 6 864 | |||
6 | 1 | 84 | 144 144 | ||
5 | 2 | 252 | 432 432 | ||
5 | 1 | 1 | 504 | 864 864 | |
4 | 3 | 420 | 720 720 | ||
4 | 2 | 1 | 2 520 | 4 324 320 | |
4 | 1 | 1 | 1 | 840 | 1 441 440 |
3 | 3 | 1 | 1 680 | 2 882 880 | |
3 | 2 | 2 | 2 520 | 4 324 320 | |
3 | 2 | 1 | 1 | 5 040 | 8 648 640 |
2 | 2 | 2 | 1 | 2 520 | 4 324 320 |
Quote: Ace2One more revision for five deads
13 *6 * combin(12,5) * (4*3*10*3) / combin(52,7)
13 * 6 ways to choose the pair
C(12,5) ways to choose ranks of the five
4*3 ways to choose suits of the five (221)
10*3 ways to assign the suits to the five
Caveat: combinatorics aren’t my strongest skill
link to original post
I also believe this is not correct.
Let me point out that if one of the paired cards can be used to make a three card flush in your hand, then you would meld the three card flush because it would give you 4 deadwood cards rather than the five deads that result from melding the pair.
Charlie's answer is correct. See my next post.
Quote: charliepatrickFirstly there are how many ways can you make 7 different ranks from 13, which is 13 12 11 10 9 8 7 / (7 6 5 4 3 2 1) = 1716.
The second number is how many ways the suits of seven different cards can happen so the suit distribution is 2-2-2-1.
The first part is that there can be four suits for the singleton, and once that has been determined then the other suits are fixed.
The next part is how those suits can be distributed onto the ranks.
The singleton can be put in one of 7 positions.
The highest of the other three suits can be put in 6*5/2 of the remaining 6 positions.
Similar the next suit can be put in 4*3/2 of the remaining positions.
The last suit is now set.
So you have 4 * 7 * 6*5/2 * 4*3/2 * 1 = 2520.
2520 * 1716 = 4324320.
7 4 6 864 6 1 84 144 144 5 2 252 432 432 5 1 1 504 864 864 4 3 420 720 720 4 2 1 2 520 4 324 320 4 1 1 1 840 1 441 440 3 3 1 1 680 2 882 880 3 2 2 2 520 4 324 320 3 2 1 1 5 040 8 648 640 2 2 2 1 2 520 4 324 320
link to original post
Charlie's answer is correct!
To have 7 dead wood in a 7 card hand there must be no flushes of 3 or more cards, and seven different ranks.
A 7 card hand with no flushes (3 or more cards) can have only one distribution of suites: 2-2-2-1 with four variations corresponding to which of the four suits has only on card.
First, ignoring pairing, let's write a formula for the number of combinations of seven card hands with no qualifying flushes, given that it must be a 2-2-2-1 distribution of cards in the four suits.
= 4* c(13,2)* c(13,2)* c(13,2)*c(13,1) = 24,676,704
This value includes hands with paired cards and less that 7 deadwood. It's simple to modify this formula to require that the 7-card hand with a 2-2-2-1 suit distribution has 7 distinct ranks; i.e., no pairing:
=4* c(13,2)*c(11,2)*c(9,2)*COMBIN(7,1) = 4,324,320
So, the frequency of a 7-deads hand = 4* c(13,2)*c(11,2)*c(9,2)*COMBIN(8,1) /c(52,7) = 0.032323012
The second part of the puzzle is for the probability of getting a hand with 5 deadwood cards. I used the first formula in this derivation as a starting point for the 5 deads calculation.
There are 10296 ways to have one rank that is the pair and five other ranks (13 * (12*11*10*9*8/5/4/3/2/1) ).
There are then 6144 (6*4*4*4*4*4) to have the suits arranged, six ways for the pair, and four for each singleton.
The same logic produces...11119680/133784560 = about 8.31%
7 | |||||
6 | 1 | 12 | 123 552 | ||
5 | 2 | 60 | 617 760 | ||
5 | 1 | 1 | 132 | 1 359 072 | |
4 | 3 | 120 | 1 235 520 | ||
4 | 2 | 1 | 840 | 8 648 640 | |
4 | 1 | 1 | 1 | 300 | 3 088 800 |
3 | 3 | 1 | 600 | 6 177 600 | |
3 | 2 | 2 | 960 | 9 884 160 | |
3 | 2 | 1 | 1 | 2 040 | 21 003 840 |
2 | 2 | 2 | 1 | 1 080 | 11 119 680 |
Btw 11119680 is exactly 1/2 of my answer
For the 1111111 rank distribution (i.e. no pairs) its essentially like creating numbers of base 4. So there are 47 lines.
For the 211111 rank distribution (i.e. one pair) the last five columns are as above, but the first two start with 12, then go 13, 14, 23, 24, 34. Thus there are 6*45 lines.
For each line I then count how many 1s, 2s, 3s and 4s there are, and create four digit numbers such as 7000, 6100, 5200, 5110 etc for the suit distribution. I then count how many of these there are.
For the pairs I looked at the simple cases e.g. 6100 is where the odd suit has to be one of the pairs, and all the other cards are in the same suit. So there are 4*3 of them.
Since the numbers added up to 6144, I didn't check the more complicated ones.

Quote: charliepatrick^ I've now developed a spreadsheet which essentially creates every combination of how the suits are given to the various ranks.
For the 1111111 rank distribution (i.e. no pairs) its essentially like creating numbers of base 4. So there are 47 lines.
For the 211111 rank distribution (i.e. one pair) the last five columns are as above, but the first two start with 12, then go 13, 14, 23, 24, 34. Thus there are 6*45 lines.
For each line I then count how many 1s, 2s, 3s and 4s there are, and create four digit numbers such as 7000, 6100, 5200, 5110 etc for the suit distribution. I then count how many of these there are.
For the pairs I looked at the simple cases e.g. 6100 is where the odd suit has to be one of the pairs, and all the other cards are in the same suit. So there are 4*3 of them.
Since the numbers added up to 6144, I didn't check the more complicated ones.
link to original post
I've read this through many many times and studied your table and I do now understand what you've done. Very impressive. You've figured out the probabilities of all the suit distributions for the "1 pair + 5 singletons" case. In the first seven columns the numerals 1-4 refer to suits of the seven cards P1,P2,A,B,C,D,E in a 'one pair+5 singletons' hand.
Again, very impressive. Your answer for 5 deads agrees with mine, but I took a simpler approach to get there.
My approach was to start with the idea that a 5 deadwood hand must have a 2221 suit distribution and then analyze how often such a suit distribution will consist of one pair and 5 singletons.
There are three possibilities to form the one pair in a 2221 suit distribution:
- the second suit has a card that pairs with one of the 2 cards in the first suit,or
- the third suit has a card that pairs with one of the 4 cards in the first two suits, or
- the only card in the fourth suit pairs with one of the 6 cards in the first three suits
Hence the combinatorial formula is:
=4*(COMBIN(13,2)* COMBIN(11,1)*2* COMBIN(10,2)* COMBIN(8,1)+ COMBIN(13,2)* COMBIN(11,2)* COMBIN(9,1)*4* COMBIN(8,1)+ COMBIN(13,2)* COMBIN(11,2)* COMBIN(9,2)*6)/ COMBIN(52,7)
= 0.083116318
And CharliePatrick's result agrees with this.
I felt the key insight was to understand that the suit distribution was 2221 for both the 7 deads and 5 deads case and then to modify that combinatorial formula for the case of no pairs and one pair.

