Probability  Dice
If you are rolling 6 sixsided standard dice what are the odds of rolling six of a kind?
Jeff B. from Miami, Florida
The answer is 6*(1/6)^{6} = 6/46,656 = 1/7,776 =~ 0.0001286 .
What is the average number of rolls until a shooter "sevens out"? I know that a 7 will appear every 6 rolls, but with comeout 711s and craps, plus the possibility of shooters making multiple points, I think the average number of rolls may be higher than expected. Is there any mathematical reference material on this?
Grshooter from Kansas City, Missouri
The average number of rolls per shooter is 8.525510. For the probability of exactly 2 to 200 rolls, please see my craps probability of survival page.
What are the odds (and frequency) in 100,000 rolls of the dice (craps/dont pass line) of losing a DP bet 2x’s.3x’s 4x’s 5x’s 6x’s 7x’s 8x’s or 9x’s in a row.
Derick from Minneapolis, USA
My craps appendix shows how to work out the odds for any one bet. There you will see the probability of losing the don’t pass bet is 2928/5940. The probability of losing n bets in a row is (2928/5940)^{n}. The frequency in 100,000 of losing exactly n can be closely approximated as 100,000 * (2928/5940)^{n+2}.
What are the odds of rolling the same number with six dice in one roll?
Kyle Hill from Colstrip, Montana
The odds of getting six of the same number with six dice is 6*(1/6)^{6}=1/7776 =~ 0.01286%.
How often can you roll a pair of dice 28 times without getting a 7? How do you figure this? Congratulation on your site, it’s great.
Arturo G. from Mexico City, Mexico
Thanks for the compliment. I take it you mean what is the probability of rolling a pair of dice 28 times without getting a 7. The probability of not rolling a 7 on any one roll is 5/6. The probability of not rolling a 7 in 28 rolls is (5/6)^{28} = 0.006066, or about 1 in 165.
Just a question about an Oriental dice game, where the players are supposed to guess which side of the die shows up. The players will first place their bets on 1,2,3,4,5,6 (like roulette) and then the "dealer" will roll 3 dice simultaneously. Payouts are 1:1 if the chosen numbers shows up once (on any of the 3 dice), 2:1 if the chosen no shows up twice, and 3:1 if the chosen number appears on all 3 dice. As the player can place any number of bets of the board, what will be the optimum number of bets to place? (assuming all my bets are equal in size)
Jansen from Toronto, Canada
The probability of three matching is 1/216. The probability of two matching is 3*5/216. The probability of one matching is 25*5/216. The probability of 0 matching is 5*5*5/216. So the expected return is 3*(1/216)+2*(15/216)+1*(75/216)1*(125/216)=17/216=7.87%. There is no optimal number of bets, you will give up an expected 7.87% of total money bet no matter what you do.
These bets can be made in both sic bo and chuck a luck.
Mr. Wizard, what is the probability of rolling two pair when rolling four dice?
Brian from St. Catherines, Canada
There are combin(6,2)=15 different sets of pairs possible. There are combin(4,2)=6 ways the dice can roll any specific two pair. There are 6^4=1296 ways to roll four dice. So the probability is 90/1296=6.9444%.
If I roll a die, my probability of rolling a six is 1/6. If I roll two dice, does my probability of rolling a six on one of them increase, or does it stay at 1/6?
Mike R. from Rosemount
If you rolled x dice the probability of getting at least one 6 is 1(5/6)^{2}. In the case of two dice this is 30.56%.
How many different ways are there of rolling 3 ones using 6 dice?
Jamie from Croydon, England
First there are combin(6,3)=20 ways you can choose three dice out of 6 for the three ones. Then each of the other three can be any of five numbers. So, the total ways are 20×5^{3}=2500. The total number of ways to throw all the dice are 6^{6}=46,656, so the probability of rolling exactly three ones is 2500/46656=0.0536. For help with the combin function see my probabilities in poker section.
What is the probability of any one die showing ONE when 3 dice are throw together. My understanding is it should be 50% chance (1/6+1/6+1/6=1/2 >50%) But in your odds table it shown 34.72%. Please help.
John C. from Singapore
The probability of rolling exactly one one with three dice is 3*(5/6)^{2}*(1/6) = 75/216 = 34.72%.
What is the probability of rolling a "pair" when tossing 4 dice?
Anthony from Toronto, Canada
The pair can be any one of 6 numbers. The other two singletons can be among the other five. So there are 6*combin(5,2)=60 combinations already. There are combin(4,2)=6 combinations of dice on which the pair can appear. The two singletons can be arrange in two ways. So there are 60*12=720 ways to throw a pair. The total number of all ways to throw the dice is 6^{4}=1296. So the probability is 720/1296 =~ 55.56%.
My friend owns a bar and has a "shake of the day" where there are ten dice in a Tupperware container, what are the odds of matching 8 out of the 10 in one shake. Thank you for your time.
August from Oshkosh, USA
The probability that if you roll 10 dice and exactly 8 numbers are the same is 6*combin(10,8)*(1/6)^{8}*(5/6)^{2} = 1/8957.952. The probability of matching at least 8 is 6*[combin(10,8)*(1/6)^{8}*(5/6)^{2} + combin(10,9)*(1/6)^{9}*(5/6) + (1/6)^{10}] = 1/8569.469.
I recently rolled, during a game of backgammon, double sixes four consecutive times. What are the odds of this happening again?
David from Sunland, USA
With every new roll the probability the next four rolls will be all double sixes is (1/36)^{4} = 1 in 1679616.
What is the probability of getting a straight by a single throw of 5 dice?
Teodoro C. Deocares from Dagupan, Philippines
There are two possible spans: 1 to 5 and 2 to 6. Each of these spans can be ordered in 5!=120 ways. There are 6^{5} = 7776 ways to roll five dice. So the probability is 2*120/7776 = 3.09%. The probability of this seems to be much higher right after I put mark 0 for large straight during a game of Yahtzee.
A fair sided die is rolled 30 times. What is the expected number of times that number 1 will come up? What is the probability that number 1 will come up it’s expected number of times?
Anonymous
The expected number of ones is 30*(1/6) = 5. The probability of exactly 5 ones is combin(30,5)*(1/6)^{5}*(5/6)^{25} = 19.21%.
If I have any give number of dice what is the probability that if I roll them all of them at least one will land on a one?
Anonymous
The probability that all the dice will not be a one is (5/6)^{n}. So the probability of at least one 1 is 1(5/6)^{n}. Let’s take an example of five dice. The answer would be 1(5/6)^{5} = 59.81%.
If I throw 36 dice what is the probability of getting at least one six?
Anonymous
1(5/6)^{36} = 99.86%
If I kept throwing and removed all the sixes each time, how would I predict the theoretical number of dice remaining after a particular number of throws?
Anonymous
Each roll the expectation is that 5/6 of the dice will remain. So the expected number of dice remaining after n throws would be 36*(5/6)^{n}. For example after 10 throws you would have 5.81 dice left, on average.
If I roll three dice, what is the probability of getting at least two numbers the same?
Anonymous
The probability all numbers will be different is (5/6)*(4/6)=20/36. So the probability at least two numbers will be the same is 1(20/36) = 16/36 = 44.44%.
If two people throw a pair of dice, What is the probability that it is the same number? Is there a formula to figure this out?
Anonymous
Yes. You simply run through all total from 2 to 12 and determine the probability of rolling each twice. So the answer would be (1/36)^{2}+(2/36)^{2}+(3/36)^{2}+(4/36)^{2}+(5/36)^{2}+(6/36)^{2}+(5/36)^{2}+(4/36)^{2}+(3/36)^{2}+(2/36)^{2}+(1/36)^{2} = 11.27%.
I recently attended a hospital fete. There was a new car as a prize if 7 dice produced 7 sixes in one throw. £1.00 a go. Odds on this must be high but how high?
Anonymous
The probability of throwing seven sixes with seven dice is (1/6)^{7} = 1 in 279,936. So the car would have to have a value of £279,936 or more for this to be a good bet. Even your average Rolls Royce is not worth this much, so I would say that was a terrible bet.
[Bluejay adds: Uh, yeah, but I think the point was that it was for charity. What’s more fun: Donating £1.00 to charity and getting nothing back but the good feeling of helping out, or donating £1.00 and getting the good feeling plus the longshot chance of winning a car?]
What are the probabilities for a 5 of a kind, 4 of a kind, 3 of a kind, full house, 2 pair, pair, straight, and nothing with the roll of five dice?
Anonymous
 Five of a kind: 6/6^{5} = 0.08% (obvious)
 Four of a kind: 5*6*5 = 1.93% (five possible positions for the singleton * 6 ranks for the four of a kind * 5 ranks for the singleton).
 Full house: combin(5,3)*6*5/6^{5} = 3.86% (combin(5,3) positions for the three of a kind * 6 ranks for the three of a kind * 2 ranks for the pair).
 Three of a kind: COMBIN(5,3)*COMBIN(2,1)*6*COMBIN(5,2) / 6^{5} = 15.43%. (combin(5,3) positions for the three of a kind * combin(2,1) positions for the larger of the singletons * 6 ranks of the three of a kind * combin(5,2) ranks for the two singletons.
 Two pair: COMBIN(5,2)*COMBIN(3,2)*COMBIN(6,2)*4 / 6^{5} = 23.15% (combin(5,2) positions for the higher pair * combin(3,2) positions for the lower pair * combin(6,4) ranks for the two pair * 4 ranks for the singleton.
 Pair: COMBIN(5,2)*fact(3)*6*combin(5,3) / 6^{5} = 46.30% (combin(5,2) positions for the pair * fact(3) positions for the three singletons * 6 ranks for the pair * combin(5,3) ranks for the singletons.
 Straight: 2*fact(5) / 6^{5} = 3.09% (2 spans for the straight {15 or 26} * fact(5) ways to arrange the order).
 Nothing: ((COMBIN(6,5)2)*FACT(5)) / 6^{5} = 6.17% (combin(6,5) ways to choose 5 ranks out of six, less 2 for the straights, * fact(5) ways to arrange the order.
Hi Wizard, I wanted to know if you can answer this. In a popular gambling game in 17th century France, a player would roll a pair of dice 24 times. He would win his bet if at least one of these rolls was a double six. There was a debate at the time over whether the probability of winning was above or below an even 50%. Can you help me?
Anonymous
Sure, this is easy. The probability of rolling at least one 12 in 24 rolls is 1(35/36)^{24} = 49.14%. So the odds favor betting against a 12. This is a clever bet because the expected number of twelves in 24 rolls is 2/3. However that does not mean the probability of a 12 is 2/3, because sometimes there will be more than one 12, and the player betting on 12 doesn’t win any more for extra twelves after the first one. If the probability of winning any given trial is p, the number of trials is n, and the probability of at least one win is w then solving for n in terms of p and w gives us...
w=1(1p)^{n}
1w = (1p)^{n}
log(1w) = log((1p)^{n})
log(1w) = n*log(1p)
n= log(1w)/log(1p)
So in your example n = log(1.5) / log(1(1/36)) = log(0.5) / log(35/36) = 24.6051. So if the probability of success is 50% in 24.6 rolls it must be slightly less in 24 rolls.
What is the probability of rolling 1,2,3,4,5,6 with six dice, six times in a row?
Anonymous
The probability of rolling 123456 with six dice in a single roll can be expressed as prob(second die does not match first die) * prob(third die does not match first or second die) * ... = 1*(5/6)*(4/6)*(3/6)*(2/6)*(1/6) = 0.015432. So the probability of doing this six times in a row is 0.015432^{6} = 1 in 74,037,208,411.
If I roll a single die 6 times, what is the probability of getting a "2" exactly 4 times?
Anonymous
Combin(6,2)*(1/6)^{4}*(5/6)^{2} = 0.008037551.
What is the probability of rolling 13 or more with 3, 4, and 5 dice, if you are allowed to keep the highest three dice in your roll?
Anonymous
Here are the probabilities:
3 dice: 25.93%
4 dice: 48.77%
5 dice: 66.13%.
Suppose we roll three fair sixsided dice. What’s the conditional probability that the first dice shows 4, given that the sum of the three numbers showing is 12?
Shikha from North Ryde
The probability of A given B is the probability of A and B divided by the probability of B. In this case the probability of rolling a 4 on the first die and then a total of 8 on the other two is (1/6)*(5/36) = 5/216. The probability of rolling any total of 12 with 3 dice is 25/216, as shown in my sic bo section. So the answer is (5/216)/(25/216) = 5/25 = 20%.
In a recent programming exercise myself and other students were asked to describe a sixsided die in code, and then use our dice to determine play simple game. The object of the game was to roll the dice until the sum of the tosses reached exactly 100. Any toss that put the total over 100 would not be added and merely added to statistics. Quickly it was determined that 17 throws would be the least amount of throws needed to reach 100. However calculating the odds of that occurring has proved elusive. Calculating the odds of a specific sequence of throws is rather straight forward, but how might one factor in both nonspecific ordering of throws, and the different ways of reaching 100 in 17 throws (16*6 + 1*4 and 15*6 + 2*5)?
from
The two ways you mention are the only ways to throw a total of 100 in 17 throws. The probability of throwing 16 sixes and one four is 17*(1/6)^{17}. There are 17 possible positions of the 4 and each sequence has a probability of (1/6)*(1/6)*...*(1/6) with 17 terms. The number of ways to get 15 sixes and 2 fives is combin(17,2) = 136. So the probability of 15 sixes and 2 fives is 136*(1/6)^{17}. So the total probability is (17+136)*(1/6)^{17}. = 1 in 110,631,761,077.
There are 3 dice, 2 are proper six sided dice, while one is a die with all sides containing a six. All the dice are in my pocket. I randomly take out a die and throw it. The result is a 6. What is the probability that the die was one of the proper dice with 6 different values?
Annojh from Toronto
Let A = Choosing the normal die
Let B = Rolling a 6 with randomly chosen die
Answer = Pr(A given B) = Pr(A and B)/pr(B) = ((2/3)*(1/6))/((2/3)*(1/6)+(1/3)*1) = (2/18)/((2/18)+(6/18)) = 1/4.
If you can roll six dice only once, what is the probability of rolling 6,6,6,6,1, and 4 in any order?
Aubrey from Kokomo
There are 6!/(4!*1!*1!) = 30 ways to arrange these numbers in any order. Another way to look at it is there are 6 positions to put the 1, and 5 left to put the 4, so 6*5=30. The probability of getting 666614 in exactly that order is 1 in 6^{6} = 1 in 46656. Multiply that by 30 for the 30 possible orders and the answer is 30/46656 = 0.0643%, or 1 in 1552.2.
My understanding of "wait time" for an event is the reciprocal of the probability of that event. I’m interested in calculating the wait time to roll consecutive 2s using one die. In a simulation I get 42 rolls on average. How do I make the connection with the probability of rolling consecutive 2s?
Lee from Andover
It is true that for single events if the probability is p then the average wait time is 1/p. However it gets more complicated with consecutive events. Let x be the state that the last roll was not a two. This is also the state at the beginning. Let y be the state that the last roll was a two. After the first roll there is a 5/6 chance we will still be in state x, and 1/6 chance we will be in state y. Let Ex(x) be the expected number of rolls from state x, and Ex(y) the expected number from state y. Then...
Ex(x) = 1 + (5/6)*ex(x) + (1/6)*ex(y), and
Ex(y) = 1 + (5/6)*ex(x)
Solving for these two equations...
Ex(x) = 1 + (5/6)*ex(x) + (1/6)*( 1 + (5/6)*Ex(x))
Ex(x) = 7/6 + (35/36)*Ex(x)
(1/36)*Ex(x) = 7/6
Ex(x) = 36*(7/6) = 42
So the average wait time for two consecutive twos is 42 rolls.
I have the same type of problem, only the expected flips to get two heads, in my site of math problems, see problem 128.
Can you tell me the odds of rolling two of the same number with two dice, three dice, and four dice? I am wondering how many dice would one have to roll at one time so that the odds are on the side of the person rolling the dice. (It does not make any difference which number is doubled.)
Mary from Minneapolis, MN
Here is the probability of getting at least one number more than once according to the number of rolls:
Probability of a Pair or More
Rolls  Probability 
2 rolls  16.67% 
3 rolls  44.44% 
4 rolls  72.22% 
5 rolls  90.74% 
6 rolls  98.46% 
What is the chance of getting a sum over 100, when rolling 20 dice? Kind regards
Terje from Stockholm
I started to use the Normal approximation to solve this, but the probability of over 100 points is too low for that method to be accurate. So I did a random simulation of 8.25 million trials and the number of trails that were 101 points or more was 127. So the probability is about 1 in 65,000.
Wizard, could you please describe the equivalent odds of the California SuperLotto Plus (1 in 41.4 Million), in terms of number of consecutive times of rolling 7 or 11? I heard it somewhere before. Most people cannot comprehend the lottery odds. But, the rolling of dice  they can relate.
Tim from Belmont, CA
Let n be your answer. The probability of rolling a 7 or 11 is 8/36. To solve for n:
(8/36)^{n} = 1/41,400,000
log((8/36)^{n}) = log(1/41,400,000)
n × log(8/36) = log(1/41,400,000)
n = log(1/41,400,000)/log(8/36)
n = 7.617 / 0.65321
n = 11.6608
So there you go, the probability of hitting the SuperLotto is the same as rolling a seven or eleven 11.66 times in a row. For those who can’t comprehend a partial throw I would rephrase as the probability falls between 11 and 12 consecutive rolls.
We are in a disagreement between workers. there is a bar down the street that has a shake a day. which is you must throw five dice at once and all five must end up being the same "like yahtzee" but he gives you three chance at it. but you must pick up all the dice all three times. so the questions is what’s the odds to do it in one shake and what’s the odds to do it in the three shakes allowed. Thanks , if you already answered this before i am sorry but i couldt find it.
Dan and co workers at maple island from Forest Lake
The probability of a fiveofakind on one throw is 6*(1/6)^{5} = 1/1,296. This is because there are six different fiveofakinds (one to six) and the probability each die will be that number is (1/6). The probability of not getting a fiveofakind is 1(1/1,296)=1,295/1,296. The probability of going three attempts without a three of a kind is (1,295/1,296)^{3}=99.77%. So the probability of getting at least one fiveofakind in three tries is 100%99.77% = 0.23%.
What is the expected number of tosses required in order to obtain at least one of each of the possible outcomes on an unbiased 6 sided dice?
Michael from Melbourne
If the probability of something is p then on average it will take 1/p trials for it to happen the first time. Obviously, on the first roll you’ll cross off one number. The probability of rolling one of the other five numbers next is 5/6. So it will take on average 1/(5/6)=6/5=1.2 rolls for that to happen. Following that reasoning to the end, the expected number of rolls is (6/6)+(6/5)+(6/4)+(6/3)+(6/2)+(6/1) = 14.7.
What is the classical probability of getting a total of 12 when 5 balanced dice are rolled?
Kennith H. from Winters
I hope you’re happy, I just added a new section answering questions such as this for 1 to 25 dice. As the fivedice table shows, the probability of rolling a total of 12 is 0.039223251028807.
Somebody bet he that he could roll a total of 6 and 8, with two dice, before I could roll a total of seven twice. This seemed like a good bet because seven is the most common total. However I lost $2500 doing this over and over. What are the odds?
Anthony from Indiana
I’m afraid you had the square side of this bet. The probability of rolling two sevens before a six and eight is 45.44%. Here are all the possible outcomes. The first column is the order of petintent rolls to the outcome of the bet, ignoring all others.
Two Sevens before Six and Eight Bet
Relavant Rolls  Probability  Formula  Outcome 
6,8  0.142045  (5/16)*(5/11)  Lose 
8,6  0.142045  (5/16)*(5/11)  Lose 
6,7,8  0.077479  (5/16)*(6/11)*(5/11)  Lose 
7,6,8  0.053267  (6/16)*(5/16)*(5/11)  Lose 
8,7,6  0.077479  (5/16)*(6/11)*(5/11)  Lose 
7,8,6  0.053267  (6/16)*(5/16)*(5/11)  Lose 
7,7  0.140625  (6/16)*(6/16)  Win 
6,7,7  0.092975  (5/16)*(6/11)*(6/11)  Win 
8,7,7  0.092975  (5/16)*(6/11)*(6/11)  Win 
7,6,7  0.06392  (6/16)*(5/16)*(6/11)  Win 
7,8,7  0.06392  (6/16)*(5/16)*(6/11)  Win 
Basically, the reason the 6 and 8 is the better side is you can hit those numbers in either order: 6 then 8, or 8 then 6. With two sevens there is only one order, a 7 and then another 7.
Two questions, please: 1) What is the probability of rolling 6,6,6,6,6,6 @ one time, with (6)6sided die? 2)What is the probability of rolling 1,2,3,4,5,6 @ one time with (6) 6sided die? Thanks! It’s killing me!
Heather from Petaluma
The probability of six sixes is (1/6)^{6} = 1 in 46656. The probability of rolling 1,2,3,4,5,6 with six dice is 6!/6^{6} = 1 in 64.8
What is the probability of getting any given number more than once if you roll a die ten times?
Ryan from Silay
1(5/6)^{10}10 × (1/6) × (5/6)^{9} = 51.55%.
In a game called Taxation and Evasion, a player rolls a pair of dice. On any roll if the sum is 7,11, or 12, the player gets audited; any other sum they avoid taxes. If a player rolls the pair 5 times, what is the probability that he avoids taxes?
George P. from Stevens Point, WI
The probability of a 7, 11, or 12 is (6+2+1)/36 = 9/36 = 1/4. See my section on dice probability basics for how I arrived at that figure. The probability of rolling anything else is 3/4. The probability of going five rolls without rolling a 7, 11, or 12 is (3/4)^{5} = 23.73%.
How many throws of a die does it take before it is likely that you have thrown a 1, 2, 3, 4, 5, and 6 at least once each? Any ideas on generalizing this for an nsided die?
Asif from Columbia, SC
Not that you asked, but let me address the mean first. For a sixsided die, the expected number of throws to get each face at least once is (6/6)+(6/5)+(6/4)+(6/3)+(6/2)+(6/1) = 14.7. For an nsided die the expected throws is (n/n) + (n/(n1)) + (n/(n2)) + ... + n. The median number of throws required is 13. The probability of taking 13 rolls or less is 51.4%, and 13 rolls or more is 56.21%.
I know you’re skeptical of dice control. I have been practicing dice setting and controlled shooting for 3 months. What is the probability of throwing 78 sevens over 655 throws randomly? Thanks for the help :)
Eric B. from Boston, MA
For large numbers of throws we can use the Gaussian Curve approximation. The expected number of sevens in 655 throws is 655 × (1/6) = 109.1667. The variance is 655 × (1/6) × (5/6) = 90.9722. The standard deviation is sqr(90.9722) = 9.5379. Your 78 sevens is 109.1667 − 78 = 31.1667 less than expectation. This is (31.1667  0.5)/9.5379 = 3.22 standard deviations below expectation. The probability of falling 3.22 or more standard deviations south of expectations is 0.000641, or 1 in 1,560. I got this figure in Excel, using the formula, normsdist(3.22).
This is about controlling the dice at Craps. You previously discussed the Stanford Wong Experiment, stating, "The terms of the bet were whether precision shooters could roll fewer than 79.5 sevens in 500 rolls of the dice. The expected number in a random game would be 83.33. The probability of rolling 79 or fewer sevens in 500 random rolls is 32.66%.... The probability of rolling 74 or fewer sevens in 500 random rolls is 14.41%."
The question I have about this bet is that 14.41% still isn’t "statistically significant" [ i.e. p < 0.05 ] , which is usually taken to mean greater than two Standard Deviations from the Mean  or a probability of less than a *combined* 5% of the event happening randomly on EITHER end of the series.
How many Sevens would have to be rolled in 500 rolls before you could say that there is a less than 2.5% chance that the outcome was entirely random (i.e. that the outcome was statistically significant) ?
Many Thanks & BTW , yours is ABSOLUTELY the BEST web site on the subject of gambling odds & probabilities that I’ve found .... keep up the good work !!!
Plexus from Warwick, Rhode Island
Thank you for the kind words. You should not state the probability that the throws were nonrandom is p. The way it should be phrased is the probability that a random game would produce such a result is p. Nobody expected 500 rolls to prove or disprove anything. It wasn’t I who set the line at 79.5 sevens, but I doubt it was chosen to be statistically significant; but rather, I suspect the it was a point at which both parties would agree to the bet.
The 2.5% level of significance is 1.96 standard deviations from expectations. This can be found with the formula =normsinv(0.025) in Excel. The standard deviation of 500 rolls is sqr(500*(1/6)*(5/6)) = 8.333. So 1.96 standard deviations is 1.96 * 8.333 = 16.333 rolls south of expectations. The expected number of sevens in 500 throws is 500*(1/6) = 83.333. So 1.96 standard deviations south of that is 83.333 − 16.333 = 67. Checking this using the binomial distribution, the exact probability of 67 or fewer sevens is 2.627%.
What is the expected number of rolls needed to get a Yahtzee?
Ian F. from Provo
Assuming the player always holds the most represented number, the average is 11.09. Here is a table showing the distribution of the number of rolls over a random simulation of 82.6 million trials.
Yahtzee Experiment
Rolls  Occurences  Probability 
1  63908  0.00077371 
2  977954  0.0118396 
3  2758635  0.0333975 
4  4504806  0.0545376 
5  5776444  0.0699327 
6  6491538  0.0785901 
7  6727992  0.0814527 
8  6601612  0.0799227 
9  6246388  0.0756221 
10  5741778  0.0695131 
11  5174553  0.0626459 
12  4591986  0.0555931 
13  4022755  0.0487016 
14  3492745  0.042285 
15  3008766  0.0364257 
16  2577969  0.0312103 
17  2193272  0.0265529 
18  1864107  0.0225679 
19  1575763  0.019077 
20  1329971  0.0161013 
21  1118788  0.0135446 
22  940519  0.0113864 
23  791107  0.00957757 
24  661672  0.00801056 
25  554937  0.00671837 
26  463901  0.00561624 
27  387339  0.00468933 
28  324079  0.00392347 
29  271321  0.00328476 
30  225978  0.00273581 
31  189012  0.00228828 
32  157709  0.00190931 
33  131845  0.00159619 
34  109592  0.00132678 
35  91327  0.00110565 
36  76216  0.00092271 
37  63433  0.00076795 
38  52786  0.00063906 
39  44122  0.00053417 
40  36785  0.00044534 
41  30834  0.00037329 
42  25494  0.00030864 
43  21170  0.0002563 
44  17767  0.0002151 
45  14657  0.00017745 
46  12410  0.00015024 
47  10299  0.00012469 
48  8666  0.00010492 
49  7355  0.00008904 
50  5901  0.00007144 
51  5017  0.00006074 
52  4227  0.00005117 
53  3452  0.00004179 
54  2888  0.00003496 
55  2470  0.0000299 
56  2012  0.00002436 
57  1626  0.00001969 
58  1391  0.00001684 
59  1135  0.00001374 
60  924  0.00001119 
61  840  0.00001017 
62  694  0.0000084 
63  534  0.00000646 
64  498  0.00000603 
65  372  0.0000045 
66  316  0.00000383 
67  286  0.00000346 
68  224  0.00000271 
69  197  0.00000238 
70  160  0.00000194 
71  125  0.00000151 
72  86  0.00000104 
73  79  0.00000096 
74  94  0.00000114 
75  70  0.00000085 
76  64  0.00000077 
77  38  0.00000046 
78  42  0.00000051 
79  27  0.00000033 
80  33  0.0000004 
81  16  0.00000019 
82  18  0.00000022 
83  19  0.00000023 
84  14  0.00000017 
85  6  0.00000007 
86  4  0.00000005 
87  9  0.00000011 
88  4  0.00000005 
89  5  0.00000006 
90  5  0.00000006 
91  1  0.00000001 
92  6  0.00000007 
93  1  0.00000001 
94  3  0.00000004 
95  1  0.00000001 
96  1  0.00000001 
97  2  0.00000002 
102  1  0.00000001 
Total  82600000  1 
Consider a hypothetical game based on the roll of a die. If die lands on 1, the player loses $1 and the game ends. If the die lands on anything else, the player’s wins $1. At this point the player may let it ride, or quit. The player may keep playing, doubling each bet, until he loses or quits. What is the best strategy?
Byron P. from Newington, CT
Speaking only in terms of maximizing expected value, the player should play forever. While the probability is 1 that the player will eventually lose, at any given decision point the expected value always favors going again. It seems like a paradox. The answer lies in the fact that some events have a probability of 1, but still may not happen. For example, if you threw a dart at a number line from 0 to 10, the probability of not hitting pi exactly is 1, but it still could happen.
However, for practical purposes, there is some stopping point. This is because the happiness money brings is not proportional to the amount. While it is commonly accepted that more money brings more happiness, the richer you get, the less happiness each additional dollar brings you.
I believe a good way to answer this question is to apply the Kelly Criterion to the problem. According to Kelly, the player should make every decision with the goal of maximizing the expected log of his bankroll after the wager. To cut to the end of this (I cut out a lot of math), the player should keep doubling until the wager amount exceeds 96.5948% of his total wealth. Wealth should be defined as the sum of the amount won plus whatever money the player had before he made the first wager. For example, if the player had $100,000 to start with, he should keep doubling up to 23 times, to a win of $4,194,304. At that point the player’s total wealth will be $4,294,304. He will be asked to wager 4,194,304/4,294,304 = 96.67% of his total wealth, which is greater than the 96.5948% stopping point, so he should quit.
Players A and B throw a pair of dice. Player A wins if he throws a total of 6 before B throws a toal of 7, and B wins if he throws 7 before A throws 6. If A begins, show that A’s chances of winning are 30/61.
Sangeeta from Mumbai, India
Let the answer to this question be called p. The probability of rolling a total of six is 5/36, and the probability of rolling a total of seven is 6/36. If you don’t understand why, please see my section on dice probability basics. We can define p as:
p = Prob(6 on first roll) + Prob(no 6 on first roll)*Prob(no 7 on second roll)*p.
This is because, if neither player wins after the first two rolls, the game is back to the original state, and the probability of player A winning remains the same.
So, we have:
p = (5/36) + (31/36)×(30/36)×p
p = 5/36 + (930/1296)×p
p * (1(930/1296)) = 5/36.
p * (366/1296) = 5/36
p = (5/36)×(1296/366) = 30/61.
How many ways are there to roll n sixsided, nondistinct dice? As stated, the dice are nondistinct, so with five dice, for example, 11356 and 16513 would be considered the same roll. With two dice, it’s trivial to determine that the answer is 21, but I can’t figure out an elegant, generalized solution.
Don from New York
The answer can be expressed as combin(n+5,n) = (n+5)!/(120×n!). Here is the answer for 1 to 20 dice.
NonDistinct Dice Combinations
Dice  Combinations 
1  6 
2  21 
3  56 
4  126 
5  252 
6  462 
7  792 
8  1287 
9  2002 
10  3003 
11  4368 
12  6188 
13  8568 
14  11628 
15  15504 
16  20349 
17  26334 
18  33649 
19  42504 
20  53130 
21  65780 
22  80730 
23  98280 
Credit to Alan Tucker, author of Applied Combinatorics.
Can you calculate what the probability is of two numbers coming up behind each other in a roll of the dice? Meaning what is the probability of two 4’s or two 6’s or two 7’s back to back? I realize that the past cannot predict the future but is there a way to calculate 7/36 X 7/36 happening back to back? I hope that makes sence.
James from Birmingham
Sure. That would be Pr(2)^{2} + Pr(3)^{2} + ... + Pr(12)^{2} = (1/36)^{2} + (2/36)^{2} + (3/36)^{2} + (4/36)^{2} + (5/36)^{2} + (6/36)^{2} + (5/36)^{2} + (4/36)^{2} + (3/36)^{2} + (2/36)^{2} + (1/36)^{2} = 11.27%.
In 180 consecutive rolls of the dice, how many times can I expect to see the following:
Two sevens in a row?
Three sevens in a row?
Four sevens in a row?
Thanks for your time :).
Melanie D. from Elizabeth City, NC
I can’t think of any useful reason to know this information, but I get asked this kind of thing a lot, so I’ll humor you.
It is a little easier getting a specified sequence of sevens starting with the first roll, or ending with the last, because the sequence is bounded on one side. Specifically, the probability of getting a sequence of s sevens, starting with the first roll, or ending with the last, is (1/6)^{s} × (5/6). The 5/6 term is because you have to get a non7 at the open end of the sequence.
The probability of starting a sequence of s sevens at any point in the middle of the sequence is (1/6)^{s} × (5/6)^{2}. We square the 5/6 term, because the player must get a non7 on both ends of the sequence.
If there are r rolls, there will be 2 places for an inside sequence, and rn1 places for a run of n sevens. Putting these equations in a table, here is the expected number of runs of sevens, from 1 to 10. The "inside" column is 2*(5/6)*(1/6)^{r}, and the "outside" column is (179r)*(5/6)^{2}*(1/6)^{r}, where r is the number of sevens in the run. So, we can expect 3.46 runs of two sevens, 0.57 runs of three sevens, and 0.10 runs of four sevens.
Expected Runs of Sevens in 180 Rolls
Run  Inside  Outside  Total 
1  0.277778  20.601852  20.87963 
2  0.046296  3.414352  3.460648 
3  0.007716  0.565844  0.57356 
4  0.001286  0.093771  0.095057 
5  0.000214  0.015539  0.015754 
6  0.000036  0.002575  0.002611 
7  0.000006  0.000427  0.000433 
8  0.000001  0.000071  0.000072 
9  0  0.000012  0.000012 
10  0  0.000002  0.000002 
Two dice are rolled until either a total of 12 or two consecutive totals of 7. What is the probability the 12 is rolled first?
Anonymous
The answer and solution can be found on my companion site, mathproblems.info, problem 201.
I’m a tabletop gamer, and was having some discussion with my friends about noncubical platonic solid dice (If you’re a big enough nerd, that means d4, d8, d12, and d20). They argued that those would be the only ones that would be demonstratively fair. I argued that manufacturing them to be fair would be entirely too difficult. Also, the only games would be craps variants rendered overly cumbersome due to the number of extra outcomes. Has any casino ever had a game that used nontraditional six sided dice?
Bayani from Carnagie, PA

If you limit yourself to the regular polygons, and want every face to have the same probability, then you are limited to the platonic solids. However, if you can lift the regular polygon requirement, then you can add the 13 Catalan solids as well.
To answer your other question, no, I have never seen a game actually in a casino that used any dice other than cubes. About ten years ago I saw a game demonstrated at a gaming show in Atlantic City that I think used a Rhombic triacontahedron, one of the Catalan solids, but I don’t think it ever made it to a casino floor. There is a game I see year after year at the Global Gaming Expo that uses a spinning top (like a dreidel), but alas, I’ve never seen that in a casino either.
If I roll three sixsided dice, what are the odds of rolling a straight and, also, what are the odds of rolling a three of a kind?
Mark from Fargo, ND
There are 6^{3}=216 ways to roll three dice. Six of those combinations will result in a three of a kind (111 to 666). So the probability of a three of a kind is 6/216 = 1/36. There are four possible spans for a straight (123 to 456). There are also 3!=6 ways to arrange the three dice in a straight. So, there are 4*6=24 straight combinations. Thus the probability of a straight is 24/216 = 1/9.
What is the average sum when rolling four sixsided dice after subtracting the lowest result (known as 4d6L)? What is the standard deviation for this roll?
Aaron from New York
The following table shows the number of combinations for all possible totals from 3 to 18.
Combinations in 4d6L
Outcome  Combinations 
3  1 
4  4 
5  10 
6  21 
7  38 
8  62 
9  91 
10  122 
11  148 
12  167 
13  172 
14  160 
15  131 
16  94 
17  54 
18  21 
Total  1296 
The mean result is 12.2446, and the standard deviation is 2.8468.
My question is based on dice odds. I know that there are six ways to get 7 and one way to get 12, but what are the chances of getting six 7’s before one 12? Are they even, and if not, how many twelves should be added to the equation to make it an even proposition?
nick
The probability of rolling a 7 is 1/6, and the probability of rolling a 12 is 1/36. The probability of rolling a 7, given that a roll is a 7 or 12 is (1/6)/((1/6)+(1/36)) = 6/7. So the probability that the first six times a 6 or 12 is rolled it is a 6 every time is (6/7)^{6} = 39.66%.
If you rephrase the question to be what is the probability of rolling five 6’s before a 12, then the answer is (6/7)^{5} = 46.27%. With four rolls it is (6/7)^{4} = 53.98%. So there is no number of 7’s before a 12 that is exactly 50/50. If you’re looking for a good sucker bet, suggest you can either roll four 7’s before a 12, or a 12 before five 7’s.
This question was raised and discussed in the forum of my companion site Wizard of Vegas.
Is there an easy way to calculate the probability of throwing a total of t with d 6sided dice?
Anon E. Mouse
Here is a handy trick, courtesy of Robert Goodhand of Somerset, UK. First put on a row six ones surrounded by five zeros on either side, as follows:
OneDie Probabilities
Dice Total  1  2  3  4  5  6  
One Die  0  0  0  0  0  1  1  1  1  1  1  0  0  0  0  0 
This represents the number of combinations for rolling a 1 to 6 with one die. I know, pretty obvious. However, stick with me. For two dice, add another row to the bottom, and for each cell take the sum of the row above and the five cells to the left of it. Then add another five dummy zeros to the right, if you wish to keep going. This represents the combinations of rolling a total of 2 to 12.
Two Dice Probabilities
Dice Total  2  3  4  5  6  7  8  9  10  11  12  
One Die  0  0  0  0  0  1  1  1  1  1  1  0  0  0  0  0  0  0  0  0  0 
Two Dice  0  0  0  0  0  1  2  3  4  5  6  5  4  3  2  1  0  0  0  0  0 
For three dice, just repeat. This will represent the number of combinations of 3 to 18.
To get the probability of any given total, divide the number of combinations of that total by the total number of combinations. In the case of three dice, the sum is 216, which also easily found as 6^{3}. For example, the probability of rolling a total of 13 with three dice is 21/216 = 9.72%.
So for d dice, you’ll need to work your way up through 1 to d1 dice. This is very easily accomplished in any spreadsheet.
How many rolls of two dice would it take to have a 50/50 chance of rolling at least one 12?
Maff
That is a classic problem in the history of the field of probability. Many people incorrectly think the answer is 18, because the probability of a 12 is 1 in 36, and 18×(1/36)=50%. However, by that logic, the probability of getting a 12 in 36 rolls would be 100%, which clearly it isn’t. Here is the correct solution. Let r be the number of rolls. The probability a roll isn’t a 12 is 35/36. The probability there are 0 12s in r rolls is (35/36)^{r}. So we need to solve for r in the following equation:
(35/36)^{r} = 0.5
log(35/36)^{r} = log(0.5)
r × log(35/36) = log(0.5)
r = log(0.5)/log(35/36)
r = 24.6051
So there isn’t a round answer. The probability of rolling a 12 in 24 rolls is 1(35/36)^{24} = 49.14%. The probability of rolling a 12 in 25 rolls is 1(35/36)^{25} = 50.55%.
If you want to make a bet on this, say you can roll a 12 in 25 rolls, or somebody else can’t in 24 rolls. Either way you’ll have an advantage at even money.
In Dice Wars, what is the probability of success for any given number of attacking and defending dice? As an attacker, what ratio has the greatest expected gain?
Anonymous
For those unfamiliar with the game, both the attacker and defender will roll 1 to 8 dice, according to how many armies they each have at that point in a battle. The higher total shall win. A tie goes to the defender. If the attacker loses, he will still retain one army in the territory where he initiated the attack. For this reason, he must have at least two armies to attack, so if he wins one can inhabit the conquered territory and one can stay behind.
The following table shows the probability of an attacker victory according to all 64 combinations of total dice.
Probability of Attacker Win
Attacker  Defender  

1 Army  2 Armies  3 Armies  4 Armies  5 Armies  6 Armies  7 Armies  8 Armies  
2  0.837963  0.443673  0.152006  0.035880  0.006105  0.000766  0.000071  0.000005 
3  0.972994  0.778549  0.453575  0.191701  0.060713  0.014879  0.002890  0.000452 
4  0.997299  0.939236  0.742831  0.459528  0.220442  0.083423  0.025450  0.006379 
5  0.999850  0.987940  0.909347  0.718078  0.463654  0.242449  0.103626  0.036742 
6  0.999996  0.998217  0.975300  0.883953  0.699616  0.466731  0.259984  0.121507 
7  1.000000  0.999801  0.994663  0.961536  0.862377  0.685165  0.469139  0.274376 
8  1.000000  0.999983  0.999069  0.989534  0.947731  0.843874  0.673456  0.471091 
The next table shows the expected gain by the attacker, defined as pr(attacker wins)*(defender dice)+pr(defender wins)*(attacker dice 1). It shows the greatest expected gain is to attack with 8 against an opponent with 5.
Net Gain of Attacker Win
Attacker  Defender  

1 Army  2 Armies  3 Armies  4 Armies  5 Armies  6 Armies  7 Armies  8 Armies  
2  0.675926  0.331019  0.391976  0.820600  0.963370  0.994638  0.999432  0.999955 
3  0.918982  1.114196  0.267875  0.849794  1.575009  1.880968  1.973990  1.995480 
4  0.989196  1.696180  1.456986  0.216696  1.236464  2.249193  2.745500  2.929831 
5  0.999250  1.927640  2.365429  1.744624  0.172886  1.575510  2.860114  3.559096 
6  0.999976  1.987519  2.802400  2.955577  1.996160  0.134041  1.880192  3.420409 
7  1.000000  1.998408  2.951967  3.615360  3.486147  2.221980  0.098807  2.158736 
8  1.000000  1.999847  2.990690  3.884874  4.372772  3.970362  2.428384  0.066365 
What is the probability of forming a Yahtzee with up to n rolls of the dice?
Anonymous
For the benefit of other readers, a Yahtzee is a five of a kind with five dice. In the game of Yahtzee the player may hold any dice he wishes and reroll the rest. He can do this up to three rolls.
The player may reroll previously held dice, if he wishes. For example, if the player's first roll is 33456 and he holds the threes and then has 33555 after the second roll he may keep the fives and reroll the threes on his third roll.
The following table shows the maximum number of dice of the same face for 1 to 20 rolls. The table shows the probability of getting a Yahtzee within three rolls is about 4.6%.
Yahtzee Probabilities
Rolls  Maximum Dice of Same Face  

One  Two  Three  Four  Five  
1  0.092593  0.694444  0.192901  0.019290  0.000772 
2  0.008573  0.450103  0.409022  0.119670  0.012631 
3  0.000794  0.256011  0.452402  0.244765  0.046029 
4  0.000074  0.142780  0.409140  0.347432  0.100575 
5  0.000007  0.079373  0.337020  0.413093  0.170507 
6  0.000001  0.044101  0.263441  0.443373  0.249085 
7  0.000000  0.024501  0.199279  0.445718  0.330502 
8  0.000000  0.013612  0.147462  0.428488  0.410438 
9  0.000000  0.007562  0.107446  0.398981  0.486011 
10  0.000000  0.004201  0.077416  0.362855  0.555528 
11  0.000000  0.002334  0.055317  0.324175  0.618174 
12  0.000000  0.001297  0.039279  0.285674  0.673750 
13  0.000000  0.000720  0.027757  0.249063  0.722460 
14  0.000000  0.000400  0.019543  0.215313  0.764744 
15  0.000000  0.000222  0.013720  0.184883  0.801175 
16  0.000000  0.000124  0.009610  0.157896  0.832371 
17  0.000000  0.000069  0.006719  0.134258  0.858954 
18  0.000000  0.000038  0.004692  0.113753  0.881517 
19  0.000000  0.000021  0.003272  0.096100  0.900607 
20  0.000000  0.000012  0.002280  0.080994  0.916714 
This question is raised and discussed in my forum at Wizard of Vegas.
I am wondering which will come up more rolling a pair of dice — an odd or even total?
relaxmax
The answer is 50/50. This will be true for any number of dice rolled, not just two.
A bit offtopic, but I've always thought an odd/even set of bets would be a good way to replace the dreaded big 6/8 bets in craps. To give the house an advantage, here are my proposed pay tables and analysis.
Odd Bet
Event  Pays  Combinations  Probability  Return 

3 or 11  1.5  4  0.111111  0.166667 
5 or 9  1  8  0.222222  0.222222 
7  0.5  6  0.166667  0.083333 
Even  1  18  0.500000  0.500000 
Total  36  1.000000  0.027778 
Even Bet
Event  Pays  Combinations  Probability  Return 

2 or 12  3  2  0.055556  0.166667 
4 or 10  1  6  0.166667  0.166667 
6 or 8  0.5  10  0.277778  0.138889 
Odd  1  18  0.500000  0.500000 
Total  36  1.000000  0.027778 
Please note that I claim all rights with this publication.
This question is raised and discussed in my forum at Wizard of Vegas.