Ana and Bastian

Ana tosses a fair coin 50 times. Bastian tosses another fair coin 51 times.

What's the probability that Bastian gets more heads than Ana does?

50 101 \frac{50}{101} 1 2 \frac{1}{2} 51 101 \frac{51}{101} 13 25 \frac{13}{25}

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

14 solutions

Jordan Cahn
Dec 4, 2018

Since Bastian only flips one more coin than Ana, he cannot have both more heads and more tails, but must have more of one. So P ( Bastian flips more heads ) + P ( Bastian flips more tails ) = 1 P(\text{Bastian flips more heads}) + P(\text{Bastian flips more tails}) = 1

But heads and tails are equally likely results for both players, the probability of Bastian getting more heads than Ana must be the same as the probability of Bastian getting more tails than Ana. Thus, the probability of either is 1 2 \boxed{\dfrac{1}{2}} .

A math problem in which logic provides the clearest explanation - thank you Jordan

jason walker - 2 years, 6 months ago

Thanks ... makes total sense, but I do wonder what the probability is that Bastian and Ana have the same number of heads. That seems like it might be horrendous to calculate.

John Williamson - 2 years, 6 months ago

Log in to reply

The probability that Ana flips k k heads is ( 50 k ) 2 50 \dfrac{{50 \choose k}}{2^{50}} and the probability that Bastian flips k k heads is ( 51 k ) 2 51 \dfrac{51 \choose k}{2^{51}} . Thus, the probability that they flip the same number of heads is k = 0 50 ( 50 k ) ( 51 k ) 2 101 0.079 \sum_{k=0}^{50} \frac{{50 \choose k}{51 \choose k}}{2^{101}} \approx 0.079

Jordan Cahn - 2 years, 6 months ago

Log in to reply

For an event that seems pretty probable, that's surprisingly low!

Parth Sankhe - 2 years, 6 months ago

Log in to reply

@Parth Sankhe 50 coins is a lot of coins! That's about an 8% chance, so it's saying, roughly, "if we both flip 50 coins, there's a little less than a 1 in 12 chance we end up with the same number of heads." That doesn't seem unreasonably small to me.

Jordan Cahn - 2 years, 6 months ago

@Parth Sankhe There is an 11.23% chance that Ana gets 25 heads, and an 11.01% chance that Bastian gets 25, so the probability that they both get 25 heads is the product of these two chances, or 1.24%

All the other probabilities will come out with lower values. By the time we add them together, I'm not surprised we'd end up with 7.9% overall probability.

John Williamson - 2 years, 6 months ago

Bonus: What's the probability that Bastian gets more heads than Ana does if he tosses his fair coin 52 times?

Uros Stojkovic - 2 years, 6 months ago

Log in to reply

a probability of 0.75

Peter Anderson - 2 years, 6 months ago

Log in to reply

See my other comment... it's a little under 0.54.

Think of a smaller example: if Ana flips one coin and Bastian flips three, the probability of Bastian flipping more heads is 1 2 × 7 8 + 1 2 × 1 2 = 7 16 + 4 16 = 11 16 = 0.6875 \frac{1}{2}\times\frac{7}{8} + \frac{1}{2}\times\frac{1}{2} = \frac{7}{16} + \frac{4}{16} = \frac{11}{16} = 0.6875 .

With more coins, the probability gets closer to 1 2 \frac{1}{2} -- in the limit, there's not such a big difference between n n and n + 2 n+2 . If I'm flipping 1 , 000 , 000 1{,}000{,}000 coins and you're flipping 1 , 000 , 002 1{,}000{,}002 coins, the probability of you flipping more heads than me is going to be pretty darn close to 1 2 \frac{1}{2} .

Jordan Cahn - 2 years, 6 months ago

k = 1 52 ( 52 k ) 2 52 j = 0 k 1 ( 50 j ) 2 50 = k = 1 52 j = 0 k 1 ( 52 k ) ( 50 j ) 2 102 0.539 \sum_{k=1}^{52}\frac{{52 \choose k}}{2^{52}}\sum_{j=0}^{k-1}\frac{{50 \choose j}}{2^{50}} = \sum_{k=1}^{52}\sum_{j=0}^{k-1}\frac{{52 \choose k}{50 \choose j}}{2^{102}} \approx 0.539

Jordan Cahn - 2 years, 6 months ago

Log in to reply

Correct! There is actually an alternative way to get the right result which is computationally much cheaper. It exploits the fact that binomial distribution tends to normal distribution for the large number of Bernoulli trials. See Arjen's solution to this problem .

Uros Stojkovic - 2 years, 6 months ago

These solutions don't make sense to me. Bastion has a greater probability of more heads because he tosses more coins. His probability goes up from its original probability by 1/2. Look at it from Ana's viewpoint. When Bastion flips a T, nothing changes. When he flips an H she looses all of the ties, and gets tied in the ones where she was a flip ahead.

Bob Christenson - 2 years, 6 months ago

Log in to reply

Rather than reiterate my argument, I'll ask this: if the probability of Bastian flipping more heads is greater than 1/2, what is the probability of Bastian flipping more tails?

Jordan Cahn - 2 years, 6 months ago

Log in to reply

It is the same probability but it doesn't matter. Ana doesn't get to flip another coin. And we're only counting heads.Tails don't penalize you. So whenever Bastion flips a coin he improves his probability of getting another head.His score can improve- hers can't.

Bob Christenson - 2 years, 6 months ago

Log in to reply

@Bob Christenson If the probability of Bastian flipping more heads is greater than 1/2, then in more than 1/2 of the possible outcomes Bastian flips more heads.

Similarly, if the probability of Bastian flipping more tails is greater than 1/2, then in more than 1/2 of the possible outcomes Bastian flips more tails.

But that means there must be an outcome in which Bastian flips more heads and more tails, which is impossible when he only flips one more coin! This seemingly paradoxical result comes from the fact that he only flips one more coin than Ana -- if he flipped 2+ more coins, you'd absolutely be right that the probability is greater than 1/2.


I'd also suggest reading @Brian Charlesworth 's solution below. His approach of looking at Bastian's first 50 throws plus one more seems more in line with your way of approaching the problem. Maybe that solution will help clarify things.

Jordan Cahn - 2 years, 6 months ago

Log in to reply

@Jordan Cahn Jordan, The number of coins doesn't matter. I'll make you a wager. You put up $50 in dollar coins and I'll put up $51. Winner takes all unless we tie and then we'll split it. Lets say that after 50 tosses we're dead even on heads flipped. I then pull out my last coin (which you don't have one of) and flip it. If it's tails then we still both still have the same number of heads so we each get $50.50 and I'm out .50. If it turns up heads then I get the whole $101.00.

Bob Christenson - 2 years, 6 months ago

Log in to reply

@Bob Christenson You just introduced three situations: You win, I win, even split. Yet the question doesn't ask "is it more likely that Bastian has more heads or has fewer heads?" It asks "what is the probability that Bastian has more heads?"

In your game, there is a 1 2 \frac{1}{2} chance that you have more heads and a 1 2 \frac{1}{2} chance that you don't. You win 1 2 \frac{1}{2} of the time, which is precisely what the question is asking.

Jordan Cahn - 2 years, 6 months ago

@Bob Christenson Except that's not right.

The question doesn't ask "Who will have more heads?" The question asks "Will Bastion have MORE heads?"

If that 51st flip is heads, Bastion wins. If that 51st flip is tails, Bastion does not win. A tie means Bastion does not have more heads, which is what the question is asking.

Rick Desilets - 2 years, 5 months ago

I understand the logic, thus, let's say Ana tosses only once, and Bastian 1000 times...I mean, how can it be that the result (1/2) is independent from the difference of tosses between Ana and Bastian? The probability for Bastian to get more than once head is 1-1/2^1000, so even if Ana gets head, Bastian has 1-1/2^1000 more chances to "win", not 1/2..

Nicola personal - 2 years, 6 months ago

Log in to reply

@Nicola Personal The key to this problem is that Bastian only flips one more time than Ana. Once he flips 2+ times more than Ana, the probability becomes greater than 1/2.

Jordan Cahn - 2 years, 6 months ago

Log in to reply

@Jordan Cahn ah, ok...would there be a formula expressing it against "n", where "n" is the difference of tosses?

Nicola personal - 2 years, 6 months ago

Log in to reply

@Nicola Personal Not a pretty one... k = 1 N + n ( N + n k ) 2 N + n j = 0 k 1 ( N j ) 2 N = k = 1 N + n j = 0 k 1 ( N + n k ) ( N j ) 2 2 N + n \sum_{k=1}^{N+n}\frac{{N+n \choose k}}{2^{N+n}}\sum_{j=0}^{k-1}\frac{{N \choose j}}{2^{N}} = \sum_{k=1}^{N+n}\sum_{j=0}^{k-1}\frac{{N+n \choose k}{N \choose j}}{2^{2N+n}} where N N is the number of tosses the first player throws and n n is the difference in tosses between the first and second players.

If someone knows a simpler formula, feel free to correct me.

Jordan Cahn - 2 years, 6 months ago

I thought that at first, but we're ignoring the probability of equal numbers of heads

Hugh Wallis - 2 years, 6 months ago

The answer is NOT 1/2. I do understand why they claim the answer is 1/2, but they the wording of the problem would need significant changes for that to be correct in this case. Other players have given different justifications for the answer 1/2, but most of them proceed from the same fundamental flaw:

  1. When Bastian and Ana have both flipped 50 coins, the probability that Bastian will have gotten more heads than Ana is 1/2. TRUE.

  2. When Bastian then flips his 51st coin, he has a 50% chance of getting heads. TRUE

  3. That means that the probability that Bastian will get more heads after 51 flips than Ana would after 50 flips is 1/2 as well. FALSE

That is not how probability works. A lot of people, including whoever wrote the problem, are getting tripped up by the "probability has no memory" thing. If you flip a coin 3 times and get 3 heads, the probability that you'll get a 4th head in a row when you flip again is obviously 50%, not 1/16. However, before you started flipping, the probability that you would flip 4 heads in a row WAS 1/16.

This problem only specifies that Ana flips 50 coins and Bastian flips 51 coins. It doesn't say that Bastian and Ana both flipped 50 coins, got 25 heads each, and then Bastian flipped another one. In that case, the probability that he would get more heads than Ana would be 50%. (You would need to add the part about them getting 25 heads each for that to work.)

The best way I can explain this is by asking you to consider an empirical simulation. If you repeated this scenario (the 50/51 coin version) 10000 times and plotted the number of heads that Ana got for each trial on the X axis and the frequency of each number on the Y axis, you would get a normal distribution with a peak (mean) of 25, with 50% of the distribution in each side of the mean. 50%, because for each trial, the probability that Ana will get more heads than 25 is 1/2. The same is true with the probability that she'll get fewer than 25, too. Bastians distribution would be the same except it would be centered on 25.5 and would have a slightly greater spread. This means that every point greater than or equal to 25 on Ana's distribution trendline would have a lower y value than the point on Bastian's trendline with the same x value. That means that most of the time this simulation is run, the Bastian would get more heads than Ana and that the probability that he'll get more heads than her is greater than 1/2. I wrote a simple computer program that carries out the scenario in this problem 10000 times and outputs the total number of 50/51 flip games for which each player got more heads than the other. I could post it somewhere if anyone's interested.

Another way to explain this is to use smaller numbers. It's not going to be a different kind of scenario and there's no logical reason that using lower numbers should make more sense, but I think many people will quickly see how obviously incorrect the answer of 1/2 is. If Ana flips 2 coins and Bastian flips 3, what is the probability that Bastian will get more heads than Ana? It should be obvious that Bastian has a greater chance than Ana does.

A Former Brilliant Member - 2 years, 5 months ago

Log in to reply

When Bastian and Ana have both flipped 50 coins, the probability that Bastian will have gotten more heads than Ana is 1/2. TRUE.

This is your first flaw. When they both flip 50 coins, the probability that Bastian will have more heads is less than 1/2. This is because he and Ana might have the same number of heads.

You are absolutely right, however, that reasoning from 50 coins and then adding another coin can be dangerous (although @Brian Charlesworth has a neat solution that avoids the pitfalls you mention). But even setting aside that reasoning, the probability is still 1/2. I'd be interested in seeing your computer program that suggests otherwise.

The crux of the problem comes down to Bastian flipping exactly one more coin than Ana: this makes it impossible for him to flip the same number of heads and tails as Ana. It also makes it impossible for him to flip more of both. So he has to flip more heads or more tails, leading to the conclusion of 1/2. As you suggested, let's consider a smaller number of coins:

Bastian flips three coins, Ana flips two. We can now consider all the scenarios:

  • Ana flips 0 heads with probability 1 4 \frac{1}{4} . Bastian flips 1+ heads with probability 7 8 \frac{7}{8}
  • Ana flips 1 head with probability 1 2 \frac{1}{2} . Bastian flips 2+ heads with probability 1 2 \frac{1}{2}
  • Ana flips 2 heads with probability 1 4 \frac{1}{4} . Bastian flips 3 heads with probability 1 8 \frac{1}{8} 1 4 7 8 + 1 2 1 2 + 1 4 1 8 = 7 32 + 1 4 + 1 32 = 1 2 \frac{1}{4}\cdot\frac{7}{8} + \frac{1}{2}\cdot\frac{1}{2} + \frac{1}{4}\cdot\frac{1}{8} = \frac{7}{32} + \frac{1}{4} + \frac{1}{32} = \frac{1}{2} As you can see, we get a probability of 1/2.

Jordan Cahn - 2 years, 5 months ago

Log in to reply

You're making the same problem I pointed out above so I'm just going to copy my response. Doesn't completely fit what you wrote, but it should still explain where you're going wrong if you can extrapolate a little.

It's not a contradiction. It would be a contradiction to say that Bastian has greater than a 50% chance of getting more tails than heads or more heads than tails in HIS set of 51 coin flips. The question is comparing the outcome of 51 coin flips to the outcome of an entirely different set of 50 coin flips. I don't remember what the percentage is, but say he has a 51% chance of getting more heads than Ana does. That means that she has a 49% chance of getting more heads than he has. That's the number that has to add up to 100%, not the probability that he'll get more heads than her and the probability that he'll get more tails than her. And the same would be true of tails: he would have a 51% chance of getting more tails than her, too. If you changed the question to ask whether he would end up with a greater proportion of heads that he flipped to tails that he flipped, than the proportion of heads that Ana flipped to tails that Ana flipped, then the probability would be 50%, but the question asks about the NUMBER of heads he'll get, not about the proportion of heads to tails within his set of 51 coin flips. Also see my answer to the other comment.

A Former Brilliant Member - 2 years, 5 months ago

Log in to reply

@A Former Brilliant Member When Bastian flips one more coin than Ana, there are only two possibilities, which are mutually exclusive:

  • Bastian flips more heads than Ana
  • Bastian flips more tails than Ana

Do you disagree? If you do, I suggest considering the 2 coin/3 coin example to see that these are indeed the only possible outcomes.

Jordan Cahn - 2 years, 5 months ago

I'll just ask you one simple question: can you say the same for tails? If you can, and I can't think of reason why you couldn't, you will end up saying that Bastian gets more heads and tails more than half the time - contradiction.

Uros Stojkovic - 2 years, 5 months ago

Log in to reply

Quick note: this wouldn't be a contradiction if Bastian flipped more than 1 more coin. But, in the given situation, it is!

Jordan Cahn - 2 years, 5 months ago

Log in to reply

@Jordan Cahn Of course, but in the described situation it's either or.

Uros Stojkovic - 2 years, 5 months ago

It's not a contradiction. It would be a contradiction to say that Bastian has greater than a 50% chance of getting more tails than heads or more heads than tails in HIS set of 51 coin flips. The question is comparing the outcome of 51 coin flips to the outcome of an entirely different set of 50 coin flips. I don't remember what the percentage is, but say he has a 51% chance of getting more heads than Ana does. That means that she has a 49% chance of getting more heads than he has. That's the number that has to add up to 100%, not the probability that he'll get more heads than her and the probability that he'll get more tails than her. And the same would be true of tails: he would have a 51% chance of getting more tails than her, too. If you changed the question to ask whether he would end up with a greater proportion of heads that he flipped to tails that he flipped, than the proportion of heads that Ana flipped to tails that Ana flipped, then the probability would be 50%, but the question asks about the NUMBER of heads he'll get, not about the proportion of heads to tails within his set of 51 coin flips. Also see my answer to the other comment.

A Former Brilliant Member - 2 years, 5 months ago

Log in to reply

@A Former Brilliant Member I hope we agree that there are two possible situations: Bastian gets more heads than Ana or Bastian doesn't get more heads than Ana - probabilities of this two sum up to 1. Since he's flipping one coin more than Ana and since the result of a single coin flip is either heads or tails, if he doesn't get more heads than Ana, it must be that he gets more tails than Ana. Hence, we have that the probability of getting more heads and getting more tails than Ana sum up to 1. Finally, there's no difference between getting heads and tails - everything that applies for heads, applies for tails too, the only possible probability is 50%.

Uros Stojkovic - 2 years, 5 months ago

True, the probability of B throwing more heads is greater than A's. But also true, the probability of B throwing more tails than A is true.

David Stupple - 2 years, 4 months ago

your reasoning if completely flawed since you're not considering the case with equal number of heads in the 50 coins each scenario.. pretty much everything you said in that long post is wrong, and the computer code you wrote must be wrong too

ric vega - 2 years, 5 months ago

Log in to reply

I left it out of my explanation because my answer was long enough already. I didn't leave it out of my computer program. It simply outputs the number of times that Bastian had more heads than Ana and divides that by the total number of trials. The inverse of the resulting proportion is the probability that Bastian will have less than or equal to heads than Ana, meaning that my explanation and my program inherently takes into account the number of times they have an equal number of heads. What do you think is flawed about that logic?

A Former Brilliant Member - 2 years, 5 months ago

Using full-blown Binomial probability theory and a spreadsheet to sum the options shows that: If Anna has 50 coins
And Bastian has ….........................50.................51...............52 coins

Prob Bastian has more Heads.....46.02%........50.00%.........53.94%

Prob A&B have Equal Heads........7.96%...........7.88%..........7.73%

Prob Anna has more Heads.........46.02%.........42.12%.........38.33%

So the answer is indeed 1/2 or 50% for the 50/51 coin case and for those who wrestle with the concern that intuitively as Bastian has more coins he should get more heads the full analysis does show the chance of him getting more Heads than Anna is indeed higher than the chance of Anna getting more Heads than him, but we can't ignore the outcome of getting equal number of Heads.

For those who like the approach of "Consider 50 coins each, which on average will be equal numbers of heads and then Bastian has one more toss which has a probability of 50% of being a Head". You got the right answer but purely by luck! :) If you use the same approach with Bastian having 52 coins you'd conclude the chances of him having more heads was 75% because in his extra two tosses it's 75% chance he'll get 1 or more heads. But that's wrong - the probability is only 54% approx for this case of 50 and 52 coins.

In the same way, you cannot automatically assume you can reduce the total number of coins and still get the same answer. It happens to work for 50/51 coins reducing to 10/11 coins or 1/2 coins but it doesn't work for 50/52 coins reducing to 10/12 or 1/3 coins. In fact if anyone can explain why this approach does work for N/N+1 coins but not other numbers I'd be interested!

I know I'm cheating by using Binomial theory and a computer but seeing the actual results set out helped me understand some of the counter-intuitive results and spot flaws in some of the approaches.

Great question though!

Steve Powers - 2 years, 4 months ago

I think John Bugden has this right. In the simplest way of thinking, Ana has an expected value of 25 heads, Bastian has an expected value of 25.5 heads. I realize you can't have half a head, but we are talking probabilities. The odds that Bastian wins are 51/50.

John McKay - 2 years, 5 months ago

Log in to reply

You can't have odds of winning greater than 1.

Jordan Cahn - 2 years, 5 months ago

Log in to reply

You can't have a probability of winning greater than 1 but you can have odds of 51/50. It means you win 51 times for every 50 times you lose. Having said that it's still the wrong answer.

Steve Powers - 2 years, 4 months ago

If the actual answer was anything close to that, it would be 51/101, but I actually don't know what the most simplified form of the answer is. I don't think it's 51/101, I think it's actually pretty complicated to calculate.

A Former Brilliant Member - 2 years, 5 months ago

No, the answer is 0.5. And this solution is the best explanation you can give to someone who thinks it's wrong.

Parth Sankhe - 2 years, 5 months ago

Yet another way to look at it: For convenience relabel the head and tail on Ana's coin, so the question becomes

"What's the probability that Bastian gets more heads than (# of TAILS that Ana gets)

[ = 50 - (# of heads that Ana gets)] ?"

which is to say, "What's the probability that the total number of heads, out of the 101 coin tosses, is greater than 50?"

Peter Byers - 2 years, 5 months ago

Let's first consider the case where they flip 50 coins only. If the probability of them having the same number of heads is p, then, by symmetry, probability of Bastian having more heads is (1 - p) / 2. Now Bastian have one additional flip to complete his 51 flips. In the end he will win if had already more heads after his 50 flips P = (1 - p) / 2 or he had the same number of heads after 50 flips and he got heads in the 51st flip P = p * (1/2). So, total probability of Bastian winning is (1 + p) /2 + p/2 = 1/2

Amr Alaa - 2 years, 5 months ago

Log in to reply

Nice approach

Muhsin Haneefa - 2 years, 5 months ago

For those seeking a new approach. It's coin toss it's almost certainly 50/50.

Oluwaseyi Lawal - 2 years, 5 months ago

What if the question was 50 flips by Ana and 52 flips by Bastian. In this case there's also the chances of both being equal. How would you desk with this scenario?

Muhsin Haneefa - 2 years, 5 months ago

Log in to reply

Here there is a chance that Bastian will get both more heads and more tails than Ana, hence probability won't be half

Parth Sankhe - 2 years, 5 months ago

I believe what is confusing people the most is the fact that Bastion indeed has a higher probability than Ana of NOT LOSING the game even though the probability of winning is the same for both. Let me explain: Imagine all the cases where Ana is 1 head in lead of Bastion before Bastian's final throw. Now this 51st throw allows Bastian a 50% chance to draw the game. The effect of this is seen in Bastian having an expected value of 25.5 vs Ana's 25. So the likelihood of Bastian getting more head's than Ana is 50%, but the chance of Bastian getting less heads than Ana is lower than the chance that Ana gets less heads than Bastian.

Mikko Kärkkäinen - 2 years, 5 months ago

Log in to reply

The probability of winning (ie more heads) is NOT the same for both Bastian and Ana - Bastian is more likely to win than Ana. Probability of Bastian winning is 1/2 but probability of Ana winning is less than 1/2 because there is also the chance of an equal number of heads. It roughly works out as Bastian 50% Equal 8% Ana 42%.

Steve Powers - 2 years, 4 months ago

What about this case? If Anna flips 2 coins, and Bastian flips 200 coins, Is probability of Bastian getting more heads than Anna = 1/2?

Sifa Serdar Ozen - 2 years, 4 months ago

Log in to reply

No it's not 1/2. The probability of Bastian getting more heads would be much higher. But what does that prove?

Steve Powers - 2 years, 4 months ago

I understood it this way. Anna has 2^50 ways to flip her coin. Bastain has 2^51. So the probability is just 2^50/2^51, which simplifies to 1/2.

Todd Bethell - 2 years, 4 months ago

This answer is wrong.

vu van luan - 2 years, 1 month ago

Log in to reply

Which part of the solution do you think is wrong?

Jordan Cahn - 2 years, 1 month ago

Log in to reply

You can see whole solution didn't mention or use about 50 times and 51 times which 2 guys tosses the coin. So it's mean that with his solution, you can get the same answer if I change the number 50 by 1; and 51 by 100? Do you see if i change number like that, the probality is not equal 1/2?

vu van luan - 2 years, 1 month ago

Log in to reply

@Vu Van Luan The solution doesn't depend on 50 and 51 flips, but it does depend on the number of flips differing by 1. So the result would be the same for 100 and 101 flips, or 1000 and 1001 flips, but not for, say, 150 and 200 flips.

I just read your solution, and it is certainly correct, but my solution is also correct.

Jordan Cahn - 2 years, 1 month ago

Assuming that two guy tosses coin 50 times in the same time. Now you can see that there were 3 situation can be happan: 1+ Ann have more head than Bastian: the probality for that is: a. 2+ Anna have the same head as Bastian: the probality for that is: b. 3+ Anna have less head than Bastian: the probality for that is: a (because Ann and Bastian is the same) (notice that a + b + a = 1) Now Bastian tosses the last time: + If it is a head: probality is 1/2 - Bastian only win in second and third situation abow. The whole probality for that is: 1/2 * (a+b) + If it is a tail: probality is 1/2 - Bastian only win in third situation abow. The whole probality for that is: 1/2 * (a) So the probality for Bastian win is: 1/2 *(a +b) + 1/2 *(a) = 1/2 * (a + b + a) = 1/2

vu van luan - 2 years, 1 month ago

Let p p be the probability that after both Ana and Bastian have tossed their coins 50 times, they have the same number of heads. Then the probability that Bastian already has more heads is 1 2 ( 1 p ) \dfrac{1}{2}(1 - p) , (the same probability that Ana already has more heads), in which case he will have more heads than Ana regardless of the result of his 51st toss. If they are tied after 50 tosses, however, then he can still end up with more heads if his 51st toss is a head, which would occur with probability 1 / 2 1/2 . So the total probability that Bastian ends up with more heads after his 51st toss is

1 2 ( 1 p ) × 1 + p × 1 2 = 1 2 = 0.5 \dfrac{1}{2}(1 - p) \times 1 + p \times \dfrac{1}{2} = \dfrac{1}{2} = \boxed{0.5}

Can you please explain the basis for 1/2(1-p)?

Orrin Ahola - 2 years, 6 months ago

Log in to reply

1-p is the probably that Ana and Bastian do not have the same number of heads. Since the coin is fair, there is then a 50% chance that Bastian has more heads (and a 50% chance that he has more tails).

Jan Verplancke - 2 years, 6 months ago

Certainly, this was the most satisfying explanation according to my opinion.

Amal Antony - 2 years, 5 months ago

We can say this game is equivalent to a game where the winner is the player with more heads after 50 tosses each; and iff it is tied then an extra toss happens to choose the winner. (Because if it were not tied in the original game, adding 1 to a player's score would not change thr winner).

Since this game has symmetrical rules and there must be a winner, the chance is 1/2 that Bastian wins

Matt McNabb - 2 years, 5 months ago

Log in to reply

interesting approach, but -even though l'm satisfied the answer is indeed 1/2- l disagree that the rules should be symmetrical between the players.. in your version of the game Anna has 1/2 chances of winning, whereas in the original game she has less than 1/2.. this is because in the original game, if Anna is ahead by one head, then the 51st Bastian coin is still flipped and it can change the outcome

ric vega - 2 years, 5 months ago

Adding 1 to a player's score CAN change the winner if, after 50 throws, Ana is winning by 1 and then on the 51st Bastian gets a head then now Ana no longer wins (has more heads).

Steve Powers - 2 years, 4 months ago

for the 51st toss, why do we take probability as (1/2)*p rather than just 1/2?

Chinmay Chhajed - 2 years, 5 months ago

Log in to reply

The 51st toss only matters if they are tied after 50 tosses, which happens with probability p p , so the probability that they are tied after 50 tosses and Bastian then tosses a head on his 51st toss (and thus wins the game) is p × 1 2 p \times \dfrac{1}{2} .

Brian Charlesworth - 2 years, 5 months ago
K T
Dec 7, 2018

Consider these two eventualities:

E 1 E_1 : B throws more heads than A

E 2 E_2 : B throws more tails than A

E1 and E2 are mutually exclusive. Suppose Ana had H heads and T tails, where H+T=50. B would need or more 52 tosses to throw H+1 or more heads and T+1 or more tails. He only has  51 tosses.

Either E1 or E2 has to be true. If both were false, the B must throw no more than H heads and no more than T tails. That is only possible with 50 or less tosses.

Combined, E 1 E_1 and E 2 E_2 must be complementary, meaning that P ( E 1 ) + P ( E 2 ) = 1 P(E_1)+P(E_2)=1 .

By symmetry, for fair coins P ( E 1 ) = P ( E 2 ) P(E_1)=P(E_2) . From which the answer P ( E 1 ) = 1 2 P(E_1)=\frac{1}{2} follows directly.

Mario Egie
Dec 10, 2018

The situation needn't be more complex. If Bastian and Ana has 50 tosses each, the probability of Bastian and Ana having equal chances of heads and tails is same. If then Bastian is given an extra toss, this toss has no memory of the previous 50 tosses and since the previous 50tosses cancels outs Ana's then there is exactly 1/2 chance of Bastian having a head, as there should be if that was the first toss.

Claudio Corbetta
Dec 14, 2018

k = 0 51 ( 51 k ) j = 0 k 1 ( 50 j ) 2 101 = 1 / 2 \frac{\sum _{k=0}^{51} \binom{51}{k} \sum _{j=0}^{k-1} \binom{50}{j}}{2^{101}} = 1/2

Winston Choo
Dec 5, 2018

Although the problem looks hard, because 50 and 51 coin tosses are a large amount, let's reduce it to 0 and 1 coin tosses.

Here, the chances of Ana getting 0 heads is 100%, because she have no coin tosses. The chances of Bastian getting more than 0 heads is obviously 50%, as he only have 1 coin toss.

So the answer must be 50%, or 0.5.

You still need to show this scales up to higher numbers.

Jeremy Galvagni - 2 years, 6 months ago

Say Anna flips 1 coin (H or T), Bastian flips 2 coins (HH, HT, TH, or HH) ½ of the time Anna has 1 H. Then Bastian has > 1 H ¼ of the time (HH). ½ x ¼ = ⅛. ½ of the time Anna has 0 H. The Bastian has > 0 H ¾ of the time (HH, HT, TH). ½ x ¾ = ⅜. ⅛ + ⅜ = ½. So Bastian has 50% chance of having more heads than Anna when he flips one more time than her. Works for 0 and 1 toss, works for 1 and 2 tosses, by induction works for N and N+1 tosses. Works for 50 and 51 tosses.

Steven Adler - 11 months, 2 weeks ago
Stefan Joeres
Dec 11, 2018

The expected value for the number of heads for Bastian is 25.5. The probability of having more heads than the expected value is 0.5. That means that the probability of throwing 26 or more heads is 0.5.

What if Bastian throws 52 coins while Ana throws 50 coins, by this reasoning, the chances that Bastian throws more heads than Ana are 1, or 100% , meaning it’s impossible for Ana to throw more heads than Bastian, but clearly this isn’t the case, Ana could throw 50 heads while Bastian throws 0.

Milenko Milicevic - 2 years, 6 months ago
Seq O
Feb 1, 2019

1 = P r o b ( H b > H a ) + P r o b ( H b = H a ) + P r o b ( H b < H a ) 1 = Prob(H_b > H_a) + Prob(H_b = H_a) + Prob(H_b < H_a) , where H b = H_b = Brian's number of heads and H a = H_a = Ana's number of heads.

Note that P r o b ( T b > T a ) = P r o b ( H b = H a ) + P r o b ( H b < H a ) Prob(T_b > T_a) = Prob(H_b = H_a) + Prob(H_b < H_a) . That is, for Brian to have more tails than Ana does, he must have fewer or the same number of heads, and vice versa.

Therefore 1 = P r o b ( H b > H a ) + P r o b ( T b > T a ) 1 = Prob(H_b > H_a) + Prob(T_b > T_a) .

Due to the symmetry between head and tail, P r o b ( H b > H a ) = P r o b ( T b > T a ) Prob(H_b > H_a) = Prob(T_b > T_a) . That is, the likelihood that Brian has more heads than Ana is equal to the likelihood that he has more tails than she does.

Therefore 1 = 2 ( P r o b ( H b > H a ) 1 = 2 \cdot (Prob(H_b > H_a) .

P r o b ( H b > H a ) = 1 2 Prob(H_b > H_a) = \frac{1}{2}
Laszlo Mihaly
Dec 11, 2018

(Updated 16/12/2018). After 50 tosses there are 3 possible outcomes: A has more than 50 heads, B has more than 50 heads, or they have 25 each. The probabilities are p 1 p_1 , p 2 p_2 and p 3 p_3 ; p 1 + p 2 + p 3 = 1 p_1+p_2+p_3=1 . Due to symmetry p 1 = p 3 p_1=p_3 and 2 p 1 + p 2 = 1 2p_1+p_2=1 . At this point B is already winning with probability p 1 p_1 and A is loosing with the same probability. When B tosses the coin again, his chance of winning increases by p 2 ( 1 / 2 ) p_2*(1/2) to a total of p = p 1 + p 2 / 2 = 1 / 2 p=p_1+p_2/2=1/2 .

if they’re both tossing 50 coins the chances Bastian gets more heads is not 50%. You’re forgetting about ties. There’s 3 categories of outcomes: Bastian gets more heads, Ana gets more heads, They get the same amount of heads. Obviously the chances of the first two outcomes happening are equal for reasons you stated, however, because there is a 3rd outcome that has some chance of happening, the chances of the other two outcomes happening must both be some amount less than 50%. Let’s say for example there’s a 10% they tie (that’s probably not true but just pretend it is for the sake of this example), that would mean the chances either of the two people get more heads than the other are 45%.

Milenko Milicevic - 2 years, 6 months ago

Log in to reply

You are right. Nevertheless, the conclusion is the the same. I will make a correction in my solution.

Laszlo Mihaly - 2 years, 6 months ago

The answer of 1/2 just doesn’t sit right in my stomach. I’m not saying that it is wrong, but it just does make sense to me intuitively. If they both tossed the same amount of coins, 50 coins let’s say, clearly their chances of getting more heads than the other are equal. If Bastian gets to toss an extra coin after the initial 50 coin tosses, shouldn’t his chances of having more heads than his opponent go up? The extra toss can only help. If he gets a tails, nobody cares, no outcome is affected. If he gets heads, suddenly he has one more head that he used to, which could matter in some scenarios. If Bastian is flipping more coins than Ana, it would stand to reason that Bastian would get more heads than Ana more often than not. In fact, Bastian should have more everything than Ana does. I just can’t wrap my head around the idea that the answer to this question is 1/2. I have so many questions about so many things. Pretend they both flip one coin, there are 4 outcomes (Both tails) (Both heads) (They get different faces) (reverse of the previous one). Bastian gets more heads than Ana in exactly 25% of these cases. What if Bastian gets to throw an additional coin after his first coin? If he gets tails, nobody cares nothing changes. If he gets heads, (which he does so 50% of the time) Only 2 outcomes are affected, both heads and both tails. His initial percent was 25%, and the extra coin flipped increases his chances of winning to 75% (50% bonus) but because he only has a 50% chance to receive this 50% bonus (he only gets the bonus when his second flip is heads) we only give him half of the bonus and say his overall chances are 50%. But do you see how his chances went up because he got to roll an additional coin? Why should this be any different with 51 and 50 coins? Again, I’m not saying that the proposed answer is wrong, maybe I’m failing to understand something, but to me, if his chances with 50 coins vs 50 coins are slightly less than 50%, and tossing an extra coin slightly increases his odds, I find it incredibly unlikely his odds are 50%.

Milenko Milicevic - 2 years, 5 months ago

Log in to reply

I updated my solution, please check it out.

Laszlo Mihaly - 2 years, 5 months ago
Alkis Piskas
Dec 13, 2018

Very simple: Both toss their coin 50 times. They both expect to get the same number of heads. Now, Bastian can toss his coin once again. His chances to get another head are 50% (1/2). This is also the probability of getting more heads than Ana.

Bill Randle
Dec 11, 2018

An unbiased coin has been tossed 101 times. It is not possible for the number of heads and tails to be equal.Thus the only two possible outcomes are the number of heads is greater than the number of tails or the number of tails is greater than the number of heads. The symmetry of the outcomes provides that each of these two outcomes are equally likely..so the probability that either tosses more heads than the other is 1/2.

good thinking sir

Sunil Nandella - 2 years, 6 months ago

By this logic the answer is still 1/2 if B had 100 throws and A had 1 throw . The scoring is the number of heads thrown by that player, not the number of heads overall vs number of tails overall

Matt McNabb - 2 years, 5 months ago

the probability that either tosses more heads than the other is 1/2.

False. Only the probability that Bastian gets more heads than Ana does = 1/2, not the probability that Ana gets more heads than Bastian does.

Peter Byers - 2 years ago

Log in to reply

I kind of getting to what you are saying. lets see, what would you actually say about the probability that Ana gets more heads than Bastian does. would it be < 1/2 or >1/2.

Sunil Nandella - 1 year, 10 months ago
Rich Person
Jan 13, 2019

They all flip 50 coins. They have an equal chance of getting heads or tails. Now the other person flips one more coin. There you go. That’s the deciding factor. 50% chance.

Marco Paolo
Dec 16, 2018

Their probabilities of getting heads or tails are the same for their first 50 flips.. until Bastian flips another coin which has a 50/50 chance of getting a head or tail.

Imagine that they'll only do 50 flips each, they get the same probabilities right? Until Bastian gets another shot of doing a flip

Antoine Merour
Dec 14, 2018

I don't know if I'm oversimplifying it but this is how I look at that question.

Since they both flipped the coin 50 times, they both have the same probabilities profile (to get any number or head or tails).

So what happens next has nothing to do with what happened previously.

On the 51st coin toss, it's as if Ana had never tossed a coin and Bastian is flipping it for the first time.

He has a 1/2 chance to get a head.

So he has 1/2 chance to get more heads than Ana.

Many of the 1/2 answers all assume that both have the same score after 50 coins. I think that if Bastian has more throws, his average heads score will be higher so his chances of scoring more heads than Ana cannot be 1/2. Great fun discussion though.

David Stupple - 2 years, 4 months ago

Let's only consider heads the total ways are when ana getting 0 head Bastian can have 0,1,2.....51 heads when ana gets 1 head Bastian again can have 0,1,2....51 heads so total probability =51 52 now for the given condition when Bastian gets 51 head ana can have heads 0,1,2....50 heads when Bastian have 50 heads ana can have 0,1,2.....49 heads this goes on upto 0 so the total number of ways according to the condition is 51+50+49...+0=(52 51)/2 now the probability is ((52 51)/2 52*51)=1/2

jerin biju - 2 years, 2 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...