Puzzling Prob Stats / Bayes problem

  • Thread starter Quantumduck
  • Start date
  • Tags
    Stats
In summary, the probability of getting a 4th heads with a fair coin is 1/9, while the probability with a biased coin is 1/2.
  • #1
Quantumduck
38
1
Hello All,
I have a problem that goes like this:I have one fair coin and one double heads coin. I pick one at random and flip the same coin three times. All three times it comes up heads. What is the probability that this same coin will come up heads a fourth time?

I said that the prob on the fourth throw is 1/2. Why?

The probability for the fair coin to throw 4 heads is .5 x .5^4 = 1/32. (the first .5 because of the random choice between the two coins.)

The probability of the double heads throwing 4 heads is .5 x 1^4 = .5

Therefore, the probability of throwing a 4th heads is .5, because it is much more likely that I have the double heads coin in my hand than the fair coin.

My question is, did I make an error in figuring this? It seems too easy, but I can not find any flaws in my reasoning (because I am too close to it)

Thanks in advance.
 
Physics news on Phys.org
  • #2
Quantumduck said:
Hello All,
I have a problem that goes like this:I have one fair coin and one double heads coin. I pick one at random and flip the same coin three times. All three times it comes up heads. What is the probability that this same coin will come up heads a fourth time?

Your universe: 1/32 + 1/2
Probability of having the double heads coin: (1/2) / ( 1/32 + 1/2 ) = 32/33
Probability of having the fair coin: (1/32) / ( 1/32 + 1/2 ) = 1/33
( of course you could get the last result by "1 - 32/33" )

Then, the probability of a 4th head is:
1.0 * 32/33 , if you have the double heads coin
0.5 * 1/33 , if you have the fair coin

So, the answer is 65/66.
 
Last edited:
  • #3
If the fair coin is chosen, the probability of 4th head would be (1/2)^4.

If the biased coin is chosen, the probability of 4th head would be 1.

So, the probability of 4th head could be
Prob(Fair Coin)*Prob(4th head on Fair Coin) OR Prob(Biased Coin)*Prob(4th head on Biased Coin)
= 1/2 * (1/2)^4 + 1/2 * 1
= 17/32

Is there a mistake somewhere?

I didn't follow Rogerio's reply, and the divisions are wrong in that.

EDIT--The mistake is that I haven't taken into account that the first three tosses result in head.
 
Last edited:
  • #4
Quantumduck said:
Hello All,
I have a problem that goes like this:I have one fair coin and one double heads coin. I pick one at random and flip the same coin three times. All three times it comes up heads. What is the probability that this same coin will come up heads a fourth time?

1/8 chance to get all heads with the normal coin, 8/8 with the double-head coin. The chance that you have the normal coin is thus 1/9 since you were equally likely to have picked up either coin. The chance of getting a tails on the normal coin is 1/2, for an overall chance of 1/18; otherwise, you have four heads. 17/18 is thus the chance of getting four heads, given the first three are heads.
 
  • #5
bala.l said:
I didn't follow Rogerio's reply, and the divisions are wrong in that.

My mistake.
Correcting the values:

Your universe: 1/16 + 1/2
Probability of having the double heads coin: (1/2) / ( 1/16 + 1/2 ) = 8/9
Probability of having the fair coin: (1/16) / ( 1/16 + 1/2 ) = 1/9
( of course you could get the last result by "1 - 8/9" )

Then, the probability of a 4th head is:
1.0 * 8/9 , if you have the double heads coin
0.5 * 1/9 , if you have the fair coin

So, the answer is 17/18.
(8/9 + 1/18)
 
  • #6
Thank you!
I noticed that the probability in the first reply were for the 5th throw coming up heads, but I understood what he was doing so I did it for the 4th throw and got ... 17/18! That rocks.
Thank you all for your help!
 

FAQ: Puzzling Prob Stats / Bayes problem

What is Bayes' theorem and how is it used in statistics?

Bayes' theorem is a mathematical formula that describes the relationship between conditional probabilities. It is used in statistics to update the probability of an event occurring based on new information or evidence. In other words, it allows us to revise our beliefs about the likelihood of an event as we gather more data.

How is Bayes' theorem different from other statistical methods?

Bayes' theorem differs from other statistical methods in that it takes into account prior knowledge or beliefs when calculating the probability of an event. This allows for a more nuanced and flexible approach to statistical inference, as it allows for the incorporation of subjective information.

What is a Bayesian network and how is it used in problem-solving?

A Bayesian network is a graphical model that represents the relationships between different variables and their conditional probabilities. It is used in problem-solving to model complex systems and make predictions based on available data. It is particularly useful in situations where there is uncertainty and incomplete information.

Can Bayes' theorem be applied to real-world problems?

Yes, Bayes' theorem can be applied to real-world problems in a variety of fields, including medicine, finance, and artificial intelligence. It has been successfully used to solve problems such as disease diagnosis, spam filtering, and predicting stock market trends.

What are some limitations of using Bayes' theorem in problem-solving?

One limitation of using Bayes' theorem is that it relies on the availability of accurate prior knowledge or beliefs. If these are incorrect or incomplete, it can lead to inaccurate results. Additionally, it can be computationally intensive to calculate the probabilities of multiple variables in complex systems.

Similar threads

Replies
57
Views
4K
Replies
2
Views
2K
Replies
45
Views
4K
Replies
41
Views
5K
Replies
5
Views
1K
Replies
11
Views
2K
Back
Top