Probabilities in a two-stages random pick

Valenxio

New member
Joined
Jul 28, 2015
Messages
3
Hi guys,

Here is the problem. Let's say my teacher has formed 5 groups of 4 persons each and I am in one of these groups. I need to calculate the odds of my teacher picking me if the method she use is: First, pick a group at random, THEN pick a person at random is the chosen group.

Do I just need to multiply the two odd ( 0.20*0.25) or is it more complicated than that? Also does that kind of method has a particular name?

Thanks!
 
Hi guys,

So here's the problem, let's say my teacher has formed 5 groups groups of 4 persons each and I am part of one of the groups. I need to calculate the odds of my teacher picking me if she use the following method: First, she pick a group at random, THEN she pick a person at random in the previously chosen group.

Do I just need to multiply the two odds (0.20 * 0.25) or is it more complex?

Thanks!

Notice 0.20 * 0.25 = 1/20
 
Hi guys,

Here is the problem. Let's say my teacher has formed 5 groups of 4 persons each and I am in one of these groups. I need to calculate the odds of my teacher picking me if the method she use is: First, pick a group at random, THEN pick a person at random is the chosen group.

Do I just need to multiply the two odd ( 0.20*0.25) or is it more complicated than that? Also does that kind of method has a particular name?

Thanks!

Just to be definitive suppose you name each person pij where i is the group number and j is the person number in the group. It wouldn't make any difference whether the i (1-5 inclusive) or j (1-4 inclusive) was picked first assuming picking a group number had nothing to do with picking a person number, that is the two events are independent. In fact, it wouldn't make any difference if you labeled all the people Pk [k=1-20 inclusive] and picked a k at random. So, yes, as long as the events are independent, you just multiply the probabilities together.
 
Thanks you for the reply!

Okay I have one more question, what are the odds than in three separated picks the same group is chosen twice in a row (regardless the person chosen)?

If I understand well what Ishuda said, the two events are still independent so do I just need to multiply ( 0.20 * 0.20)?
 
Okay I have one more question, what are the odds than in three separated picks the same group is chosen twice in a row (regardless the person chosen)?

This problem is similar, but actually quite a bit more complex to solve, so you'd need a different method of approaching it. There are three separate picks, but only two of them need to be the same. We're given the further stipulation that the two groups that are the same must be picked back-to-back. So, just reasoning it out, we have two scenarios which satisfy these conditions:

Group 1 and Group 2 are the same
Group 2 and Group 3 are the same

To make the concept simpler, let's consider flipping coins. We're going to flip three coins, and we want to find the probability of getting two of the same face in a row. If we want coins 1 and 2 to both be heads, how many possible outcomes are there? Notice that it doesn't matter whether the first coin is heads or tails. So there are two outcomes for the first coin. But then the second coin has to the same as the first. Regardless of what face the first coin was, only one outcome will work. And, going further, we also don't care what face the third coin is. Now overall, how many outcomes does the first scenario have? Well, there are two outcomes for the first coin, one for the second, and two again for the third, so we have:

2 * 1 * 2 = 4 outcomes

Applying similar logic for the second scenario, when the second and third coins are the same, we note that the first coin doesn't matter, nor does the second, but the third coin must be the same as the second, so we have:

2 * 2 * 1 = 4 outcomes

There are 4 outcomes in the first scenario, and 4 more in the second scenario, for a total of 8 outcomes which satisfy the given conditions. However, we're not quite done. Consider the possibility that all three coins are the same. If that's the case, then those possibilities were counted in both of our scenarios. That means we counted them twice, and need to subtract off that many possibilities. So, then, how many ways can it occur that all three coins are the same? The first coin tells us which face we're looking for, and then each subsequent coin must match that face, so we have:

2 * 1 * 1 = 2 outcomes

Taking our 8 outcomes from before and subtracting the 2 outcomes where the scenarios overlap, we have 6 scenarios which satisfy our conditions. And since probability is just favorable outcomes divided by total outcomes, we have: 6 / (2 * 2 * 2) or 6/8, which is 0.75

Now can you apply a similar reasoning to your original problem where there are five groups to pick from?
 
Thanks you for the reply!

Okay I have one more question, what are the odds than in three separated picks the same group is chosen twice in a row (regardless the person chosen)?

If I understand well what Ishuda said, the two events are still independent so do I just need to multiply ( 0.20 * 0.20)?

Yes, as one way to do it and only as a starting point. But that would be for a particular group and would only count the first two picks. You still need to deal with the third pick. A better way, IMO, is outlined below.

One thing you do need to be careful of though is how you count twice in a row. For example, suppose you choose group 1, group 1, group 1 as the three picks. Do you count that as a 'two in a row' or distinguish as a 'three in a row' and say it doesn't count. In the example given by ksdhart, the 'three in a row' was counted as a 'two in a row' and that is why the answer was 6/8 where as, if the 'three in a row' was not counted as a 'two in a row', the answer would be 4/8=1/2
HHT<---------Yes
TTH<---------Yes
HTT<---------Yes
THH<---------Yes
TTT<---------No?
HHH<---------No?
HTH<---------No
THT<---------No

The way I would go about it is to first find out how many way don't satisfy my conditions. Then, the ways that do satisfy my conditions are those that are left [if m fail then n-m succeed]:
Case (1) no matches two in a row
1st choice: any of the five groups = 5 choices
2nd choice: can't match the first choice = 4 choices
3rd choice: can't match the second choice = 4 choices
Case (2) all the same group (all three match)
1st choice: any of the five groups = 5 choices
2nd choice: match the first choice = 1 choice
3rd choice: match the second choice = 1 choice
Case (3) total number of choices (doesn't matter)
1st choice: any of the five groups = 5 choices
2nd choice: any of the five groups = 5 choices
3rd choice: any of the five groups = 5 choices

Put it together
 
Last edited:
Top