Advanced (Or simple?) balls in a jar probability

Click For Summary

Discussion Overview

The discussion revolves around the probability of selecting specific ball types from multiple jars, with a focus on combinatorial methods and probability distributions. Participants explore the implications of having 20 unique ball types distributed across 8 jars, each containing 5 unique balls.

Discussion Character

  • Exploratory
  • Mathematical reasoning
  • Debate/contested

Main Points Raised

  • One participant questions the probability of observing a particular ball type in a jar, suggesting various methods including simple fractions and combinatorial approaches.
  • Another participant asserts that the probability is not simply 1/5 and raises concerns about the implications of different numbers of ball types.
  • Some participants propose that the probability of a specific ball being in a jar could be calculated as 1/20, assuming independence.
  • One participant suggests that the probability of a ball being in at least one jar can be calculated using binomial distribution, contingent on the first question's answer.
  • Another participant provides a combinatorial approach, stating that the chance of selecting a specific ball type among the 5 chosen is 1 - (19C5/20C5), leading to a probability of 1/4.
  • Some participants challenge the assumptions of independence and replacement in their calculations, suggesting alternative probabilities based on different interpretations of the problem.
  • There is a discussion about the uniqueness of ball types in jars, with some participants clarifying that a single ball cannot be in more than one jar or in the same jar more than once.
  • One participant notes that the total number of balls exceeds the number of unique types, leading to confusion about the distribution.

Areas of Agreement / Disagreement

Participants do not reach a consensus on the probability calculations, with multiple competing views and interpretations of the problem remaining unresolved.

Contextual Notes

Participants express uncertainty regarding the assumptions of independence and replacement in their probability calculations. The discussion also highlights the complexity of the problem, with differing interpretations of how ball types are distributed across jars.

csb
Messages
2
Reaction score
0
If I have 8 jars, each jar contains 5 unique ball types. However, I know that I have 20 unique ball types out there. So, I have balls labelled from B1, B2, B3, ...B20 to put into 5 jars.

Let's say

Jar 1 : B1, B3, B4, B7, B12 (5 unique balls)

jar 2 : B9, B7, B11, B15, B18 (5 unique balls)

jar 3 : B20, B1, B5, B17, B13 (5 unique balls)

jar 4 : B4, B5, B6, B3, B17 (5 unique balls)

jar 5 : B3, B10, B2, B14, B16 (5 unique balls)

and so on...

My question :

  1. What is the probability to observe one particular ball type in a jar ?
  2. What is the probability to observe one particular ball type in at least n jars ?
Is there a well-known probability distribution or combinatorial method that might relate to this ?

For question 1, can I say that the probability is 1/5 ? Or is it 5C1/20C5 ? Or should I use more complicated thing like Stirling number of the second kind for this problem ? Totally not sure...

For question 2, once I know the probability from question 1, I can use binomial distribution, right ?

Thanks
 
Physics news on Phys.org
The answer is simple, but it is not 1/5. What if there were 5 different ball types? Or 6?
 
I think the answer to number 1 is 1/20. the p of the ball being in a jar is 5/20 times p of picking the ball 1/5. assuming independence.

for number two you need the answer to number one times the p that the ball is in another jar*you pick that jar*you pick that ball

but you also need to accoount for the p that the ball is in the space again but in the same jar that you already picked, this can make my answer for number 1 incorrect so I would say it is not an easy problem. 5C1/20C5 is not right it is way to small.
 
There 20 unique ball types in total. If you take one jar, how many unique ball types are there in it? And how many unique ball types are there in total? So, assuming that any ball type is equally likely to be in the jar, what is the probability that any given ball type is one of the types that is in the jar?
 
The probability that any ball type is one of the ball types in the jar is 1 because they are all a ball type.

Is the question what is the probability that a particular ball is in at least one jar?

I thought it was; if you distribute 20 unique balls uniformly into 5 distinct jars, 5 balls in each jar, each ball put in a jar independently with p=1/20, what is the probability that you will choose one particular ball out of a jar?
 
I'm assuming question 1 is equivalent to: I randomly select 5 out of the 20 available types. What is the chance that type 1 is one of the 5 selected?

The answer to that question is 1 - (19C5/20C5), which is 1/4. That's because there are 20C5 ways to select the 5 types, and 19C5 ways to select the 5 types without choosing type 1.
 
Na because that's without replacement. you can repeat balls so it is 1-19^5/20^5 = .2262
 
davidmoore63@y said:
I'm assuming question 1 is equivalent to: I randomly select 5 out of the 20 available types. What is the chance that type 1 is one of the 5 selected?

The answer to that question is 1 - (19C5/20C5), which is 1/4. That's because there are 20C5 ways to select the 5 types, and 19C5 ways to select the 5 types without choosing type 1.
That is the right answer, but it is more easily arrived at by considering that there are 5 mutually exclusive 1/20 chances for a given ball type to be in a given jar: the answer is 5 x 1/20 = 1/4.
 
MrAnchovy said:
That is the right answer, but it is more easily arrived at by considering that there are 5 mutually exclusive 1/20 chances for a given ball type to be in a given jar: the answer is 5 x 1/20 = 1/4.

19C5/20C5 is not correct because it doesn't consider cases such as

B2 B2 B3 B3 B4
B2 B2 B2 B2 B2
or
B2 B3 B4 B5 B5
 
  • #10
csb said:
each jar contains 5 unique ball types
 
  • #11
O god dam
 
  • #12
So binomial with p=.25
 
  • #13
not advanced
 
  • #14
csb said:
If I have 8 jars, each jar contains 5 unique ball types. However, I know that I have 20 unique ball types out there. So, I have balls labelled from B1, B2, B3, ...B20 to put into 5 jars.

Let's say

Jar 1 : B1, B3, B4, B7, B12 (5 unique balls)

jar 2 : B9, B7, B11, B15, B18 (5 unique balls)

jar 3 : B20, B1, B5, B17, B13 (5 unique balls)

jar 4 : B4, B5, B6, B3, B17 (5 unique balls)

jar 5 : B3, B10, B2, B14, B16 (5 unique balls)

and so on...< Snip>

It seems this gives you a total of 25>20 unique balls.
 
  • #15
WWGD said:
It seems this gives you a total of 25>20 unique balls.
But only 20 unique ball types. The uniqueness of balls in a jar is trivial: a single ball cannot be in more than one jar, nor in the same jar more than once. There are 8 jars in total so a total of 40 (unique) balls.

Note that there are balls of ball type B1 for instance in both jar 1 and jar 3.
 
  • #16
MrAnchovy said:
But only 20 unique ball types. The uniqueness of balls in a jar is trivial: a single ball cannot be in more than one jar, nor in the same jar more than once. There are 8 jars in total so a total of 40 (unique) balls.

Note that there are balls of ball type B1 for instance in both jar 1 and jar 3.

Ah, O.K, maybe it may be clearer to use the notation ## B_{1, i_1}, B_{2,i_2},..., B_{20, i_{20}} ## to denote the ball types, where ##i_j## ranges over the number of balls of type ##j##.
 
  • #17
MrAnchovy said:
But only 20 unique ball types. The uniqueness of balls in a jar is trivial: a single ball cannot be in more than one jar, nor in the same jar more than once. There are 8 jars in total so a total of 40 (unique) balls.

Note that there are balls of ball type B1 for instance in both jar 1 and jar 3.
So you can't have the same ball type in the same jar

but you can have repeat of ball types for each jar

and there are 40 total different balls in 8 jars for some reason
 
  • #18
As the first part of the problem has been shown to be trivial, and the OP has not returned to show any work on the second part it is probably not worth discussing any further.
 

Similar threads

Replies
235
Views
15K
  • · Replies 9 ·
Replies
9
Views
2K
  • · Replies 6 ·
Replies
6
Views
4K
  • · Replies 1 ·
Replies
1
Views
4K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 24 ·
Replies
24
Views
3K
  • · Replies 125 ·
5
Replies
125
Views
20K
  • · Replies 9 ·
Replies
9
Views
10K