All about flooble | fun stuff | Get a free chatterbox | Free JavaScript | Avatars    
perplexus dot info

Home > Probability
More Power 2 the Cards (Posted on 2007-10-01) Difficulty: 3 of 5
Part of my solution to Power 2 The Cards involved the likelihood of finding at least one match among the 88,889 groups of 5 digits in each of the three 444,445-digit integral powers of 2.

Of course one match is a gross understatement.

1. How many of the 100,000 possible groups of 5 digits would one expect to find represented 0 times, 1 time, 2 times, etc. within such a large number, assuming that it's digits can be considered stochastically random, (the integer equivalent of a normal number)?

2. As a consequence, what fraction of the 88,889 5-digit groupings into which a 444,445-digit integer can be divided would you expect would have no matching 5-digit group elsewhere in the number (keeping strictly to the 5-digit bounds, non-overlapping)?

  Submitted by Charlie    
Rating: 5.0000 (1 votes)
Solution: (Hide)
Each 5-digit group has a small probability of being in a given slot within the large number, but there are 88,889 such slots. This is the condition for which the Poisson distribution was designed. Any given group is expected to occur 88,889/100,000 times. The Poisson distribution with that mean has a distribution as given in the following table, also shown with value multiplied by 100,000, to indicate how many you'd expect to occur 0 times, 1 time, etc.:

     prob      freq   accounting
                        for
0  0.411111834 41111      0             
1  0.365433198 36543  36543                
2  0.162414958 16241  32482                
3  0.048123011  4812  14436                
4  0.010694016  1069   4276                
5  0.001901161   190    950              
6  0.000281654    28    168              
7  0.000035766     4     28             
8  0.000003974                        
9  0.000000392       ______
                      88883

Rounding errors (including ignoring those with more than 7 occurrences) account for being 6 short of 88889.

So 36,543/88,883 of the groups could be expected to have no matches elsewhere in the number. That's about a 41% probability that a given group has no matches elsewhere in the number.

More precisely, we could use the actual binomial distribution.

  0.00001^I*0.99999^(88889-I)*C(88889,I)
       |
I      V
0 0.411110007   
1 0.365435228   36544   36544
2 0.162415657   16242   32484
3 0.048122616    4812   14436
4 0.010693674    1069    4276
5 0.001901033     190     950
6 0.000281622      28     168
7 0.000035760       4      28
8 0.000003973   
9 0.000000392           _____
                        88886

The 36544/88886 is still about a 41% probability that a given group of 5 has at no match anywhere else in the overall number, taking group-of-5 boundaries into consideration with no overlap.

Comments: ( You must be logged in to post comments.)
  Subject Author Date
There are no comments yet.
Please log in:
Login:
Password:
Remember me:
Sign up! | Forgot password


Search:
Search body:
Forums (0)
Newest Problems
Random Problem
FAQ | About This Site
Site Statistics
New Comments (3)
Unsolved Problems
Top Rated Problems
This month's top
Most Commented On

Chatterbox:
Copyright © 2002 - 2024 by Animus Pactum Consulting. All rights reserved. Privacy Information