There are 6 postmen in city, each assigned to deliver 3 letters to 3 different families within the day. Before the postmen set off to deliver, they are first dispatched to their respective designated receiving counters, where their designated letters, 3 each, to be delivered are waiting. However, by some stroke of fate, the algorithm for the postmen's counter assignment and the letter groupings' counter assignment have gone haywire, and each postman will be wrongly assigned a receiving counter, and that each counter will have a wrong group of 3 wrongly addressed letters. How many ways can this peculiar occurence happen?
Details:
Each postman is assigned 3 letters, and distinct addresses written on the 3 envelopes. These postmen know nothing of these information prior to receiving the letters.
There are only 6 receiving counters in the city.
All 6 postmen are assigned the wrong receiving counters.
Each group of 3 letters are dropped at the wrong receiving counters.
No group of 3 letters are received by the correct postman.
For each group of 3 letters, each of the proper address for each letter is only among the 3 envelopes together with the group. However, none of the letters are placed on the proper envelopes.
How the 18 letters are distributed among the 6 groups is no longer considered, as no problem was encountered so far in distributing them equally, so what letters are supposedly assigned to particular postmen are already predetermined. In other words, the letters are already properly grouped according to how they were supposed to be grouped.
For example:
Expectation: Postman 1 is dispatched at Receiving counter 1, receiving Letter Group 1 with letters (capital letter) and addresses (small letter) Aa, Bb, Cc.
Reality: Postman 1 is dispatched at Receiving counter 4, receving Letter Group 2 with letters and addresses Dd, Ee, Ff.
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.
A Latin rectangle consists of a k × n matrix, all of whose entries are integers between 1 and n , with the property that none of the k rows of n numbers contain a repeated digit, and none of the n columns of k numbers contain a repeated digit. A normalised Latin rectangle is a Latin rectangle whose first row is the numbers 1 , 2 , 3 , … , n (in order) and whose first column iis the numbers 1 , 2 , … , k (in order). Let L ( k , n ) be the number of normalised k × n Latin rectangles.
Say that a k × n Latin rectangle is semi-normalised when its first row is 1 , 2 , 3 , … , n (in order) (with no restriction on the shape of the first column), and let S ( k , n ) be the number of semi-normalised Latin rectangles. Any semi-normalised Latin rectangle gives us a solution to the posting problem, with the elements in the second row telling us which receiving counter Postman 1,2,3,4,5,6 is assigned to, and the elements in the third row telling us which receiving counter Letter Group 1,2,3,4,5,6 is assigned to. For each Letter Group, there are 2 ways of sorting the letters into envelopes so that no envelope is in the correct envelope. Thus the number of possible solutions to the problem is S ( 3 , 6 ) × 2 6 It is also clear that S ( k , n ) = ( n − 1 ) ( n − 2 ) L ( k , n ) since any normalised Latin rectangle can be renumbered to form a seminormalised Latin rectangle in ( n − 1 ) ( n − 2 ) ways as follows:
Thus the number of posting arrangements is 2 6 × 2 0 × L ( 3 , 6 ) . It is a standard result that L ( 3 , 6 ) = 1 0 6 4 , which tells us that the number of posting rearrangements is 6 4 × 2 0 × 1 0 6 4 = 1 3 6 1 9 2 0 .
It is worth noting that the answer is also 6 4 times the number of ordered pairs ( π , σ ) such that both π and σ are derangements of 1 , 2 , 3 , 4 , 5 , 6 , and moreover such that π − 1 σ is also a derangement.