You are on page 1of 3

Weird Dice Author(s): JOSEPH GALLIAN Source: Math Horizons, Vol. 2, No. 3 (February 1995), pp.

30-31 Published by: Mathematical Association of America Stable URL: . Accessed: 26/08/2013 08:56
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact

Mathematical Association of America is collaborating with JSTOR to digitize, preserve and extend access to Math Horizons.

This content downloaded from on Mon, 26 Aug 2013 08:56:17 AM All use subject to JSTOR Terms and Conditions



a person with ho Consequently, tels on the orange property St. James a per ordinary pair of dice. Well, there are five possibilities for the two faces:

you are in the middle of Imagine a game of and some monopoly one substitutes a pair of dice la with beled 1, 2, 2, 3, 3, 4 and 1, 3, 4, 5, 6, 8 for the standard dice. Does this change the game in any way? The first thing that comes tomind is that the probabilities for the vari ous sums might be different. Supris ingly, this is not the case. Figure 1 re

Virginia (4 spaces past jail) isworst

off. In fact, this small change in the probabilities for certain doubles us ing weird dice causes St. James Place tomove up from the 10th most fre quently landed space to the 6th most landed on space whereas Virginia

with the weird dice! Moreover, son that owns the maroon property

Place (6 spaces past jail) isbetteroff

Next, we consider the product of the two polynomials created by using the ordinary dice labels as exponents:

(5,1), (4,2), (3,3), (2,4) and (1,5).

(x6+ x5+ x4+ x3+ x2+ x) x(x6 + x5 + x4 + x3+ x2 + x).

Observe that using the distribu tive property we pick up the term _ x6 in this produce in precisely the follow
JU , JU _,,_.._, I**

veals that the frequencies of the pos sible sums are exactly the same for either pair of dice (1 way to roll a sum of 2; 2 ways to roll a sum of 3, etc.) of Because this

ing ways: x5 -x1, x4

JU , Ju x5. Notice

Ju ,

dice does not effect the game. But recall that one of the rules is that of monopoly if a player is in jail he/she rolls doubles then the player advances the number of spaces shown on

onemight conclude that the change in

4 5

^p^l j?^ 2 ~3 ~ QTJ

p?, I?J


l ?IU ?1

* * I



5 6 7 8

[~T| |yjj
_ 1**1_





7 8

7 8

8 9

the correspondence between pairs of la bels whose sums are whose

_ dence sums

^~ 7







p-ri ^?^ ryyi 7

products are x6. This correspon

is one-to-one, and all dice?

6 and pairs of terms

(not visiting) and

[X] _^_


10 11


10 11

9 10 ff7*] U ! I_I_I_




and it is valid for all including our weird dice and any other


the the dice. With standard dice each of the sums 2, 4, 6, 8, 10 and 12 has a 1/36 chance of occurring as the re sult of a double. With the weird dice the sums 2 and 8 have a 1/36 chance of occurring; the sums 4, 10 and 12 have a 0 chance; and the sum 6 has a 1/18 chance.

JOSEPH GALLIAN is professor of math ematics at the University of Minnesota Duluth. He recendy won one of the MAA's Distinguished Teaching Awards.

in the rankings. These weird dice raise two inter esting mathematical questions: How were the labels derived and are there other weird labels consisting of posi tive integers? It is possible to answer these questions with a simple analy sis. To do so we begin by finding a way tomodel summing the faces of a pair of dice. First, let us ask ourselves how we may obtain a sum of 6, say, with an

drops from 24th place to 27th place

servation about products mials, thismeans that

desired probabil - ities. So, letai, a2, 03, 04, a5, a$ and 6i, 62, 63* &4> &5> &6 be any two lists of positive inte ger labels for a pair of cubes with the property that the probability of sum with these rolling any particular dice is the same as the probability of rolling that sum with ordinary dice labeled 1 through 6. Using our ob of polyno

dice that yield the

(x6+ x5+ x4 + x3+ x2 + x) x(x6 + x5+ x4+ x3+ x2 + x)


1995 MathHorizons February

This content downloaded from on Mon, 26 Aug 2013 08:56:17 AM All use subject to JSTOR Terms and Conditions

Illustration byGregNemec t=

(xai + xa* + x*3 + xa* + x"5 + xa*)

both forms for P(x). P(l) and P(l) = = lai + la2 +

That +

is, lae = 6



x(xbl + x62 + x63 + xb4 + xb5 + xbe). Now all we have to do is solve this equation for the a's and 6's. How can we solve one equation in 8 un knowns? You didn't learn this in high school but you did spend much of your efforts factoring. So, let's fac tor the left-hand side of the equa tion. The polynomial x6 + x5 + x4 + x3 + x2 + x factors uniquely into ir reducibles as

P(x) =x6 + x5+ j4 + x3 + x2+ x,

so the die labels are 6, 5, 4, 3, 2, 1? an ordinary die. When t = 2,


that r = 1 and Clearly, this means 5 = 1. What about q? Evaluating in two ways just as we did for P(0)


= x8 + x6+ x5 + x4 + x3+ x,


x(x + l)(x2 + x + l)(x2 X+ 1)

so that the left-hand side of the equa tion has the irreducible factorization

2, the smallest pos could roll with the labels for dice would corresponding be 3 since q is the smallest of the a's and the smallest permissible b is 1. Because the sum of 2 is possi ble with ordinary dice this violates our assumption about the dice hav = 2 ing equal probabilities. Thus, q is not permissible. We have now re duced our list of possibilities for q9 r, 5, and t to q = 1, r = 1, s = 1, and t= 0,1,2. Let's consider each of these possibilities for t in turn. When t = 0,

= hand, if q sible sum one

shows thatq ^ 0. On theother

so the die labels are 8, 6, 5, 4, 3, 1? the other weird die. Thus we have derived the weird dice labels and proved that they are the only other pair of dice that have this property. We invite the reader to investigate the analogous questions for the tetra hedron, octahedron, dodecahedron, and icosahedron dice. For the an swers see [1] and [2].

x2(x + l)2(x2 + x + l)2(x2 x + l)2.

This means that these factors are the irreducible factors of only possible = xai + xa2 + xa3 4- xa4 + xa* + P(x) xae. Thus, P(x) has the form

+ l)r(x2 + x + l)*(x2 - X+ 1)*, afl(x

where 0 < g, r, s, t < 2. To further restrict the possibilities in for q, r, s, and twe evaluate P(l)


= x4+ x3+ x3+ x2+ x2+ X,

so the die labels are 4, 3, 3, 2, 2, 1? one of our weird dice.

Duane. (1979). "Re the faces of Mathe dice," numbering vol. 52, 312-315. maticsMagazine, 2. Gallian, Joseph A. and David Rusin. (1979). "Cyclotomic polyno mials and nonstandard dice," Discrete Mathematics, vol. 27, 245-259.

References 1. Broline,

MathHorizons 1995 31 February

This content downloaded from on Mon, 26 Aug 2013 08:56:17 AM All use subject to JSTOR Terms and Conditions