

04202007, 07:06 PM


puzzler


Join Date: Aug 2004
Location: UK


Math trivia
I shall post occasional mathematical 'factoids' here. Please feel free to add your own.
Brun's constant and the Pentium bug: Viggo Brun's constant is the sum of the reciprocals of the twin primes. He proved that this (probably infinite) sum converged to a constant in 1919.
(Twin primes are the prime numbers that only differ by two, such as 5 and 7 or 71 and 73. There are conjectured to be an infinity of twin primes, though this has yet to be proven. A large pair are:
33218925 * 2^169690 +/ 1, which have 51,090 digits each.
The only number that appears in two twin pairs is 5 [3, 5] [5, 7])
Brun's constant is 1/3 + 1/5 + 1/5 + 1/7 + 1/11 + 1/13 + 1/17 + 1/19 + ... and has a value of approximately 1.902161
While using a computer to calculate Brun's constant in 1994, Thomas Nicely discovered the infamous 'Pentium Bug'. The chip gave occasional errors when performing floating point division operations due to five errors encoded into a lookup table inside the chip. This cost Intel millions to fix.

04212007, 09:10 PM


puzzler


Join Date: Aug 2004
Location: UK


Re: Math trivia
International Standard Paper sizes and square roots: The most common size of printer paper (in Europe anyway) is A4 (about 8.3 x 11.7 inches or 210 x 297 mm). A piece of A4 is twice the area of an A5 sheet and half the area of an A3 sheet.
If you want to be able to cut a sheet of paper in half like this and get two smaller pieces with the same proportions as the original, then the height to length ratio has to be the square root of two (about 1.414).
ISO paper sizes begin with A0 which is defined as having an area of one square metre. When A0 is cut in half you get A1 and so on until after 4 divisions you end up with A4 which is therefore 1/16 of a square metre in area.
It's possible to buy sheets of paper larger than A0. One of twice the area is called 2A0 and one twice as big as that is called 4A0. This system of naming the larger formats doesn't fit well with the naming scheme of the smaller sizes  the person who came up with those large format designations was no mathematician.

04222007, 11:30 PM


Not as smart as Adam


Join Date: Apr 2007
Location: Queensland
Gender: Male


Re: Math trivia
Quote:
Originally Posted by ceptimus
It's possible to buy sheets of paper larger than A0. One of twice the area is called 2A0 and one twice as big as that is called 4A0. This system of naming the larger formats doesn't fit well with the naming scheme of the smaller sizes  the person who came up with those large format designations was no mathematician.

Urmmm, why? Correct me if I'm wrong but if if 2A0 is twice the size of A0 then 4A0, being twice the size of 2A0 is logically four times the size of A0. I agree that it doesn't fit the smaller names, but from a mathematical point of view the larger size names work perfectly.
__________________
Don't pray in my school and I won't think in your church.

04232007, 09:34 PM


puzzler


Join Date: Aug 2004
Location: UK


Re: Math trivia
The flaw with the naming arrangement is that it changes from a 'power' series (for the smaller sizes) to a multiplicative one for the larger. I agree that the naming scheme for the larger sizes makes perfect sense  it just doesn't fit well with the naming scheme for the smaller sizes.
In the smaller sizes, the number is a (negative) power of 2: A sheet of An (where n is 0, 1, 2, 3, 4, 5, 6, etc.) has an area of 2n square metres.
If the same naming system were used for the larger sizes, then a sheet of 2A0 would be called A1 and a sheet of 4A0 would be A2, etc.
On the other hand, if we used the 'large' naming scheme throughout then the sheets should be called (in decreasing size order):
4A0, 2A0, A0, 1/2A0, 1/4A0, 1/8A0, 1/16A0, etc. (1/16A0 would be what we call A4)

04232007, 10:19 PM


weird citrus golem


Join Date: Jul 2005
Location: New Zealand
Gender: Male


Re: Math trivia
While I agree it's a confusing system for naming paper sizes, your post actually makes it clear that it's not unmathematical. The multiplicative term precedes the 'A', and the negative exponent follows it. As a result, there's more than one way to refer to each paper size:
1/2A1 = A0 = 2A1 = 4A2
1/4A2 = 1/2A3 = A4 = 2A5
It's not unlike the notation used in scientific calculators, e.g. 3.782e+36

05122007, 12:34 PM


puzzler


Join Date: Aug 2004
Location: UK


Re: Math trivia
Benford's law and the detection of fraud: If you take some apparently random numbers, such as the amount of money in a bank account, or the number of times a thread at has been viewed, what would you expect the first (leftmost) digits of those numbers to be?
Your first thought might be that all digits are equally likely, and as we don't normally bother to write leading zeros, that the remaining digits 1 to 9 would each occur one ninth of the time. The surprising thing is that this is not the case: 1 occurs much more often than that, with 2 occurring less often than 1, and so on down to 9 which occurs with the lowest frequency.
One of the first people to report this fact was the nineteenthcentury astronomer, Simon Newcomb, who noticed that books containing tables of logarithms always showed more wear on the pages of numbers starting with '1'. Benford investigated this further, and in 1938 concluded that the first digit is d with a probability of log 10(1 + 1/ d).
The table below shows roughly how often each digit appears as the first digit.
Code:
digit frequency
1 30.1%
2 17.6%
3 12.5%
4 9.7%
5 7.9%
6 6.7%
7 5.8%
8 5.1%
9 4.6%
Fraudsters who are not aware of Benford's law are likely to generate false receipts and invoices with numbers that don't agree with the law and this allows their fraud to be easily detected.

05132007, 11:11 PM

God Made Me A Skeptic


Join Date: Jul 2004
Location: Minnesota


Re: Math trivia
I first encountered Benford's Law in Warren Weaver's most excellent probability book, Lady Luck.
__________________
Hear me / and if I close my mind in fear / please pry it open
See me / and if my face becomes sincere / beware
Hold me / and when I start to come undone / stitch me together
Save me / and when you see me strut / remind me of what left this outlaw torn

05152007, 05:43 AM


El Caganer


Join Date: Mar 2007
Location: in ur bethlaham


Re: Math trivia
Quote:
Originally Posted by ceptimus
the first digit is d with a probability of log10(1 + 1/d).

Huh!
So sum ( n=1 to 9) [log 10(1 + 1/ n)] = 1, then. That's cool too!

05152007, 06:24 AM


weird citrus golem


Join Date: Jul 2005
Location: New Zealand
Gender: Male


Re: Math trivia
Yeah, it is cool. When you work it out you get:
sum (n=1 to 9) [log10(1 + 1/n)]
= log10(1 + 1) + log10(1 + 1/2) + log10(1 + 1/3) + ... + log10(1 + 1/9)
= log10((2/1) * (3/2) * ... * (10/9))
The numerators and denominators cancel out, leaving log10(10) = 1

05242007, 09:18 PM


puzzler


Join Date: Aug 2004
Location: UK


Re: Math trivia
Different sorts of infinity: This might be slightly heavy for a 'trivia' thread, but I'll try and keep it light.
Infinity does tend to boggle the mind at first, but mathematicians encounter it pretty often and tend to become immune. Cantor showed that there are different classes of infinity  in a loose sense, some infinities are 'bigger' than others.
Cantor called the 'smallest' infinity alephnull  this is the infinity of the integers: 1, 2, 3, 4, ... but (and this may seem paradoxical) the same infinity holds for the primes: 2, 3, 5, 7, 11, 13, 17, ... or all the even integers, or all the numbers divisible by onemillion, or all the rational fractions.
It seems strange at first  obviously there are twice as many integers as there are even integers, and there are a million integers for every one that is a multiple of a million  and there are a 'lot more' fractions than there are integers. How can we say that all these infinities are the same?
Georg Ferdinand Ludwig Philipp Cantor said that any set of things that could be put in a onetoone correspondence with the counting numbers has the same 'cardinal' alephnull. He then went on to use his famous 'diagonal' proof to show that not only are there are higher infinities than alephnull, but there are an infinite number of them.

05252007, 05:05 AM


Member




Re: Math trivia
Math has a whole 'nother language.

05252007, 06:46 AM


El Caganer


Join Date: Mar 2007
Location: in ur bethlaham


Re: Math trivia
Quote:
Originally Posted by ceptimus
One of the first people to report this fact was the nineteenthcentury astronomer, Simon Newcomb, who noticed that books containing tables of logarithms always showed more wear on the pages of numbers starting with '1'.

Simon Newcomb's brother was the greatgrandfather of physicist William Newcomb, who is best known in freethinking circles as the deviser of Newcomb's Paradox.

05252007, 08:18 PM


mostly harmless


Join Date: Jul 2004
Gender: Male


Re: Math trivia
Quote:
Originally Posted by ceptimus
Different sorts of infinity: This might be slightly heavy for a 'trivia' thread, but I'll try and keep it light.
Infinity does tend to boggle the mind at first, but mathematicians encounter it pretty often and tend to become immune. Cantor showed that there are different classes of infinity  in a loose sense, some infinities are 'bigger' than others.
Cantor called the 'smallest' infinity alephnull  this is the infinity of the integers: 1, 2, 3, 4, ... but (and this may seem paradoxical) the same infinity holds for the primes: 2, 3, 5, 7, 11, 13, 17, ... or all the even integers, or all the numbers divisible by onemillion, or all the rational fractions.
It seems strange at first  obviously there are twice as many integers as there are even integers, and there are a million integers for every one that is a multiple of a million  and there are a 'lot more' fractions than there are integers. How can we say that all these infinities are the same?
Georg Ferdinand Ludwig Philipp Cantor said that any set of things that could be put in a onetoone correspondence with the counting numbers has the same 'cardinal' alephnull. He then went on to use his famous 'diagonal' proof to show that not only are there are higher infinities than alephnull, but there are an infinite number of them.

I love the infinites and transfinites. The wiki on the alephs is rather well done.
__________________
Through with oligarchy? Ready to get the money out of politics? Want real progressives in office who will work for the people and not the donors?

05272007, 05:20 PM


[thanks] whisperer


Join Date: Jul 2004
Location: England/Miisaland
Gender: Male


Re: Math trivia
Great thread, cep, but "math"? Have you become American?

05272007, 05:23 PM


puzzler


Join Date: Aug 2004
Location: UK


Re: Math trivia
Nice to see you posting again Joe!
I make concessions in my use of language to cater for our largely American readership.

10042007, 08:22 PM


puzzler


Join Date: Aug 2004
Location: UK


Re: Math trivia
Nontransitive (unfair) bingo cards: In this much simplified bingo game, only the numbers 1 to 6 are used. The winner is the first person to complete a horizontal row.
The amazing thing is that, with two players, this game is unfair! Let your opponent choose a card first, then you choose the one 'before'. In the long run A beats B, B beats C, C beats D, and D beats A! These cards were devised by the famous Donald E. Knuth.

10042007, 10:28 PM


weird citrus golem


Join Date: Jul 2005
Location: New Zealand
Gender: Male


Re: Math trivia
Clever. Let me see if my intuition is right about this  the way it works is that the two numbers on a losing card are spread across two rows on a winning card. Considering A & B, for example, there's a better chance of either a 1 or 3 showing up before both the 2 & 4, than there is of both the 5 & 6 showing up (also prior to the 2 & 4).

10052007, 09:26 PM


puzzler


Join Date: Aug 2004
Location: UK


Re: Math trivia
I don't know now.
I tried to explain it, but found my mind was boggling on both cylinders after a while...
I googled "nontransitive bingo" hoping to find a clear explanation, but the first result was this thread! (way to go, Google!)
So I figured there are only 6! ways of drawing 6 bingo balls (that's only 720 ways) so I wrote a little program to check which cards win for each of the possible games. I might have a bug in my program, but I found that the cards aren't nontransitive B is the best card, D is the worst.
Code:
Player 1 Player 2 Player 1 Player 2
card card wins wins
A B 336 384
A C 348 372
A D 384 336
B C 408 312
B D 372 348
C D 372 348
I don't understand now. I found out about this from one of Martin Gardner's books. Maybe there is a misprint, but more likely I made a mistake somewhere.

10052007, 10:25 PM


weird citrus golem


Join Date: Jul 2005
Location: New Zealand
Gender: Male


Re: Math trivia
Maybe you need to take into account that there aren't really 6! possibilities? What I mean is that, a game can be over in as little as 2 draws, and by considering each of the 6! possibilities, many of which would never occur in a real bingo game, you're effectively weighting the probabilities.
Take a game of A vs B, for example. Should we assign equal weight to:
(1,2,3,4,5,6)
(1,2,6,5,4,3) and
(1,6,5,4,3,2)?
Or should we allow for shorter sequences and assign equal weight to:
(1,2) and
(1,6,5)?

10052007, 10:30 PM


puzzler


Join Date: Aug 2004
Location: UK


Re: Math trivia
Thanks fragment. That makes sense.
I'll hack my program when I get time, so that it works out the results your way.
Edited to add: Now I've thought about it a bit more, I'm still not sure. Each of those 720 possible ways for the balls to come out is equally likely? My program just runs through all the possibilities and finds which card wins  once a card has won a game, it doesn't care about any remaining numbers. In the card A versus card B example, B wins 384 of the possible games and A only wins 336.
Last edited by ceptimus; 10052007 at 10:42 PM.

10052007, 11:06 PM


weird citrus golem


Join Date: Jul 2005
Location: New Zealand
Gender: Male


Re: Math trivia
Actually, I think I'm wrong now. Even if the numbers aren't drawn, you need to consider each 6! to have equal weighting so that you get the right probabilities, i.e. 1/6 for first draw, 1/5 for second, etc.
One question  what happens when two cards get a row at the same time?

10062007, 08:43 PM


puzzler


Join Date: Aug 2004
Location: UK


Re: Math trivia
Thanks fragment! Of course! I wasn't checking for drawn games! I tweaked the code, and the cards are indeed nontransitive with A B C D A
Code:
Player 1 Player 2 Player 1 Player 2 Drawn
card card wins wins games
A B 336 312 72
A C 288 288 144
A D 312 336 72
B C 336 312 72
B D 288 288 144
C D 336 312 72
So if you pick the card 'before' your opponent's card, then in the long run for every 30 games played, you should win 14, lose 13 and draw 3.
At least I didn't doubt Martin Gardner or Donald E. Knuth. I knew it must be a bug in my code.

10062007, 11:20 PM


weird citrus golem


Join Date: Jul 2005
Location: New Zealand
Gender: Male


Re: Math trivia
Thanks for working that out cep! Saved me from having to do it...

10142007, 12:36 AM


Member


Join Date: Jul 2004
Location: Lebanon, OR, USA
Gender: Male


Re: Math trivia
Here's a curiosity I recently discovered: all the possible isohedral dice. What is an isohedron? It is a polyhedron where all the faces are alike, though that need not be the case for the edges or vertices.
For all edges and all vertices alike, one gets the five Platonic solids:
tetrahedron
cube / hexahedron
octahedron
dodecahedron
icosahedron
But not surprisingly, there are a LOT more isohedra; the mathematics is given in these pages:
Wikipedia on Dice
Mathpuzzle.com: Fair Dice (has pictures of all possible ones)
Math Games: Fair Dice
Klaudius's Properties of Dice (the mathematics behind them)
Wolfram Mathworld: Isohedron
Using Klaudius's classification, there are these possible dice:
Number of Faces, Number of Sides per Face
2*n  3 (n = 4, octahedron)
2*n  4 (n = 3, cube)
4  3 (tetrahedron)
12  3, 4, 5 (5sided: dodecahedron)
20  5 (icosahedron)
24  3, 3, 4, 5
30  4
48  3
60  3, 3, 4, 5
120  3
I've indicated which ones are regular polyhedra. Of the 4sided ones, the 2*n, 12, 24, and 60 have kiteshaped faces and the 2*3=6, 12, and 30 ones have rhombusshaped faces. And in general, the 5sided ones; 12, 24, and 60, have two mirrorimage variants.
The triangularface ones can be constructed by using the faces of regular polyhedra as the bases of pyramids. A nsided face can be turned into a 2*nsided one by adding vertices at the centers of the edges. And the 2*nsided triangular family can be constructed in the same way, but with using an isolated ngon with a pyramid on each side.
ngon: 2*n
tetrahedron: 12, 24
cube, octahedron: 24, 48
dodecahedron, icosahedron: 60, 120
The count: 2 infinite families and 16 extra ones. If one counts variants with moreregular and lessregular faces separately, one gets even more possibilities.
Only some of these possible dice have been produced by commercial dice makers, but you can get them at stores that sell supplies for pencilandpaper roleplaying games.
Some dice makers have made rollingpin dice (extruded regular polygons); these are the only kind that have an odd number of useful faces.

10292007, 10:41 AM


Member


Join Date: Jul 2004
Location: Lebanon, OR, USA
Gender: Male


Re: Math trivia
If you like mathematics and programming, check out Project Euler, which contains big set of mathematics problems presented as a challenge to programmers  who can write programs that solve these problems that are (1) elegant and (2) fast?
The problems range from Add all the natural numbers below 1000 that are multiples of 3 or 5 to Intersections.
I myself do not have much interest in participating, though others here might.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)


Thread Tools 

Display Modes 
Linear Mode

Posting Rules

You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
HTML code is Off



All times are GMT +1. The time now is 02:45 AM.



