# Let's discuss Mathematics

Discussion in 'Science & Technology' started by ParadigmShifter, Mar 16, 2009.

1. ### SS-18 ICBMOscillator

Joined:
Mar 5, 2006
Messages:
15,266
Location:
Here and there

Joined:
Apr 4, 2007
Messages:
21,810
Location:
Liverpool, home of Everton FC
Actually, that expression comes up during the proof for Fermat's Little Theorem, spoiler here:

http://mathworld.wolfram.com/FermatsLittleTheorem.html

Note x|0 (x divides 0) is true for all non-zero x.

EDIT: For questions such as yours, the answer is usually something to do with Fermat's Little Theorem or its corollary, Wilson's Theorem.

3. ### oagersnapEmperor

Joined:
Mar 2, 2004
Messages:
1,415
Location:
South Schleswig
The problem with your method is that you essentially make a yearly average of the first quarter, a yearly average of the three last quarters, and then simply average the two. This is wrong, because that makes the average of the first quarter become just as important as the average of the three last quarters when calculating the overall average.

Joined:
Apr 4, 2007
Messages:
21,810
Location:
Liverpool, home of Everton FC
Back home in Sci/Tech!

I suppose the move to OT was good for thread publicity.

On topic: you shouldn't really use "average" as a term in the maths thread - the wikipedia page lists 14 different types of averages

Joined:
May 2, 2003
Messages:
1,156
deleted

6. ### UltraworldEmperor

Joined:
May 2, 2003
Messages:
1,156
originally it is someone else's question

7. ### UltraworldEmperor

Joined:
May 2, 2003
Messages:
1,156
but it is per year.

8. ### Gabryel KarolinGammelgädda

Joined:
Oct 21, 2003
Messages:
1,583
Location:
Sweden, Europe
Seriously, forget about the quarters, they don't matter. Think of it as two unspecified timeperiods which together is one full year.

9. ### oagersnapEmperor

Joined:
Mar 2, 2004
Messages:
1,415
Location:
South Schleswig
So what? When you add (average expenses per employee for jan-mar) + (average expenses per employee for apr-dec), you will automatically get (average expenses per employee for jan-dec) which is the number you're looking for. There's no need to "scale" anything.

10. ### nc-1701bombombedum

Joined:
Oct 28, 2005
Messages:
4,025
Location:
America
When we apply Fermat's Little theorem we get (1 + n)^p - n^p - 1 == 1+n -n -1 ==0 mod p
where == denotes congruence modulo p.

11. ### tycoonistDeity

Joined:
Sep 22, 2006
Messages:
3,480
Location:
London
Just from some scribblings on a piece of paper, doesn't this work for p is any natural number (ie not just primes..)

Oh and I didn't use fermat's little theorem, just expanded binomially, then simplified.

Joined:
Apr 4, 2007
Messages:
21,810
Location:
Liverpool, home of Everton FC
(1+1)^4 - 1^4 - 1 = 16 - 1 - 1 = 14 == 2 (mod 4)

13. ### AtticusDeityRetired Moderator

Joined:
Aug 20, 2006
Messages:
3,666
Location:
Helsinki, Finland
Dutchfire is visiting ParadigmShifter in Liverpool to play Scrabble with him. They're going to determine the starter of the first game by tossing a coin, and just as soon as Dutchfire has reached a one euro coin from his pocket, ParadigmShifter stops him and says: "No, we can't use that coin, we're in the UK, we must use pound sterling", and takes a coin out of his pocket. He calls tails and is about to toss the coin when Dutchfire grabs his hand. "I saw what you did there!", he ejaculates. "The coin isn't fair. Due to your queens fat face on the other side, it's bound to land tails more often". ParadigmShifter is infuriated by the insult, but swallows it because of the embarrassment of getting caught: the coin indeed lands more often tails.

Can you help them to use PS's coin to determine the starter of the game in a fair way?

14. ### Miseisle of lucy

Joined:
Apr 13, 2004
Messages:
28,622
Location:
London, UK
Do you know how many times it falls on tails vs heads? If so then you can just toss it 10 times and if tails comes up more than the expected number times then tails wins.

P.S.
lol

15. ### SouronThe Dark Lord

Joined:
Mar 9, 2003
Messages:
5,947
Location:
(GMT-5)
What comes to mind is to spin the coin instead of tossing it. Then hit the coin from the side to knock it over. This way the side it falls on should depend on the silhouette of the coin, instead of the weight, and silhouette is the same from both sides. A problem with this approach is that it would not work if you hit the coin from the top, as in that case the heavier side is likely to be face up.

16. ### AtticusDeityRetired Moderator

Joined:
Aug 20, 2006
Messages:
3,666
Location:
Helsinki, Finland
They have no knowledge of the probability of heads vs. tails.

They also refuse to use any methods that aren't reducible to coin toss since they want a method that can be generalized to other mediums too (die for example).

You're on the right track that they are going to toss the coin multiple times.

17. ### Miseisle of lucy

Joined:
Apr 13, 2004
Messages:
28,622
Location:
London, UK
Well you could "calibrate" it by flipping 100 times or something, to figure our roughly the probabilities, then flip it 10 times and see which beats the calibrated probabilities.

18. ### AtticusDeityRetired Moderator

Joined:
Aug 20, 2006
Messages:
3,666
Location:
Helsinki, Finland
That won't do either, they want an exact method. There is at least one (if we admit that there in general is a fixed probability for the coin (and that's the reason I find this thing fascinating)).

19. ### Miseisle of lucy

Joined:
Apr 13, 2004
Messages:
28,622
Location:
London, UK
Can you flip it 100 times, and then each player guesses whether after 100 more tosses, the number of tails comes out higher or lower than the first 100?

20. ### AtticusDeityRetired Moderator

Joined:
Aug 20, 2006
Messages:
3,666
Location:
Helsinki, Finland
What if they both want to guess the same?
How do they determine which one gets to guess first?

Hint: It isn't predetermined how many times the coin is tossed. Except that the number is even.