
Economics Economics Forum  Financial Mathematics, Econometrics, Operations Research, Mathematical Finance, Computational Finance 
 LinkBack  Thread Tools  Display Modes 
October 25th, 2013, 12:53 PM  #1 
Senior Member Joined: Jan 2013 Posts: 209 Thanks: 3  Simplest complete definition of why prediction markets work
For each game of choosing 0 or 1, which agents play many of at once, each cycle, each agent gambles part of its money on 0 or 1. After all gambles have been made for the cycle, for each game, the category (0 or 1) with the least money gambled on it wins and the most money loses. All that money is won by those who gambled on the smallermoneyed category. There is only 1 possible strategy that can consistently win this game: Prediction of prediction of prediction... deeper than the other players are predicting eachother. Therefore only players who understand the market on a larger scale, how other players think, and who cooperate to the extent it increases prediction ability of the group, can win over those who do not. If gambling on each game costs a small percent or flat rate (or whatever function) to pay for computing resources or friction or whatever costs are involved, and that money is moved into the opposite of the streaming bits of input data (like from a camera or text or whatever), then only those which cause the streaming data to be predicted will get the most money recursively. This leaves open the possible algorithms, or even just a slight adjustment of the number of 1 bits as a standard deviation for example, which could choose to gamble 0 or 1 in each game. Its a definition of all possible things that make prediction markets work, not specificly what those things are. If you find one, this will measure it. I have some ideas in mind, as my research continues. Just writing about this part of the theory. 
October 25th, 2013, 06:49 PM  #2 
Senior Member Joined: Jan 2013 Posts: 209 Thanks: 3  Re: Simplest complete definition of why prediction markets w
Just realized something very important about the above theory and the specific "a slight adjustment of the number of 1 bits as a standard deviation for example" which is my recent theory of physics. The prediction market kind is Fermions, particle/wave types which push against eachother so they rarely occupy the same quantum state (those few "quantum numbers" which Pauli Exclusion is about, like an electron cant be in the same part of an atom or anywhere in space as another electron). The new math operator I thought of for general computing, able to represent Conways Game Of Life, Rule 110, and permutations especially those related to Bose Condensate and curved space as overlapping entanglement, is simply an even size set of bit vars which are held to have half the bits be 0 and half 1. Fermions resist overlap. Bosons naturally overlap, as the sets of even size of bit permutations are for, as I thought of that math operator to represent brainwaves in the curved space of how neurons connect to eachother. Fermions cause eachother to balance (the market stabilizes at a buy/sell price). Bosons are more like constraints that agree with eachother on the parts they overlap. The main thing I'm missing before being able to simulate this new combination of math (while I've simulated many parts individually and thats how I know these kind of things) is how the bosons ("even size set of bit vars which are held to have half the bits be 0 and half 1") are prevented from stacking on eachother very high without adding any extra benefit that the constraints are for (pushing bit vars between 1 and 0 so gravity occurs in balance and intelligence results from removing randomness). It appears that the energy level of atoms, as an electron (a fermion) goes up a level per photon (a boson) or down a level when emitting a photon, and generally defines speed or higher orbit, may be what I'm looking for. My sets of even size bit permutations should somehow stack on eachother to change the "prediction market kind is Fermions" view of those bits. I'll keep thinking about it. What do you all think? This is why fermions are "odd" and bosons are "even". Viewed statistically, fermions are bell curve (as we see particles as a bell curve of positions/momentums) and bosons are almost a bell curve except somewhere along their path (many bits which sum to a standard deviation, as pascals triangle describes) they have to come back to the same position they started (same number of 0 bits and 1 bits), standard deviation 0. Since pascals triangle derives bell curve, which I say as ...the standard deviation of the sum of C^2 coin flips, counting tails as 1 and heads as 1, equals C... then there are about (how close is it?) squared more paths through pascals triangle (fermion) than there are paths which end at its center (boson). Theres also lots of math fitting bell curves being made of many other bell curves, soliton waves being bell curve magnitude (varying phase, as in fourier and WaveSim.java (search for it) implementation of 1d schrodinger equation), the fourier of a bell curve (real part of complex) being another bell curve (by magnitude, phase varying), the fractional fourier transform being periodic (repeats after pi number of fouriers, probably related to fractal nature of zeta function's "universal holomorphic" property being divisible by pi, while surprisingly having only integers left in its definition) and said to be somehow "bosonic" (Wikipedia page says)... but I'm getting away from the important simpler thing here... E_boson = M_fermion C_pascalsTriangle^2 / pythagoreanTheorem, normally written as E = M C^2 / sqrt(1  (V/C)^2). As I've said many times but am only recently learning enough to speak the languages of math and science, the universe in total cancels to zero and is balanced from all angles and is simply the set of all possibilities of high dimensional waves, nothing but a bunch of circles/spheres/hyperspheres flowing through eachother. The universe looks like an infinite dimensional bell curve, which mathematically is equal to an infinity1 dimensional hypersphere and a infinity^2 dimensional hypercube (as the Wikipedia page on hypercubes shows how their corners start to look like bell curves in higher dimensions). My main problem with this stuff is how to build some kind of game or puzzle, something on screen, to explain it in terms our complicated minds can understand. You cant even simulate a particle moving through space until you derive the space and derive what it means to be a particle, and how those things form dimensions which an observer can see them in, because if I just tell it draw at x and y on screen, its limited to 2d or 3d, but every particle/wave is its own dimension. Maybe the bosons forming Conways Game Of Life or Rule 110 would be a fun toy to see how hard it is to run Conway in reverse. Its a symmetric operator, but that doesnt mean its equally easy to push in all directions. Some paths have more possible states than others, the core issue of why normal computing doesnt reverse. Its a tangled mess, and to design quantum computers to do those kind of things is a step in the wrong direction. 
October 25th, 2013, 10:00 PM  #3  
Global Moderator Joined: Nov 2006 From: UTC 5 Posts: 16,046 Thanks: 938 Math Focus: Number theory, computational mathematics, combinatorics, FOM, symbolic logic, TCS, algorithms  Re: Simplest complete definition of why prediction markets w Quote:
 

Tags 
complete, definition, markets, prediction, simplest, work 
Thread Tools  
Display Modes  

Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
What is the simplest form of this  naufalzhafran  Elementary Math  2  March 9th, 2014 01:20 AM 
Need Help with Prediction  martor54  Advanced Statistics  2  July 3rd, 2012 04:02 AM 
Trading the markets  Probability quetion  Hero  Advanced Statistics  6  November 13th, 2010 07:20 AM 
simplest  rose  Algebra  6  June 23rd, 2010 12:30 AM 
Resource request: highly inefficient markets  CRGreathouse  Economics  4  April 15th, 2009 09:22 PM 