Using maths to see the future

[alext87]

Everything has a pattern e.g. if a coin is flipped there is an equal chance of tails and heads if the coin is fair.

If coin flipped 10 times you would expect 5 heads and 5 tails, in reality, randomness stops this from happening. However, if the coin is flipped a million times then there will be a 50:50 split between heads and tails.

 

Does this mean in theory that you could travel in time as there must be a pattern to everything? And as you got further away from present it would be more accuracy as randomness would have less of an effect?

 

THIS IS ONLY IN THEORY!!

 

Are there any problems, in theory, to this?

[Dave]

You say 'everything has a pattern'. This is not true; there are some truly random events (for example, radioactive decay). Even if everything did have a pattern, it wouldn't really qualify as time travel, rather future predictions based on known events. This already happens in weather forecasting, and they don't usually get it right.

 

Even in a mathematical sense, you can't state a theorem and prove it from lots and lots of evidence; you have to prove it completely from a logical standpoint.

[alext87]

Ok radioactive decay has randomness but it has a half live which can be predicted. Though the radioactivity given off in 1second may appear to be random if 200hours was taken then it would be predictable.

[blike]

I guess the real question is: Are there any true constraints, other than the processing capability, to using physical laws to predict the future?

[Dave]

There's a little thing called the Heisenburg Uncertainty Principle stopping you; it basically says that you can never know both the velocity and position of a particle at any instance in time.

 

alext87: half-life is still only an approximation. There's absolutely no guarantee that after x number of hours the radioactivity will have halved exactly (which is what you're after). You've got to realise that 1) there's far too many variables for all the computers in the world to calculate trajectories of particles 2) it's impossible anyway and 3) there may be objects/collisions that we just don't know about.

 

Even if we overcame those obsticles, we'd need to obtain the starting 'values' for every single particle, so overall it's theoretically and physically impossible.

[swansont]

I guess the real question is: Are there any true constraints, other than the processing capability, to using physical laws to predict the future?

You have to actually know the laws and variables. There are systems that are chaotic, and so very sensitive on initial conditions and correctness of the model. Processing power doesn't really extend the capability solve those equations, because a wrong answer diverges rapidly as you iterate or add terms to the model. Everybody thought that bigger and better computers would help us predict the weather, but it hasn't really - I think better observations and better data have done more than more computer power.

[Tesseract]

Being able to predict long term effects and find patterns in random sets , is called chaos theory. It also says that there can be completely random events from normal equations.Given the theory to be true, it would provide support both for predicting the future and not being able to do it.

[Dave]

As an example of this, take the double pendulum. It works exactly asyou think it should, as long as you don't elevate it above an angle of 90 degrees (I think). There's a nice site on it over at wolfram:

 

http://scienceworld.wolfram.com/physics/DoublePendulum.html

 

It's an example of a very simple system that will show signs of chaotic behaviour quite easily.

[alext87]

Computer power does matter because i am only talking about it in theory! There are chaotic behaviours but there are far more predictable patterns and natural laws so it could be done to predict events to a certain extent in the very long term e.g. the weather system in 20 years time.

[Radical Edward]

the patterns are probabilistic ones, and the trials are generally independent. so over time, you will get to about a 50:50 head to tail ratio, however you will never be able to say if a particualr coin is going to land heads or tails.

[alext87]

No but i am talking about predicting over a long period of time! not just 1 coin but many!!!!!

[YT2095]

Computer power does matter because i am only talking about it in theory! There are chaotic behaviours but there are far more predictable patterns and natural laws so it could be done to predict events to a certain extent in the very long term e.g. the weather system in 20 years time.

I agree to an extent. but what you`de end up with would be a boat load of likelyhoods and one of them would be exactly right, and on occasion non would be.

basicly little better than we have now, although as Dave mentioned regarding the weather, they NEVER get it right for our area even below chance! probability, and Dave only lives up the road and I can see his town from my window. so unless you can factor in EVERYTHING, the likelyhood of an accurate and predicted outcome is very slim, and not entirely dependant on raw "Computing Power"

[Dave]

(I should also point out that weather prediction systems require a couple of supercomputers just to process trivial data such as air pressure, temperature, humidity, etc).

[DimShadow7]

Just wanted to add my two cents into the picture. The main problem with extending the law of large numbers (saying that the actual occurances of x divided by the total number of occurances approaches the probility of x occurring as total number of occurances increases) is only true if each event is completely independent of each other. In the case of time, this is simply not true. When one event occurs it then influences every other event after it. So there is less randomness in the long run of things and more of a directional movement.

[MandrakeRoot]

I dont think that there will be less randomness in the future then there is now. I dont see why flipping a coin would influence future flippings of coins for instance.

 

If you considere many coins then already now you can say something about the ratio. With the same number of coins in the future, your "prediction" will nto be more accurate then it is now.

 

Weather predictions have become more accurate in the last years, but it is still very hard to predict very accurately. Ten years ago the grid used to predict global weather patterns was about 100 km * 100 km, and now it is approx 10 km * 10 km, so tehre is some progress. Predictions will never be always correct you know, that would be contrary to the concept "random". Since if you can always predict accurately then it is not random.

 

Mandrake,

[DimShadow7]

The fact that flipping a coin would not effect future coin flips was my point. On the other hand, say that in the (extreme but point proving )choice between killing someone and not killing them. By not killing them I have allowed them the ability to make decisions that may influence other people in some way or other, which will then domino into a greater scheme of things. By killing them I have virtually nullified all of those possibilities, and those things that could have happened had that person lived, can no longer happen. Thus each event is dependent on the events before it, and therefore there is the directional movement in time, as apposed to the completely isolated and random system that flipping coins poses.

[MandrakeRoot]

You are right. Though i believed we were talking of physical systems beyond the influence of human nature ?

Because you can then also include the possibility of stopping the choatic pendulum,making it perfectly predictable.

 

Mandrake