>>16539433
>If you know the initial conditions exactly, any chaotic system should be completely predictable. Indeed, if you can't predict a system you know the exact initial conditions for, it's not chaotic.
>>16539426
>>16539475
This explains it perfectly. A chaotic system is by definition deterministic. If you replicate *exactly* the same initial conditions, a chaotic system will behave *exactly* the same way when you try to duplicate it's behavior. Unfortunately in the real world its physically impossible to measure *all* initial conditions *perfectly*. If you measure the temperature of a specific spot and it's even 0.000000000001% off, eventually your simulation will wildly deviate from the actual real world behavior. That's why we try to predict weather only on short time scales, and at longer time scales we only try to predict trends.

Non-deterministic systems *do* exist. If you have an unstable isotope, you can say -statistically- when it is likely to decay. You can make a computer model that predicts how a certain amount decays over time. If you measure which atoms decayed at which time however, your model will NEVER match up with the real thing, even if you *did* have an absolute *perfect* measurement of the state of the entire system. If you physically replicated the same thing again, somehow matching *perfectly* everything to the first run, it will still behave slightly differently. This is because radioactive decay is fundamentally non-deterministic. But on average, the behavior of the system as a whole can still be predicted with very high accuracy.

Back | Go to board

Board:

/sci/

Thread:

Science is settled, chud.