Comparison of Two Models

[NortonH]

My previous thread was closed because i dared to stray from one subject and mention the other.
So this thread is specifically created for the purpose of discussing the comparison and contrast between climate models and the models you might use to calculate lotto numbers.

Numbers of variables.
The lottery model contains 36 identical pingpong balls.
A climate model requires hundreds of parameters.
https://wattsupwiththat.com/2011/06/30/earths-climate-system-is-ridiculously-complex-with-draft-link-tutorial/

https://wattsupwiththat.com/2012/01/21/the-ridiculousness-continues-climate-complexity-compiled/#comment-872002

Starting positions
The initial positions of the 36 balls can be determined very accurately. Less than 0.1mm precision.
The hundreds of climate parameters are less easy to measure.

Models
The ping pong balls can be described by models which require few laws of physics. Newtonian laws of motion, conservation of mass, momentum and angular momentum pretty much covers it.
The climate models are harder with the hundreds of variables all interacting in different ways

Results
We are trying to use the current state of the ensemble to predict the future trajectory of the system. That means that in ten seconds time when we select a ball we want to know which one it will be.
In climate we want to know in ten years time the state of the parameters, paticularly temperature.

Many people get confused at this stage and start talking about taking averages of thousands of different possble scenarions. A Monte Carlo simulation for example.
But that is not what we are trying to do.
 This problem is about determining the next single trajectory, not the average of all possible trajectories and that is a major difference.

Conclusion.
Given that we can model the climate very accurately and made loads of useful predictions why has nobody bothered to model the lottery?
 

[Sensei]

1) lottery numbers randomized this week are practically disconnected from numbers randomized the next week, and two weeks later. Somebody who would place numbers on timeline graph should see no correlation between them. Each number should have equal chance to be picked. The more serious lottery, the better balls and equipment are checked whether their mass, size and properties of plastics etc. etc. are variable as less as possible.

2) humans are measuring energy that is arriving to the Earth from the Sun. In the old times it was done on the surface of Earth and measured 1050 W/m^2 in the noon. Later it was done using balloons. And in modern times after launching rockets, satellites can be do it from orbit (1360 W/m^2). This gives information about Sun's activity and input of energy which is second by second heating Earth. It's extrapolated to the entire surface of the Earth which is pointing toward the Sun. Thermometers (and other equipment) are placed on the entire Earth. Satellites are monitoring temperature of surface from the orbit. All these data are put on timeline graph. Physical parameters can be interpolated and extrapolated using various curves (starting from linear interpolation).

https://en.wikipedia.org/wiki/Interpolation

https://en.wikipedia.org/wiki/Extrapolation

There are correlations between today temperature,pressure,humidity etc. other weather parameters, and the last week, the last month, the last year parameters.

I would compare "climate change predictions" to predicting "chess game winner" in the middle of game (when we have gathered enough data). You can extrapolate data from observations of chess playfield to predict the all possible outcomes in advance.

Edited November 17, 2018 by Sensei

[John Cuthber]

I have a model of the UK national lottery.
I predict that it will draw 7 balls with numbers between 1 and 59.

 

Does my model produce reasonable predictions?