# Chaos Numbers

1. Jul 12, 2013

### msticky

I have a list of chaos numbers
Average each 10 results as display below
Sum each 10 results as display below
Average the 30 results as displayed below

here’s my question:
I believe that I can see as the average goes up we have larger sums
So I think I can predict something from the last 9 numbers to tell me about the tenth?

Could I have some feedback about what you can see if anything.

0
3
7
2
1
3
1
2
6
8 3.3 33
3
1
1
3
3
2
4
1
2
5 2.5 25
6
13
3
14
3
3
4
7
6
2 6.1 61
7
6
2
6
11
3
3
2
1

3.96666666666667 3.96666666666667

2. Jul 12, 2013

### Staff: Mentor

Where do your numbers come from?

The sum of 10 numbers is always 10 times the average of them.

Guessing that it will be close to the average of the first 9 is usually not completely wrong (but it can be). It depends on the origin of your random numbers.

3. Jul 12, 2013

### msticky

Lets just say that you can make any amount of arguments about the source, I did call it chaos, but finding the answer to my question and then testing against the data will be the right answer.

As for the 10th number I know the answer.

I feel that the tenth number could be calculated from the examples above?

4. Jul 13, 2013

### Staff: Mentor

Do you think this will be true for all sets of 10 numbers?
Does not look so chaotic if that's the case.

5. Jul 14, 2013

### msticky

Each group of results like above are different in lag.

If you take the results above for 1 test and average the results value then you get the middle. If you average the result values into groups of halves, quarters or thirds you see that you will get different averages with each average for the same result values. Like the above sample the averages are all different.

What I feel you can find in this chaos set of numbers is a lag that clearly shows that there is a change in the averages but what is a elegant way of extracting this value. If we look at the above sample the lag in the result values show a rapid decrease in performance in the latter stage although this is not reflected in the overall average.

So where an expected result may be predicted on the overall average in reality the result would be delayed because the average over the last result values show the average has mostly double?

Imagine a line mark out every 3.96 of some distance, volume or other type.

Now mark out the actual result along that line:
We would start to mark out a rate of 3.3 for 10 results and then there is a slight decrease to 2.5

We would now see that the actual is above the average and in front for a period.

To keep the result within the overall average we must have a correction which is displayed in the last 10 results with a decrease in result

What mathematics can I use to extrapolate this behaviour?

6. Jul 14, 2013

### Staff: Mentor

That is a property of random numbers. They are random, and not the same everywhere.

???
Which "lag" (between what?), which performance (some interpretation of the numbers you did not explain?), some time-ordering of the values (?).

Sorry, I have no idea what you are asking here. I think there is at least a lot of context missing.

7. Jul 14, 2013

### msticky

Thanks anyway