Many strokes can be arranged in icons with four strokes in the icon 4 etc.
Then a given total T can be counted in e.g. 4s by repeating the process ‘form T take away 4’, which can be iconised as ‘T-4’ where the repeated process ‘form T take away 4s’ can be iconised as ‘T/4’ making it possible to predict the counting result through a calculation using the ‘recount-equation’ T = (T/b)*b, where the number T/b is called a per-number describing the total and the bundle-size.
Thus counting a total of 8 in 2s there are T/b = 8/2 = 4.
In survey, numbers change unpredictably. We can set up a table accounting for the frequency of the different numbers. From this we can calculate the average level and the average change. The average level can then be used as the winning probability p in a game that is repeated n times. By counting the different possibilities it turns out that there is a 95% probability that future numbers lie within an interval determined by the average level and change.
When adding stacks OnTop, overloads can be removed through internal trade where a full stack of 10 1s is traded with one 10-bundle using the restack-formula T=(T-b)+b.
Per-numbers must be transformed to totals before being added.
Thus 2days at 6$/day + 3days at 8$/day = 5days at (2*6 + 3*8)/(2+3) $/day.
And 1/2 of 2 cans + 2/3 of 3 cans = 3 of 5 cans = 3/5 of 5 cans.