I normally count my numbers in intervals that based off the power of 2, then using iteration for anything that isn't a number represented by the power of 2.
For example, the usual power of two:
1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024, 2048, 4096, 8192, 16384, 32768 etc.
To get lets say 73, I pick the biggest number that's closest in range to it: 64.
I then iterate until I get 73 by using progressively smaller numbers, for example:
- 1st iteration: 64
- 2nd iteration: 64+8 = 72 (as 16 and 32 obviously don't fit, I can only use numbers that are 8 or smaller in order to fit)
- 3rd iteration: 72 + 1 = 73
I find it much easier that way than simply counting it in linear intervals, ie. 1, 2, 3, 4 or multiples like 4, 8, 12, 16. However this does not apply to certain things like counting money, as regular intervals are easier to use. That said, I still iterate by the power of 2 when guessing how much there is and then comparing it to the actual amount to see how far off I was.