Examples
Examples
Try the starting example
- Variants: name, relative weight
- A,50 B,30 C,20
- Deterministic hash bucket count
- 10000
- Users assumed uniformly hashed
- 1000
- Inclusive bucket ranges by variant
- A: 0–4999; B: 5000–7999; C: 8000–9999
- Buckets allocated by variant
- A: 5000; B: 3000; C: 2000
- Expected users by variant
- A: 500; B: 300; C: 200
- Total allocated buckets
- 10000
Try another scenario
- Variants: name, relative weight
- A,1 B,1 C,1
- Deterministic hash bucket count
- 5
- Users assumed uniformly hashed
- 10
- Inclusive bucket ranges by variant
- A: 0–1; B: 2–3; C: 4–4
- Buckets allocated by variant
- A: 2; B: 2; C: 1
- Expected users by variant
- A: 4; B: 4; C: 2
- Total allocated buckets
- 5
Calculation method
Calculation method
Inclusive bucket ranges by variant
Allocate integer buckets by largest remainders, then assign contiguous ranges in input order beginning at zero.
Buckets allocated by variant
Normalised weights × bucket count, rounded by largest remainders; input order breaks ties.
Expected users by variant
Users × variant bucket count / total buckets, an expectation rather than observed assignment.
Total allocated buckets
Sum integer bucket allocations; every bucket must belong to one variant.
Symbols and units
- a
- Variants: name, relative weight
- b
- Deterministic hash bucket count
- c
- Users assumed uniformly hashed
Use the units shown. A list keeps the order in which you enter its values. The symbol ^ means “raised to the power of”.
Assumptions & limits
Assumptions & limits
- Plans integer bucket boundaries and expected exposure under a uniform stable hash. It does not hash user identifiers, enforce privacy, guarantee exact user counts or provide experiment significance.
Are these inputs private?
These calculations run in your browser. A shared link contains the values you choose to share. Only questions submitted to the separate AI finder are sent to OpenAI.
Further reading
OpenStax: arithmetic and geometry ↗
Background reading; this does not imply endorsement or independent review of SumFizz.