A/B Testing Password Strength Calculator Examples
Worked examples show how length, character variety, and predictability can change an A/B password comparison.
These examples use design details only, never real passwords. They show why a longer or more varied design can improve the estimate, while predictable structure can reduce it.
Predictable mixed-character design versus random-looking design
A 12-character design with a 20% penalty is compared with a 16-character design with a 5% penalty.
Input Summary
Password A
12 characters; letters and numbers; 20% penalty
Password B
16 characters; letters, numbers, and symbols; 5% penalty
Calculation Breakdown
- 1Calculate A12 × 5.95 × 0.8057.1 bits
- 2Calculate B16 × 6.55 × 0.9599.6 bits
- 3Compare99.56 − 57.1242.4 bits
Result Summary
Compare
42.4 bits
A/B Testing Password Strength Calculator
Password B estimates 99.6 bits versus 57.1 bits for Password A.
Long lowercase design versus shorter varied design
A 20-character lowercase-only design with a 10% penalty is compared with a 14-character letters-and-numbers design with a 10% penalty.
Input Summary
Password A
20 characters; lowercase only; 10% penalty
Password B
14 characters; letters and numbers; 10% penalty
Calculation Breakdown
- 1Calculate A20 × 4.70 × 0.9084.6 bits
- 2Calculate B14 × 5.95 × 0.9075.0 bits
- 3Compare74.97 − 84.60−9.6 bits
Result Summary
Compare
−9.6 bits
A/B Testing Password Strength Calculator
Password A estimates 84.6 bits, about 9.6 bits above Password B.
Same length, different predictability
Two 15-character letters-and-numbers designs are assigned different penalties.
Input Summary
Password A
15 characters; letters and numbers; 45% penalty
Password B
15 characters; letters and numbers; 5% penalty
Calculation Breakdown
- 1Calculate A15 × 5.95 × 0.5549.1 bits
- 2Calculate B15 × 5.95 × 0.9584.8 bits
- 3Compare84.79 − 49.0935.7 bits
Result Summary
Compare
35.7 bits
A/B Testing Password Strength Calculator
Password B estimates 84.8 bits, compared with 49.1 bits for Password A.
How to Read Your Results
Compare the estimated entropy values first; the higher value is the stronger result within this model.
A positive entropy difference means Password B is higher, while a negative difference means Password A is higher.
Relative strength change is measured against Password A, so it is most useful when A is the baseline design.
Use the predictability penalty consistently and conservatively when a design contains recognizable structure.
Treat results as a comparison of designs, not a security guarantee.
Assumptions & Important Notes
- The examples use the calculator's fixed character-variety estimates.
- Penalties are illustrative estimates rather than measurements of specific passwords.
- No actual password text is entered or analyzed.
- The examples assume the two designs are otherwise comparable.
Related Examples
Frequently Asked Questions
Do these examples use real passwords?
No. They use only length, selected character variety, and predictability estimates.
Why can lowercase-only Password A beat a more varied Password B?
In the second example, A is sufficiently longer to offset B's higher per-character estimate.
What does a negative difference mean in an example?
It means Password A has the higher estimated entropy because the result is calculated as B minus A.
Should I copy an example structure exactly?
No. Examples are for understanding the calculation. Avoid familiar or reused password structures.
Ready to calculate your own result?
Use the live calculator with your own inputs, timing, and preferences.