The utility judges abcdefghijklmnopqrstuvwxyz to be a strong password even though it clearly has a pattern. I understand that since this meter is data-driven, it cannot be expected to identify such patterns. Anyhow,
- Is it feasible to improve the model to understand (such and other) patterns along with those in the training data?
- Does it actually make sense for a password meter to have this capability?