Skip to content
2xKit

How Password Strength Is Actually Measured (Entropy, Not Just Length)

Why 'P@ssw0rd!' is weaker than a long string of random lowercase words, and what entropy actually measures.

Quick answer

Password strength is measured in bits of entropy, roughly the base-2 logarithm of the number of guesses an attacker would need to try, which depends on both the size of the character pool used and the password's length, not on how random it looks to a human. A password like "correct horse battery staple" can have more entropy than "P@ssw0rd!" despite looking simpler; check any password's real entropy with the Password Strength Checker.

It's tempting to judge a password's strength by eye, does it have symbols, numbers, mixed case, and it looks strong. But real password strength is a mathematical quantity called entropy, measured in bits, and it depends on the size of the character pool a password draws from and its length, not on whether it happens to look scrambled to a human observer.

What entropy actually measures

Entropy answers one question: how many guesses would an attacker who knows the password's format need to try, on average, before finding it? If a password uses a pool of N possible characters per position and is L characters long, the total number of possible combinations is N^L, and entropy in bits is log2(N^L), or L × log2(N). A 8-character password using only lowercase letters (a pool of 26) has about 8 × log2(26) ≈ 37.6 bits of entropy. Adding uppercase, digits, and symbols expands the pool to around 95 characters, so the same 8-character length jumps to about 8 × log2(95) ≈ 52.6 bits, a real but modest improvement.

Length matters more than pool size once you compare it directly: extending that same 95-character-pool password from 8 to 16 characters roughly doubles the bits of entropy to about 105 bits, a far bigger jump in guessing difficulty than switching from lowercase-only to a full symbol set at a fixed short length. That's the mathematical reason a long passphrase of random lowercase words routinely out-scores a short password stuffed with symbols.

Why 'P@ssw0rd!' scores badly despite looking complex

Entropy calculated from character-pool size assumes each character is chosen independently and unpredictably. Real password crackers don't brute-force character by character, they use dictionaries of common words, known substitution patterns (@ for a, 0 for o, ! at the end), and lists of billions of previously breached passwords. "P@ssw0rd!" technically draws from a large character pool, but because it's a well-known word with well-known substitutions, it falls in the first few thousand guesses of any competent cracking tool, its theoretical entropy massively overstates its real-world resistance.

A password strength tool worth trusting checks both: the raw theoretical entropy from length and character variety, and pattern-based red flags like dictionary words, keyboard walks (qwerty, asdf), and repeated characters. The Password Strength Checker evaluates both dimensions and estimates realistic crack time rather than just reporting a character-pool number that can be misleading on its own.

Generating passwords with real entropy

The most reliable way to get a genuinely high-entropy password is to generate one randomly rather than inventing one by hand, since human-created passwords consistently cluster around predictable patterns even when people are actively trying to be random. The Password Generator produces cryptographically random passwords at a chosen length and character set, and the Passphrase Generator builds long, memorable multi-word passphrases using genuinely random word selection, which is often easier to type and remember than a random symbol string at equivalent entropy.

Frequently asked questions