Love calculator —
name compatibility, the classic way.

Behind the magic

The classic love calculator algorithm, explained.

A brief history

The love calculator is one of the internet's oldest traditions. Before social media, before smartphones, there were simple HTML pages that took two names and returned a percentage — generating equal parts laughter, arguments, and hope across dial-up connections the world over. The core algorithm dates to the earliest days of the web and has been implemented countless times in countless languages.

The specific version most people remember and the one we use here — is the "letter-frequency pair-reduction" algorithm. It looks mathematical, produces wildly different results for similar names, and gives absolutely identical results every time for the same pair. It has no scientific basis whatsoever, which is precisely what makes it so charming.

How the algorithm works

The algorithm has four clean steps:

1. Normalise the names. Convert both to lowercase and strip everything that isn't a letter. Then sort them alphabetically. This is what makes the result symmetric (Alice + Bob = Bob + Alice, always). Concatenate the sorted pair into one combined string.
2. Count letter frequencies. For every letter a–z, count how many times it appears in the combined string. Discard letters with a count of zero. The result is a compact array of non-zero counts.
3. Pair-reduce the array. Repeatedly match the first element with the last, the second with the second-to-last, and so on, summing each pair. If a sum has two digits, reduce it to a single digit by adding its digits. Repeat until only two numbers remain.
4. Form the percentage. Read the final two numbers as a two-digit percentage: (first × 10 + second). Clamp to 1–99 and you have your love score.

For example: "rose" + "jack" becomes the combined string "jackrose" after sorting. Letter counts: a=1, c=1, e=1, j=1, k=1, o=1, r=1, s=1 → array [1,1,1,1,1,1,1,1]. Pair-reducing: [1+1, 1+1, 1+1, 1+1] = [2,2,2,2] → [2+2, 2+2] = [4,4] → result: 44%.

Why the same pair always gives the same result

Pure determinism. There are no random numbers, no timestamps, no hidden state. Given the same two names, the algorithm executes the same steps and arrives at the same answer every single time. The only inputs are the letters in the names — not how you feel about someone, not the date, not the phase of the moon.

Similarly, swapping the names never changes the score, because the algorithm sorts both names before combining them. "Alice + Bob" and "Bob + Alice" produce identical combined strings, identical frequency arrays, and identical results.

What the score actually means

Nothing, scientifically. Compatibility between two people is shaped by shared values, communication styles, life goals, humour, timing, and a thousand other unmeasurable factors. No algorithm working only on name strings can capture any of that.

What the score does offer is a conversation starter, a moment of shared amusement, and a reason to try your name against a dozen other names until you find one with a suspiciously high percentage. That's the real purpose and it has worked beautifully for decades.

A note on name input

The algorithm uses only the letter characters in each name. Spaces, hyphens, apostrophes, and accents are stripped (with Unicode letters preserved for their base characters) before the calculation runs. Middle names, last names, nicknames; they all affect the result differently, so feel free to experiment with different combinations.