Friendship Calculator by Name

From two names to eight friendship qualities

It compares two names across six measurements, turns those into eight friendship qualities, and gives the pair a named Friendship Personality based on whichever two qualities come out strongest.

If you tell it how long you have known each other, or how old you both are, those figures nudge particular qualities: a long friendship raises the dependable ones — trust, loyalty, emotional support, reliability — while being close in age mostly raises fun and similarity. The result shows each adjustment as an exact number.

It is a word game with a friendship theme. The measurements are real properties of the text; the friendship interpretation on top of them is for fun.

How it differs from the love version

These two calculators look similar and share almost no arithmetic.

The love calculator by name measures vowel patterns, distinct-letter set overlap, alphabet distance between initials, and longest common subsequence. It is built around sound and structure.

This one measures positional matching, consonant balance, initial bonding and ending match. Only shared letters appears in both. The emphasis is different because the things people notice about a friend's name are different — matching initials, names that rhyme at the end, names that feel like a matched set.

It also has something the love version does not: optional inputs that actually modify the output. That makes it the only one of the two whose result depends on anything beyond spelling.

Who plays it

Friends messing about, mostly. It is the sort of thing that gets run in a group chat, screenshotted, and argued about for ten minutes, which is precisely the right use.

The Friendship Personality is what people actually share. A percentage is forgettable; "The Partners in Crime" is a label people adopt. Groups tend to run every pairing and collect the archetypes, which produces a surprisingly entertaining map of a friend group.

Younger users in particular treat it the way earlier generations treated magazine quizzes — not as information, but as a prompt. That is fine, and it is why the tool is careful to be playful rather than authoritative.

How to use it

1. Enter both names. First names work best for this one, since friends rarely think of each other by surname.
2. Add the years you have known each other if you want the loyalty and trust side to reflect it.
3. Add both ages to bring fun and similarity into play.
4. Press Check, and read the Friendship Personality first — it is the part worth screenshotting.

What each input means

The two names

Required, minimum two letters. Case and spacing are normalised; accents and non-Latin scripts are preserved and counted properly.

Years you have known each other

Optional, 0 to 100. The effect saturates at twenty years, so a twenty-year friendship and a forty-year one receive the same boost. Below ten years the modifier is negative, which is not a judgement — it simply means a newer friendship has not yet accumulated the thing being measured.

Both ages

Optional, 1 to 120, and only used when both are supplied. What matters is the gap, not the ages themselves. The effect saturates at twenty years apart.

Gender

Optional, displayed on the result, excluded from the scoring. There is no rule about which combinations of genders make better friends, so the calculator does not invent one.

The six measurements

Shared letters

The Sørensen–Dice coefficient over letter multisets — twice the matched letters divided by the total letters in both names. The one measurement this shares with the love version, because it is the core of every name-matching game ever played.

Positional match

How many letters land in the same position when the names are lined up from the start, divided by the shorter name's length. "Sam" and "Sal" match in two of three positions. "Sam" and "Mas" contain the same letters and match in none.

This is stricter than the subsequence measure the love calculator uses, and it rewards names that genuinely look alike rather than merely sharing an alphabet.

Consonant balance

Each name's consonant proportion, compared. The love calculator looks at vowels; this one looks at consonants, which weights the comparison towards the harder, more percussive parts of a name.

It is not simply the inverse of the vowel measure once other characters are involved, and it produces a noticeably different ranking on names with clusters — "Bridget" and "Brendan" fare better here than on a vowel comparison.

Initial bond

A shared first letter scores the maximum. Anything else falls back to alphabet distance, scaled down so it can never rival an actual match.

This is the most deliberately generous measurement in the set, and it is that way on purpose: matching initials are something friends notice, name themselves after, and put on group chat titles. Rewarding it is true to how the game is actually played.

Name length comparison

One minus the difference in letter count over the longer name. Names of equal length score full marks.

Ending match

How many final letters match, reading backwards until they stop, divided by the shorter name's length. "Jenna" and "Anna" share "n", "n", "a" at the end. This is what makes rhyming names score well, and it has no counterpart in the love calculator.

The eight friendship qualities

Each quality blends three of the six measurements, with its own weights.

The blends are chosen for flavour. Loyalty leans on the initial bond and the ending, because those are the parts of a name that stay fixed; similarity leans on shared letters and consonant balance, because those describe how alike the names are overall. It is a metaphor, and it is labelled as one.

What the eight qualities actually mean

The numbers are a game, but the categories are not. These eight words describe things that genuinely vary between friendships, and the labels are worth reading on their own terms.

Trust

Whether you can say something unflattering about yourself and know it will not travel. Trust in friendship is mostly about discretion, and it is usually established early and broken once.

Loyalty

Whether they defend you when you are not in the room. Distinct from trust: someone can keep your secrets and still fail to stand up for you. Loyalty is the quality people most often discover they were wrong about.

Communication

How easily the conversation moves, and whether it survives silence. Some friendships need constant contact; others resume mid-sentence after a year. Both count as good communication, which is why frequency alone is a poor measure of it.

Fun

Whether you enjoy each other's company with nothing else going on. It is the most underrated quality on this list, because it is the reason friendships start and often the reason they continue when nothing else obliges them to.

Understanding

Whether they read you accurately — spotting when you are annoyed, or joking, or covering. It is what makes a friend easy to be around when you are not at your best.

Similarity

How much you have in common: humour, background, tastes, pace. It makes friendships easy to start and is far less important than people assume for keeping them. Plenty of long friendships are between people with almost nothing in common.

Emotional support

Whether they are useful when something goes wrong. This is where friendships get sorted quite sharply, because it costs something. It is also the quality people are least able to predict in advance.

Reliability

Whether they turn up when they said they would. Unglamorous, easy to overlook while everything is going well, and the single quality most cited by people describing a friendship they value.

How the optional inputs change things

This is the part that makes the calculator more than a spelling comparison, so it is worth understanding.

Years known is converted to a strength between 0 and 1, reaching full strength at twenty years. That strength is centred — so ten years is neutral, more is positive, less is negative — and then scaled by each quality's sensitivity. Loyalty is the most sensitive, followed by reliability, emotional support and trust. Fun and similarity ignore it entirely.

Age closeness works the same way, reaching full strength at identical ages and zero at twenty years apart. Similarity is the most sensitive, then fun. Trust, loyalty and reliability ignore it.

The direction of both is defensible even inside a word game. Trust and loyalty genuinely are things that accumulate over years — that is close to what the words mean. Shared cultural reference points and a similar life stage genuinely do make things easier to find funny together. The modifiers do not make the calculator accurate, but they do make it point somewhere sensible.

The friendship personalities

The archetype comes from your two strongest qualities, looked up in a fixed table. The lookup is order-independent, so trust and loyalty gives the same result whichever ranks first.

Where a combination has no named archetype, a fallback is built from the single strongest quality, so every pair gets something rather than a blank. The result says which two qualities produced the archetype, so you can see the reasoning.

Worked example: two names

Worked example: adding the extras

What actually makes friendships last

Since this page has spent several thousand words on a word game, it is worth being straight about the real thing for a moment. None of what follows comes from the calculator; it is ordinary observation, offered because it is more useful than a percentage.

Frequency beats intensity, up to a point. Friendships that survive tend to involve regular low-effort contact rather than occasional grand gestures. A message every couple of weeks does more than an annual dinner.

Someone has to do the organising. Adult friendships rarely maintain themselves. In most lasting friendships one person is slightly more willing to suggest a date, and the friendship survives because they keep doing it.

Repair matters more than avoidance. Friendships that last are not the ones without friction; they are the ones where something is said afterwards. The ability to say "that landed badly" and move on is worth more than never causing offence.

Shared history compounds. This is the one thing the years-known modifier gestures at honestly. Long friendships are genuinely different, not because time is magic, but because accumulated context means less has to be explained.

Proximity does most of the work early on. Most friendships begin because two people were repeatedly in the same place. When that stops — a job change, a move — friendships need deliberate effort or they fade, and that fading is normal rather than a failure.

Mapping a whole friend group

The archetypes are more interesting across a group than between two people, and the exercise takes about ten minutes.

Run every pairing in the group and write down the personality each pair gets. With six people that is fifteen pairings, and what emerges is a map: who ends up as "The Reliable Duo", who are "The Partners in Crime", which pairs get something unexpected. The results are arbitrary, which is exactly why the argument about whether they are right is entertaining.

A variant worth trying is to have everyone predict their own archetype first, then run it. The gap between how people see a friendship and what a letter comparison produces is reliably funny, and it makes the point about what the tool is measuring better than any disclaimer.

The years-known field adds a further layer in a group setting, because it is the one input that reflects something real. Running the same pair with and without it shows how much of the result is spelling and how much is the modifier — usually the answer is "mostly spelling", which is worth seeing.

One suggestion if you do this in a group: run it with nicknames as well as full names. Watching a pair jump twenty points because someone goes by a shorter name removes any lingering temptation to take the numbers seriously.

Why name games stick around

Matching names is not a modern invention and it has never been confined to romance. Friendship versions of the same ritual show up everywhere, usually among children and usually involving initials.

Matching initials in particular carry real social weight. Friend groups name themselves after them, group chats get titled with them, and two people who share a first letter will mention it within about a minute of meeting. That is the reason this calculator treats a shared initial generously rather than as one signal among many — the game is being modelled as it is actually played, not as a neutral comparison.

Rhyming and matching endings do similar work. Names that end alike get treated as a pair by other people, which is why the ending match measurement exists at all. There is no reason two friends called Jenna and Anna should be closer than any other pair, and yet a name pattern like that genuinely does get commented on.

What all of these have in common is that they are prompts for attention rather than sources of information. A name game gives two people a reason to compare themselves and a vocabulary for doing it. The comparison is arbitrary; the attention is not. That is a decent account of why a tool like this is worth building carefully even though its output means nothing.

What it cannot tell you

Whether two people are good friends. It has two names and, at most, a duration and two ages. None of that describes a friendship.

Which of your friends is closest. It can rank your friends by how much their names resemble yours, which is not the same question and is not an interesting one.

Anything about compatibility as a real property. The eight qualities are recognisable English words attached to blends of letter measurements. The words are meaningful; the numbers under them are not.

If you want something that reflects a real friendship, the friendship calculator asks you to rate ten aspects of it and builds a profile from your answers. Its output is only as good as your honesty, which is a far better foundation than spelling.

If a low score bothers someone

It happens, particularly with younger users, and it is worth addressing directly rather than hiding behind a disclaimer at the bottom of the page.

A low number here means two names are spelled differently from one another. That is the entire content of the result. It carries no information about whether two people like each other, how much they have been through together, or whether the friendship is any good.

The quickest way to prove this to someone is to run the same pair again using nicknames, or middle names, or surnames. The score will move, sometimes by twenty points or more, without anything about the friendship having changed in the intervening thirty seconds. Once someone has seen that happen, the original number stops looking like a verdict.

There is a broader point here about tools of this kind. A calculator that produces a confident-looking percentage invites people to treat it as a measurement, and the more polished it looks the more that happens. That is why this page shows its measurements openly, says plainly what they are derived from, and repeats that it is a game. A tool built for entertainment should be enjoyable, and being enjoyable includes not leaving someone feeling worse about a real friendship.

Common misunderstandings

Thinking a low score says something about the friendship

It says the names share few letters. Nicknames will change it substantially.

Expecting the same archetype every time

The archetype follows the top two qualities, and the optional inputs move those. Adding years known can legitimately change the personality.

Assuming a negative modifier is a criticism

Below ten years, the years-known modifier is negative simply because it is centred at ten. A three-year friendship is not being marked down as a friendship; a number is being centred.

Comparing with the love version

Different measurements, different weights, different modifiers. The two have no reason to agree.

Reading too much into the archetype

It is a label drawn from a small fixed table, chosen by two numbers that came from spelling. It is meant to be shareable, not accurate.

Using surnames for one and not the other

Length comparison and positional match both suffer. Keep the style consistent.

Assuming a high score means a close friendship

Two complete strangers with similar names will score well here. The number describes spelling alone and has no opinion whatsoever about whether the two of you have ever actually met.

Limitations

Positional match and ending match both compare from a fixed end of the name, so they are sensitive to spelling in ways that do not correspond to how names sound. "Catherine" and "Kathryn" sound nearly identical and match in almost no positions.

Consonant balance uses the five Latin vowels to decide what a consonant is, so it is not meaningful for scripts that do not work that way.

Initial bond is deliberately generous to shared initials, which means two friends called Sarah and Sam will always do reasonably well on loyalty regardless of anything else.

Both modifiers saturate, so a fifty-year friendship and a twenty-year one are indistinguishable to the calculator.

The archetype table is finite. Uncommon combinations fall back to a single-quality description, which is less characterful.

Frequently asked questions