Page 10 — Compatibility

Three axes, and the rule that produces them

You came for the percentage. It does not exist — pairing four-letter codes and following what happens to the couples has been done, it is very little, and it supports no ranking at all; anyone publishing a round number per pairing is making it up. This page has been wrong twice, and both corrections are here. The first version used a single index of function overlap, and an overlap index rewards sameness: first place was always another INTJ. The second split the ruler in two — recognition, how easily you are understood without translating, and complement, how much the other person supplies that your stack does not generate alone — and fell into the symmetrical error: with nothing measuring what daily life costs, it recommended your opposite on all four letters. Hence the third axis. Livability measures how much of ordinary life needs no negotiating.

The rules are published before the numbers and recomputable by hand in a minute: it is the only honest way to build a page like this. No type occupies all three corners at once, and that is not a defect in the arithmetic — it is the shape of the sixteen stacks. The mirror understands you without translating and shares every one of your preferences, and still lands mid-table, because it extends you in nothing. The ESFP holds your four functions in nearly inverted order and is your opposite on all four letters: it comes last. And the ENFP, which the entire internet sold you as destiny, is not first, or near it. The order is a reading order, not a verdict. The evidence section, which comes before the table on purpose, explains why none of it predicts anything about your life.

These three axes are a construction of this site. They measure nothing. Pairing MBTI types to predict the satisfaction, duration or quality of a relationship has been attempted, and the little that exists supports no ranking at all: in a study of 62 couples in clinical care, type similarity between spouses produced no significant difference in marital satisfaction, positive affect or conflict regulation (Kong, 2010). The foundation that safeguards the use of the instrument — which is not the company that publishes it — is explicit that it was not designed to select people and that it does not measure ability, competence or skill. The four-function stacks the rules operate on come from Grant, Thompson and Clarke (1983): they are one model’s convention, not an experimental finding, and type dynamics is precisely the part of the MBTI with the weakest evidential record. Read recognition as vocabulary overlap between two heads, and complement as a theoretical bet about what is missing from yours. It is not a prediction, it is not advice, and it is not permission to end anything.

What the research shows

The best news on this page is that the numbers do not matter

There is one study large enough to settle the argument, and it is not about the MBTI — it is about personality in general, which is where the data live. Dyrenforth, Kashy, Donnellan and Lucas (2010) used nationally representative samples of married couples in Australia, the United Kingdom and Germany — 23,250 people, around eleven thousand six hundred couples — to separate three things that almost always arrive tangled: the effect of a person’s own traits, the effect of the partner’s traits, and the effect of the similarity between the two. The third is what pages like this one sell. With the first two accounted for, it explained less than 0.5% of the variance — in relationship satisfaction and in life satisfaction too. Less than half of one per cent. What predicted satisfaction was each person’s own traits: a person’s own accounted for around 6% of the variance in relationship satisfaction; the partner’s, for 1% to 3% — and there the heaviest were agreeableness, conscientiousness and emotional stability.

Notice what that does to the recognition axis in particular. It is, by construction, a measure of similarity — and similarity is precisely the variable that barely showed up. The complement axis fares no better: nobody has measured that either, and the notion that opposites complete each other has, in the literature, even less support than the notion that likes understand each other. Both axes on this page are in the same boat, and the boat is a theoretical construction.

The nuance comes from Montoya, Horton and Kirchner (2008), a meta-analysis of similarity and attraction. It does attract: the effect is robust when what is measured is initial attraction, especially in designs where one person rates another they have not really met. In existing relationships, the effect of actual similarity stops showing up, and what goes on predicting there is perceived similarity. Which is to say: similarity predicts well who you will find interesting in the first conversation, and badly how the two of you will be doing in year three. The table below, and every compatibility list ever handed to you as an INTJ, measures the first thing and sells it as the second.

Translated into your life, without hedging: you do not need to find an ENFP. There is no right person waiting with the correct four letters, and you owe nobody an explanation for having fallen for an ISFJ. The useful question was never which letters — it is whether both people are mature enough, and maturity here has a practical, uncomfortable definition: knowing what you feel before anyone asks, saying it before it turns into filed resentment, and changing the plan when the data change. None of those three is a function of a four-letter code. All three are a function of work done — which is, point for point, what the Journey page is about.

And to close the circle honestly: almost none of this tested type. What exists is little, small and unreplicated, and nothing in it supports a ranking by four-letter code. Anyone trying to do better would hit an earlier problem first: readministered a few weeks apart, the instrument returns at least one different letter for close to half of people in the classic studies, and type dynamics — the very basis both of this page’s rules are built on — has never assembled consistent evidence. The Myers & Briggs Foundation is explicit: the instrument was not designed to select people and does not measure ability or competence. A compatibility page is precisely that, selection and prediction. Which is why it appears here with the rules printed over the top of it, rather than as a lone number on a screen.

  • Dyrenforth et al. (2010), 23,250 married people in Australia, the United Kingdom and Germany: with each person’s own traits accounted for, similarity between partners explains less than 0.5% of the variance in relationship satisfaction — and in life satisfaction.
  • What predicted satisfaction was each person’s own traits: their own, around 6% of the variance; the partner’s, 1% to 3% — the latter largest for agreeableness, conscientiousness and emotional stability.
  • Montoya, Horton & Kirchner (2008): actual similarity predicts initial attraction and stops predicting in existing relationships — there, only perceived similarity holds up.
  • The few studies that have paired MBTI types and measured satisfaction support no ranking at all. These two axes measure nothing either.
  • The Myers & Briggs Foundation: the instrument was not designed to select people and does not measure ability, competence or skill.

The two axes, type by type

The mean of the three axes, less almost half the distance between the highest and the lowest — you can check the arithmetic against the three figures on the same row. The deduction is the reason there are three axes: recognition without complement is a mirror, complement without livability is a daily effort, and neither is a match. Nobody comes near 100, and nobody scores zero.

INTJ × ENTJ

Stack: Te · Ni · Se · Fi

45Combined
70Recognition
38Complement
70Livability

How the two numbers were made

Ni 1st ↔ Ni 2nd+ +28 The same long-range vision, running second for them.
Te 2nd ↔ Te 1st+ +21 The same criterion of effectiveness, in their command seat.
Fi 3rd ↔ Fi 4th+ +14 The same private compass, even more mute in them.
Se 4th ↔ Se 3rd+ +7 Your inferior is their tertiary, one rung less blind.
Dominant Te — extraverted+ They go towards the world first. But no Fe: nobody here manages the temperature of the room.

What draws you in

There is no translation to do. You say something hard at full speed and it lands as information, not as aggression. It is the specific relief of not having to wrap anything — and of watching somebody take your half-finished idea and already be moving people with it.

Where it grinds

Command. Two people who have already decided, and the same function in first and second place means you arrive at the same place along routes with swapped priorities: you want the model closed before moving, they want to move in order to discover the model. Each reads the other as reckless, for different reasons.

What they need from you

To know the plan exists before you have finished it. The ENTJ reads silence as a vacuum and a vacuum as an invitation — if you do not say you are thinking, they will already have decided for two, and not out of bad faith: out of horror at stalled time.

What you need from them

Their willingness to start with the model at seventy per cent, which is the exact antidote to your habit of losing years perfecting the plan. And the social front they run for free, which would cost you the entire day.

How this pairing usually fails

It turns into a company rather than a relationship. Two competent operators, a full diary, a clean division of responsibilities — and Fi third in one and fourth in the other, which means <em>nobody</em> in this pairing holds the function that says the feeling part. The years pass with remarkable efficiency and one day one of you notices that the last conversation about the relationship itself was in the previous decade.

Who has already answered this

Three traditions, three ideal partners, and no measurement

The question “what is the INTJ’s ideal match” has been answered confidently by several models, and this is where it gets interesting: they do not agree with each other. They do not disagree over detail — they disagree over the middle letter. If three formalised systems point at three different people, the question does not have an answer hidden somewhere: the question is malformed. The disagreement is worth looking at closely, because it dissolves the subject better than any footnote could.

Keirsey

ENFP — any Idealist

Keirsey works with four temperaments rather than function stacks, and the INTJ is always a Rational (NT). The pairing advice in Please Understand Me II (1998), inherited from Please Understand Me (1978), is that Rationals do best with Idealists (NF) — the same abstraction, the opposite handling of people. That is, ultimately, where almost all of the fame of INTJ × ENFP comes from.

It is one author’s clinical hunch, presented as the fruit of decades of counselling and never submitted to an outcome test. Keirsey, incidentally, is also the one who warns hardest about the arrangement: he named the “Pygmalion project” after the tendency of each partner to reform the other in their own image, which is exactly what an INTJ does to an ENFP once the enthusiasm becomes an item on the audit.

Socionics

ESFP — the SEE

Socionics has the most formalised duality theory of any of them, assembled by Aušra Augustinavičiūtė from the 1970s on an eight-function structure. And it does not agree with Keirsey. The INTJ’s functional analogue is the <strong>ILI</strong> — introverted intuition in the base position, extraverted logic as the creative function, the same Ni–Te pair — and the ILI’s dual is the <strong>SEE</strong>, the sensory-ethical extratim, the ESFP analogue. Not the ENFP: the ESFP. The ideal partner, in that tradition, is the person who lives in their body and says what they feel on the spot.

And there is a disagreement inside the disagreement, worth knowing before citing this anywhere: the correspondence between socionics and the MBTI is disputed, and the J/P letters mean different things in the two systems. Mapped by stack, the INTJ becomes ILI and the dual is SEE (ESFP). Mapped by letters, the INTJ becomes LII, whose dual is ESE — the ESFJ analogue. Two routes inside one model, two different ideal partners, and the ENFP appears in neither. Duality has never been tested on real couples in a controlled design; the correspondence between the systems itself is estimated at something near 30% of cases.

Community consensus

ENFP

In forums, subreddits and MBTI videos, INTJ × ENFP circulates as the “golden pair” with a technical-sounding argument: the two supposedly share the same functions in swapped order. It is worth checking that claim against the stacks, because it is false. You are Ni · Te · Fi · Se; the ENFP is Ne · Fi · Te · Si. The shared functions are two, Te and Fi, with their priorities reversed — and the perceiving axis, where each of you is strongest, is inverted in attitude.

Folklore, and it should be treated as such: there is no publication, no data and no instrument behind it. Where the genealogy is traceable it runs back to Keirsey by way of the internet — NT with NF — and not to any measurement. One true thing survives in the middle of it: with Fi second, the ENFP genuinely is one of the few types who read your flat tone correctly. That is an observation about mechanism, not evidence about outcomes.

Official MBTI

None

The tradition with the strongest claim to the acronym is the one that refuses to answer. Myers published no table of couples in Gifts Differing (1980); the 1998 manual carries no compatibility coefficient between types; and the Myers & Briggs Foundation states explicitly that all types are equally valuable, that results are never to be used to limit anyone, and that the instrument was not designed to select people, because it does not measure ability, competence or skill.

It is the dullest of the four positions and the only defensible one. What the official literature offers is vocabulary for describing friction after it shows up — not a criterion for choosing somebody beforehand. And Pittenger’s (2005) warning closes the subject: the more generic the description of a pairing, the truer it reads to whoever is already in one.

Three models, three different ideal partners — ENFP, ESFP and, inside the same socionics by the other mapping route, ESFJ — plus a fourth tradition that refuses to name anyone. None of the three guesses has been measured: no cohort, no follow-up, no outcome. When independent models disagree like that about a question none of them has tested, the disagreement is the finding. That, and not the table below, is what you should take away from this page.

The rules

How each number was produced

The rules operate on the four-function stacks — Ni · Te · Fi · Se in your case — following Grant, Thompson and Clarke’s convention. Two properties of those stacks make the sum simple and checkable. First: every type carries exactly one function from each axis — one intuition, one sensing, one thinking, one feeling. Second: the two perceiving functions in a stack always hold opposite attitudes, which means no type carries Se and Ne at the same time. Recognition walks your four functions and asks where each one sits in the other stack. Complement walks three things your stack does not supply and asks how much of each the other person has. Livability does not look at functions at all: it looks at the four letters, which is where the cost of a Tuesday lives. All three run 0 to 100, and no type leads more than one.

Starting point, on all three axes0Nothing is given. An earlier version of this page opened every pair at 40 so the table would sit inside a tidy range, and that was cosmetics: three zeroes tell the truth.
Recognition · weight of each of your functionsNi 40 · Te 30 · Fi 20 · Se 10The axis asks how much of you the other person also runs, and your four functions are not worth the same: Ni is the command and Se is the remainder. The weights sum to exactly 100, which is the ceiling of the axis.
Recognition · discount for positional distance×1 · ×0.7 · ×0.4 · ×0.1Same position in the stack pays the full weight; one place apart pays 70%; two, 40%; three, 10%. Te second in both is the same criterion at the same moment of the decision; Te second against Te fourth is the same tool with one of them nearly mute.
Recognition · function of inverted attitude×0.35Ne is not Ni and Ti is not Te — but neither is it absence, and this line used to pay ×0. Zeroing it put INTP, ENTP, ISFJ and ESFJ on exactly the same figure, which claimed an INTP is as unrecognisable to you as an ESFJ. Nobody believes that. A third of the weight: reasoning from internal consistency and reasoning from effectiveness are both still reasoning, in a way neither shares with somebody reading the room.
Complement · Se, the body and the now30 · 50 · 28 · 6Your inferior function, by where it sits in their stack — and <strong>the peak is the auxiliary, not the dominant</strong>. This line used to pay most for the person who leads with your inferior, and that was the page’s central error: theory calls it maximum complement; a shared kitchen calls it translating every day. Se in second place gets you out of the house on a Saturday. Se in first place lives there, and you are the visitor.
Complement · Ne, the option kept open18 · 30 · 17 · 4Applies to the stacks that carry Ne instead of Se — the two never coexist. Ne substitutes for part of the service: it reopens the plan Ni has closed, but it returns nobody to their body. Same curve, lower ceiling.
Complement · Fe, warmth said outward18 · 30 · 17 · 4The one judging function your stack holds in no version at all. Zero for anyone running Fi: you already have Fi, and more Fi solves nobody’s problem of nothing being said out loud.
Complement · extroverted dominant+10Flat, no scale. Somebody has to go towards the world first, and your stack opens with an introverted function that only moves once it has converged. It paid +20 and was cut for the same reason as the Se curve: going first helps, going always exhausts.
Livability · preferences in commonI/E 30 · J/P 30 · N/S 25 · T/F 15The axis that was missing. The other two answer "does this person understand me" and "does this person extend me"; neither answers "what is a Tuesday like", and a ranking built from those two alone recommended your opposite on all four letters. The weights are what each preference costs an INTJ, not symmetry: introversion decides how much of the week is recovery, J/P decides whether a plan is a plan. T/F last, the most negotiable of the four.
The combined figuremean − 0.45 × (highest − lowest)The mean of the three axes, less almost half the distance between the best and the worst. The deduction is what stops the mean recommending you to yourself: the mirror scores 100 · 6 · 100 and would ride up on two of them. A pairing has to be worth something on all three. <strong>It is a reading order, not a verdict</strong> — and nobody on the table clears 45.

The weights were chosen, not derived. They came from no data at all: they came from what type theory claims about awareness and control at each position in the stack — and that theory is, within the MBTI, exactly the part with the least empirical support. Try it: swap the 0.7 rung for 0.5 and the ENTJ falls from 70 to 50, level with the ISTJ, and half the reading of this page changes meaning. Two further things deserve saying out loud, because they are the rule confessing what it is. First: recognition reaches 100 exactly once in sixteen, and it does so in the mirror — the ceiling of that axis is a warning, not a prize. Second: complement never reaches 100, and the real maximum is 82, at the ESTP. That is not a compressed scale; it is structure. A dominant Se and a dominant Fe do not fit in the same stack, so the thing you most lack and the second thing you most lack will never be delivered by the same person. No four-letter code solves you.

Sources

Not every claim about type carries the same weight. The tag beside each work says where it comes from — and what counts as evidence varies a great deal between them.

  1. Academic psychology Dyrenforth, P. S., Kashy, D. A., Donnellan, M. B., & Lucas, R. E. (2010). Predicting relationship and life satisfaction from personality in nationally representative samples from three countries: The relative importance of actor, partner, and similarity effectsJournal of Personality and Social Psychology, 99(4), 690–702. 23.250 pessoas (≈ 11.600 casais casados) na Austrália, Reino Unido e Alemanha.
  2. Academic psychology Kong, S. S. (2010). Relationships between MBTI psychological type and marital satisfaction, divorce proneness, positive affect, and conflict regulation in clinic couplesJournal of Korean Academy of Nursing, 40(3), 336–348. PMID 20634625; artigo em coreano. 62 casais de clínica. Sem diferença significativa em satisfação conjugal, afeto positivo e regulação de conflito conforme a semelhança de tipo — mas com diferença significativa na propensão ao divórcio dos maridos conforme a semelhança no índice Sensação/Intuição.
  3. Academic psychology Montoya, R. M., & Horton, R. S. (2013). A meta-analytic investigation of the processes underlying the similarity-attraction effectJournal of Social and Personal Relationships, 30(1), 64–94.
  4. Academic psychology Montoya, R. M., Horton, R. S., & Kirchner, J. (2008). Is actual similarity necessary for attraction? A meta-analysis of actual and perceived similarityJournal of Social and Personal Relationships, 25(6), 889–922. DOI 10.1177/0265407508096700.
  5. Methodological critique Pittenger, D. J. (2005). Cautionary comments regarding the Myers-Briggs Type IndicatorConsulting Psychology Journal: Practice and Research, 57(3), 210–221.
  6. Official MBTI Myers, I. B., & Myers, P. B. (1980). Gifts DifferingConsulting Psychologists Press. O subtítulo “Understanding Personality Type” é da reedição Davies-Black de 1995.
  7. Official MBTI Myers, I. B., McCaulley, M. H., Quenk, N. L., & Hammer, A. L. (1998). MBTI Manual: A Guide to the Development and Use of the Myers-Briggs Type Indicator3ª ed., Consulting Psychologists Press. Amostra nacional representativa, N = 3.009.
  8. Official MBTI The Myers & Briggs Foundation (s.d.). Ethical Use of the MBTI® Assessment · All Types are Equally Valuablemyersbriggs.org, acesso em 2026. Duas páginas distintas.
  9. Official MBTI The Myers-Briggs Company (s.d.). The MBTI assessment (themyersbriggs.com)Empresa que publica e licencia o instrumento, distinta da Myers & Briggs Foundation, que zela pelo uso ético.
  10. Specific model Augustinavičiūtė, A. (1998). Socionika: VvedenieAST / Terra Fantastica. Obra escrita em Vilnius, anos 1970–80.
  11. Specific model Grant, W. H., Thompson, M., & Clarke, T. E. (1983). From Image to Likeness: A Jungian Path in the Gospel JourneyPaulist Press.
  12. Specific model Keirsey, D. (1998). Please Understand Me II: Temperament, Character, IntelligencePrometheus Nemesis.
  13. Specific model Keirsey, D., & Bates, M. (1978). Please Understand Me: An Essay on Temperament StylesPrometheus Nemesis. O subtítulo “Character and Temperament Types” é das edições a partir de 1984.
  14. Specific model Wikipedia (2026). Socionicsen.wikipedia.org, acesso em ago. 2026. O próprio artigo carrega aviso de que depende excessivamente de fontes ligadas ao tema.
  15. Specific model Wikisocion (2021). Intuitive Logical Introtim (ILI) · Logical Intuitive Introtim (LII)wikisocion.github.io. Arquivo estático de wikisocion.net, congelado num backup de março de 2021; o site original saiu do ar em agosto de 2021.