Men are not repelled by being sorted. They are obsessed with it
What men reject is unverifiable plus an appeal to feelings. The awkward part is that our own three-axis code is on the wrong side of that line more than we would like.
$ eldr --cite mbti-thinking
68.6% of men report a preference for the Thinking type on the MBTI. That figure comes from Hammer and Mitchell's 1996 normative sample in the Journal of Psychological Type; the Myers-Briggs Company's own national representative sample puts it lower, at 56.5%, and the existence of two different answers to the same question is itself a fact about the instrument that I will come back to. Either way: four letters, assigned by a questionnaire, with no clinical standing — and men will put the code in a work bio, in a dating profile, in an argument about why they are like this.
Hold that next to the fact that the same men will not go near the category next door, which performs a structurally identical operation: answer some questions, receive a symbol, receive a narrative about yourself, discuss it with others who share the symbol.
The difference is not the sorting. It is what the sorting claims to be.
The claim
Men do not reject symbolic self-categorization. They reject two specific properties: unverifiable, and addressed to feelings. Remove those two and the same man who rolls his eyes at one system will voluntarily memorize his code in the other.
Verifiable, cheaply, and the test is sitting in plain sight: the popularity of fantasy sports projections, trading signals, handicapper picks, fitness readiness scores. Every one of those is a confident external number about an uncontrollable outcome — exactly the product men supposedly refuse to buy — and every one is presented as a measurement rather than as an insight about your inner life. The mechanism sells. The register decides who buys.
Which is a strong claim with an obvious problem attached, and the problem is us.
The problem, which is us
This product assigns a code. Three binary axes off three of the fourteen inputs: evidence-led or gut-led, strikes or waits under pressure, decides alone or by council. Eight files, a conviction number and a decision-drag figure in days.
If the argument above is correct, then the code is doing the same job the four letters do — it is a self-narrative artifact that men will find satisfying and shareable — and the fact that it is dressed in monospace rather than in a chart is a matter of register, not of epistemics. Saying so out loud is uncomfortable and skipping it would be dishonest.
So here is where the line actually falls, and it is thinner than the marketing would like.
- The code describes your answers, not you. It is computed from what you typed in the last two minutes and it changes when your answers change. That is a feature in the same way a thermometer reading is a feature: nobody claims it is your identity.
- It makes no prediction. No compatibility, no forecast, no claim about what will happen. It changes how hard the counter-case pushes and how the window is set. That is its whole operational effect, and if it had no operational effect it should not exist.
- There is no fourth axis and there never will be, because a taxonomy that keeps growing is a taxonomy being tuned for how much people enjoy it.
The case against the code existing at all
MBTI is a bad thing to cite in your defense. Its test-retest reliability is poor, its factor structure does not survive scrutiny, and psychometricians have been saying so for decades. Arguing "men accept our categorization because they accept MBTI" is an argument that men accept a measurement that does not measure well — which is evidence about marketing, not about instruments. And your own three axes are eight buckets off three self-report questions, which is if anything a weaker instrument. The intellectually honest move is to delete the code and keep the position.
I think that objection is close to unanswerable on the psychometrics. The retest problem is real and well documented, and any product citing MBTI as a validity argument is citing popularity.
The defense is narrower than the objection assumes. The code is not offered as a measurement of a stable trait; it is offered as a summary of how this file was answered, which is why it is allowed to be unstable. It earns its place by changing the output — a gut-led man who waits under pressure gets a different window and a harder counter-case than an evidence-led man who strikes. If the code did not change the read, the objection would be entirely correct and it should be deleted.
What would falsify its usefulness: if reads with the code suppressed produced the same positions, confidence numbers and windows as reads with it. That is testable with the data the intake already collects. It has not been tested, and until it has, the code is a hypothesis with an interface.
Why this is in the case files rather than in the marketing
Because the case files are the section where the evidence is examined, including the parts of it that undercut the product. A page that cited 68.6% and stopped there would be using a statistic to flatter the reader. The statistic is real, the inference about register is defensible, and the nearest thing this product has to a personality taxonomy is on weaker ground than its position engine is.
If you want the part that survives all of this, it is one sentence: the register decides who will use an instrument, and the instrument still has to show its work.
And one consequence of that sentence, for anyone building something adjacent. The register is not a coat of paint you apply at the end. Monospace over a claim that cannot be checked is worse than the version that never pretended — it borrows the credibility of measurement without accepting the obligation, and men are unusually good at detecting that particular theft, having been sold it before.
Stuck on a decision like this?
Talk it through in ELDR Chat. It argues both sides and gives you a straight answer and your next move. Two questions free, no sign-up.