About · The standard behind the bench

Built for the exam that exists, to a standard you can check.

But For is an independent study product and publication for the NextGen UBE, the bar exam that began replacing the MBE era in July 2026. One product, one price, and an editorial standard published right here so you can hold us to it.

What this is

A drilling bench built format for format to the real exam: original multiple-choice in both NextGen formats (including six-option select-twos scored with the exam's real partial credit), integrated question sets where the facts unfold and earlier answers lock, both performance-task formats on a real clock, AI grading of written work that quotes its evidence, and a ledger that tracks confidence on every answer and turns your confident-wrong record into gap-training sessions.

And a publication: the sourced record of how this exam actually behaves in the world, starting with the first administration, every claim labeled by how well it is corroborated.

Plainly, what it is not: no videos, no live classes, no tutors. The bench sits beside whatever outlines you already use. It does one job, the doing layer, at the highest standard we can hold.

What the exam is actually testing

NCBE names the skills openly: issue spotting and analysis, legal research, legal writing, investigation and evaluation, counseling, negotiation and dispute resolution, client relationship management. Most prep treats that list as a table of contents. We treat it as a specification, and reverse-engineer the cognitive operations underneath each format:

C1
Legally significant fact filtering.Every stem plants facts that matter and facts that merely glitter. The select-two's near-miss distractors and the sets' unfolding records train the separation reflex, because the exam's first hidden question is always "which of these facts changes the answer?"
C2
Issue recognition under time.Spotting the doctrine a fact pattern summons, before the options frame it for you, is pattern recognition, and pattern recognition is built by volume in the exam's own grammar, not by rereading outlines.
C3
Rule-to-fact mapping, both registers.The NextGen asks you to apply provided law objectively, even against your memorized instincts, and to argue persuasively when the task demands advocacy. The bench drills both registers and labels which one each task rewards.
C4
Authority discipline.The legal-research task tests whether you can rank what binds over what persuades, and what a case held over what it said in passing. It's a skill the MBE never touched and the bench treats as first-class.
C5
Triage and efficiency.Nine hours, three sessions, a clock that punishes perfectionism. Time targets on items and the PT's real 60-minute deadline train the judgment of when a point is earned and it's time to move.
C6
Calibration: knowing what you know.The most dangerous error class is the confident miss, because it never feels like a gap. Confidence is captured on every answer so the miscalibration itself becomes visible, measurable, and trainable.

How the bench builds it

The mechanisms aren't decoration; each one implements a principle from the learning sciences, chosen because it survives scrutiny:

Retrieval practice over rereading. Answering under exam conditions is the act that builds durable memory; the testing effect is among the most replicated findings in cognitive psychology. The bench is retrieval from the first minute; there is nothing to passively watch.

Format-true encoding. Skills transfer best to the conditions they were practiced in. Six-option select-twos scored with real partial credit, records that grow mid-problem, a research library with traps. Practicing in the exam's exact grammar means exam day runs on patterns your brain has already laid down, not on translation from a format that no longer exists.

Interleaving and desirable difficulty. Mixed-subject lanes force the harder discrimination ("which doctrine is this?") that blocked, one-subject drilling never asks. It feels worse and works better; that trade is deliberate.

Elaborated feedback, immediately. Every option carries an explanation of why it seduces or fails, and every item ends in a one-line discriminator built to be remembered: the pattern anchor you'll actually retrieve in the room.

Metacognitive calibration. Sure or Unsure, on every answer. Review opens on the misses you were sure about; the ledger tracks the calibration gap over time; gap training rebuilds your weakest ground from your own record. The aim is an examinee whose confidence is information.

What we do not claim: shortcuts, predicted questions, or guaranteed outcomes. The claim is narrower and stronger: the mechanisms are the ones the evidence supports, aimed at the skills the examiners publish, drilled in the format they deliver.

The rules everything is held to

R1
Experience, never content.Examinees are bound to confidentiality about what the exam asked. We hold ourselves to the same line, voluntarily and absolutely. We do not solicit, record, or repeat exam content. Ever.
R2
Every claim gets labeled.Official statements are marked official. Press reports carry the number of independent outlets behind them. Inferences say they are inferences. If we can't source it, we don't say it.
R3
Original, always.Every scenario, statute, and option is written in-house in our fictional State of Meridian and similarity-gated against NCBE's official samples before it ships. Nothing licensed, nothing farmed, nothing echoed.
R4
Provisional until earned.Score bands and difficulty estimates stay labeled provisional until real calibration data earns the label's removal. No week-one difficulty claims; no predicted questions; no manufactured urgency.
R5
Splits get flagged.Where jurisdictions genuinely divide, the materials say so instead of pretending the majority rule is the only rule.
R6
The grader shows its work, and yields.AI grading is rubric-bound, quotes the answer's own words as its evidence, and stays assistive: you can overrule it, and your call stands on the record.

How the materials are made

Every batch of questions passes six gates before it can ship, each one scripted, each one a hard stop:

The six gates.

Structure: exact format integrity for every item: stem, options, keys, a teaching explanation for every option, and a one-line discriminator worth remembering.
Doctrine audit: the new batch is diffed against a regenerated map of everything already in the bank, so coverage deepens instead of repeating.
Naming: every coined name is checked against the entire corpus; collisions are renamed before delivery.
Key balance: answer letters are distributed and audited, so patterns can't be gamed.
Import acceptance: the full corpus re-imports with zero errors, or the batch is not done.
Scope citation: every keyed doctrine maps to an explicit line of NCBE's published content scope, verified against the source. Doctrines the outline does not list are rejected and logged, however tempting; the reject ledger is part of every batch's record.

Authorship is in-house, to a law-faculty standard, under one editorial direction, and every batch lands as a draft that must survive founder review before members see it.

The company, plainly

But For is an independent publication and study product of Ink Forms Rivers, d/b/a But For. We are not affiliated with, sponsored by, or endorsed by the National Conference of Bar Examiners or any bar examining authority; "NextGen UBE" is referenced solely to identify the examination.

One product, one price: founding membership, $199 once, everything through the February 2028 administration. Never a subscription. Checkout by Stripe; we never see your card.

support@butfor.ai reaches a human who reads every message.

The fastest way to judge us is the product itself: the three-question calibration is free, the grader will grade an answer you write, and the report shows the editorial standard in action. Founding access is there when the evidence convinces you.