How to Train Your Brain for Sharper Thinking: A Practical RFT Guide

Part of the guide: Affiliate Marketing

When I was 19 I spent two years on a mission in England, and I accidentally ran the most aggressive brain-training experiment of my life. Up at 6:30am, three hours of study, then twelve hours on street corners (9:30am to 9:30pm) talking to anyone who would stop, including Jehovah’s Witnesses who would calmly take apart everything I thought I knew. I would walk home at night with my head full of contradictions I could not resolve. And then, days later, something would click. A connection I did not have before. That is the real answer to how to train your brain: you load it past the point of comfort with the right kind of problem, on purpose, again and again, until your wiring reorganizes around it. This guide shows you the science underneath that, and the exact protocol to do it deliberately instead of by accident.

You are reading this because you have noticed something you do not like. The connection that used to fire in half a second now takes three. You walk into a meeting sharp and walk out fuzzy. A younger colleague sees the pattern in the data before you do, and you feel it. Most people call this “brain fog” and reach for more coffee, which is treating a wiring problem with a fuel additive.

Here is the part that actually stings. You were raised on the idea that intelligence is fixed. You got an IQ number once, or a test score, and you filed yourself away under it. So the quiet thought “maybe I can get sharper” feels a little naive, a little beneath you. Meanwhile you can feel yourself getting duller, and you cannot accept that either. That tension is the real problem, and it keeps capable people stuck. Let me walk you through the mechanism, the evidence (the honest version, with the caveats), and a protocol you can start today.

In this guide

1. How to train your brain without wasting months on games that do not transfer
2. What Relational Frame Theory actually is (the mechanism, not the buzzword)
3. How to train your brain with RFT: the step-by-step protocol
4. What the research really shows (the honest version)
5. The compounding case (why this is worth your hours)
6. Do this today

How to train your brain without wasting months on games that do not transfer

Start with the bad news, because it protects you. Most “brain training” does not make you smarter. It makes you better at the brain-training game, and that is it.

Researchers have a precise way to talk about this. They split improvement into two kinds:

• Near transfer. You practice a task and you get better at that task, or at tasks that look almost identical. Practice matching shapes, get faster at matching shapes. Real, but narrow.
• Far transfer. You practice one thing and your general ability rises, so you improve at tasks you never trained on. This is the holy grail, and it is rare.

The reason this matters: almost every popular brain game produces near transfer and almost no far transfer. Lumosity is the cautionary tale. In 2016 the U.S. Federal Trade Commission fined Lumos Labs $2 million for advertising that its games sharpen everyday cognition and stave off decline, claims the company could not back. The dual N-back craze, the speed games, the memory matrices: people get visibly better at the exercises and the gains mostly stay locked inside the exercises. You become a world-class N-back player and a normally-thinking human.

So the smart question is not “which brain game is fun?” The question is “which kind of training has actually shown far transfer in peer-reviewed work?” When you filter for that, the list gets very short. One of the few approaches that keeps showing up is built on Relational Frame Theory.

What Relational Frame Theory actually is (the mechanism, not the buzzword)

Relational Frame Theory (RFT) comes out of behavioral science. Strip the jargon and the core claim is simple and a little startling: the thing we call “reasoning” is mostly one skill running underneath everything, the skill of relating things to other things. Same and different, opposite, more and less, before and after, part of, caused by, from my perspective versus from yours.

RFT calls this derived relational responding. “Derived” is the load-bearing word. It means you can respond to a relationship you were never directly taught. Three pieces make it work, and once you see them you cannot unsee them.

The three moves that make derived relations work

Mutual entailment. Teach a direction, get the reverse for free. If you learn that a “florin” is worth more than a “ducat,” you instantly know a ducat is worth less than a florin. Nobody taught you the second sentence. Your brain derived it.

Combinatorial entailment. Chain relations and get the new ones automatically. If a florin is more than a ducat, and a ducat is more than a groat, then you know a florin beats a groat and a groat is the runt, without ever comparing them directly. That is reasoning. It is also exactly what an IQ test measures when it asks you to complete an analogy.

Transformation of stimulus functions. This is the deep one. Once two things are in a relationship, the meaning of one bleeds into the other. Tell a child a nickel is “worth more” than a dime in a little game, and the child will reach for the nickel even though the dime is physically bigger. The word “more” rewrote which coin felt valuable. This is the engine of insight and generalization. It is how learning one fact changes your behavior toward a hundred things you were never trained on.

Put those together and you get the non-obvious insight that makes RFT different from a memory game. Relating is a generalized skill, not a content skill. It sits under language, math, analogy, planning, and perspective-taking. When you raise the fluency of that one underlying skill, you raise the floor on everything built on top of it. That is a structural reason to expect far transfer, instead of just hoping for it. The dual N-back trains a narrow loop. RFT trains the substrate.

This is also why my confused-mind nights in England worked. I was forcing my brain to hold relations that did not fit together, opposing claims, contradictions, comparisons I could not collapse, and the new connections that formed were relational connections. I just did it the slow, painful, accidental way. You do not have to.

The relational frames you are actually training

RFT researchers describe families of frames. The exact count is open (the theory says the number is limited only by the verbal community that trains them), but a structured program tends to drill a stable core set to fluency. Here is the working map, with what each one powers in real life.

Relational frame	What it relates	Plain example	What it powers in your work
Coordination (sameness)	A is the same as / equal to B	“profit” = “earnings”	Definitions, synonyms, recognizing the same idea in new clothes
Distinction	A is different from B	revenue is not profit	Spotting false equivalences, precise thinking
Opposition	A is the opposite of B	asset vs liability	Trade-off analysis, devil’s-advocate reasoning
Comparison (more / less)	A is greater or lesser than B on a dimension	margin A > margin B	Ranking, prioritization, sizing decisions fast
Hierarchy (part / whole)	A is a member or part of B	a SKU is part of a product line	Categorization, systems thinking, org design
Temporal (before / after)	A comes before or after B in time	build the list before the launch	Sequencing, planning, cause-and-effect chains
Spatial (here / there, over / under)	A is positioned relative to B	this layer sits above that one	Mental models, diagrams, navigation
Causal / conditional (if / then)	A produces or depends on B	if price drops, then volume rises	Forecasting, strategy, debugging
Deictic (perspective)	I / you, now / then, here / there	how this looks from the buyer’s seat	Negotiation, empathy, reading a room

A serious program does not stop at the simple version of each. It also trains relating relations (analogies are literally a relation between two relations) and flipping between frames at speed. CortexSurge, the tool I built around this, drills 19 frames to fluency across its zones precisely because the real-world payoff lives in the combinations, not the singles.

How to train your brain with RFT: the step-by-step protocol

This is the part you can act on. The research procedure that produced the headline results is called SMART (Strengthening Mental Abilities with Relational Training). I will give you the mechanics of it and then the practical way to run your own version.

Step 1: Get a real baseline before you touch a single exercise

You cannot tell training from placebo without a before-number. Take a fluid-reasoning assessment, write down the score, and date it. This sounds obvious and almost nobody does it, which is exactly why people argue forever about whether brain training “works.” Measure first. Then you get to know, not believe.

Step 2: Train the right relations, in the right order

SMART does not start with everything at once. In the original procedure the first half of training (roughly the first 29 stages) uses only sameness and opposite relations. The second half (the next 26 or so stages) moves to more-than and less-than. There is a reason for the order. Coordination is the first frame a child masters, opposition builds on it, and comparison is harder still. You are recapitulating the natural build order of the verbal mind, on purpose, at higher intensity. Start with same and opposite, earn comparison, then layer in the rest.

Step 3: Train to fluency, not just to “correct”

Here is the mistake almost everyone makes, and it is the difference between results and a hobby. They train until they can get the answers right, then they stop. Wrong target. The thing that correlates with the IQ and reasoning gains is fluency, which is accuracy plus speed. Not “can I derive this relation,” but “can I derive it in under two seconds, automatically, without grinding through it.”

The SMART criterion makes this concrete. You do not advance a stage by getting a few right. You advance by producing a streak of consecutive correct responses (16 in a row in the standard protocol) at that level of complexity. The streak forces the response to become reflexive instead of effortful. In a well-built app this shows up as an accuracy threshold (around 85 percent at a given level) before it raises the difficulty on you. If your tool lets you coast, it is not training you. The discomfort is the point.

Step 4: Climb the complexity ladder deliberately

Fluency at one level is not the goal. The goal is fluency that holds while the problems get harder. SMART runs through 55 stages of increasing relational complexity. Early stages relate two or three items. Later stages stack multiple relations, mix frame types in a single problem, and add the entailment steps so you have to derive the answer rather than recall it. The progression matters more than the volume. Doing easy reps forever does nothing. Living right at the edge of what you can do, where you are at maybe 80 to 90 percent, is where the rewiring happens. This is the same edge I was sitting on, miserably, on those English street corners.

Step 5: Dose it like training, not like cramming

Your brain consolidates between sessions, not during them. The protocols that worked spaced the load. Here is the dosage that the research and my own use converge on:

Variable	Research protocol (SMART)	Practical version for a working adult
Session length	30 to 45 minutes	15 to 25 focused minutes
Frequency	1 to 3 times per week	4 to 5 times per week
Total span	around 10 weeks	2 to 3 months for clear gains
Advance criterion	16 consecutive correct per stage	roughly 85% accuracy, then level up
Stages / levels	55 stages of rising complexity	adaptive levels (thousands, gamified)
First “click”	varies	often around 15 to 25 hours in

Notice the shape. Short, frequent, and finished before fatigue makes you sloppy. I learned this the expensive way running my own business at full tilt. I work in strict 90-minute segments with real breaks between them, because a tired brain practicing relations just grooves in slow, sloppy relations. Train fresh or do not train.

Step 6: Measure far transfer, not your score on the game

Come back to Step 1. After 8 to 12 weeks, retake a fluid-reasoning test you did not practice on. That gap, baseline to retest, is the only number that matters. Your in-app score going up only proves near transfer. The retest proves whether the skill left the building. Track it. If it moved, you have evidence. If it did not, change the variable (usually it is fluency, you trained for “correct” and stopped).

Step 7: Keep a maintenance dose

Fluency is a physical property of your wiring, and wiring follows use. You do not need the full ramp forever, but a couple of short sessions a week holds the gains. Think of it the way a serious lifter thinks about strength. You earned it under load. You keep it under lighter, regular load.

What the research really shows (the honest version)

You are skeptical and you should be. The whole brain-training category earned that skepticism. So here is the evidence laid out straight, wins and warts, because a number you cannot trust is worse than no number at all.

Study	Who / how	What it found	The caveat (read this)
Cassidy et al., 2016	Children and adolescents, online SMART training over several months	Reported full-scale IQ gains averaging roughly 23 points; gains also on numerical and verbal reasoning aptitude	Experiment 1 had no control group, so practice and maturation are not ruled out there
Amd & Roche, 2018	Socially disadvantaged children, Bangladesh	Measurable gains in fluid intelligence after relational training	Small sample; specific population, so generalization is limited
McLoughlin et al., 2020	Active-controlled feasibility study (a real comparison group, chess, doing something else)	Non-verbal IQ gains from relational training; gains explained variance in educational attainment	Feasibility study, modest size; “feasibility” means promising, not settled
SMART replication / controls, 2023	Typically developing children, controlling for baseline IQ and attention	Full-scale IQ gains shrank substantially once baseline differences were controlled	This is the cold-water study; some of the early effect was baseline and attention, not pure cognitive lift
May et al., 2022	Systematic review and meta-analysis of SMART across studies	Positive effects on cognitive and academic measures overall	Effects were heterogeneous; methodology across the field is uneven

So what is the fair conclusion? Not “RFT training adds 23 IQ points,” full stop. The honest version is this: relational training is one of the few brain-training approaches with a credible mechanism for far transfer and repeated peer-reviewed signal that it improves reasoning and fluid intelligence, and the strongest claims soften when you control properly. That is still a remarkable position to be in. Most of the category cannot show far transfer at all and got fined for pretending. This one has a theory for why it should transfer, and a research record that mostly points the right way. You train it because the mechanism is sound and the downside is a few sharp hours a week, not because anyone can promise you a number.

If you want the wider map of the tools I actually stand behind and how they fit together, start here.

The compounding case (why this is worth your hours)

Run the math on a small, durable edge. Suppose deliberate relational training makes you reliably faster at seeing the pattern, the analogy, the if-then, the trade-off. Not a miracle, just sharper by a step. Now compound that across every meeting, every negotiation, every strategy call, every problem you debug, for the next twenty years. A high earner does not get paid for hours. You get paid for the quality and speed of the connections you make. Sharpen the connection-making itself and you have moved the lever under everything. I am wired the way you are. I have built and run my own things, and what keeps me at the desk is the same quiet fear you feel right now: a connection you used to make instantly showing up a half-second late. People like us do not coast. That is the bond, and it is exactly why this protocol takes for us when it bounces off everyone looking for a shortcut.

There is an identity underneath this too, and it is the real prize. Most people walk around believing their intelligence is a fixed allotment they were issued at birth and are slowly spending down. When you train your brain on purpose and watch a retest move, that belief breaks. You stop being someone whose mind was inherited and start being someone whose mind is forged. That self-concept, “I can make myself sharper, deliberately, and I have the data,” changes how you approach every hard thing for the rest of your life. The IQ points are nice. The identity is the asset that compounds.

One reference experience can flip a belief in seconds. I watched a Brazilian jiu-jitsu grappler tap out a karate master I was certain could not be touched, in 14 seconds, and a certainty I had carried for years was just gone. Your first clean retest result does that to “intelligence is fixed.” Go get the reference experience.

Do this today

Stop reading and take three actions before you close this tab:

1. Take a baseline fluid-reasoning assessment and write the number down with today’s date.
2. Run your first 15-minute session on same and opposite relations, and push until the responses feel automatic, not just correct.
3. Put four sessions on your calendar this week like they are client calls, because they are, with the most important client you have.

The brain you think with in five years is being built by what you make it do this week. Make it relate hard things, fast, on purpose.

Related reading: Brain Fog Is Costing You Money: How to Think Clearly Under Pressure, Is IQ Fixed? What the Science Actually Says in 2026, The High Earner’s Daily Routine for Mental Performance.