The Four Types of Curriculum

By OSJJ

Open Source Jiu-Jitsu

THE CURRICULUM SERIES · PART TWO

The Four Types of Curriculum

Linear, rotating, flow, and positional. What each one is, how it works, and how each holds up against what we actually know about how humans build skill.

Justin Hall / Co-Founder, Open Source Jiu-Jitsu

In the last article the argument was that you need a curriculum at all, a real path rather than a stack of classes. This is the follow-up that was promised: which kind. There are four types, four fundamentally different shapes a curriculum can take, and they are genuinely different, not four flavors of the same thing. Within any one of them you could build many different curricula, but the shape itself, the order it imposes and the room it suits, is what this article compares.

I do not want to grade them on taste, though, because taste is how most of these debates go and taste settles nothing. I want to grade them the way you would grade anything that is supposed to produce a result, against what is actually known about how people learn. So before the four types, here is the measuring stick.

How to judge a curriculum, scientifically

Strip away the Jiu-Jitsu for a moment. A curriculum is a machine for turning a beginner into an expert, and the science of learning gives us four fairly settled things such a machine has to do well. These are the yardsticks I will hold each method against.

• How it organizes knowledge. This is the big one, and it is the least understood. Decades of research on expertise show that experts and beginners do not just know different amounts, they organize what they know differently. A beginner stores isolated moves, surface features, this technique and that technique. An expert stores knowledge around problems and positions: given this situation, here are my goals and my options. A curriculum that builds knowledge the way experts actually hold it produces usable skill far faster than one that hands over a pile of techniques the student has to reorganize on their own, years later, if they ever do.

• How it builds durable memory. People do not learn a skill by seeing it once. Memory consolidates through spacing and return, meeting a thing, losing it, meeting it again, and keeping a little more each time. A method that revisits the important material on purpose builds durable skill. A method that shows something once and moves on builds the illusion of it.

• How it manages difficulty. Beginners have almost no spare working memory on the mat. A good method meets the learner near the edge of what they can already do, one manageable step beyond, and scaffolds up from there. Pitch it too far ahead and you do not challenge them, you overload them, and overloaded people quit.

• How well it transfers. This is the one that separates a good drilling student from a good grappler. Skill that only works in the cooperative rep and falls apart the moment a resisting stranger does something unexpected has not really been learned. What you want is skill that transfers to live, chaotic, novel situations. And there is a catch worth knowing: the practice that transfers best often feels less efficient in the moment. Smooth, predictable, one-thing-at-a-time practice feels productive and can produce brittle skill, while practice that mixes things up feels harder and builds skill that lasts. Keep that in mind, because it complicates the word efficient in everything that follows.

Four yardsticks: organization, memory, difficulty, transfer. Now the four types.

1. Linear progression

The purest method. Every lesson builds directly on the one before it, in a fixed order, each new piece resting on the piece just laid down. Lesson two assumes lesson one. Lesson twenty assumes the nineteen under it. It is how a textbook is built, and it is, in the abstract, the most efficient way to take one specific person from zero to competent, because nothing is wasted and nothing is out of order.

Against the yardsticks it scores beautifully on difficulty management, it is scaffolding in its purest form, each step exactly one step beyond the last, and very well on initial memory, because everything reinforces what came just before. Its weakness is hidden in that same smoothness. Perfectly linear, perfectly predictable practice is exactly the kind that can feel efficient while producing skill that is a little brittle under live chaos, and it says nothing about knowledge organization on its own.

But the real problem with linear progression is not scientific. It is logistical, and it is fatal for a group program. Linear only works if every student is on the same step, which means you must control who is in the room and where they are. The moment a new white belt walks into lesson fourteen, the sequence is broken for them, and in a normal academy new people walk in every single week. This is why linear progression is not a group curriculum at all. It is a private-lesson curriculum.

And that is exactly where you should use it, deliberately. Build a true linear progression for your private students, a real, ordered path built around that one person. This is the honest justification for what private lessons cost. It is not only that the student gets your personal attention, it is that one on one is the only setting where a fully linear path can actually be delivered, a progression built for that one person and no one else. That is something a group can never give: a client who can see the path laid out for them specifically and feel each session build on the last toward where they are going. In a group, the logistics make it impossible. In a private, it is the whole point. If your privates are just a group class with one student, you are underdelivering and underpricing at the same time. Build them a linear path they can see is theirs, and the premium justifies itself.

A note on the examples that follow: each one illustrates the shape of a method, how the pieces are arranged. The modular and flow examples use our own curriculum’s structure. The linear and positional examples are illustrative, with placeholder positions to show the arrangement. In every case the point is the structure, not a prescription of what you must teach.

Shape: a locked chain. Each link rests on the one before it, and the order cannot move.

Pull out any session and every session beneath it loses its footing. A student cannot walk in at Session 4, which is exactly why this shape belongs in a private, where there is only ever one student and the chain is built for them alone.

Why linear stays in the private lesson

Before the other three methods, it is worth being clear about why linear does not leave the private lesson, and the reason is not that it is better, it is that it needs a condition a group cannot provide. Linear only works when everyone starts at the beginning and moves through in order.

Think about how every other serious subject is taught. You take Biology 101, and it is a prerequisite for Biology 102, which is a prerequisite for the one after that. Everyone in the room enrolled at the same time, everyone has the same background, and nobody wanders into week nine of organic chemistry having never taken week one. The whole institution is built to protect the sequence. A university can teach linearly because the environment guarantees the conditions the sequence needs.

Jiu-Jitsu has none of that. There is no admissions office, no semester, no prerequisite anyone can enforce. A brand-new white belt walks through the door on a random Tuesday and needs a class that works for them today, in the same room as a four-stripe blue belt who has been there three years. Nobody enrolled together and nobody can be turned away until the next intake, because there is no next intake. The academy has to teach everyone, at every level, arriving at every possible moment, all at once. A locked sequence cannot survive that, which is why linear lives only in the private lesson, the one corner of the academy where you control who is in the room.

The other three methods, rotating, flow, and positional, are the real answers to the question a group actually poses: how do you teach a room where you cannot control who walks in or when. Each one handles that differently, and each one is a legitimate, intelligent solution in its own right, not a runner-up to something better. The rest of this article is an honest analysis of what each one trades away and what it keeps.

2. The modular rotating curriculum

The workhorse of large academies. You define a fixed set of techniques, distribute them across a set number of lessons, and cycle through the lessons on a loop. The content inside each lesson is deliberate and connected, this lesson is a coherent unit about a specific thing, but the order in which a student encounters the lessons is essentially arbitrary. Join in week one or week thirty, you simply start the loop from wherever it happens to be and eventually see all of it.

Picture each lesson as a puzzle piece. On the day you receive it, it is a good, complete piece on its own. Over months of collecting pieces in no particular order, the student gradually assembles the whole picture in their own head. That is the method's genius and its limit at the same time.

Its great strength is accessibility and scale. Anyone can join at any time, which is the actual condition of a real academy. It survives multiple instructors, because each lesson is self-contained and does not require the teacher to know what the class did three weeks ago. It is the easiest of the four to run and to hand off, which is why big schools with rotating coaching staff lean on it. On the memory yardstick it quietly does something clever, too: because the material is spread out and mixed rather than blocked, it gets some of the benefit of spacing and mixing almost by accident, which can aid transfer.

Its weakness is knowledge organization and efficiency of learning. Because the order is arbitrary, the student is left to assemble the deep structure themselves, one stray piece at a time, and many never fully do. It is, learning-for-learning, the least efficient of the four, precisely because it leaves the most important work, organizing the knowledge into a usable whole, to chance and to time. It works. It just works slowly, and it asks the student to be their own architect.

There is a deeper cost hiding inside that arbitrariness, and it is the real detriment of the method. Spacing helps the lessons a student does hit, but it cannot fix which lessons they hit, or how often. In a modular loop the only path to mastering any single lesson is repetition. A lesson sticks once a student has met it enough times, and not before, and the loop hands out those repetitions by attendance rather than by need. Which lessons a student drills over and over and which they barely see comes down to which days they made it to class against wherever the loop happened to be. A student can train hard for eighteen months and, through nothing but the luck of the schedule, hit Lesson 5 eight times and Lesson 16 twice. Nobody planned that, and nobody would choose it, and the curriculum has no way to notice or correct it, because it does not track what any one student has and has not seen. Mastery ends up distributed by chance, and the thin spots stay hidden until the day they surface in a roll.

Shape: a loop of self-contained lessons. Each lesson stands alone, pairing a standing technique with one or two ground techniques, so a student may enter at any lesson.

This example is one of our modular fundamentals curriculums, a thirty-six lesson build that carries a student from white to blue belt. Each lesson pairs a standing technique with one or two ground techniques and lists its own variations, and because no lesson depends on the one before it, the order can rotate and a student can enter at any point without losing the thread. That tolerance for open enrollment and drop-in attendance is the whole point, and giving up week-to-week continuity to stay accessible to a mixed, changing group is the whole cost. The connective tissue between lessons is left for the student to assemble over repeated cycles.

3. The flow curriculum

This one we call flow. Technically it is a sequential-rotating hybrid, and that name tells you exactly what it is doing: it tries to keep the coherence of linear while keeping the come-anytime accessibility of rotating.

Here is the mechanism. The lessons build on and flow from one another, position to position, week to week, the way linear does. This week's material grows out of last week's and sets up next week's, so a student who is present for the whole arc experiences something close to a real progression. But unlike true linear, the prior lessons are not mandatory. They help, a student who was there last week has an advantage this week, but a student joining today can still fully participate today. The path has a direction and a flow, and yet the door is open at every point along it.

That is a harder thing to build than either of the methods above, and it is worth being honest about why. It demands that every lesson do two jobs at once: advance the arc for the people following it, and stand on its own for the person who just walked in. Designing material that does both, week after week, takes a real depth of understanding of the art and of how the positions connect. This is not a curriculum you can throw together. The sophistication of a well-built flow curriculum is exactly this dual-purpose design, every lesson advancing the arc and standing alone at the same time.

Against the yardsticks, flow is the attempt to have most of it. It recovers much of linear's coherence and memory-building, because the material actually connects and revisits, while keeping rotating's accessibility, because you can join anytime. It is more efficient for learning than pure rotating and more accessible than pure linear. The cost is entirely on the builder: it is the most demanding of the four to construct and to teach well, and a flow curriculum built badly collapses into a rotating one that is just harder to follow.

Flow also softens the problem that undoes modular. Because a stretch of time is organized around one connected theme, a student's attendance lands on related material no matter how often they make it in. Life keeps them to one session that week and they still walk away with the meat of what is being taught, because that single class sits inside a coherent arc rather than floating alone. Come four times and they absorb it fully. Either way they are not gambling on which unrelated lesson the loop happened to be on. The coverage is uneven, as it always is when people attend on their own schedule, but it is uneven within a connected subject instead of scattered at random, which is a very different and far more forgiving kind of uneven.

Shape: a connected chain with open doors. A position of the week drives the drilling, and each week hands off to the next, yet a newcomer can still join any week.

This example is one of our flow fundamentals curriculums, a thirty week build where the position of the week drives the drilling and techniques return through the same positions across the cycle. Each week runs a two-lesson pairing, taught Monday and Tuesday, then Wednesday and Thursday, with Friday and Saturday reviewing and combining the two. The real work lives in the Notes column, the handoffs that thread one week into the next: roll up into half guard after last week’s armbar to keep the flow, pass to side mount after the sweep, roll out from turtle to recompose guard. A student who trains consistently builds one continuous thread rather than a set of disconnected lessons. But because each week is still a complete pairing in a single position, the newcomer who walks in this week gets a full, standalone class. That double duty, a real thread a newcomer can still drop into, is the hard part, and it is why this shape asks the most of the person building it.

4. The positional curriculum

The last one is different in kind from the first three, and I think it is the most interesting, because it is the one that starts from the science rather than arriving at it. The first three are mostly answers to the question, in what order do we teach the techniques. The positional method asks a different question: what if the position, not the technique, is the unit of the curriculum.

Here is how it works. You build the entire curriculum around positions. For each position, you define two things, and only two. First, the goal: what am I actually trying to achieve from this spot. Second, the methods: the handful of available paths to that goal. That is the whole structure, one goal and its short menu of methods, repeated for every position. A student does not collect a scattered pile of techniques. They learn a position by understanding its single purpose and the few ways to fulfill it, and then the next position, and the next, each one a complete little problem stated as a goal and the methods that serve it.

Go back to the first and most important yardstick, how a curriculum organizes knowledge, and you can see why this one may be more than just another option. This is precisely how experts already hold the art in their heads. An expert does not think in a list of two hundred techniques. They think, I am here, my goal is this, my options are these. The positional method builds the student's knowledge in the exact shape an expert's knowledge already has, from the very first day, instead of handing them isolated pieces and hoping they reorganize them into expert structure years later. It front-loads the organization that every other method leaves the student to discover on their own.

There is a name for this. What an expert has and a beginner lacks is what learning researchers call a schema, an organized mental map of a problem and its solutions. The classic finding is that experts sort problems by their deep structure, the situation underneath, while beginners sort by surface features, the visible move. A technique is a surface feature. A position is deep structure. Building the curriculum around positions is, in plain terms, building the student a schema from day one. An educator would call this a schema-based, problem-centered curriculum: schema-based because it builds knowledge in the shape an expert’s already has, and problem-centered because each position is taught as a problem to solve rather than a technique to memorize. That is the real reason the positional method has a claim to being the most scientific of the four.

The positional method carries the same forgiving quality as flow, for the same reason. A position owns a block of time, so a student who trains once that block still gets the heart of that position, and one who trains through the whole block absorbs it in depth. Their exposure lands on one coherent problem rather than on whatever disconnected piece a loop happened to be showing, so the unevenness of real attendance does the least possible damage.

One design decision inside the positional method deserves its own mention, because it is where the method is most often gotten wrong: how fast the positions rotate. A popular version of position-based teaching is the position of the month, or of two months, where the room deep-dives a single position for weeks and then sets it aside and moves on. The deep focus feels thorough, and it is not a bad instinct, but it stumbles on the memory yardstick. When a position is then abandoned for months before it comes back around, the gap is so wide that it stops functioning as spacing and starts functioning as forgetting. The student loses most of what they built before they ever revisit it. The stronger version rotates the positions on a much tighter interval, on the order of every couple of weeks, so that each position returns while it is still fresh and the student keeps compounding instead of rebuilding. The lesson from the science is that there is a window: revisit too soon and you waste reps, revisit too late and you are teaching it again from scratch. A position-based curriculum lives or dies on landing inside that window, and a slow deep-dive rotation usually lands well outside it.

The same interval logic applies to anything you run alongside the main curriculum. Most programs carry a few supplemental threads, takedowns, leg locks, a competition track, and these are exactly where owners let the spacing go slack, teaching takedowns once a quarter and wondering why nobody can shoot. A supplemental thread that surfaces every few months is decorative. One that returns on a tight interval, on the same order as the core rotation, actually builds. If a thread is worth running at all, it is worth running often enough to stick.

It scores strongly on transfer for the same reason. A student who has learned a position as a set of goals and methods is equipped to solve the position live, including in situations the drills never showed them, because they understood the purpose rather than memorizing a sequence. And it is naturally accessible, since positions are modular, a student can enter at any position. The demand it places on the builder is conceptual: you have to be able to articulate, clearly and correctly, what the true goal of each position is and what the real methods for achieving it are, which requires understanding the art at the level of principles, not just moves.

There is one more reason I place this method at the top, and it has nothing to do with beginners. Because the unit is the position and not the technique, an advanced student can stay inside a single position and mine it all the way to the bottom, learning not only its goals and methods but how to manipulate it, how to present the opponent with dilemmas. I owe this framing to Sean Applegate, who put it to me plainly: the ability to create dilemmas is what separates a brown belt from a black belt, and it is where he spends most of his time with his own pros. That idea reshaped how I think about the ceiling of this method. A dilemma is the highest expression of control. Instead of chasing one attack the opponent can simply defend, you arrange the position so that every option they have left is a losing one. Defend the choke and you give up the arm. Save the arm and you give up the back. You are no longer attacking a technique, you are attacking their choices, forcing them to solve a problem that has no good answer. That is only possible once a practitioner understands a position completely, every branch, every reaction, every door in and out, and the positional method is the only one of the four built to take a student that deep into a single position. The beginner learns the position’s goals. The advanced student learns to weaponize the whole thing. One structure serves both, which is exactly why I consider it the most advanced of the four.

Built well, this is, to my mind, the method with the highest ceiling, and it is a genuinely underused way to think about teaching Jiu-Jitsu. It needs no special format or game to work. It is simply a different and, I would argue, more scientific answer to what a curriculum is made of.

Shape: a tight rotation of position-problems. Each position is taught as a goal and the methods that serve it, and positions return before they fade.

First, a single position, dissected the way the method treats every position:

Then the positions themselves rotate on a tight interval, each position held for one short block with its top and bottom taught together, and each returning while it is still fresh:

Every position is a small, complete problem with a purpose and a short menu of solutions, which is how an expert already holds the art in their head. And because the rotation is tight, a position comes back around while the student still remembers it, landing inside the window where returning builds skill instead of rebuilding it.

Teaching a position with its top and bottom in the same block has a second payoff beyond keeping the rotation tight. You can drill the two sides against each other. One partner works the control while the other works the escape, one works the submission while the other works the counter, so both students learn at the same time and the drilling carries live resistance instead of being cooperative. The escaping partner is genuinely trying to escape, which gives the controlling partner something real to hold, and each side comes to understand its half by feeling the other. Learned weeks apart, control and escape are two techniques. Learned as a paired block, they are two sides of one position, which is closer to how the position actually behaves when it is live.

The four types against the four tests

Early on I named four tests any teaching method has to pass: how well it organizes knowledge, how it builds durable memory, how it manages difficulty, and how well it transfers to live rolling. The method-by-method walk above touched each of these in passing. Here they are gathered in one place, every type scored against every test. Read down a column to compare the methods on a single test, or across a row to see the shape of one method.

One thing jumps out of the grid. Only the positional method scores strong on all four tests. Linear is superb at managing difficulty and weak at durable memory, because a one-way chain rarely comes back to what it taught. Rotating is the easiest to run and the weakest on the science, since the student is left to organize and connect the pieces alone. Flow is strong nearly everywhere, which is why it is the hardest to build. And positional is the only one with no soft spot, which is the clearest version of the argument for studying it hardest, whatever else you run.

The four side by side

So which one

There is no single winner, which is the honest answer, but there is a right one for each situation. For your private students, build linear, and price it accordingly, because you are delivering the one structure a group can never get. For a large academy with a rotating cast of instructors who need something reliable and easy to hand off, the modular rotating curriculum is the sane choice, and its inefficiency is the price of its scale. For a serious group academy with a head coach who can do the hard design work, flow gives you most of linear's coherence without closing the door on new students. And the positional method is the one I would tell any owner to study hardest, whatever else they run, because organizing knowledge the way an expert already holds it is the closest thing to a cheat code the science offers, and almost nobody in our sport is teaching that way on purpose.

You can also blend them, and good academies do. A rotating schedule whose individual lessons are built positionally. A flow curriculum for the main class and a linear path for privates. The methods are tools, not tribes. What matters is that you chose your structure on purpose, for reasons you can name, and that the reasons are about how your students actually learn rather than about what is easiest for you to deliver. That is the whole difference between running a curriculum and running classes.

Sequence is only one axis

Everything above answers a single question: in what order do you teach. But the strongest curricula do something the four types alone do not describe. They add a connective layer, a through-line of principles that ties the positions together into one coherent game, so a student is not just meeting positions in a sensible order, they are learning how the whole thing fits and why.

That connective layer is a second axis, independent of sequence, and it is the real difference between a curriculum and a system. It is big enough to be its own conversation, and it is exactly where this series goes next.

Coming next

Part One of this series made the case for having a curriculum at all. This one, Part Two, named the four types and judged them. Part Three takes up the second axis just raised: the difference between a curriculum and a system, and how the best academies turn a good sequence into a coherent game. Further along, the series turns practical, with a build guide for each of the four types in turn, linear, rotating, flow, and positional, walking through how to actually construct it from a blank page. If one of the four already sounds like the fit for your academy, that is the one to watch for.

Working with OSJJ

Building any of these well is real work, and curriculum is one part of what we do. Open Source Jiu-Jitsu builds curricula in all four structures, linear, rotating, flow, and positional, matched to how your academy actually runs, and we help owners choose the right one and construct it from a blank page. It sits alongside the rest of the work, the operations, finances, and systems that keep an academy running, because a strong curriculum and a strong business are equally part of what makes a school last. If any of that is something you want help with, reach out at opensourcejiu-jitsu.com.

Open Source Jiu-Jitsu

Open Minds. Open Mats. Open Source.

Back to the Free Library