How to Build a BJJ Game (Without Copying Someone Else's)

"Build your game" is advice everyone receives around blue belt and almost nobody receives instructions for. Here is a version that produces something usable.

A game is a chain, not a collection

Watch anyone good and you will notice they keep arriving at the same handful of positions. Not because they are limited, but because everything they do funnels there. They have a place they take the roll, a small set of things they do from it, and answers for the common reactions.

That is a game: a chain of connected positions, three or four links long, where you already know the branches. It is not a list of favourite techniques, and it is not a highlight reel of things that worked once.

Beginners build lists because lists are easier to assemble. Lists fall apart under pressure because there is no connection between the items, so every exchange starts a new search.

Start where you actually end up

Do not pick a game from a video. Pick it from your own rolls.

Spend two weeks noting where your rolls go, without trying to change anything. Where do you end up on top? Where do you end up on the bottom? Which position produces most of your best moments, however few?

Almost everyone finds a natural tendency: a guard they gravitate toward, a passing style that suits their build, a top position they hold well. Build from there. A game grafted onto a body it does not suit is why so many people spend a year on a guard that was never going to work for their hips.

Bodies do matter here, and less than people claim. Longer limbs suit some guards, shorter and stronger builds suit pressure and close control. That is a starting bias, not a rule.

The four questions per position

For each position in your chain, you need four answers. If you cannot state them, the position is not yours yet.

1. How do I get here? Two entries, not one, ideally from different situations.
2. What is my main attack from here? One, initially.
3. What do they do to stop it, and what do I do then? The counter to the counter is where a game becomes real.
4. What do I do if it fails entirely? The exit that leaves you in a neutral or better position rather than in trouble.

That is the difference between knowing a technique and owning a position. Every one of those four answers is a branch, which is why a game is really a decision tree with a favourite route through it.

Build the defensive game first

Nobody wants to hear this, and it decides how fast the rest arrives.

If your escapes from mount, side control and back control are unreliable, your offensive game never gets to run, because you spend your rounds underneath people. A practitioner with solid escapes and one guard beats a practitioner with five attacking systems and no defence, every time, at every level below advanced.

So the first chain is usually defensive: survive here, escape to there, arrive in your guard. Only then does the attacking chain get built on top.

Three chains that work as starting points

Not prescriptions, examples of the shape:

• Bottom, gi: closed guard, break posture, angle, hip bump or armbar, and back to closed guard when it fails.
• Bottom, no-gi: half guard with an underhook, come up to a single leg, top position, and back to half guard if the single leg stalls.
• Top: knee cut passing, side control, knee on belly, mount, with a defined answer to the frame that stops the knee cut.

Notice how short they are. Three or four links each, each with a return path. That is a complete game for a first year of building.

When to add

Add a new position when the current one is boring, which is the honest signal that it works. If you keep arriving in your position and the same three responses keep working, you are ready. If you are still failing there sometimes, adding something new is avoidance wearing the clothes of ambition.

Also add when a specific opponent type reliably breaks the chain. That is a real gap, and the addition is targeted rather than random.

What does not justify adding: seeing a great video, a seminar, or a competitor doing something spectacular. Those are how people end up with fifty half-learned techniques and no game.

Writing it down keeps it honest

A game plan that exists only in your head drifts. Write the chain, one line per position, with the four answers. One page.

Then check it against reality every month: which link actually held, which one keeps breaking. That check is the whole reason to have written anything down, and it turns a plan into a curriculum, which is exactly what a two-list notebook is for. Between sessions, run the branches as decisions rather than rereading them, because a game you can recite and cannot retrieve under pressure is still a list.

Building a BJJ game