Cross Ashi Garami — Defending

What This Is

This is the saddle from underneath — the side of the position nobody chooses. Your leg is crossed and trapped, the attacker sits perpendicular to you with both of their legs committed to one of yours, and the inside heel is the thing they want.

The saddle is the strongest leg entanglement in the sport, and the reason is not the finish. It is that the attacker holds it almost for free. Their legs do the work, their hands stay available, and they can wait. You cannot. That asymmetry — cheap to hold, expensive to survive — is what this page is about.

Nothing here is a rescue from a finished saddle. Every reliable answer happens before the position is complete, which is why the page spends more time on the moments before than on the position itself.

Safety First

If you are new to leg entanglements, defend this position with a partner who will release rather than finish. There is no version of learning it by surviving it.

The Two Sides Compared

The same position, read from each side. Advantage is one number and it hides most of what matters here — the interesting asymmetry is in what the position costs each player and how long it lasts.

The key difference: the saddle is the back control of the leg game — the attacker spends almost nothing to hold it and the defender has almost no time to leave.

Read the energy row alongside the time row. Most bad positions are expensive for both players, and that shared cost is what gives a defender their window: the attacker tires, or overreaches, and something opens. The saddle removes that. It is cheap for the attacker precisely because their legs hold it, so waiting is free for them and ruinous for you.

How You End Up Here

Almost never directly. The saddle is where a leg-lock sequence arrives rather than where it starts, and the two common paths matter because they are the last places the position could still have been prevented.

Conceding the inside line

Most saddles are built from an entanglement you were already in — ashi garami or single leg X — when the attacker crosses to the inside. That crossing is the position being made. Once their outside leg travels over yours and lands, the saddle exists and this page begins.

The shoulder roll and the Imanari entry

An attacker rolling underneath arrives already crossed. There is no gradual concession to notice here, which is what makes these entries dangerous to a defender whose reads are calibrated on slower entanglements.

The Invariant in Action

This is the whole position from your side. The attacker has taken the contested space, and every technique on this page is an attempt to get some of it back or to leave before it matters. A defender who accepts the inside space and then hunts for a clever escape has agreed to the terms of the exchange.

This is why the entry deserves more attention than the position. Which entanglement you are in decides which submissions exist; conceding the cross is what changes the answer from “outside heel hook, and I have time” to “inside heel hook, and I do not”.

This is the reason the heel is the thing you hide. The foot is the handle — deny the handle and the knee cannot be turned, regardless of how good the rest of their control is.

This fixes the order of every escape on the page. Space has to be made before anything can move through it — clear the knee line first, extract second. A defender hauling on a trapped leg is trying to move through space that does not exist yet.

This is the clock. The knee is being attacked across its natural range rather than along it, so the danger zone arrives with very little travel. That is the mechanical reason the deadline is the attacker reaching the heel rather than anything you feel.

Escape Mechanics

Hide the heel first, always

Before anything else, the heel goes where the hand cannot reach it — turned in toward your own centreline, foot flexed, shin rotated so the heel is behind rather than beside. A saddle with no available heel is a control position rather than a submission, and control you can work against. This buys the time every other option needs.

Deny the second leg

The saddle finishes when the attacker gets both of your legs — the trapped one and the free one — because a defender with a free leg has a frame, a base and a direction to travel. Keep the second leg out of the entanglement and moving. This is the single most useful thing to do and the thing most often forgotten while attention is on the trapped foot.

Come to your side, toward the attacker

Flat on your back is where the finish is easiest. Turning toward the attacker unwinds the rotation the heel hook needs and shortens the lever between their hips and your knee. Turning away increases both.

Clear the knee line, then extract

Extraction is the exit, and it only works after the knee line is clear. Pulling on a trapped leg while the knee is still crossed feeds the rotation. Clear first, straighten second, leave third.

What Causes Escapes to Fail

Working on the leg instead of the heel. A defender who spends the first two seconds trying to pull the leg out, and only then hides the heel, has usually given away the only two seconds they had.

Turning away. It is the instinct, and it is the direction the submission wants. Every degree away from the attacker is a degree of rotation you have donated.

Feeding the second leg. Bringing the free leg in to help — to push, to frame, to base — is how a controlled saddle becomes a finished one.

Waiting for a signal. The heel hook does not announce itself. The cue to act is the attacker reaching for the heel, not any sensation in the knee.

Counter-Offensive Options

An attacker in the saddle has committed both legs to one of yours, which is a large investment in a small target. That investment is a real counter-attacking opportunity and it is the least-taught part of defending here.

The honest framing: counter-attacking from a saddle you are losing is a worse plan than leaving. But an attacker whose heel control has failed — because you hid it early — has a position they must maintain with a defender who is no longer in immediate danger, and that is where the exchange becomes two-sided.

See the Junny lock for the catalogued example of this idea from a different entanglement: attacking the reaping leg of someone who is reaping you.

Common Errors

Treating it like the outside entanglements. The defensive habits that work against an outside ashi — booting the foot, sitting up, wrestling the leg free — are slower than the inside heel hook. The position is different in kind, not degree.

Waiting to be sure. By the time a saddle is unambiguous it is finished. Acting on the crossing rather than the arrival is what separates defenders who escape from defenders who tap.

Toughing it out. There is no leg-lock equivalent of surviving a choke. The tissue does not warn you first.

Drilling Notes

Drill the entry, not the position. Start your partner in ashi garami and let them cross — the useful repetition is recognising the moment the position changes character, which a static saddle start never teaches.

Drill the heel hide as a reflex before drilling any escape. It is the step that buys every other step its time, and under pressure it is the first thing to disappear.

Agree the release before you begin. A partner who finishes to prove the position works has taught nothing that a partner releasing at heel contact does not teach more safely.

Ability Level Guidance

Proficient. Not because the defence is complicated — it is four ideas — but because the timing window is narrow and the cost of being late is a knee. This is a position to meet under supervision with a cooperative partner well before meeting it in a roll.

Related Pages