Back Defence — Harness

What This Is

This page covers escape from harness back control — the over-under grip configuration where the attacker’s over-arm threads over one of the defender’s shoulders and grips across to the other, while the under-arm goes under the opposite armpit and meets the over-arm’s wrist (typically in a gable grip). Harness is mechanically distinct from seatbelt: the seatbelt’s two arms cross in front of the defender’s chest, while the harness’s arms form a closed loop around the defender’s torso that is extremely difficult to strip.

For the top-perspective page, see: Harness Control. The harness is popular among advanced no-gi practitioners because it is robust in scrambles — the gable grip does not release under normal defensive pressure, and the closed-loop geometry survives bridging and hip movement better than a seatbelt does. However, the harness’s strength is also its limitation: it is a holding grip, not a finishing grip. The attacker cannot apply an RNC or a back triangle directly from harness; they must transition to a seatbelt or a straitjacket first. That transition is the defender’s primary escape window.

Harness escape therefore has two characteristics that distinguish it from seatbelt escape. First, the defender can afford to spend more time under harness control because no strangle is directly threatening — the chin tuck is still important but it is not the ticking clock it is under seatbelt. Second, the defender’s escape strategy is often to force or invite the attacker’s transition to a finishing grip, then defend the transition. The escape is less about breaking the harness grip and more about exploiting the transition moment.

Also Known As

The Invariable in Action

The gable grip is one of the strongest two-handed grips in grappling because the fingers do not interlock (there is no crushing grip weakness) and the palm-to-palm contact reinforces under load. Attacking the grip directly is usually unproductive — two-on-one pulls do not open a correctly-set gable. The grip opens most reliably at moments when the attacker adjusts their body position, which creates micro-gaps in the grip that the defender can exploit. The escape plan therefore tracks body position changes rather than grip-strength contests.

The harness is structurally inert for strangling because both arms are committed to the torso loop — no arm is available to thread under the chin. The attacker must release at least one hand of the gable to reach the neck. When they do, the grip opens for a moment. A prepared defender times their escape to this transition, not to the static harness hold. Inviting the transition (by making the attacker want to finish) is often more productive than trying to escape the harness itself.

Harness is mechanically asymmetric. The over-arm threads over one shoulder (call it the over-arm side); the under-arm enters under the opposite armpit (the under-arm side). The over-arm side has the attacker’s shoulder pressed against the defender’s upper back; the under-arm side has more rotational freedom because the attacker’s body is further from the defender there. Granby rolls, hip rotations, and turn-to-face all work better toward the under-arm side than the over-arm side. Identifying the under-arm side is the first step of every harness escape.

Defence and Escape

The harness escape strategy is: identify the under-arm side, force or invite the attacker’s grip transition, then execute one of several exits available during the transition.

Priority 1 — Chin Tuck Baseline

Even under harness (where no strangle is directly set), the chin stays tucked. The attacker’s transition attempts will often aim for the neck — a relaxed chin invites those attempts to succeed. Tucked chin baseline means that when the transition begins, the defender’s neck is already protected.

Priority 2 — Identify the Under-Arm Side

Feel which arm of the attacker is over your shoulder and which is under your armpit. The under-arm side is your escape direction. All subsequent escape steps — Granby rolls, hip turns, rotations — go toward that side. This identification must happen before any escape motion.

Priority 3 — Force the Transition

Active harness defence forces the attacker to commit to a finishing attempt. Methods include: heavy pressure toward the attacker’s chest (they must reposition to escape crush), rotation toward the under-arm side (they must transition grips to follow the rotation), or driving to turtle (they must choose between releasing harness to maintain back control and holding harness while conceding position).

Priority 4 — Escape the Transition

When the attacker releases one hand of the gable grip — to begin an RNC thread, to regrip to a seatbelt, to reach for a straitjacket — the grip opens. The defender has one to two seconds to execute an exit: Granby roll (under-arm side), hip turn (under-arm side), turtle drive, or direct face-out rotation depending on what the attacker’s transition leaves available.

Priority 5 — Belly-Down Bail-Out

If the harness is deep and the transition is not coming, belly-down is available as a last-resort escape route. See: Turtle Recovery. Harness does not work well in belly-down because the attacker’s arms are trapped around the defender’s ribs rather than the neck, and the belly-down position mechanically pins the attacker’s own hands.

Named Harness Escape Techniques

Common Errors

Drilling Notes

Ability Level Guidance

Related Pages