Dogfight

What This Is

The dogfight is a neutral scramble position that arises from half guard when the bottom player stands up to their knees or comes up to a kneeling position. Both players are now in a kneeling or partially standing configuration with one player’s leg still captured in the half guard trap. Neither player has positional dominance. Neither player has full base. The underhook battle is the entire contest: whoever secures the underhook on the trapped-leg side controls that hip, and controlling the hip determines what happens next.

The dogfight is not a position either player wants to be in for long. It is a transition — a moment of neutrality between the guard-and-pass dynamic below and whatever resolves above. The bottom player who stands up to dogfight has escaped the danger of being flattened and passed; the top player who follows the bottom player up has prevented the guard recovery that would result from simply being left behind. Both players are now in a scramble where technique and timing determine the winner, not the pre-existing positional advantage.

Because the dogfight is genuinely neutral — neither player has structural advantage until the underhook is won — it is a critical position for practitioners who rely on half guard. The stand-up to dogfight is often the first step in converting half guard from a defensive survival position into an offensive scramble. But a practitioner who stands up to dogfight without a plan — without knowing what to do when the underhook is won or lost — will find themselves standing in a scramble they are not equipped to resolve, and the top player will take them back down, re-establish the pass, and be further ahead than before.

The Invariable in Action

The dogfight makes INV-07 explicit as a contest. The moment both players arrive at the kneeling position, both players are reaching for the underhook — both players are trying to establish the connection that will give them control. The underhook is the connection: an arm under the opponent’s armpit on the trapped-leg side that makes contact with the opponent’s back and controls their shoulder and hip. Without the underhook established, neither player has meaningful connection; they are just kneeling near each other with one leg configuration remaining. The player who establishes the underhook first has closed the distance on the critical structure and acquired control before the opponent can. This is why the underhook fight in the dogfight is not incidental — it is INV-07 made into an explicit battle.

The dogfight is a balance contest before it is a technique contest. Both players are in a partially standing configuration with compromised base — one leg is still captured in the half guard trap. Neither player has full structural stability. INV-13 governs what happens when the underhook is won: the player with the underhook must use it to destabilise the opponent before committing to a specific attack. The underhook alone, without disrupting the opponent’s balance through it, does not produce a back take or a single leg — it just produces a stalemate with one player’s arm deeper than the other’s. The destabilisation — the shoulder drive, the hip bump, the weight transfer — is what converts the underhook advantage into an actual attack. Without it, the opponent’s balance allows them to defend whatever follows.

The dogfight is a hip height contest expressed through the underhook battle. Winning the underhook is valuable precisely because it enables the shoulder drive that elevates the practitioner’s hips relative to the opponent’s. The back take from the dogfight works because the underhook-side drive raises the attacker’s hip above the opponent’s hip, allowing the swing behind — the back take is, mechanically, a height advantage being converted to a positional advantage. The player who remains lower is the player who is being taken behind.

The shoulder drive in the dogfight does not directly produce the back take — it forces a reaction, and that reaction opens the back. When the bottom player drives the underhook shoulder into the top player’s chest, the top player’s instinct is to base wide and resist. That basing motion is the opening: as the top player’s weight shifts outward, the bottom player’s hip swings behind. The player who understands this is not trying to overpower the top player through the shoulder drive — they are using the drive to elicit the reaction that opens the swing.

The dogfight’s leg trap is the bottom player’s structural tether. Releasing it deliberately — stepping the trapped leg out and converting to a wrestling stance — is not a concession; it is a positional reset that can deny the top player the leverage they are expecting from the trap. When the top player’s strategy depends on the leg being captured (preventing the stand-up, creating the angle for re-flattening), voluntarily releasing the trap breaks that strategy. The bottom player who releases on their own terms can convert to a single leg or a neutral wrestling position; the bottom player whose trap is stripped by the top player is reacting, not acting.

Entering This Position

From Lockdown — Stand-Up

The primary entry. The bottom player in the lockdown posts the top foot, drives the hip up off the mat, and comes to the near knee in one continuous motion. Both players are now in the dogfight configuration: one player’s leg is still trapped, both players are kneeling or partially standing, and the underhook battle begins immediately. The lockdown’s calf hook should be released as the bottom player stands — holding the lockdown through the stand-up makes the movement mechanical and slow. The stand-up is the transition; the dogfight begins the moment both knees leave the mat.

From Half Guard Bottom — Direct Stand-Up

Without the lockdown, the bottom player can still come to the dogfight from standard half guard. With the underhook already in place, the bottom player posts the near foot and drives off the mat. This is faster than the lockdown entry but requires the underhook to already be won — standing up without the underhook into the dogfight is entering a neutral scramble with a disadvantage already in place.

From Underhook Half Guard — Knee Drive

When the bottom player has the underhook in the half guard and the top player’s pressure lightens — they base back to defend a sweep, or they post a hand to stop an attack — the bottom player can drive the near knee into the mat and come to the dogfight position directly. The momentary lightening of pressure is the opening: the bottom player does not wait for a better opening but takes the base-recovery moment immediately.

From This Position

The dogfight resolves in one of several directions depending on who wins the underhook battle. The bottom player and top player have different option sets, though the position is genuinely neutral until the underhook is decided.

Bottom Player — Wins the Underhook

Back take: With the underhook under the top player’s armpit, the bottom player drives their shoulder into the top player’s chest, disrupts the top player’s base (INV-13), and reaches the far arm around the top player’s waist to take the seatbelt. The body follows the underhook — the bottom player swings behind the top player as the seatbelt is established. This is the highest-value exit from the dogfight with the underhook: the back is the best position in grappling, and the dogfight underhook is a direct path to it when used correctly.

Single leg attack: The underhook controls the top player’s shoulder and limits their ability to sprawl. The bottom player who drives the underhook shoulder through and ducks their head to the outside of the top player’s leg can take the single leg directly from the dogfight. The top player’s leg is already compromised by the half guard trap; the single leg extends that control from the ground up.

Pull back to guard: If the back take and single leg are both defended, the bottom player can pull the top player back down into half guard and reset. The underhook prevents the top player from advancing during the re-engagement. This is not a failure — it is controlled repositioning.

Top Player — Wins the Underhook

Re-flatten to half guard: With the underhook under the bottom player’s armpit, the top player drives the elbow down and flattens the bottom player back to the mat. The dogfight is over; the top player is back in a top half guard passing position with the underhook advantage. From here, the pass continues.

Advance to front headlock: The top player with the underhook can drive the bottom player’s head down — the bottom player’s head is close to the top player’s body during the dogfight — and establish the front headlock as the bottom player’s posture collapses. From the front headlock, the top player has multiple attacking options. See the Front Headlock page.

Hip switch to side control: If the bottom player’s base is already compromised during the underhook battle, the top player can use the underhook to hip switch — swinging their hips over the bottom player — and establish side control directly from the dogfight position.

Common Errors — and Why They Fail

Error: Standing up to dogfight without a plan for the underhook battle. Why it fails: The dogfight is resolved by the underhook. A bottom player who stands up and then looks around to figure out what to do next will find the top player’s underhook already in place. The stand-up and the underhook fight must begin simultaneously — the bottom player is fighting for the underhook as they come up, not after they arrive. Correction: The underhook fight begins during the stand-up. As the bottom player posts and drives up, the near arm is already reaching for the underhook position. The arrival in the dogfight and the underhook entry are one movement, not two sequential actions.

Error: Stalling in the dogfight without committing to the underhook battle. Why it fails: INV-07. A player who does not actively fight for the underhook creates distance rather than connection. The dogfight drags on as both players hold positions without resolving the underhook, and eventually one player tires or the situation produces a scramble that neither player controls. In competition, prolonged stalling without engagement produces stalling penalties. Correction: The underhook battle must be active and continuous. Try, fail, adjust, try again — but never stop fighting for the underhook. A practitioner who cannot win the underhook directly should be trying to strip the top player’s underhook, which is functionally the same fight.

Error: Attempting the back take without first destabilising the top player through the underhook. Why it fails: INV-13. A balanced top player with the underhook lost will simply base and prevent the swing behind. The back take requires the top player’s hip to be already moving — the bottom player drives the underhook shoulder through, disrupts the top player’s base, and then swings behind. If the swing behind is attempted without the prior destabilisation, the top player posts and the back take stalls. Correction: The sequence is always: win underhook → drive shoulder through → feel top player’s balance disrupted → swing behind and establish seatbelt. Do not skip the shoulder drive. It is not cosmetic.

Error: Abandoning the half guard trap during the dogfight stand-up. Why it fails: The leg trap is the bottom player’s structural control during the dogfight. Releasing the trap before establishing a new control (underhook, single leg, back take) gives the top player full freedom to step back and reset to a passing position. The bottom player has stood up into a scramble they no longer have any leverage in. Correction: Maintain the leg trap through the dogfight. The trap is released only when another control replaces it: the single leg, the back take, or a deliberate return to guard. It is never released without a replacement.

Drilling Notes

Ecological Approach

Dogfight resolution game: Both players start in the dogfight position (one player’s leg trapped, both kneeling). The bottom player scores by completing a back take (seatbelt established for three seconds) or a single leg sweep to top position. The top player scores by completing the pass to side control or mount. No submissions. Run ninety seconds, switch. This game forces both players to develop real-time underhook fighting skills and tests whether each player has a plan for winning the position.

Systematic Approach

Phase 1 — Stand-up entry. From lockdown or underhook half guard, the bottom player drills the stand-up to dogfight. Top player is cooperative and follows the movement. Focus: underhook fight begins during the stand-up, leg trap is maintained through the movement, side position is preserved as the bottom player comes up. Twenty repetitions.

Phase 2 — Underhook win to back take. From dogfight, both players fight for the underhook. When the bottom player wins it (cooperative for now), they drill the shoulder drive and back take sequence. Focus: shoulder drive first (INV-13), then swing behind, then seatbelt. Twenty repetitions.

Phase 3 — Underhook loss response. From dogfight, the top player wins the underhook (cooperative). The bottom player drills stripping the underhook — underpumping or creating space — before the top player can use it to flatten. This is a defensive skill that is just as important as the offensive sequence. Twenty repetitions.

Phase 4 — Dogfight resolution game (ecological), as above.

Ability Level Guidance

Also Known As

Related Pages

Concepts layer