The knee shield is taught almost everywhere as a way to not get flattened — a defensive frame that buys time. Craig Jones rebuilt it into the opposite: a single half-guard position that threatens a sweep, a back-take, and a submission at once, so that defending any one of them opens another. This is the three-threat frame, codified in Make Z Guard Great Again.
At ADCC 2022 in Las Vegas, in the under-99-kilogram bracket, Craig Jones settled into the position he is most associated with — a half guard with his top shin framed across his opponent's torso like the letter Z — and the match against Joao Costa was over almost before it registered as having begun. Jones did not sweep. He did not dive at a leg, the thing the whole arena expected from the sport's most famous leg-locker. From the knee shield he trapped an arm, swung underneath it, and finished a rolling reverse armbar — a Choi bar — in roughly eighteen seconds.
It was, by the event's own tally, one of the two fastest submissions of the entire tournament, behind only an eleven-second heel hook from Gordon Ryan on the other side of the bracket. And it came from the last place a scouting report would have flagged as dangerous. Costa was not caught in a leg entanglement or a scramble. He was caught inside a frame that most of the grappling world teaches as a way to survive being on the bottom — a place to stall until you can recover full guard.
That is the whole argument of Jones's Z-guard game in a single exchange. The knee shield is not where you wait. It is where you attack. The shin across the torso is not a wall; it is a launch point that threatens three different things at once, and the submission that ended Costa was simply the threat he chose not to defend.
This is not a story about a fast finish. It is a story about a position, and about what happens when someone decides a defensive frame was never defensive at all.
The Z-guard's value is not any one attack but the unity of three: from the same shin frame Craig Jones can elevate and sweep, wrestle up to the back, or enter the legs — and because all three start identically, defending one opens the next. The knee shield stops being a place to survive and becomes a place to hunt.
The knee shield is old, and it belongs to no one. A bottom half-guard player wedging the top shin across the passer's hips or ribs — a frame that keeps the chest off him and preserves space — appears across decades of half-guard instruction under a handful of names: knee shield, knee-elbow frame, and, when the shin sits diagonally across the torso and the body forms a Z, the Z-guard. As a defensive tool it is close to universal. What was not universal, until relatively recently, was treating it as a primary place to launch offense rather than a brake to slow a pass.
Craig Jones's route to it runs through the same coach who shaped the rest of his game. He came up in Adelaide, then made the decisive move to Melbourne and under , a physiotherapist by training and one of the most analytically minded coaches in the sport. Giles built a program organized around modern, systematized no-gi — half guard, leg entanglements, and the connections between them — at a time when much of the grappling world still treated bottom positions as damage control. Giles promoted Jones to black belt in . The full arc of that partnership, and Jones's career and later business disruption, is covered on his athlete page, ; what concerns this article is one position and what he did to it.
What makes the Z-guard a system rather than a personal habit is that Jones wrote it down. In 2022 he released Make Z Guard Great Again for BJJ Fanatics, a six-volume instructional built entirely around converting the knee shield from a defensive frame into an attacking hub. The volumes are organized the way the game actually runs: retention and how to keep the frame, counters to the standard passes that dismantle it, then the attack families — kimura and triangle, the Choi bar and armbar and omoplata, and only then the back-takes and leg entries — and finally the transitions against opponents who stand up rather than kneel. The title is a joke; the structure is not. It is the clearest single codification of the position as offense that the instructional market has produced.
The Z-guard also appears, from the other direction, in Jones's leg-lock system — there it is one of two entries that feed his heel-hook game from underneath. This article is deliberately the reverse view: the position on its own terms, with the leg entanglement treated as one of its three exits rather than its purpose. The two overlap at exactly one seam, the moment the frame collapses into a leg entry, and cross-link there.
A section in the parent half guard article can tell you what a knee shield is and why a bottom player frames a shin across the top player's body. What it cannot tell you is why, in Jones's hands, that same frame threatens to end the match three different ways from a single starting shape. That is the system, and it has four parts: the frame, the three threats, the attack tree hanging off them, and the retention that keeps the whole thing standing.
Everything begins with where the shin sits. The Z-guard Jones plays is a low knee shield: the top shin is framed across the opponent's midsection with the knee driving toward the far hip, not a high knee jammed into the chest or throat. A high knee shield is easy to see and easy to attack — the passer can pin the knee to the mat, staple it, and pass over the top. A low, active shin does something different: it wedges into the hip line, keeps the opponent's weight off the chest, and — critically — preserves the underhook or the inside position on the far side. The frame is not there to push the opponent away. It is there to hold a specific distance at which all three threats remain live. Lose the distance, or let the knee ride high, and the position reverts to the stall everyone else plays.
The core of the system is not a move. It is a geometry in which three attacks share one starting position, so that the opponent cannot pre-commit to defending any of them without opening the others.
The point is the simultaneity. A passer who drives forward to smother the sweep exposes his arm to the Whizzer and his posture to the wrestle-up. A passer who postures back to defend the wrestle-up unweights the shin and hands over the elevation. A passer who fights the leg entry gives up the space to come up. No single one of the three is unstoppable; together, defended one at a time, they form a position that keeps generating the next attack.
This is where the Z-guard system diverges most sharply from the leg game it is often collapsed into. A great deal of Jones's Z-guard offense is upper body. Once the opponent reacts to the frame, the arm nearest the shin becomes available, and a family of joint attacks opens that has nothing to do with the legs:
These are the tools Make Z Guard Great Again spends its middle volumes on, and they are why the position cannot be filed as a leg-lock entry. A knee shield that only threatens the legs is a knee shield the opponent defends by protecting his legs. A knee shield that threatens the near arm, the head, and the legs is the reason the frame is dangerous even against someone who has read the scouting report.
None of the above matters if the passer simply removes the knee shield, and the standard passes exist precisely to do that: the hip-switch to pin the shin, the "headquarters" position to control the near leg, the leg-weave to fold the frame. The first working volume of Jones's instructional is a counter for each — how to recover the shin when the hip switches, how to re-frame when the knee is pinned, how to keep the underhook when the weave comes. Retention is not a separate skill bolted onto the attacks; it is the attacks. Every recovery of the frame re-arms the three threats, which is why a competent Z-guard player can hold the position for long stretches and why the passer, failing to clear it cleanly, keeps feeding new openings. The frame you keep is the frame you attack from.
By 2026 the knee shield is one of the most common bottom positions in no-gi grappling, played at every level from the first competition stripe to ADCC finals, and a large share of how it is now taught — as offense rather than a stall — traces to the window in which Jones codified it and sold the manual cheaply. Make Z Guard Great Again put a full attacking system on a screen for the price of a couple of seminars, and a generation of hobbyists learned the frame as a launch point because that is how the most-watched instruction on it presents it.
The claim worth correcting is the one that files the Z-guard as a purely defensive relic in an era of dynamic open guards. It is the same misread the position invites in the gym: because the knee shield looks like a survival frame, it gets treated as one. What elite half-guard play in 2026 shows is the opposite — the frame is where a wrestling-literate bottom player generates the most offense with the least risk, precisely because the three threats it carries force the passer to solve three problems at once. The knee shield did not survive the leg-lock era by being defensive. It survived by becoming the cleanest place to start a wrestle-up, a sweep, or a leg entry from the same shape.
What remains distinctly "Jones" is less any single technique than an emphasis: the low frame taught as an attacking distance, the three threats taught as a unit, and the upper-body attack tree taught alongside the leg entry rather than instead of it. The leg-lock branch of his game, and the East-Coast systematization it grew alongside, is the subject of the heel hook and the leg-lock system articles. The Z-guard is the platform all of that launches from — and, on its own, a complete attacking position.
Isolating bouts decided specifically by the Z-guard — rather than by the leg game it can feed — is harder than it sounds, because the position's sweeps and wrestle-ups often lead to points and passes rather than a single clean finish that a highlight can label. One match, however, is unambiguous, and it is the position's clearest demonstration on the biggest stage. Scores, times, and rounds below are flagged for source-check against event records; see Known open items.
White belt learns the frame and nothing else: shin low across the torso, knee toward the far hip, and do not get flattened. If a white belt can hold the knee shield against steady pressure without letting the knee ride high or the underhook disappear, the position's first and hardest lesson is already in place.
Blue belt learns that the knee shield is not a rest. Blue belt is where the frame stops being a place to hide and becomes a place to threaten — the first sweep off the elevation, the first wrestle-up attempt off the Whizzer, the realization that the shin is a lever and not a wall.
Purple belt starts to see the three threats as one thing. The purple belt no longer chooses between sweeping, wrestling up, and entering the legs; the choice is made by whichever the opponent defends first. This is the rank at which the Z-guard becomes legible as a connected system rather than a bag of separate moves.
Brown belt adds the upper-body attack tree and the retention that keeps it live. The brown belt hunts the kimura, the Choi bar, and the triangle off the same frame that feeds the sweep, and — just as important — recovers the shin against the hip switch and the weave so the threats keep re-arming. The frame becomes a position held for minutes, not seconds.
Black belt plays the position as a two-way conversation and steers exchanges into it. The black belt enters the Z-guard from failed sweeps, from bottom half, from a stalled pass, and reads which of the three threats the specific opponent is least equipped to stop. The knee shield is no longer where he ends up; it is where he chooses to fight.
Coral belt teaches all of it at once and can explain, without notes, why the low frame beats the high one and which pass each retention counter answers.
1. Playing the knee shield as a stall. The foundational error, and the one the whole system is built to punish. Treating the frame as a place to rest surrenders its entire value; if you are not threatening a sweep, a wrestle-up, or a leg entry from it, you are simply waiting to be passed. The Z-guard is offense or it is nothing.
2. Riding the knee too high. A high knee shield — shin on the chest or jammed into the throat — is easy to pin, staple, and pass over. The frame Jones plays sits low, across the midsection, wedged into the hip line. High feels safer and is not; it invites the exact passes the position is meant to defeat.
3. Committing to one threat. A player who telegraphs "I am sweeping" — or only ever hunts the leg — lets the opponent defend the single thing he offers. The position works because the three threats are held in reserve together. Show all three; let the opponent's defense pick which one finishes.
4. Losing the underhook or inside position. The frame without the far-side inside control is a frame the passer flattens. Much of the retention game is a fight to keep the underhook, and the attacks that follow depend on it. Give up the inside position and the knee shield collapses into bottom side control.
5. Skipping retention and drilling only the finishes. The most-copied part of the system is the flashy Choi bar; the neglected part is recovering the shin when the hip switches or the weave comes. A Z-guard player who cannot keep the frame never gets to the attacks the frame launches. Learn to keep it before learning to finish from it.
At open mat this week, start every round in the knee shield — low shin across the torso, knee toward the far hip — and give yourself exactly three permitted outcomes: elevate-and-sweep, wrestle up with the Whizzer, or enter a leg entanglement. Nothing else. No recovering full guard, no bailing out. When the frame gets flattened, and it will, reset and start again. You are training the position as offense, which is the only way it works.
Then isolate retention alone. Have a partner run the standard passes — the hip switch, headquarters, the leg-weave — while your only job is to recover the shin and the underhook each time. No attacks. Twenty passes defended. This is the unglamorous half of the system, and it is the half that decides whether you ever reach the other one.
Finally, isolate the upper body. From the frame, hunt only arm attacks — kimura, Choi bar, triangle, armbar — and forbid yourself the legs for three rounds. This forces the half of the Z-guard that separates it from a leg-lock entry, and it is the half that caught Joao Costa in eighteen seconds. Add the leg entries back only once the frame, the sweep, and the arm attacks feel like one position rather than four.
The Z-guard is a half-guard position first and a Craig Jones system second — a knee-shield frame that he rebuilt into an offensive hub.
The map shows the position's shape: rooted in half guard, distinct at the frame and its three threats, shared with the leg game only at the moment the shin collapses into an entanglement. What makes it a system is not a secret technique but an emphasis — the frame taught as a place to attack, and three threats taught as one.
Publication date will be set at Sanity ingest. This log will be appended to over the life of the article. Any factual corrections will be dated and specific.
Publication draft prepared 2026-08-07. Awaiting hero image (needs permissioned photo — placeholder pre-launch). Two-pass edit not yet complete. Source-check pass required before status: scheduled in Sanity.
Known open items: