For most of jiu-jitsu's history, the knee wedged across a passer's hip was a stalling tactic — something you did for two seconds before getting flattened. The Z-guard turned that shin into a frame with range, catalogued in its own instructionals, with its own sweep, back-take, and leg-entry tree. It is not bottom half guard's rest stop. It is a different fight.
For most of the sport's history, nobody would have paid for an instructional about this shape. A bottom player's shin, angled across the top of a passer's hip, with the two legs of the guard player forming a rough letter Z when seen from above — that frame existed in half-guard curricula for decades as a stopgap. You threw the knee up, bought a beat of space, and used it to scramble somewhere else: back to full guard, up to the dogfight, anywhere but there. Nobody taught it as a destination, because nobody believed you could live in it.
That belief cracked open between roughly 2018 and 2022, when the position most schools still filed under "half guard, miscellaneous" started getting its own name on a product page. Craig Jones released a solo instructional built entirely around it in 2019. Bernardo Faria began teaching it as the entry point to his bottom game rather than a checkpoint inside it. Felipe Pena Preguiça, a multiple-time black-belt world champion, built a standalone knee-shield system in 2022, the same year Jones released a second, longer solo title on the position. Four instructionals, four different competitive résumés, one shared premise: the frame everyone had been taught to abandon in two seconds was actually a place to fight from indefinitely.
That is the whole case for treating the Z-guard as its own position rather than a paragraph inside half guard. It is not what happens when bottom half guard goes right. It is a structurally different guard — longer range, angle-driven, built to manage distance rather than survive a pin — that happens to share a parent with the position everyone already knows.
The Z-guard's shin frame does not just buy time inside bottom half guard — it converts the position from a pin you survive at close range into a long-range guard with its own sweep, back-take, and leg-entry tree, playable indefinitely rather than escaped from in two seconds.
The knee shield itself is old and belongs to no one in particular. A bottom half-guard player wedging the top shin across a passer's hip or ribs to hold distance appears across decades of grappling instruction, under a scatter of names — knee shield, knee-in half guard, and, when the shape reads as a Z from above, Z-guard. Community sources commonly credit the specific "Z-guard" name to Leo Kirby, a black belt associated with Marcelo Garcia's lineage, though — as with most terminology in a sport this old — the underlying frame long predates any single naming credit, and that attribution should be read as a naming claim, not an invention claim.
The mechanical foundation belongs to , who rebuilt bottom half guard around the far-side underhook after a career-threatening knee injury in the late 1980s. That story, and the underhook half guard it produced, is the subject of the parent article. What concerns this article is narrower: the moment the , as opposed to Correa's underhook game generally, got treated as a stand-alone frame worth its own system.
Lucas Leite was an early and influential bridge. His pressure-and-angle coyote half guard, built through the 2000s, typically starts from a knee shield before the underhook is secured — using the frame not as a stall but as the on-ramp to his sweeping system. Bernardo Faria, a five-time IBJJF World Champion, went further: in his instructional material aimed at grapplers without an athletic advantage — commonly marketed as his Jiu-Jitsu for Old Guys series through BJJ Fanatics — Faria teaches the knee shield as the easiest way to begin playing half guard at all, an entry point rather than a fallback, built around sweeps his academy has since documented under names including the Old School Sweep and the kicking sweep (the latter shown in material from his academy credited to black belt Joseph Cashin).
Lachlan Giles, the Melbourne-based coach who promoted Craig Jones to black belt in 2016, systematized the knee shield's defensive half — his Half Guard Anthology and a course specifically titled Defensive Knee Shield treat the frame as a retention tool, the shape a grappler should be able to hold against pressure before ever attacking from it.
Jones is the figure most responsible for the position's offensive codification. He released The Z Guard Encyclopedia through BJJ Fanatics in 2019 — an early example of a BJJ Fanatics title built around a single frame rather than a whole guard family — and followed it in 2022 with a second, six-volume release, Make Z Guard Great Again, covering the frame's attacks, retention, and passing counters in more depth. That second instructional is the subject of a separate bjj.media article, Craig Jones's Z-Guard System, which covers his specific three-threat framework; this article treats the underlying position the two instructionals share. The same year, Felipe Pena Preguiça — a three-time black-belt gi World Champion, including the 2021 IBJJF Absolute title — released his own three-volume No Gi Knee Shield Guard, built around an unorthodox sweep set that differs in flavor from both Faria's and Jones's approach.
Four practitioners, four instructionals, a shared four-year window. That is what "codified" means for a position that used to have no name at all.
Start with the shape, because the shape is the entire argument for treating this as a distinct position rather than a subsection of half guard.
The frame: a diagonal shin, not a flat pin. In a passed, pinned bottom half guard — the position the parent article and its companion recovery piece both treat in depth — the top player's chest is welded across the bottom player's sternum, and the whole exchange happens at zero range. The Z-guard replaces that zero range with several inches of standing structure. The bottom leg hooks around the opponent's near leg at the knee or thigh, and the top leg's shin angles diagonally across the opponent's hip or lower torso, knee driving toward the far side. Viewed from above, the two legs and the torso trace a rough Z — hence the name. Some technical breakdowns distinguish a low Z-guard, where the feet connect and clamp around the opponent's leg for a stronger structural lock, from a more generic knee shield, where the shin merely rests across the body without the same foot connection; the terms are used loosely and often interchangeably across instructional material, and this article treats "Z-guard" and "knee-shield half guard" as the same family, following the most common community usage, while noting the stricter distinction exists.
Distance, not proximity, is the organizing variable. Where the parent half guard's underhook game is fundamentally a close-range fight — win the arm, stay glued to the opponent's chest — the Z-guard's shin is a spacer. It holds the opponent's hips and chest away from the bottom player's own torso, converting a smash-range exchange into a frame-range one. This single change is what makes the position playable indefinitely rather than escaped from in two seconds: a bottom player who has never been taught to attack from bottom half can still be taught to live safely behind a well-maintained shin, because the frame itself, not a leg lock on the opponent, is doing the defensive work.
Grips still matter, and they still route through the far side. The Z-guard does not discard Correa's underhook lesson; it delays it. With the shin holding range, the bottom player's near-side hand typically fights for a collar, wrist, or sleeve control on the opponent's far side, and the far-side arm hunts the same underhook that defines the parent position — but from a distance at which the fight for it is calmer and less desperate than it is when flattened. A Z-guard player who never wins the underhook is not automatically passed the way a flattened bottom-half player is; the frame buys time the underhook fight in bottom half does not have.
The attack and sweep tree. Because the position holds range rather than surrendering it, its offense fans out further than a pinned bottom half guard's does. Four families recur across the instructional material surveyed above, independent of which coach is teaching them:
The unifying idea, stated plainly: a flattened bottom half guard has one job — recover the underhook or lose the position. A Z-guard has four live exits, because the shin has bought the range to hold more than one threat open at once.
By 2026, the knee-shield frame is closer to a universal habit than a specialist's tool — it is one of the first things a competitor reaches for the moment a more dynamic guard collapses, in both gi and no-gi. What changed over the past decade is not that the frame exists; it is that a growing share of the sport's instructional output treats it as a place to attack from rather than merely hide behind.
The claim worth correcting is the one that files the Z-guard as a purely defensive relic, useful only for surviving a smash pass until a scramble arrives. The instructional record argues the opposite. Faria markets it as an entry point for grapplers who cannot out-athlete their opponents. Pena, a gi specialist with an unorthodox competitive style, built a standalone system out of it in no-gi. Jones, whose broader reputation rests on the leg-lock game, has released two separate solo titles on this one frame rather than folding it into his leg-attack material. None of the three is trying to survive from the position; all three are trying to win from it.
What remains genuinely distinct about the Z-guard, relative to the parent half guard, is the range at which the whole exchange happens. Bottom-half recovery — covered separately by bjj.media as its own escape sequence — is fundamentally about closing a gap the passer has already won: frame, bridge, shrimp, and fight the underhook back from a flattened start. The Z-guard skips that fight by never fully surrendering the range in the first place. It is why coaches increasingly teach the knee shield as the first shape a guard player should default to when a pass begins, rather than the frame they reach for only after already being flattened.
Isolating bouts decided cleanly by the Z-guard specifically — as opposed to the broader half-guard or leg-lock exchanges it can feed into — is harder than for a single-motion submission, because the position's value is usually a sweep or a sustained frame rather than an isolated finish. One match stands out as unambiguous.
Beyond that single finish, the position's competitive footprint is real but diffuse: Faria's and Pena's world-championship careers both run substantially through knee-shield entries into their broader bottom games, but neither has a single, cleanly-dated match in which the sweep sequence from the Z-guard specifically — as opposed to their half-guard or deep-half games more broadly — can be verified without access to round-by-round footage. Rather than assert a specific sweep, opponent, and timestamp that cannot be checked, this article leaves that competitive record general and flags it below.
White belt learns the frame exists and that it is not a stall to be abandoned in a panic. The first lesson is structural: shin diagonal across the hip, knee toward the far side, foot active — and staying there without immediately trying to scramble to something else.
Blue belt learns the frame has range. The realization that reorganizes the position is that the shin is not just a wall against the pass but a lever — the first elevation sweep, the first sense that distance can be converted into offense rather than only defense.
Purple belt learns to hold more than one threat at once. Rather than choosing between the sweep, the wrestle-up, and the leg entry in advance, the purple belt learns to let the opponent's reaction to the frame decide which exit opens first.
Brown belt adds the upper-body attack tree without giving up the frame that makes it possible. Kimura, triangle, and armbar variants become live the moment the opponent reaches or postures, and the brown belt has learned to recover the shin against the standard passing counters — the hip switch, the knee weave — so the position survives contact rather than collapsing on first pressure.
Black belt treats the Z-guard as a chosen entry rather than an accident. It becomes the frame a black belt steers toward from a failed sweep, a stalled pass, or a collapsing open guard, because it is where their offense is densest and their risk is lowest.
Coral belt can explain, without notes, which coach's material teaches which sweep, why the low shin beats the high one, and where the frame's naming history is solid versus folkloric.
1. Riding the knee too high. A shin jammed at the chest or throat rather than the hip is easy for a passer to pin and step over. Fix: keep the frame low, across the hip line, where it holds structure instead of inviting a smash.
2. Playing it as a stall rather than a range. A bottom player who inserts the shin and then does nothing has recreated the old two-second habit the position was built to replace. Fix: treat the frame as a live lever from the moment it lands — threaten the sweep, the wrestle-up, or the leg entry, not just survival.
3. Losing the far-side grip fight. The shin alone does not finish anything; without a hand fighting for the collar, wrist, or eventual underhook on the far side, the passer can wait the frame out. Fix: the shin buys time for the hands to work — never let the lower body carry the whole position alone.
4. Letting the frame collapse under sustained pressure. A knee shield that gets flattened by repeated hip-switching or weaving reverts to ordinary pinned bottom half guard, with all of that position's problems. Fix: drill the specific recovery counters — re-inserting the shin against the standard passes — as its own skill, not an afterthought to the attacks.
5. Treating the Z-guard and a flattened bottom half guard as the same emergency. They call for different responses. A flattened bottom half guard needs the frame-bridge-shrimp-underhook recovery sequence described in the parent position's companion article; a live Z-guard needs range management and grip fighting. Confusing the two — panicking out of a healthy frame, or trying to sweep from a position that has already gone flat — wastes the position's central advantage.
At open mat this week, start every round already inside the Z-guard — shin diagonal across the hip, foot active, near-side grip established — rather than working up to it. Give the top player one job: pass. Give the bottom player three permitted outcomes and nothing else: sweep off the elevation, wrestle up to an overhook, or enter a leg attack. No bailing to full guard, no submissions outside the frame's own tree. When the shin gets flattened, reset and start again rather than scrambling out of trouble by other means.
Then isolate retention alone. Have a partner run the standard counters — the hip switch, the knee weave, sustained forward pressure — while the only job is to keep the shin alive and recover it the instant it is threatened. No attacks. This is the unglamorous half of the position and the half most hobbyist training skips.
Finally, isolate the upper-body tree by itself. From the frame, hunt only arm attacks — kimura, triangle, armbar variants — and forbid the legs for several rounds. This is the half of the position that separates it from a pure leg-lock entry, and it is the half that produced the fastest documented Z-guard finish on record.
The Z-guard is a half-guard position first — a frame that shares its parent's trapped-leg structure but manages distance rather than surviving a pin.
Every one of these was reached by someone who decided a two-second stopover was worth staying in. That decision is the map.
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 2026-08-08. 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: