Keenan Cornelius did not just cross the worm guard's lapel tether with the reverse De La Riva hook — he named the result, flowcharted it, and then called it his favorite guard to attack from. Reverse de la worm binds the same leg twice, once with a hook and once with fabric, and the two locks reinforce rather than compete.
Somewhere in the process of filming the Lapel Encyclopedia, Keenan Cornelius arrived at a guard he liked better than the one that made him famous.
By his own account — the line that recurs across the secondary coverage of the release — the worm guard was no longer his favorite tool for attacking from the bottom. A different entry was: one that kept the worm guard's lapel tether but married it to the inside hook of the reverse De La Riva, on the same leg, so that the fabric and the frame were no longer two separate problems for the opponent but one compounding one. Cornelius called it reverse de la worm, gave it its own chapter, and told viewers plainly that it was "the most powerful of the guards from the worm guard systems and the one he currently prefers to attack from."
That is an unusual thing for an inventor to say about a variation of his own invention. The worm guard is the position with the origin story — the lapel pulled loose under pressure against a ranked passer, the name supplied by a training partner, the word that entered the sport's vocabulary within a year. Reverse de la worm has no equivalent debut match on record. What it has instead is something arguably more telling for a system built to be taught rather than merely performed once: its own flowchart, its own chapter, and — the clearest sign a position has been taken seriously by the people who have to answer it — its own dedicated section in Cornelius's passing instructional, built specifically to kill it.
This is a story about the moment a family of guards produced a member that its own creator ranked above the original.
The worm guard ties one leg with fabric. The reverse De La Riva ties one leg with a hook. Reverse de la worm ties the same leg with both at once — and the two controls do not compete for space, they reinforce each other, which is why its own inventor calls it the strongest guard in the family he built.
Reverse de la worm did not arrive as a standalone invention with its own debut match. It arrived as the second-generation product of a system that was already searching itself for improvements — and its birthplace is documented precisely because Cornelius filed it inside the same reference project that catalogued the rest of the family.
The parent system is well established. Keenan Cornelius built the worm guard in the Atos Jiu-Jitsu competition room around 2014, under André Galvão, and the lapel-under-the-leg tether that resulted became the founding mechanic of an entire category of gi guards — a fuller account of that origin, the 2014 Pan American debut against Murilo Santana, and the Galvão naming story lives on the and does not need repeating here. What matters for this position is what happened after: once the worm guard worked, Cornelius kept searching the same design space — different fabric, different limb, different starting frame — for every problem the original entry exposed. Reverse de la worm is one of those answers, and by his own repeated framing, the best one.
The name follows the family's naming logic exactly. "Worm guard" describes the lapel snaking under a limb; "reverse de la worm" describes the same lapel mechanic layered onto a reverse De La Riva leg frame — the inside-hook guard, itself a later addition to the De La Riva family, credited in community history to Caio Terra and popularized on the competitive stage by Rafael Mendes and the Miyao brothers (see the reverse De La Riva profile for that lineage in full). Cornelius's contribution was not inventing either half. It was noticing that the two structures — a fabric tether and a leg hook — could be run on the same leg without fighting each other, and that doing so produced a guard more resistant to escape than either half alone.
The codification is where the position's paper trail becomes concrete rather than anecdotal. Reverse de la worm is documented as one of the named members inside the Lapel Encyclopedia (Keenan Online, 2019) — the reference-scale instructional that treats the whole lapel family as a taxonomy rather than a single technique, and which sits alongside the earlier, narrower Worm Guard instructional (c. 2015) as the family's two codification events. Independent secondary coverage of the Encyclopedia's release describes a dedicated reverse-de-la-worm chapter of roughly thirty minutes, illustrated with its own flowchart diagramming entries, defenses, and attack branches — the same treatment given to the worm guard and squid guard as standalone members of the catalog, not folded into either as a footnote. A later review of the Encyclopedia catalogued it among the roughly 166 lessons spanning "worm guard, squid guard, and reverse de la worm" as three of the release's distinct named sections.
The strongest evidence of the position's discreteness, though, is not the attack-side chapter. It is the defense-side one. Cornelius's passing instructional — marketed under the Lapel Slayer name — carries its own section labeled for the position specifically, opening with the line "Opponent playing reverse de la worm on you? Here are your options," followed by named counters: a foot-pop escape and a backwards-roll counter Cornelius attributes to Lucas Lepri, the Alliance competitor and coach known for pressure passing. A passing curriculum does not build a bespoke section, with a bespoke named counter, for a guard that is indistinguishable from its neighbors. It builds one because the position presents a distinct enough problem to require a distinct answer.
That is the origin in full: not a match, but a search. Built inside the Atos-to-Cornelius worm-guard lineage, assembled from the reverse De La Riva's hook and the worm guard's tether, named for the fusion, chaptered and flowcharted in the Lapel Encyclopedia, and important enough to earn its own countermeasures in Lapel Slayer.
Reverse de la worm is defined by two simultaneous points of control on the same leg — a hook and a fabric loop — and understanding why that combination behaves differently from either ingredient alone is the entire mechanical spine of the position.
1. The reverse De La Riva hook. The base frame is the reverse De La Riva: the bottom player threads a hook to the inside of the opponent's lead leg, mirroring the classic De La Riva's outside-leg wrap. As on the parent guard, the shin sits against the inside of the opponent's shin or thigh, and the knee of the hooking leg angles across the body. This hook alone is what makes the position immune to the knee-cut pass that beats classic De La Riva — a property it inherits directly from its parent (see Reverse De La Riva for the full mechanical case).
2. The lapel, freed and routed under the same leg. This is the addition that makes it a worm-family guard rather than a plain reverse De La Riva. The bottom player pulls a lapel loose from the belt and threads it under the same hooked leg, reclaiming the tail in the gripping hand. The published distinction between this position and its closest sibling, the ringworm guard, is a single variable: where exactly the lapel weaves relative to the opponent's shin. Cornelius's own material describes the difference as being "literally in the placement of the lapel, with the lapel underneath your leg but in front of your opponent's shin" — a small routing change that nonetheless changes which sweeps and back-takes are live. That level of fine-grained differentiation between named family members is itself evidence the family is a genuine taxonomy and not a marketing exercise: two guards that differ by one weave of fabric still earn separate names, separate flowcharts, and separate chapters, because the small change produces a materially different game.
3. Why the two controls compound instead of compete. A plain reverse De La Riva hook is, like its parent, only as strong as the grip holding it — sleeve, ankle, or belt, managed by an arm that tires and can be stripped. A plain worm guard tether, absent any hook, controls a leg but concedes the passer's upper-body posture and base entirely. Reverse de la worm removes both weaknesses at once, because the hook and the tether are doing different jobs on the same limb: the hook manages the opponent's knee line and blocks the knee-cut, while the lapel — a length of fabric anchored to the opponent's own torso — carries the load of keeping that leg immobilized without depending on the strength of the gripping arm. The passer cannot free the leg from the hook without first dealing with the fabric, and cannot strip the fabric without first dealing with the hook, because both structures are wrapped around the same point. This is the mechanical claim that makes the position non-derivative of either parent: a section of a reverse De La Riva article can describe the hook, and a section of a worm guard article can describe the tether, but neither can describe what happens when both are locked onto one leg simultaneously.
4. Two-handed lapel management, one-handed hook maintenance. Because the lapel is carrying the structural load, the bottom player's gripping hand is managing a tail rather than fighting a strength contest — the same principle that runs through every member of the lapel family. The free hand and the hook-side leg's knee shield handle posture and distance, borrowed directly from the reverse De La Riva's on-the-side, never-flat requirement (see Common Mistakes below).
Family A · The armbar. Independent coverage of the position's Technique Thursday demonstrations describes a "double attack from reverse de la worm guard" built around an armbar finish — the tether and hook combining to strip the opponent's base while the upper body is exposed. This is a direct submission route the plain worm guard's control-and-sweep tree does not typically offer at the same speed, because the reverse-DLR frame already occupies the leg the opponent would otherwise use to base out of the arm attack.
Family B · The back-take. Secondary instructional coverage frames a specific back-take sequence off the position, describing it as combining "the worm guard and a clever inversion" to manipulate the opponent's posture into a back exposure. As with the reverse De La Riva's own reverse-berimbolo family, the inversion is only safe once the hook and the lapel are both locked — inverting into an unlocked position sends the bottom player spinning under a set opponent rather than behind one.
Family C · The sweep, inherited from both parents. Because the position sits directly on top of the reverse De La Riva's sweep-and-X-guard tree and the worm guard's off-balance-and-elevate tree, a bottom player who cannot finish the armbar or complete the back-take retains the sweep options of both parent systems — the position does not have to produce a single named "reverse de la worm sweep" to remain dangerous, because it inherits two guards' worth of off-balancing leverage on one immobilized leg.
Every other reverse de la worm attack documented in the secondary material — variations on the double attack, entries into the ringworm's slightly different lapel route, escapes back to plain reverse De La Riva when the tether is stripped — is a member of one of these three families, not a fourth.
By 2026, reverse de la worm sits where a successful second-generation invention is supposed to sit: no longer a curiosity buried in an encyclopedia's back pages, but a position with its own recognized counters, taught on both sides of the exchange.
The clearest marker of that maturity is the existence of the Lapel Slayer passing material aimed specifically at it. A guard that passers can safely ignore does not get a bespoke section with a named counter attributed to a specific competitor's passing style. The foot-pop escape and the Lepri-attributed backwards-roll counter exist because enough opponents were running into the position with no rehearsed answer that Cornelius's own passing curriculum had to supply one — the same arc the broader lapel guard family went through as a whole (see the lapel guard system profile for how that counter-literature matured across the entire family, not just this member).
Within the family itself, reverse de la worm now occupies a specific tactical niche rather than a general one. Where the base worm guard is the entry point taught first, and the squid guard is the answer to an opponent who clears the tethered leg early, reverse de la worm is the version competitors reach for when they already favor a reverse-De-La-Riva-heavy bottom game and want the fabric tether layered on top of a frame they already trust. Cornelius's own stated preference for it — "the one he currently prefers to attack from" — reads less like a marketing line and more like the honest report of a competitor who tested every route in the family and settled on the one that gave him two locks instead of one.
As with the rest of the family, the position is gi-exclusive. The lapel half of the structure has no no-gi analog, so outside the jacket the position collapses to a plain reverse De La Riva — still a legitimate guard, but missing the compounding control that gives reverse de la worm its name and its case for existing as a separate entry.
Reverse de la worm's documented record is instructional rather than competitive. Unlike the worm guard's dated 2014 Pan American debut against Murilo Santana, no publicly verifiable competition match currently traces a specific result to this specific variant of the family by name — the secondary coverage available discusses it exclusively in flowchart, chapter, and Technique-Thursday-demonstration terms, not scoresheet terms. Rather than assert a competition instance that cannot be sourced, this section is held to the position's two verifiable instructional milestones.
White belt should not attempt this position at all. It requires fluency in two guards — worm and reverse De La Riva — that are themselves not beginner material, and a white belt reaching for it will produce a tangled uniform and an open passing lane, not a guard.
Blue belt is still building the two prerequisite skills separately: the reverse De La Riva hook and knee shield on one hand, basic lapel-freeing mechanics on the other. Attempting to combine them before either is independently solid under resistance is the most common way this position is attempted and abandoned at this rank.
Purple belt is where the fusion becomes realistic. A purple belt who already carries a working reverse De La Riva can start layering the lapel tether onto it, learning the sequencing the position demands: hook first, so the leg is already occupied, then the lapel, so the fabric reinforces rather than replaces the frame.
Brown belt can hold both controls simultaneously against a resisting opponent and start working the family's exits — back to plain reverse De La Riva when the tether is stripped, across to the ringworm's alternate lapel route when the standard weave is defended, out to squid guard entirely if the tethered leg is cleared before either control locks.
Black belt treats it the way Cornelius himself describes it: not a novelty variation but the preferred attacking route within a family of options, selected because two simultaneous controls on one leg out-perform either control alone, and deployed specifically against opponents whose base reverse-De-La-Riva defense has no answer prepared for a lapel arriving on top of it.
1. Attempting the fusion before either half is solid. The single most common failure. A player without an independently reliable reverse De La Riva hook, or without clean worm-guard lapel-freeing mechanics, cannot combine the two under resistance — the attempt produces a half-set hook and a half-threaded lapel, both weak. Fix: drill the reverse De La Riva hook and the worm guard tether separately to competence before layering them.
2. Sequencing the lapel before the hook. Threading the fabric first, before the reverse-DLR frame is established, gives the opponent a free leg to clear before either control is meaningful. Fix: hook first — the reverse De La Riva frame occupies the leg — then route the lapel under the leg that is already controlled.
3. Confusing the lapel route with the ringworm's. Because the two guards differ by a single weave of fabric relative to the opponent's shin, a player who is imprecise about the route ends up in neither guard cleanly, with a control that neither flowchart accounts for. Fix: know specifically which side of the shin the lapel passes on for this guard versus its near-twin, and check it deliberately rather than by feel until it is automatic.
4. Gripping the tail like a collar. As with every member of the lapel family, the entire mechanical advantage disappears if the hand squeezes the fabric in a death grip instead of letting the tether carry the load. Fix: the hand manages the tail; the loop, anchored to the opponent's own torso, does the holding.
5. No answer to the known counters. A player who drills only the entry and never the reply to a foot-pop or a Lepri-style backwards roll is unprepared the moment a passer who has studied Lapel Slayer meets the position. Fix: drill the position's own documented counters from the top as well as the bottom — knowing how it is meant to be passed sharpens how it should be held.
At open mat this week, do not attempt the fusion cold. Split the session in two. First isolate the reverse De La Riva hook and knee shield alone against a same-rank partner, exactly as you would for the parent guard, until it holds under moderate resistance. Then, in a second round, add the lapel: once the hook is set and stable, free the lapel and route it under the same leg, reclaiming the tail, while your partner tries only to strip the fabric — no passing, no sweeps. The goal is a specific sensation: the hook managing the knee line while the tether, not your gripping hand, carries the leg's control.
Once both controls hold independently, run a third round with your partner allowed to defend at half speed, including the foot-pop escape. Practicing against the position's own documented counter is the fastest way to feel which sequencing errors — lapel before hook, a death grip on the tail, an imprecise weave — actually cost you the guard. Repeat weekly until the fusion stops feeling like two guards stacked awkwardly and starts feeling like one.
Reverse de la worm sits at the exact junction of two families this map already covers separately.
The map is the point: this position exists only because two separate branches of it — a hook lineage and a lapel lineage — were deliberately grafted onto the same limb.
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:
parent is set to de-la-riva, matching the convention already used on the sibling worm-guard and lapel-guard-system articles, because a lapel-guard family hub does not yet exist as a bjj.media page. If one is created, or if reverse-de-la-riva is judged the more precise parent given this guard's dependence on that frame, repoint at ingest.