By the BJJ Sportswear Editorial Team | Last reviewed: May 2026 Data sourced from IBJJF World Championships, ADCC 2022–2024, and analysis of 7,500+ competitive submissions
Opinions about the “best” BJJ submission are everywhere in this sport. Walk into any gym and you will hear a different answer depending on who you ask — the purple belt who just discovered heel hooks, the black belt who has finished every match with the same bow and arrow for fifteen years, or the coach who swears nothing beats a well-timed guillotine.
The problem with opinions is that they are shaped by personal experience, which is limited. The data is not.
This guide ranks the most effective BJJ submissions using verified competition statistics from the IBJJF World Championships, ADCC, and professional no-gi events covering more than 7,500 finishes. We separate gi and no-gi because they are genuinely different sports with different answers. And we explain not just what finishes matches, but why — so you can make informed decisions about where to invest your training time.
Table of Contents
Why the Data Matters More Than Your Coach’s Opinion
Every submission you spend serious drilling time on is an investment. You have a finite number of hours on the mat, and drilling a low-percentage technique at the expense of a high-percentage one is a costly mistake that compounds over years.
Data from elite competition cuts through bias. It does not care which technique looks impressive on social media or which one your favourite athlete is known for. It tells you, across thousands of matches at the highest levels of the sport, what is actually finishing opponents who know what they are doing.
Two important caveats before we get into the rankings:
Gi and no-gi are different sports with different answers. The bow and arrow choke — one of the highest-percentage finishes in IBJJF gi competition — does not exist in no-gi. Heel hooks — dominant in ADCC — are mostly banned in IBJJF gi. The data below is separated accordingly.
Success rate and attempt rate are different things. The guillotine choke is the most attempted submission in competition but has only a 9.3% finish rate. The Ezekiel choke is attempted rarely but finishes at over 60% when it does appear. Raw frequency is not the same as effectiveness.
The Numbers: Gi Competition (IBJJF)
The following data draws from IBJJF World Championship results, primarily the 2023 black belt adult divisions — the most competitive and data-rich bracket in gi grappling.
1. Rear Naked Choke — 45% of IBJJF Worlds Finishes
This might surprise you. The rear naked choke is a no-gi staple, but even in gi competition, it dominates finish statistics because back control is the most dominant position in all of grappling — and the RNC is its primary weapon.
According to IBJJF competition data, the rear naked choke accounted for 45% of submissions at the 2023 IBJJF World Championships. When you achieve back control, your opponent has no offensive options — they can only survive. That fundamental asymmetry is why the rear naked choke consistently tops every dataset regardless of ruleset.
Training implication: If the RNC finishes 45% of matches, your back-taking system deserves more drilling time than almost anything else in your game.
2. Armbar — 21% of IBJJF Worlds Finishes
The armbar is the most versatile submission in the sport. It can be attacked from guard, mount, side control, and back control. It works across all weight classes, genders, and belt levels. At the 2023 IBJJF Worlds, it accounted for 21% of all finishes.
What makes the armbar particularly valuable from a training investment perspective is its transfer: every hour you spend refining your armbar from guard also improves your guard retention, your hip movement, and your understanding of posture control. The return on investment extends beyond the submission itself.
3. Bow and Arrow Choke — 89% Success Rate When Attempted
Raw frequency does not tell the whole story. The bow and arrow choke is attempted less often than the armbar or triangle, but when elite competitors attempt it, it finishes at an extraordinary rate.
BJJ Blog’s analysis of IBJJF Worlds 2019 data found the bow and arrow successful in 17 out of 19 attempts — an 89-90% finish rate. That is among the highest of any submission at elite level. The reason: it is only attempted from a fully established back control position with a collar grip already secured. By the time you attempt it, you have already won the positional battle. The submission is largely a formality.
Training implication: The bow and arrow is a gi-exclusive, position-dependent submission. Investing in it means first investing in your back-taking and seatbelt control game.
4. Triangle Choke — Varies Widely by Format
The triangle choke is one of the most recognizable submissions in the sport and produces fascinating data splits by format. At the IBJJF Worlds 2019, 28 out of 45 triangle attempts were successful in gi competition — a 62% success rate. In ADCC 2022 no-gi, only 7 out of 25 succeeded — 28%.
The gi differential is significant. Grip-fighting to establish a triangle is easier with a collar grip. No-gi defenders have more freedom to stack and posture. If you primarily compete in gi, the triangle deserves serious drilling time. If you compete no-gi, the investment returns less.
5. Collar Chokes (Sliding, Cross Collar, Ezekiel)
Gi-specific choking options using the lapel and collar represent a significant category of finishes. The sliding collar choke topped the 2022 IBJJF Worlds black belt male division data as the single most common finish. Collar chokes are the “101” techniques of gi jiu-jitsu — drilling them is never wasted time.
The Ezekiel choke is the statistical outlier in this category. At ADCC 2022 it appeared only twice — and both times it finished. In the 2023 Brasileiros it accounted for 10% of submission finishes. It is rare precisely because it requires a specific positional setup from inside the guard, but its finish rate when attempted is exceptional.
The Numbers: No-Gi Competition (ADCC)
No-gi data tells a different story — one heavily shaped by the Danaher Death Squad’s systematic development of leg lock systems and the resulting defensive evolution across the entire sport.
1. Inside Heel Hook — 21% of ADCC 2023 Finishes, 34.85% Success Rate
The inside heel hook is the highest-percentage submission in elite no-gi grappling. At ADCC 2023, it accounted for 21% of all finishes. Its success rate of 34.85% is significantly higher than the outside heel hook’s 14.08% — and the difference comes down to anatomy.
The inside heel hook attacks the knee’s ACL, LCL, and MCL in addition to the ankle, applying rotational force to the most vulnerable angle of the knee joint. Minor rotation can cause serious injury before opponents feel meaningful pain, which is why it finishes at a higher rate and why it carries serious responsibility in training. The heel hook guide covers the safety protocols that must accompany this technique.
Gordon Ryan, Craig Jones, and the broader Danaher-influenced no-gi game have made inside heel hooks mainstream at elite level. Defensive awareness is now universal at black belt level — meaning the “easy heel hook era” is over, but the technique remains dominant because the leg lock entry system is simply too structurally sound to abandon.
2. Rear Naked Choke — 20% of ADCC 2023 Finishes
In no-gi, the RNC climbs to near-parity with the inside heel hook. Without the gi jacket creating friction and grip options, back control is even more positionally dominant in no-gi. And once you have the seatbelt with hooks in, the RNC is the fastest path to a finish.
The clean split between RNC and inside heel hook at the top of ADCC data reflects modern no-gi strategy: elite competitors focus on two pathways — secure back control and finish with the RNC, or enter leg entanglements and finish with the inside heel hook.
3. Armbar — 15–18% of No-Gi Finishes
The armbar remains highly relevant in no-gi but drops slightly from its gi percentage. The reason: no-gi defenders can posture more freely without lapel grips restricting them. However, the armbar’s versatility means it is still the most broadly useful joint lock in the no-gi arsenal.
4. Guillotine — Most Attempted, 9.3% Finish Rate
This is the most instructive data point in the entire dataset. The guillotine is the most attempted submission in competitive grappling — and it finishes at only 9.3%.
This matters enormously for training decisions. Grapplers instinctively reach for the guillotine because it presents itself frequently during scrambles and shot defense. But elite defenders have well-developed guillotine defense. If you are spending a large portion of your submission drilling time on the guillotine without a sophisticated system behind it, the data suggests you are working on a lower-return investment.
This does not mean the guillotine is a bad technique. It means the standard arm-in guillotine that most practitioners drill is not a reliable finisher at high levels. The high-percentage version requires understanding of proper head position, the high elbow squeeze, and specific setups — not just the opportunistic grab.
What the Data Tells Us About Gi vs No-Gi
The submission data makes the practical differences between gi and no-gi concrete rather than abstract.
In gi, collar chokes and back-position attacks dominate because the lapel creates additional control and finishing options unavailable in no-gi. The bow and arrow, sliding collar choke, and cross collar choke are exclusively gi techniques with some of the highest finish rates in the sport.
In no-gi, leg locks — primarily the inside heel hook — have become structurally essential at any serious competitive level. The Danaher-influenced leg lock system has fundamentally changed modern no-gi grappling in a way that cannot be ignored.
If you compete exclusively in gi, you can deprioritize heel hook offense (though defense remains essential). If you compete in no-gi, ignoring leg locks is no longer a viable competitive strategy.
Success Rates by Belt Level: Where Data Changes
The submission landscape shifts significantly as belt level increases. This matters for how you should approach training at different stages.
White and blue belt: Armbars and basic chokes dominate. Matches frequently end from fundamental techniques because defensive gaps are large. As a white belt, focusing on armbar, RNC, and triangle from standard positions is the highest-return training investment.
Purple and brown belt: Submission diversity increases. Triangle chokes and back-takes become more refined. Leg lock awareness becomes essential even in gi-focused training. Submission chaining — linking one attack to create another — becomes the norm.
Black belt: Fewer matches end by submission overall. When finishes do happen at elite level, they are almost always from back control or highly refined, deeply drilled attacks. Black belts are not finishing with new techniques — they are finishing with techniques they have drilled thousands of times until the execution is mechanical.
The practical implication: do not try to learn 20 submissions. Learn 3 to 5 deeply, understand the positions that produce them, and drill the setups until they are automatic.
The Submissions the Data Says to Deprioritize
Not every popular technique earns its drilling time. Based on competition statistics, several submissions have finish rates too low to justify significant training investment for most practitioners.
Flying submissions (flying armbar, flying triangle): High visual appeal, very low success rates at any competitive level. Extremely high percentage of attempts fail, often ceding position.
Buggy choke: Gained significant social media attention but accounts for less than 3% of finishes in competition with a very low success rate. An interesting supplement once your core game is established — not a foundation.
Toe holds: Effective at lower belt levels where leg lock defense is underdeveloped, but success rate drops sharply at black belt and elite no-gi level.
Inverted armbar: Mechanically complex, requires specific positional setup, and finishes at a low rate compared to the standard armbar available from identical or similar positions.
These techniques are not worthless — they have value as occasional attacks and positional transitions. But if you are deciding where to spend your next 100 hours of drilling, the data clearly points elsewhere.
How to Use This Data in Your Training
The purpose of submission statistics is not to make your game mechanical or predictable. It is to help you make informed decisions about where your drilling time produces the highest returns.
Three practical applications:
1. Build your A-game around high-percentage submissions. Choose two or three from the top of the data for your format (gi or no-gi) and drill them until they are deeply ingrained. The kimura is worth noting here — it ranks in the top five for both gi and no-gi, making it one of the most format-agnostic submissions in the sport.
2. Understand the positions that produce finishes, not just the submissions. The RNC is dominant because back control is dominant. The bow and arrow is dominant because it only appears after full back control is established. The pattern is consistent: position first, submission second. Your positional game will always determine your submission options.
3. Track your own data. The best practitioners keep mental or written records of what they attempt and what finishes. Your personal dataset will not match the elite competition data exactly — but tracking it reveals your own patterns and where you are leaving percentage on the mat.
Frequently Asked Questions
What is the most effective BJJ submission overall?
In gi competition, the rear naked choke accounts for 45% of IBJJF World Championship finishes — making it the single most effective submission across the sport. In no-gi, the inside heel hook and rear naked choke are statistically near-equal, each accounting for approximately 20–21% of ADCC finishes. Across both formats combined, back control and the RNC represent the highest-percentage path to a submission finish.
Which submission has the highest success rate when attempted?
The bow and arrow choke has among the highest individual success rates at elite level — approximately 89% when attempted in gi black belt competition, based on IBJJF Worlds data. The Ezekiel choke showed a 100% success rate at ADCC 2022 (two attempts, two finishes), though the sample size is small. Both techniques share a common trait: they are only attempted from fully dominant positions, which naturally produces high finish rates.
Are leg locks worth learning for a gi competitor?
For offense — it depends on your competition format. Heel hooks are banned in IBJJF gi for all belt levels. Straight ankle locks and knee bars are permitted and appear in competition data at meaningful rates. For defense — yes, unconditionally. Leg lock awareness is essential regardless of format because understanding the mechanics of leg attacks is the only way to develop reliable defense against them.
Why does the guillotine get attempted so often if it has a low finish rate?
The guillotine presents itself constantly during scrambles, takedown defense, and guard passes — situations that happen in almost every match. Its high attempt rate reflects how frequently the position arises, not a calculated decision that it is a high-percentage finish. The lesson from the data: just because an opportunity presents itself frequently does not mean the technique finishes frequently. A systematic guillotine setup produces very different results from an opportunistic grab.
How does belt level change which submissions work?
At white and blue belt, armbars and basic chokes dominate because defensive gaps are large. At purple and brown belt, submission diversity increases and back-take systems become more refined. At black belt, overall submission rates actually decrease — matches are more likely to go to points because both competitors are technically sophisticated. When finishes do happen at elite level, they are almost always the result of deeply drilled, highly specific attacks from established positions rather than opportunistic grabs.
Should I train submissions I am not trying to finish with?
Yes — and the data supports this. Understanding a submission mechanically, even if it is not part of your primary game, improves your defense against it and often reveals setups that transition to your primary submissions. The kimura is a clear example: many practitioners use it primarily as a control and transition tool rather than a finish, and it appears in the top five of both gi and no-gi finish data as a result.
The Bottom Line
The data across more than 7,500 elite-level submissions delivers a consistent message: master back control and its primary finish (the RNC), build a reliable armbar from multiple positions, and develop a format-appropriate leg lock game if you compete no-gi.
Collar chokes in gi and inside heel hooks in no-gi represent the highest-percentage additions once your foundational game is established. Everything else — flying submissions, low-percentage exotic attacks, social media techniques — can wait until your core game is deeply drilled and reliably producing finishes under pressure.
Position creates submission. That truth has not changed in the seventy-plus years since Brazilian jiu-jitsu was formalized. The data simply confirms what the best coaches have always taught.
Data sources: IBJJF official competition results | ADCC official | BJJ Equipment statistics study (7,500+ submissions) | BJJ Blog submission analysis (7,567 submissions, 2022–2023)
Last reviewed: May 2026. Statistics updated annually as new competition data becomes available.
