Ideally, every roll would start with both opponents standing. Potential collisions with other people rolling nearby make this impractical on a crowded mat. Additionally, takedowns often result in wear and tear on the body. Most gym rolls have an implicit agreement that both opponents will avoid standing at the same time.

The naïve compromise is to start rolls with both opponents on their knees. The assumption is that this allows both participants to fight for top position while avoiding the potential injuries associated with stand up grappling.

The problem is that both-opponents-on-their-knees is a contrived situation that doesn’t occur in competition. It doesn’t even occur during a friendly gym roll unless you purposely start there. One opponent simply needs to stand to gain an immediate positional advantage over their kneeling opponent.

Even worse is that the both-opponents-on-their-knees situation keeps the bad aspects of jiu-jitsu standup (both opponents neutralising each other in a tight clinch) and loses the good ones (mobility to gain favourable angles).

A better way is to start in a guard position, the most common being closed or seated. It is a more effective use of training time to start in the guard, than to struggle on the knees for a minute or so before ending up in the guard anyway. In competition, most time is spent in the guard so your training time should reflect that.

An alternative approach (especially for the more experienced practitioner) is to start in an inferior position. This allows you to practise your escapes directly, rather than allowing your opponent to pass and establish dominant position before beginning an escape. Unless you are specifically working on the timing of your escapes, allowing your opponent to establish dominant position so you can practise your escapes just builds bad habits.

When we learn a new move in jiu-jitsu, we want to learn it correctly. We want to perform it the right way.

A move is a collection of concepts and tradeoffs. There is no right way to always perform a move in the general sense. The specific way to perform a move will depend on the environment it is performed in.

We’ll use the knee-slide pass as as example. The knee-slide (alternatively knee-slice, knee-cut or esgrima) pass is a way of passing the half guard. It is conceptually simple – start in half guard, pin your opponent’s upper body and then pull your leg free to end up in side control.

Below are multiple demonstrations of this pass. Note the way that different instructors emphasise different aspects of the pass. Some are so different that it is easy to consider them different passes.

Here is Saulo Ribeiro teaching the pass. Saulo emphasises heavy pressure and preventing your opponent from regaining the closed guard.

Now consider Rafael Mendes. He is much lighter than Saulo. It is easier to sweep a lighter opponent than a heavier one. Rafael emphasises keeping his weight off his opponent to avoid the sweep.

Steve Campbell takes a more conceptual approach. He notes that by twisting your opponent’s spine, it will be easier to extract your trapped leg. Steve’s gi focused approach twists the spine by pinning the hips first and turning the shoulders second.

Marcelo Garcia also shows the spine twisting approach to this pass. Marcelo pins the shoulders first and then turns the hips. This approach works both gi and no-gi while Steve’s hips-first approach favours the gi.

There is no right way to perform a move. If the move works, then you did it the right way. The environment dictates what is correct. You may trade pressure for mobilitity or vice-versa. Gi grips allow you to sacrifice positional stability for positional advancement. Consolidating your gi and no-gi game trades possibilities for reduced cognitive load.

Being good at jiu-jitsu means being able to make the correct tradoffs to be successful against your opponent.

Gi or no-gi is a polarising question in BJJ circles. Most people have a preference for one or the other. Calling it a preference is a little understated. Religious fervour would be more accurate.

The main difference between gi and no-gi is that you can grip the clothing. Essentially, gripping the gi acts as a strength multiplier. In BJJ, there is a strange hangup about strength. “You’re strong”, is a backhanded compliment in BJJ rather than a genuine compliment as in other sports. A core tenet held by most BJJers is that strength doesn’t matter, technique beats strength.

Now we can see why the gi or no-gi question raises such a passionate response. It’s touching people’s core beliefs about BJJ. People will tend to favour whichever style gives them the most success in their rolling as it reinforces their belief in BJJ being about technique rather than strength.

A grappling contest is not a contest of the athletes’ purely technical ability. It’s a contest of their technical ability combined with their physical attributes (strength, flexibility, explosiveness, etc).

Let’s see how the gi (strength multiplier) affects this contest.

If you are grappling against someone of similar skill, the contest comes down to attributes. Adding the gi will favour the stronger competitor. Conversely, removing the gi will favour the weaker competitor.

If you are grappling against someone of different skill, adding the gi will favour the more skilled competitor. Removing the gi will favour the less skilled competitor.


The gi provides a benefit against:

  • stronger but less technically skilled opponents
  • weaker but similarly skilled opponents

No-gi provides a benefit against:

  • weaker and more technically skilled opponents
  • stronger but similarly skilled opponents

Your preference for gi or no-gi is likely to be affected by the opponents that you regularly compete against. Here’s a cheatsheet to determine whether gi or no-gi will give you the advantage against an opponent. Consider your regular training partners and your preference for gi or no-gi. Notice a correlation?

Grappling Cheatsheet

As a coach, my job is to provide a training environment where people can learn BJJ safely and effectively. But there’s more to it than that. My job is to ensure that they learn how to keep themselves safe while training and competing.

These two things are not the same.

The most important thing to understand in BJJ is that your safety is your responsibility. Make sure you can keep yourself safe. Don’t rely on rules or referees to keep you safe. The following video shows that you can’t rely on rules and referees.

When you train in the gym, you are friends with the people you are training with and you know that they will look out for your safety. In a competition, or when training at another gym, you can’t afford to rely on your opponent to look out for you.

This means that you must train with illegal moves in the gym. When you train heel hooks, slicers, knee reaping, spinal locks etc with training partners you trust, you begin to understand those moves. When you understand the moves, you know when to tap and when to continue working your escape. More importantly, understanding the moves allows you to see the move coming and allows you to avoid getting caught right from the start. Just because a move is illegal doesn’t mean that someone won’t try to apply it on you.

Basic risk assessment. The worst injury that could happen to you is a spinal injury. Your game should be designed around this risk. Everytime you are in a position where your opponent can pick you up is a potential spinal injury. This means jumping guard, armbars from underneath, spider guard, closed guard. If you play in these positions you should expect to be picked up and you must know how to bail safely. A more prudent approach is to just avoid these positions entirely.

BJJ is a combat sport. Most of the time it’s a friendly sport. But in competition where emotions are high, it can turn into a fight. Your safety is your responsibility.

The majority of BJJ competitions follow the IBJJF ruleset or a variation of it. The objective in a BJJ match is for two competitors to battle to be the first to submit the other. This simple objective sounds like fun for the competitors and promises excitement for the spectators. So why aren’t IBJJF matches like this? Why are they boring to watch, confusing to score and not a lot of fun to compete in?

If you’ve never competed under any other ruleset, competing under IBJJF rules is fun. Junk food is also delicious when you’re hungry. Our expectations are higher than that. We expect our skills to improve as we train. We expect our coaches and training partners to improve as well. Likewise, we should expect our competition rules to improve.

The IBJJF is aware that the rules have problems, which is why the rules are tweaked frequently but to no avail. The tweaks just add complexity and further obscure the underlying problem of the ruleset.

The problem with the IBJJF rules is with its fundamental assumption. When the rules were being formed, BJJ matches were won by the traditional strategy of: takedown/sweep → pass guard → knee on belly → mount/back → submit. Not all matches end in submission, so a natural solution is to award points to competitors based on their progression in this strategy. We see this with the points allocation.

  • 2 points takedown
  • 2 points sweep
  • 2 points guard pass
  • 3 points knee on belly
  • 4 points mount or back

(The points for guard passing and knee on belly were later switched in one of the first rule tweaks).

This point allocation (and the mindset behind it) is the root cause of the failure of the IBJJF rules. The rules were initially intended to measure a competitor’s progression to the submission, but instead reward a competitor for following a particular strategy. Deviation from this strategy is penalised. This is a subtle point. Let’s look at examples of how the rules affect, rather than measure a match.

Movements progressing faster towards the submission aren’t rewarded as much as slower ones. Takedown straight to side control 2 points vs takedown to guard then guard pass 5 points. Sweep to side control 2 points vs sweep to guard then guard pass 5 points. Reversal from side control 0 points vs reguard and sweep 2 points.

The failure of the IBJJF rules is that it rewards and penalises strategies instead of the submission. This mindset allows the sport to become politicised. Consider the political effects of the current IBJJF ruleset:

  • strategies involving leg attacks are restricted to make them almost useless
  • particular guards are encouraged by disallowing slams from guard
  • non-traditional guards are discouraged – specific rules for 50/50 position, anti leg reap rules

Let’s consider other grappling rules and how they guide the strategies used by competitors.

ADCC. No points in the first half of the match. This allows competitors freedom in the strategies they choose. The points are based on IBJJF with some tweaks. “Clean” sweeps and takedowns that bypass the guard are recognised. Pulling guard is penalised. This is an improvement upon the IBJJF rules, but there is still restriction on successful strategies once the points come into play.

FILA Grappling. Heavy wrestling influence in that points are scored for positional control and escapes. Guard is not emphasised with any special rules for sweeps or guard passing. Obtaining positional dominance is a clear winning strategy as a 10 point lead wins the match (technical superiority). These rules encourage grappling, movement and control rather than the submission.

Eddie Bravo Invitational. A submission only ruleset. Matches that don’t end in submission are decided in overtime. Overtime is alternate rounds of starting in dominant position and attempting to submit/escape. Fastest submission/escape wins. Submission-only rules don’t favour particular strategies. The overtime rules favour a strategy of strong positional control, but this is negated by offering prize money for submissions during the round.

Recall that the objective of a BJJ match is to get the submission. Rules can affect the outcome of a match as competitor’s play to the nuances of the ruleset. Good rulesets only measure. Bad rulesets interfere.

The top player’s power hand is his most important tool in the half guard top position. Even if he is not attacking with it directly, he is using it for balance and defence. Removing the use of his power hand leaves him at a serious disadvantage. We examined fighting the power hand in week 6. We continue this topic here.

The frame that we established last week will be our staging platform for attacking. The top player has limited options for what he does with his power hand. He can:

  1. Grab the bottom player’s nearside hand, head or collar
  2. Grab or push the bottom player’s nearside knee
  3. Post on the ground or otherwise hide the power hand

These limited options make it much easier for the bottom player to prepare a response.

  1. The bottom player must stretch back to create distance and not leave her nearside hand within reach of the top player’s power hand. If the top player reaches for her head or collar, his arm will become extended and it is easier for her to grab his power hand.
  2. If the top player pushes the bottom player’s nearside knee, she can firmly cover his power hand with her hand. This will gain her a second or so before he can free his hand.
  3. If the top player is hiding is power hand, there is nothing the bottom player needs to do as the power hand is no longer a threat.

In all cases, the bottom player has neutralised the top player’s power hand, and is then able to sit up and gain head position.

Until now, the bottom player has been reacting defensively to the top player’s counters. Even though her goal is an offensive one (gain head position), she has been going about it with a defensive mindset. Her thoughts have been on countering the obstacles that the top player is putting in her path. The top player is the aggressor as he is deciding which obstacles to present to the bottom player.

The top player has the advantage because there are many possible obstacles that he can use (sitting back, pressuring the far deltoid etc). The more possibilities he has, the more options the bottom player must mentally process and the slower her reaction time will become. To regain an advantage and to switch to a more aggressive mindset, the bottom player must reduce the options available to the top player.

The bottom player can create a frame (structure) with her limbs that restricts the options available to the top player. The structure we are using is similar to Robson Moura’s 93 guard but with the inside knee blocking the top player’s far hip instead of her far hand grabbing the pants (no-gi compatible). The key points of this structure are:

  • Near-side knee/shin blocks the top player’s far hip as we covered in week 4.
  • Far-side knee/shin block the top player’s chest or near-side shoulder/biceps.
  • Hands are holding the top player’s upper body and pulling him towards her shins (pull-push like in spider guard).

The knee/shins create a shield that prevents the top player from being able to effectively crush. It may be difficult to transition from a half guard to this structure, so the bottom player may need to use stiff arm (locked elbow) frames to hold her opponent away as she positions her knees.

Faced with this structure, the top player’s immediate concern is not in preventing the bottom player from gaining head position. His attention must first be on dismantling her frames. There are only a small number of ways that he can do this. If the bottom player is aware of the likely behaviour of the top player, she has the advantage.

A strong frame gives the bottom player an advantage is it gives her respite from being crushed, time to think, and a staging platform to launch prepared attacks.

This week was used to review what we’ve covered so far and to ensure we understand what the key points are.

Key points:

  • The main objective for the bottom player is to gain dominant head position.
  • Control the centre of mass after obtaining head position.
  • The main objective of the top player is to prevent the bottom player from gaining head position.

The bottom player must counter the top player’s defenses and gain head position. The main defenses of the top player that we have looked at so far are:

  • using hand/arms to prevent the bottom player closing the gap to gain head position
  • smashing pressure on the far deltoid
  • sitting back on his heels to increase the gap the bottom player must cover.

It is important to understand that the top player will use these defenses in combination and will switch between them quickly. The bottom player must be able to quickly recognise how the top player is defending to be able to counter the defence and ultimately gain dominant head position.

In the fight for head position, the obvious thing for the top player to do is to use his hands and arms to push the bottom player and prevent her from closing the distance to gain head position (head to chest). Regardless of which hand is dominant, his outside (crossfacing) arm will be able to apply more strength in preventing the bottom player from sitting up as it is able to apply force higher on her body and hence exert more leverage. This outside hand is his power hand, and it is the hand that the bottom player will need to defeat before being able to gain head position.

The top player’s power arm is at it’s strongest when his elbow is close to his ribs. It is weaker when it is extended, or when it is drawn back so his hand is close to his chest.

The main problem for the bottom player is that as she starts to bring her head closer to the desired position, she is moving her head towards the top player’s power hand in its strongest position. She is effectively running into a wall.

It is advantageous for the bottom player when the power hand is either fully extended or fully drawn back. A good tactic for her is to have the top player extend his power hand. She can then restrain it in some fashion before moving her head into good positioning. When the bottom player has control of the top player’s power hand, he can no longer use it to defend against her gaining head position.

One way to entice the top player to extend his power hand:

  1. Move your head far away from the top player. Use z-guard (knee shield – outside shin in top player’s hip) to provide a foundation then arch your back to move your body away.
  2. With your body extended, the knee shield loses strength. Elevate the knee to block against the top player’s chest/shoulder instead of his hip. This provides a frame to help prevent him crushing you with his bodyweight.
  3. The top player now needs to use his power hand to remove this frame. Wait for him to bring his hand within range, then grab and extend it.
  4. Extend your top leg. This removes the frame and creates an easy path to sit up and gain head position.

Until now, our focus has been on what the bottom player is attempting to do. The top player as been providing resistance, but that resistance has been unskilled. We shall now begin dealing with a skilled opponent.

The first objective of a skilled top player who finds himself in half guard is to gain good posture. Specifically, he will bring his butt close to his heels. This gives him good base without needing his arms to support himself. Good base allows him to easily use his arms and move his torso to begin to attack.

This butt to heels posture is bad for the bottom player. Even if she manages to win the head control battle, it will be difficult for her to gain control of the top player’s centre of mass because the limb orientation makes this awkward.

The objective of the bottom player is for the top player to lift his butt away from his heels. We will look at a leverage based way of achieving this.

To begin, the bottom player will triangle her legs and place the outside foot on the floor. She triangles her legs so that the top player’s inside knee becomes pinned. A joint that is pinned can act as a fulcrum.

With gi jiu-jitsu, the simplest way to force the top player’s butt to lift is to grab the front of his belt and yank forwards.

No-gi needs a more technical approach. The bottom player will apply upwards (towards the sky) pressure with her inside leg. The force generated by the upwards pressure is not enough to lift the top player’s butt, but it has other purposes. Firstly it forces the top player to actively sit down harder to counteract this upwards pressure. Secondly it engages the muscles that the bottom player will use to sit up.

By actively sitting down, the top player is in-effect, trying to rotate his body backwards around his pinned knee. This is not ideal for him, as he wants to prepare for the bottom player fighting for head position. He wants the ability to quickly rotate forwards to use bodyweight to block the top player’s attempt to sit up. By rotating backwards, he has to first stop and then reverse direction before he can fully apply force in a forward direction.

By engaging her muscles, the bottom player will be able to sit up faster as her muscles are primed for use.

So the effect of the upwards pressure is to slow down the top player, and to speed up the bottom player. The bottom player will use this speed differential to sit up and hug the top player’s torso. With a tight hug, she can push off her outside foot (which is flat on the floor) and pull the top player forwards so his butt separates from his heels.

Now the battle for head position can proceed.