Tuesday, May 22, 2012

BJJ Statistics Decoded

I previously posted a link to a great article on BJJ statistics. It was created by bishopbjj.com in a series of articles then compiled into one resource by bjjeasteurope.com.

I found this to be a very helpful and useful resource. But in looking at the numbers, it's not the whole story. Meaning how much of it is based on trends and how much of it is based on effectiveness? Are moves being used because they are more effective, or because they are more popular? Moves are seasonal in BJJ and there are trends in BJJ from fashion, terminology, and moves. Every year a new gi line is the "it" gi. Last year everyone did the "hang loose" sign. This year everyone screams "oss."

BJJ is more about lifestyle than any other martial art and so it tends to move more in terms of what is more fashionable.

Meaning BJJ is not purely based on logic, there is a lot of psychology and the mind of a BJJ player though they try to always self correct and find efficiency, also enjoys the esoteric side of this sport/art.

There is also a lot of trying to beat your opponent with something they are not familiar with. As things get popular there is a short window of effectiveness before people become wise to it. Like 50/50 in years past. Also looking at trends can help you prepare yourself in what to expect from your opponent in that tournament year which is something Lloyd Irvin is known for.

Let's get to the stats:

So the obvious one is, submissions occur most often when the player is already up on points. The point system is based around dominant moves, and as you get to more dominant and dominant positions, the submission becomes easier. As well if a player can earn more points than the other player, they are probably better to begin with so the submission comes as no surprise.

The top positions for submission being back, mount, and cross side. Makes sense, as those are the only original positions of BJJ.

Now the top submissions gets interesting. The top submission is the armbar. Which makes a lot of sense because it is a very versatile move that can happen from many different positions, situations, with many variations. The close second is the choke from back. Which is also obvious because the majority of all the submissions occurred from the back.

Then you have cross collar, triangle, footlocks, and neck chokes rounding out the bottom. This is where it gets interesting, does that mean those moves are less effective? Or does it just mean those moves can only occur from specific situations. Like triangle will typically happen from guard. So does that mean it's less effective than an armbar? Or does it just mean there are less opportunities to use it? We have no idea what the ratio of attempts to finishes were for such a move. Like cross collar choke. It is one of the least occurring submissions, but when it is attempted from mount, what is the finish ratio? So you can have a move that could be less frequent but more effective. Meaning it's not easy for player A to get to mount, but when he does he finishes 100% of the time with the cross collar.

Moving on to sweeps. Most sweeps occurred from de la riva or half guard. Very few occurred from butterfly. Does that mean butterfly sweeps don't work or does it mean no one plays butterfly anymore? Going to tournaments, how often do you see anyone using butterfly? Maybe just Marcello Garcia nowadays.  But it was used quite often 10 or more years ago. Now people may have stopped using butterfly as much because it's less effective. Or because the opponent never allows a butterfly situation to occur because their teacher used to get swept against it so much when he/she competed. What is telling though is that out of all the guards, de la riva is the only guard where you can complete a sweep against an opponent where they are not only standing but they are postured.

Butterfly would only work when your opponent is on his knees and posture broken. This then may be more of an issue of BJJ players no longer passing from the knees and de la riva is a reaction to this.

Which then leads us to most passes occurring from half guard. If you attempt to pass standing, there is a good chance of being swept with de la riva. So then it would make a lot of sense to move the fight into the half guard and initiate your pass from there. So with the data like I mentioned, kneeling passes are the least popular form of passing.

When you look at guards used to submit, de la riva is not the top guard. It is used to sweep opponents who are standing. Spider guard is the most often used for submission. Which also makes sense, as de la riva is the easiest way to sweep a standing opponent, spider guard is the easiest way to break a standing opponents posture down and submit them (both relying heavily on the sleeves).

Closed guard you either need to jump into closed guard or eventually get your opponent to kneel, or they kneel to pass. We have no idea though what the finish ratio of submissions once an opponent is in the closed guard. I would speculate the ratio would be the highest, because once in the closed guard, it is out of all the guards designed specifically for submissions, keeping posture broken, and keeping your opponent kneeling. Players like Roger Gracie or Kron Gracie are not so great in their ability to sweep or finish from every position. They are great at once getting an opponent into the most effective position, to have a high closing ratio. Meaning like a true salesman, when they do get in front of the right person, they can close the sale every time.

Now whoever is up on points usually wins or scores a submission. So this leads into why most fights occur with one player pulling guard or both in a lot of instances. Pulling guard does not give you any disadvantages, and gives you a chance to score the first point by sweeping. Probably from de la riva or half guard. Throws or takedowns will occur if both players are willing to engage in throws. But with always someone trying to pull guard, it becomes a difficult task. So does it pay off? More than half the time they will win the match.

It's not who is better, it's who is better in the time allotted. And with a certain amount of time burned after the first points being scored, it becomes a race, with the person who scored first usually winning out. The other person has to first even it out, then surpass their opponents score in a limited amount of time. So 82% of the time whoever scores first wins.

This is of course all according to the 2012 Pan Ams for black belts. It will be interesting to see what the purples were using, as those may be the moves being used most often in a few years when they become black belts. Like berimbolo for instance.


  1. Great post! HBJJ for life.

  2. http://www.innerbjj.com/2012/02/can-bjj-become-cult.html



Affiliate links are used and I may receive a commission if you click.

Inner BJJ is a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for sites to earn advertising fees by advertising and linking to amazon.com.