The Pareto Principle AKA 80/20 Rule in grappling

[Aesopian]

Maybe TangoMF is too modest to post this (or he just forgot) but he put up an interesting article on the Jiu Jitsu Lab over the weekend:

The Pareto principle and progress: playing the percentages in Brazilian Jiu Jitsu | The Jiu Jitsu Laboratory

You can see my replies at the bottom of his post. It's a fun topic to debate. Here is one more comment I made to him personally:

I liked the article overall, especially since you shared my same skepticism of the 80/20 "rule" actually being a rule. It's okay as a way of shifting focus on to those things that are most important for getting a good result.

I've had an idea for an article for a long time that has never come together because my own ideas aren't solid enough to communicate. It has to do with evaluating ones understanding of BJJ's concepts and technique against principles like parsimony and reductionism.

For example, when presented with many different ways to do a technique or get at a submission, rather than being overwhelmed the many variations, you should look for an underlying theory that ties them together. A theory that encompasses and explains more situations than any other is probably the best one.

I first ran into this when I got Rigan's book The Triangle. It shows more triangles (and reverse triangles, or inverted triangles, or whatever you want to call them) than anyone could ever need. How could I remember all that? I can't. But once I saw that a triangle is simple one arm in, one arm out and catching the head, then I can explain all triangle techniques using that concept.

That applies to many trigger positions for things like armbars, darce chokes, omoplata and so on.

Don't know if there's any real meat to this line of thinking, and I'm hesitant to pin philosophy of science on to BJJ when I don't have the academic chops to know if I'm even understanding the subject well enough.

 

[padirk165]

link not working for me

 

[mixmaster staph]

I think there's a lot more to it than what's in the article, but it's still a nice starting point for people who might not see it that way.

 

[Aesopian]

Works fine for me. Try again.

 

[Uchi Mata]

The Pareto principle is a useful thought construct, but as you observed shouldn't be taken too literally. What separates low and high level guys aren't so much the moves they do as how well they do them. Judo provides an even more salient example, as most ippons come from 4-5 throws and 1-2 pins, with an occasional straight armbar or loop choke out of all the techniques in Judo. You could easily win the Olympics with only what you learn in the first 3 months of Judo, if you could do it well enough.

You mentioned triggers for moves, which I think are very important and not taught explicitly often enough. So much of the time the difference between winning and losing is just seeing opportunities as they arise and quickly taking advantage of them. I'd be curious if you have any developed thoughts on that.

 

[BruceLeeMMA]

Maybe TangoMF is too modest to post this (or he just forgot) but he put up an interesting article on the Jiu Jitsu Lab over the weekend:

The Pareto principle and progress: playing the percentages in Brazilian Jiu Jitsu | The Jiu Jitsu Laboratory
.

LOL, I thought this was some new wave system that American Kickboxing Academy was using.

 

[Shemhazai]

Interesting read. I also thought Danaher's explanation of the applicability of this concept, as well as his criteria for what constitutes a high percentage technique, was a bit difficult to grasp. Especially the part about a high percentage technique having to be doable for a practitioner of any level, well ... I'm just not sure if that holds true.

I'll take two examples of moves that I personally like, that are also signature moves of current world champions: Marcelo Garcia's North-South choke and Roberto 'Cyborg' Abreu's helicopter/tornado sweep. In my opinion, both of these moves are what I consider 'long term investment' techniques, meaning that for most people, they will take years to become proficient at, but once proficiency is attained, and the move finally 'clicks', it will have a high success rate, as evidenced by Marcelo and Cyborg hitting these moves left and right against world class opponents. In the case of the NS choke, the most difficult part - especially when performing it in the gi - seems to be developing the sensitivity needed to troubleshoot the move during the setup and feel when the choke is 'on'. In the case of Cyborg's sweep, it's a matter of fluidly transitioning to the inverted position while controlling posture (in this case especially without the gi), and also of recognizing openings for other attacks as the opponent counters, either to pursue these or to threaten with them to once again be able to hit the sweep. These are skills I certainly wouldn't expect any 6 month white belt to master, yet when faced with competition footage of guys like Marcelo and Cyborg destroying people with these techniques, I don't think anyone in his right mind would claim that learning them would be a waste of time.

 

[Normot]

I agree with Shemhazai, uchimata is a hard throw to learn, I worked on it for close to a year before I ever threw anyone who wasn't an extreme beginner. Now I hit it often; a similair story for thousands of Judoka.

 

[Jiujitsunoob]

I think the Pareto principle is really just a rough estimate of a bell curve distribution that pops up in naturally in a lot of circumstances.

The article is interesting, but I feel the Pareto principle can't really used for anything more than a thought experiment.

I mean if you take the Mendes brothers game, there is probably a select few techniques they use (lets say 20% of their moves) that show up with greater frequency (lets say a frequency of 80%).

More complicated moves (ie the berimobolo) don't necessarily fall in the 20% category for everyone.

And with game theory, the second someone puts a technique in the 'low-percentage' category and stops training it, somebody comes along and exploits the fact that nobody is familiar with the defenses.

Still a fun article to read.

 

[TangoMF]

I think the Pareto principle is really just a rough estimate of a bell curve distribution that pops up in naturally in a lot of circumstances.

It's absolutely a question of distribution, but I guess the whole idea of Pareto's work is that some important distribution's aren't Gaussian but log-normal (actually log4).

But whatever, like I said in the article - I'm not even sure if this can apply to Jiu Jitsu, but it was fun to think about. Thanks everyone for the interesting comments, keep them coming.

Oh, and thanks Aeso for posting this.

 

[seoi]

I think the Pareto principle is really just a rough I mean if you take the Mendes brothers game, there is probably a select few techniques they use (lets say 20% of their moves) that show up with greater frequency (lets say a frequency of 80%).

More complicated moves (ie the berimobolo) don't necessarily fall in the 20% category for everyone.

Still a fun article to read.

I think the mendes brothers are exception. I liked the idea and its really a nice methodology. I always heard that "Even if you know a lot, when you need, you use the basic, that you learned while you were bluebelt." Today as blackbelt my strong positions are my judo (that i learned as kid, and the techniques that I used to used while blue/purble belt".

The problem with Pareto is that BJJ is not static. In the first world championships there were a lot of submission by clock choke and today they are almost extinct.

 

[buhriyon]

You may be able to apply the Pareto principle in regards to individual techniques, but BJJ is way to complex to make the rule useful. I believe that the principles such as hip movement, bridging to make space, shrimping and all the other things that can usually only be learned by experience and for some it just clicks all of a sudden.

 

[SonOfZion**]

makes sense, now i know why i get smashed 80% of the time.

 

[Rod1]

I think the mendes brothers are exception. I liked the idea and its really a nice methodology. I always heard that "Even if you know a lot, when you need, you use the basic, that you learned while you were bluebelt." Today as blackbelt my strong positions are my judo (that i learned as kid, and the techniques that I used to used while blue/purble belt".

The problem with Pareto is that BJJ is not static. In the first world championships there were a lot of submission by clock choke and today they are almost extinct.

I think the lack of clock chokes comes from people doing judo less and less and they now know how to efficienctly ride the point system to victory instead of having more "traditional" matches.

 

[Shemhazai]

I always heard that "Even if you know a lot, when you need, you use the basic, that you learned while you were bluebelt."

I hear this a lot, but the problem with this idea is that it makes it impossible to distinguish "traditional" from "basic". I've heard BJJ traditionalists call the arm drag an "advanced" move, yet within a broader grappling perspective, it's perhaps one of the most fundamental moves there is.

 

[seoi]

I hear this a lot, but the problem with this idea is that it makes it impossible to distinguish "traditional" from "basic". I've heard BJJ traditionalists call the arm drag an "advanced" move, yet within a broader grappling perspective, it's perhaps one of the most fundamental moves there is.

traditional bjj guys don't know how to do half guard. Today with Internet you can learn everything, a lot of guys now are able to do arm drag in the white belt division, but it was common to see it just on brown and black belt fights. So I think its advanced, because it involves a lot of timing, but perfectly easy to do by a well coordinated white belt. Maybe in America is even easier to find white belt doing because of the wrestling background that some kids have (pure guessing).

 

[Aesopian]

I hear this a lot, but the problem with this idea is that it makes it impossible to distinguish "traditional" from "basic". I've heard BJJ traditionalists call the arm drag an "advanced" move, yet within a broader grappling perspective, it's perhaps one of the most fundamental moves there is.

This and your other post make great points.

Everyone "knows" what the basics are, but no one can completely agree on them. We all get that some moves are more important than others, and some are better to teach to beginners than others. Maybe they are what a founder like Helio laid out as his curriculum, or maybe they aren't.

Someone once posted about how he teaches the butterfly guard to beginners for their first trial class. This was controversial, but his argument was that it was less awkward than making them close guard on strangers, and it taught them to make space and safely stand up, which he felt this was an important lesson especially if they never came to another class.

Evaluating techniques based on intelligent criteria is good. That's what Tango's idea of using the Pareto principle helps you do, even if the 80/20 ratio is made up. I like Danaher's rules too, though they have their limits (like the north-south choke example above.)

The traditional standard of teaching moves that work for a weaker, smaller person against a bigger, stronger opponent is one I agree is valuable. But what are its limits? How much smaller and weaker or bigger and stronger are we talking about? Everything breaks down at some point.

Roy Harris once wrote an interesting article about how he doesn't believe it's possible to train without attributes, though he includes sensitivity and timing as attributes (which some debate.) Michael Jen, a peer of Harris', makes similar points about how no one truly trains with zero strength and flexibility unless they are a corpse. The goal is, of course, to limit and reduce the amount of strength, flexibility and other physical factors as much as possible.

One definition for "the basics" that people aren't always aware they are using is "those moves I learned first." I actually have no problem with this approach as long it's successful. It works if they were good moves and your instructor knew what he was doing. But it doesn't answer why you learned them first. Did your instructor have a thoughtful reason, or was it just what he learned first too?

I have my personal definition for what "the basics" are, and I'm sure it will change and deepen with time and experience, as you'd expect with any complex topic like this.

 

[BeRGLeZ]

I hear what you're saying ...
Remembering techniques individually was not always completely viable for me initially, as I tended to merely try to 'rehash' the details together as I was taught.
The more experience I attained, the more I realized it has to be so much more than that - being able to conceptualize the technique and understand its purpose/role and how to implement it are key.
One big example for me was attacking the back - I went from having quite rubbish back control to significantly better back control in a short space of time - when I had a light bulb moment. Once I could facilitate a better understanding of the key points of control, the submissions and recovery/positional dominance naturally followed.

I've always enjoyed (and recommend this approach) playing with a sub and trying to utilize it from as many scenarios and angles as possible. Remembering 500 individual variations of the sub is far too difficult - as you said, you need an adhesive to bind them all together and often it's a concept that drives the minor modifications here and there, as essentially it's the same move, just differing in application depending on the context.

 

[DanOfEarth]

Roger and Marcelo are the two best BJJers ever IMO. They tend to agree with the principle, therefore, I tend to agree with it.

 

[mixmaster staph]

I actually think that the importance of the "effectiveness" of techniques is seriously exaggerated. What I think is far more important is the way you divide your focus amongst those techniques (by that I mean learning to submit from mount before you can pass wouldn't make a great deal of sense) and the way those techniques synergise.