Evolving Jiu-Jitsu

Sometimes I get distracted when rolling.  I know that I’m supposed to be controlling people and submitting them, but there’s often too many interesting things happening and it’s easy to get sidetracked.  I’ve always known at a base level that I need to be working towards control and submission rather than exploring off on a tangent, but it wasn’t until today that I realized I already have a more formalized way of thinking about how to divide my time on the mat between known good things and the occasional sidetracking that is fun, but not necessarily beneficial.

Today, I rolled with maybe eight or nine people with whom I have rarely or never rolled.  With the lower ranks, I experimented with shifts in position to see what they would do.  With higher ranks I tried techniques that I’ve been working on and doing well with against my normal opponents.  I was often surprised, especially with the lower ranks.

As I traveled home, I started thinking about scenarios that I encountered that were unusual for me.  Some of them were forced on me, while others were simply accidents of circumstance.  While there were many things I thought about, the details don’t matter much for what I’m talking about now.  As I tried to organize my thoughts, I was struck by the realization of the similarity between evolutionary biology and what happens on the mat.  Relating jiu-jitsu to evolutionary biology also helped me clarify just how much I should be experimenting and playing.

I’ve always had an interest in genetics and biology.  I even have a book on genetic algorithms which is a translation of evolutionary theory into a generic framework for solving hard, but optimizable computer problems.  One of the basic tenets of both computer and biological genetic optimization is that a population of individuals will tend towards an optimal state by discarding individuals that aren’t suited for the environment.  The population in the context of jiu-jitsu is simply a collection of techniques.  Things that work well are kept, while things that don’t are discarded.

Like a biological population, mixing of techniques from different sources gives you a more robust set of techniques.  Everything is tested and its fitness in the environment is determined.  If you are weak, strong guy techniques are attempted, then discarded.  If you are not flexible, techniques that require flexibility are ignored.  Yet if a technique fits will with your attributes, both physical and mental, you’ll latch on and make that technique part of your technique population.

As an aside, you can also consider techniques that are used against you in the frame of a antibodies.  If it’s harmful, repeated exposure will build up your defenses and you’ll in effect be inoculated against submissions and positions that are harmful to you.  But this is a different angle from what I want to cover.

A very important aspect of genetic algorithms and biological systems is that every once in a while, something random will happen.  Without getting into fitness functions and local minima and such, suffice it to say that sometimes the population gets stuck thinking it is optimal when it’s not.  It’s like a herd of goats going up a foggy mountain and getting to a small rise that is higher than everything else around it for 50 feet, but just a little further out at 60 feet is the path that leads up the rest of the way to the top of the mountain.  If the goats can only see 50 feet out due to the fog, they could reasonably think they were at the top of the mountain, and there’s no need to do anything else.  They can stand there and be proudly obnoxious as only goats can.

One fix is in random chance.  By nudging a member of the population out beyond what they can see, one of two things will happen.  Most of the time, the goat will fail to find a good path and bad things will happen.  Like falling to a grim death.  At least the grizzlies will benefit due to the random mutation.  Once in a while, however, a few goats will find the new path towards a better environment and once again the population will end up progressing towards a new optimal state.

The goats are your techniques.  You refine what you know and get to a happy place where you can sweep and submit lots of people, but every once in a while you need something random to happen.  Something that makes you wonder what if there’s something more to that unusual thing that just happened.  Most of the time, you realize that it’s just nutty, and proper/normal technique would either work better, or disable the random scenario.  The goat falling to its death is like a black belt submitting you when you try a novel escape.  Yet once in a while, it may be that you stumbled on something valuable for your game and you find a path up the mountain (and past a black belts guard).

These random events should be infrequent.  Too much random mutation and you end up with cancer.  In jiu-jitsu you could spend all your time experimenting with neat new techniques, yet fail to optimize your fundamentals.  That is just as harmful to your growth as cancer is to any biological organism.  The random mutations should be occasional, and easily discarded.  So learning an advanced technique when you are a white belt isn’t necessarily bad, but it can be if that’s most of what you do during open mat.

It’s important to give these random unusual circumstances a chance.  The mat should be a laboratory where you refine known good thesis and occasionally experiment with techniques that are probably not going to work well, but might have a slim chance of being beneficial.

Another aspect of evolving biological systems that we can draw a lesson from is when there is a steady state and populations are stagnating.  This is like a jiu-jitsu player, or even a school that is stuck in a rut.  Sometimes the best thing for a biological population is to have the environment change drastically.  While the dinosaurs certainly didn’t approve of their demise, the changing environment made it possible for us to exist and develop jiu-jitsu.  As far as I’m concerned, that’s a good trade, even if I would like a pet velociraptor.

For day-to-day mat work, this means that putting yourself in an uncomfortable position can help you develop your game and evolve past a temporary rut.  Add stress to the system, and see what happens.  Maybe you need to work on your standing guard passes.  Maybe it’s escaping side control.  Unlike the dinosaurs, you won’t go extinct.  But that constant stress of dealing with a bad environment will acclimate you to the situation, and it will help you develop when you aren’t developing in other areas.  Once you get kicked out of your rut, your whole game may begin to progress towards a new optimized state.  Your game will adapt and evolve just like biological systems.

You can also add stress to the system by exploring outside of your normal techniques.  Take a seminar from someone with a different style.  Explore the games of people with different guard styles than you are used to.  Force yourself to be be at least competent in an area of jiu-jitsu that you don’t think you’ll use on a day to day basis.  If your environment is in a steady state, it’s easy to get comfortable.  Even random perturbations may not be enough to get you out of a comfort zone.  Change your environment and see what happens.

Biological evolution is a powerful template to apply to various problems.  It can be used in computers, it can be used in research, and it can be used in your jiu-jitsu.  Hold on to techniques that work and refine them for most of your mat time.  Refine your working theory like a scientist in a lab.  Once in a while, look for the random chance circumstance that just might lead you to a new level of technique, but realize that these random things are often negative and only good for identifying what is wrong with them so that you can discard them.  If you feel like you are stagnating, then maybe it’s time to alter your environment.

Evolve your jiu-jitsu, and watch out for goats.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.