Skip to content

judo

Tighten the loop in judo

Hey what's up? So I have a new blog post today and once again it's about judo. My reflection today is from a podcast I heard recently about the employement of the OODA loop in the armed forces.

OODA loop quickly

The OODA loop (observe, orient, decide, act loop) is a decision-making model developed by United States Air Force Colonel John Boyd in the early 1970s. [...] The approach explains how agility can overcome raw power in dealing with human opponents.

So the OODA loop is a decision-making loop, and I think the link between raw power and agility to defeat an opponent applies interestingly to judo... in a different way than one could imagine

Judo isn't about decisions, it's about reaction

Judo is fast, like really fast and especially in competition (see shiai for more details). Thinking about the technique and executing it is very rarely how it goes: you see an opening or an opportunity and you react by reflex with a technique or situation from training. A lot of it is making your grip and your position as comfortable as possible for your best technique to be effective. This cannot work if you have to think during the fight.

The average reaction time of a human is 0.25 seconds when dealing with something visual, so if you need to see or feel the technique, think about it's application and enact it, you are looking at close to half a second to execute a throw. In half a second, your opponent as already recovered their balance.

Impose a decision loop on your opponent

I surmise a theory, that if you are able to impose a decision loop like the OODA on your opponent, you force them to stop "reacting and executing" and more thinking and acting. It costs them time, energy and increases their stress as well as creates opportunity for you.

When your opponent has to react to your ko-uchi/ouchi and then to take a step back, you are already making them extend their time and create a decision loop:

  • "Do I defend? what is he going to do next?"
  • "How can I attack? I feel his ko-uchi is particularly vicious, maybe I should pay attention to that"

Generating charos

The core of the OODA loop in armed forces is to increase one's loop to maximize actions in real battle scenarios, while disrupting your opponent's. Disrupting your opponent's decision can rely on several factors:

  • Disabled detection (think taking out radar stations, creating smoke and such, so Observe)
  • Broken communications (Orient and Decide)
  • Stop fuel lines (Act)

I believe that by pushing your opponent with precise attacks and a good kumi-kata/disruptive techniques, you are effectively introducing chaos in a similar way with your opponent. I can say that I have experienced such situations even with lower belts than me, attacking aggresively and decisively against me, even though they were technically worse.

This is something to try, but this quick reflection was interesting for my part.

Let's go for 2026!

What's new?

Alrighty, let's go for 2026, honestly I have had quite some fun working again on this site, and I'm looking forward to 2026.

Judo competition picture

This is me above before my second match (and last) of my competition in December 2025. I need to compete more and encourage those interested to try it as well. It's loads of fun.

Right now I took some time to think about building a page and a space for judo, mostly focused on the Czech Republic side of it. Let's see how it goes. The objective is to consolidate what there was on Notion and what we worked on with Kuba for my blue belt's exam in November 2025.

Technology?

The blog

The website took some time to update, but I think I start to be satisfied with how it looks. It's based on Material for Mkdocs, so a tad simpler to use than the previous one I had a couple of years back with Netlify and Gatsby. The biggest improvement is also the GitHub Actions that are very fast to deploy.

Another thing to do to improve the website, beyond content, is to integrate some form of comment system. I'm going to likely try Giscus. But I need some time to set it up.

Other stuff

I haven't been idle on the other fronts but haven't posted for quite some time... So right now I've learnt a few things in Airflow, like dags and so on; as well as, connecting to LLMs through APIs, which is useful.

2 years ago already I did an internal bootcamp on LLMs where I scored th highest for it in terms of points. Eh, happy about it, but I'm certainly no expert.

I have some ideas to write, perhaps about distributed computing and setting up teams, but quite honestly right now I'll focus on stuff I like. For the moment: judo!