Are There Opportunities to Use OODA Loops in Your Software Project? | Hacker News Reader

Are There Opportunities to Use OODA Loops in Your Software Project?

spin.atomicobject.com

9 pointsphilk1012 comments

I'm familliar with OODA but I've wondered how well the theory holds up in reality - that is, is there research that supports it?

I've found that the concept - appealingly memorable and elegent to me and apparently others - has spread elsewhere; I've heard military people use it like it's assumed, but I haven't found any research showing that people actually work that way or even that the concept is helpful (I haven't looked extensively either).

OODA loop is just a catchy acronym for the scientific method. Gather data, make hypotheses, choose one to test, actually test it; repeat.

And do it faster than your enemy, in the context of the military setting in which it was developed.

Also recommend the book Science, Strategy and War which is a PhD thesis to answer the GP’s question on research.

Thanks for the recommendation.

https://www.taylorfrancis.com/books/mono/10.4324/97802030888...

John Boyd is often known exclusively for the so-called ‘OODA’ loop model he developed. This model refers to a decision-making process and to the idea that military victory goes to the side that can complete the cycle from observation to action the fastest.

This book aims to redress this state of affairs and re-examines John Boyd’s original contribution to strategic theory. By highlighting diverse sources that shaped Boyd’s thinking, and by offering a comprehensive overview of Boyd’s work, this volume demonstrates that the common interpretation of the meaning of Boyd’s OODA loop concept is incomplete. It also shows that Boyd’s work is much more comprehensive, richer and deeper than is generally thought.

They are very different as far as I know:

The second step, Orient, does not include making hypotheses. It's about forming a picture or concept in your mind of the situation in which you have to act.

There is nothing in OODA about hypothesizing or testing; you are quickly forming a final conclusion and acting - it's almost the opposite of science, in that regard. It was created for fighter pilots; a failed hypothesis would often result in death and be your last thought.

Science and aerial combat are very different: Science has time to be methodical and it experiments to uncover an absolute truth; in aerial combat you must make final, 'good enough' decisions as quickly as possible, because taking time results in failure.

I’d argue the hypothesis is the third step Decide. The Orient step is framing of the situation and the most important part of the OODA loop versus the actual hypothesis step.

I think the major difference between the OODA Loop and Science seems to be that OODA Loop is in a way backwards looking. You take in the knowledge of the past and your enemy’s knowledge and make a hypothesis you execute on based on that assumed knowledge to not die. Science you predict a future state where the knowledge is unclear or unknown.

It holds up pretty well, mostly because it is so general. It is ingrained in a lot of things, and its basically just another restatement of the classic control loop - something that used from rockets to therapy. There isnt that much remaining of colonel Boyds work besides some old presentations, but it is very much grounded in cybernetics and systems theory.

Good points about the control loop. Is there a fundamental abstraction of it that generally applies - a 1st law of control loops, so to speak?

IDK, Ashby's Law of requisite variety comes to mind. In order for something to control/cope with something else it needs to have the capacity to do so. It seems super trivial when stated like that, but actually has profoundly deep implications.

OODA loop is what powers Kubernetes. The controlller lies at the heart of everything, the API is a Desired State Management, autonomic OODA controllers attempt to fulfill and maintain that desired state. https://kubernetes.io/docs/concepts/architecture/controller/

You can write "cloud native" systems in this pattern. You can also write Custom Resources to extend this pattern to your own systems, and run them via Kube (optional).

I also really enjoyed this submission yesterday on Teleo-Reactive systems, systems which monitor environment/state & adjust their planning & potentially objectives as the environment/state changes. https://web.archive.org/web/20220815134119/https://teleoreac... https://news.ycombinator.com/item?id=43346783

> (diagram for 'Orient')

(This diagram seems to match Wikipedia's, which claims it's from Boyd or based on Boyd.)

The components of 'Orient', the analytical step, are a bit odd: There's no training, skill, knowledge, or intuition; even more critically there's no emotional input, including the influence of present emotion (anger, anxiety, fear, calm, etc.) and our feeling about different options - which is most of how humans orient and decide, especially when choosing quickly.

The diagram lists only culture, genetics, new information, prior experiences, and analysis & synthesis. The last three are unavoidable; the first two seem almost ideological to focus on, given the obvious ones I mention are excluded.

Possibly, Boyd says otherwise (we don't have the primary source), or somehow the categories include the inputs I mentioned.

Also,

> Over time, he formalized his ideas of warfare in his classic “Patterns of Conflict” lectures that inspired the Maneuver Warfare doctrine of the Marine Corps.

Afaik, maneuver warfare goes back to WWII or maybe before. Boyd was born in 1927 and became prominent in ~1960s.

OODA applied to software projects is basically the kernel of "agile" programming.

Iterate quickly, responding to feedback at each iteration.

Reply on news.ycombinator.com