Thread by @RealGeneKim

Taking notes from this great short video on USAF Colonel John Boyd, his Energy Maneuverability theory, OODA loops, and more.



I've read the awesome book "Boyd: The Fighter Pilot Who Changed the Art of War" decades ago, but...

amazon.com/Boyd-Fighter-P…

...but I'm trying to refresh my memory on his teachings, and view it through the lens of how it informs how we create and exploit optionality, maybe against an adversary, but also on how one creates and operates systems.

(From Boyd to Steve Blank/Eric Ries & Dr. Carliss Baldwin)

But before I re-read the books on Boyd, I wanted to get a high-level survey of his material, and had so much fun watching this video — which included footage of multiplayer dogfight sims!

On Boyd: After Korean War, he became an instructor, he became known as "40 Second Boyd"

He bet that anyone behind his plane (an F-86), he could get always get behind them within 40 seconds (and thus have superior fighting position).

No one ever won that bet.

Prior to Boyd, air-to-air combat viewed as too dynamic to boil down to fixed # of moves/countermoves

In contrast, Boyd believed that AAC existed within real-world constraints: limitations by fighter plane, pilots; such as max turn rates, vertical ascent rates, etc.

As captain, he wrote Wrote Aerial Attack Study — "options constrained by range/maneuvering limitations"

Before I forget, here are some links:

Aerial Attack Study: oplaunch.com/resources/aeri…

Energy Maneuverability Study: archives.gov/files/declassi…

YouTube channel w/online multiplayer flight combat sim (I Fly Central), including typical stern attack:

Aerial Attack Study allowed expression of how Boyd could so consistently get behind aircraft he was in front of:

Execute hard climbing turn allows 1) gaining potential energy by gaining altitude, 2) slower speed, 3) no longer within other's view, 4) tighter turn to get behind

..and 5) you have shot at adversary at the top of this maneuver, called the "high speed yo-yo". (Fun video showing this done successfully at 6m 24s — so cool you can do this on video. Can't do this in a book! :)

He's doing some commentary on sim pilots, their techniques...

Next, Boyd describes how to counter high speed yo-yo — "once attacker is in climb, relax g loading and enter shallow descent; when attacker is in downhill portion, turn into them hard. Even if you don't get shot, you'll have opportunity for separation as paths cross"

(cue demo!)

(Demo of countering two attempts at high speed yo-yo maneuvers at 9m)

Criticisms: 1) Boyd took pilots' skill out — not true, as all these maneuvers req considerable skill and timing; 2) Boyd didn't invent anything new — but first written attempt to exhaustively codify moves

Okay, now he describes famous OODA loops: Observe, Orient, Decide, Act.

In dogfight: observe enemy below/in front; orient by known info of you and adversary; decide to dive given all known info, and act upon in

Observe enemy brake turn, Orient by knowing you can't match turn..

Decide knowing that they lost lots of energy in turn, so execute yo-yo turn, and act upon it.

You've executed OODA loop twice, adversary executed loop once; they're behind you in OODA loop; you are driving the encounter;

The person who executes OODA loop faster forces reaction

...this explain how pilots think during dogfight, & how to act to gain advantage. The most critical part is Orientation: "it's critical to know what adversary can and cannot do. i.e., if they're in a P-47 at 200mph, we know they can't pull up into an Immelmann turn, informing Act

"I don't want to spend lots of time in Decide phase, b/c I want to go thru loop quickly; thus importance of understanding moves and countermoves."

Okay, that's his explanation of OODA.

"So up to now, most of these scenarios assume identical planes against each other. What happens if adversary is so superior, that they can sustain brake turn, inspite of extra speed."

"This was what happened in Vietnam. F-4 was new gen fighter, optimized for beyond visual range

Rules of engagement required visual identification, putting F-4 at huge disadvantage: powerful radar and longer range missiles. F-5 might have been better suited, but didn't have range to get into theater.

At disadvantage at MiG-15. New tactics were needed. Enter John Boyd.

He wrote Energy Maneuverability: he teamed up with lieutenant with background in mathematics. Diagrams like showed F-4 at disadvantage at all altitudes and speed.

MiG 15 out-turns F-4 nearly everwhere in envelope: at say 500 knots: MiG turns at 5+g vs F-4 at 2.7g.

Only advantage that could be identified using "energy fighting" — "this is diff to understand [for someone with only algebra background like me], but it's a way to manage your potential & kinetic energy. Those who conserve or gain energy more have advantage."

(Nice explanation)

Here's what Boyd came up with: EM theory, with lots of backup mathematics.

Summary: MiG has maneuverability advantage almost everywhere; F-4 has advantage of time, because of range; enter fight with energy advantage.

Boyd came up with rules for F-4 pilots

Rule was: at subsonic flight, climb at mach 0.83 to 20K feet, and fly flat enabling increased speed, maximizing energy.

Super easy for pilots to use this rule.

Apparently these tactics used until Gulf War(s), and ultimately to build better aircrafts (F-15, F-16, and F-18)

Insights/wonderings:

- design of US aircraft during 1950s-60s availed fewer options for pilots against adversaries in Vietnam-era dogfights (engagements happened at visual range, forced "turn and burn" tactics) — being unable to out-turn or out-climb leaves few winning choices

- design of 1950s fighters were optimized for "altitude and speed, flying straight" — 

- definition of optionality: the sum value of all options created by a decision

(Nice example in this article: medium.com/impression-ven…)

- EM theory identified the low optionality available to the F-4 vs. MiG-15: the prerequisite conditions for F-4 to win required entering with and/or sustaining energy advantage.

- OODA loop: the larger the optionality available, the broader the range of Orient/Decide/Act.

- a system with low optionality (e.g., high coupling, high cost of change) is likely going to have longer OODA loops, allowing competitors to out-innovate/out-execute/out-compete

(Mixing units here? optionality is scalar countable number; "longer OODA loops" implies duration??)

- at extreme, system may have very low optionality compared to peers (e.g., annual software release vs 10 deploys/day; 1 month car model changeover vs. 30m changeover; 1 fixed car production vs. multiple car model production; 60 line side store changes/day vs. 6 LSS chgs/day...

... cont'd: ability to sustain ability to keep developing new features quickly, easily over time vs. features taking longer and longer to implement over time ala @mik_kersten @cutlefish b/c of tech debt

@mik_kersten @cutlefish - driving to the point: enabling optionality is a critical capability in software architectures, unless we're in situation where's there's no variability, no uncertainty, no changes in customer needs, and no need for improvement... ever.

@mik_kersten @cutlefish - Four Steps To The Epiphany and Lean Startup describe opposite conditions: we may not know who customer is, channel to market, business model, technical feasibility. Those are when we need optionality, and we are willing to pay for ability to defer decisions (options).

- ah, an optionality of a fighter could be described as the spherical shape around the aircraft, constrained by its thrust, gravity, rate of turn, etc.

I think 1950s fighters probably would have visibly much fewer options available to them than, say, an F-16.

Correction 1: unit of optionality is the sum of all options, so I’m guessing units is likely measured in dollars, or sum probability of win.

Correction 2: optionality of fighters represented above is only for turning. Doesn’t factor in max altitude, or other domains in which…

…a 1950s fighter would win. Plotted a certain way, I’m sure we could show that 1950s had far bet options than MiG-15 (eg, beyond visual range missiles, operating altitude above MiG-15 maximum, etc)

Correction 3. Much of optionality of fighter constrained by its design…

…in other words, it can’t be changed by the pilot. However, EM also changed optionality — altitude at start of engagement, how one manages energy can add or remove options (Eg, on ground at standstill, you have nearly zero options)

Reflection 1. Using fast vs slow integrated problem solving categories:

Slow: creating EM framework, creating doctrine and training, rolling it out across pilot force, establishing training.

Fast: training, and combat operations

Slow: creating new class of fighters