Solving complexity through simplicity - explaining with equations is enlightening

/

By Duncan Anderson. To see all blogs click here.

Summary: You solve complexity through simplicity. 

  • The goal with complexity is to reduce the problem down to a few key variables you can reason with.

  • You need to figure out how the variables interact and how you can be wrong. 

  • The world is far too complex to be able to understand at full resolution. 

  • Complexity => simplicity

    • Complexity = problem space you cannot reason with

    • Complexity => Simplicity = key variables * how they interact 

  • All models are wrong (aka are abbreviations of the world), but some are useful. 

Jingle: if you can't make an idea into an equation (model) you don't understand it deeply. 

  • Trying to understand the world through making models is literally some of the best fun I know of. 

  • When you understand something deeply, you’re able to cater your explanation to the recipient. This is an example.

    • In this example, the neuroscientist has changed his vocabulary and syntax (his language) to accommodate his listener

    • Models allow us to use the same language to express a complex idea to the majority.


Idea Iteration Articulations: 

  • L1: verbal description only

  • L2: equation

  • L3: equation + taxonomy for each variable

  • L4: visual representation of equation (aka model)

  • L5: machine with flowchart and speed and quality metrics for each node

  • L6: L5 + having numbers that flow through each node

Examples of simple “L2: equation” models:

  • Progress = Plan * Execution

  • Trust = Time * Consistency

  • Art = Expression + Meaning

  • Despair = Suffering - Meaning

  • Anxiety = Uncertainty * Powerlessness

  • Disappointment = Expectations - Reality

  • Good life = Enough Money + Friends + Health + Purpose

  • Happiness = wanting what you have / having what you want

  • Emotional Intelligence = Self-awareness + Self-regulation + Social skill + Empathy + Motivation

  • Attention = Focus - Distractions

  • Pain + Reflection = Progress

  • I love little juxtaposition equations as well. Ie change one variable and what happens (aka flexing the model)? I often find it easiest to explain something by comparing it to something else or saying what it isn’t. 

    • Depression vs Sadness

      • Depression = Unhappiness * Not knowing why you are unhappy

      • Sadness = Unhappiness * Knowing why you are unhappy

    • Emotional health (see blog)

      • Emotional health != Feeling only positive emotions

      • Emotional health = Feeling the full spectrum of emotions * In a healthy way

    • Boredom

      • Boredom done well => Relaxation

      • Boredom done poorly => Anxiety

    • Startups

      • Great Team * Bad Market => Market’s reputation in tact

      • Weak Team * Great Market => Market’s reputation in tact

      • Great Team * Great Market => Something special happens

    • Habits vs addictions

      • “First you make your habits, then your habits make you.”

      • Addiction done well => Habit

      • Habit done poorly => Addiction

    • Truth vs Lies

      • “You have nothing to fear from the truth… but that doesn’t mean the truth won’t hurt.”

      • Lie = Sweet in the beginning + Sour in the end

      • Truth = Sour / Sweet in the beginning + Sweet in the end

    • Suffering vs Growth

      • Hardship + No purpose = Suffering

      • Hardship + Purpose = Growth

  • Unidirectionality is also a fav thing of mine. Axioms that are mathematically incompatible have an ugly beautifulness that makes me chuckle.

    • Feeling * Reason => Meaning

    • Reason * Meaning => Feeling

    • Smile because you are happy… Or smile and it will make you happy!

    • There is a future blog on this :)

      • Passion * Work => Happy

      • Making the world better * Work => Happy + Passion

  • *Aside: I've decided I'm going to try make a blog that is only equations. Sounds like fun!

  • Comment: 

    • Seriously, anything, ANYTHING!!!! Can be communicated in an equation / model. 

    • Duncan’s blogs + you = happy :) 

    • No one is saying that these equations are perfect and explain everything, but do they help you understand something you didn’t before? 

      • Good equation => increases understanding

      • Good equation => shows the world from a different point of view (lens) 

      • Bad equation => confuses things

I love love LOVE trying to articulate things. “All I do is articulate and rearticulate.” Does this make me a AAA human? 

  • For the purpose of this blog let’s say ideas come in 3 sizes: Small, Medium and Large. 

  • IMO for all Medium or Large ideas you should at minimum be able to express them “L1: verbal description only” + “L2: equation”. 

  • What level do you go to? Well that is the fun. 

  • Examples and process below.

++++++++++++++++

Details

Process for trying to articulate an idea better: 

  • 1. Is the idea and Medium or Large? If Yes then go to ‘2’. 

  • 2. Try and push the idea up “Idea Iteration Articulation” levels as far as you can (ie from L1=>L6 if you can)

  • 3. Then find the optimal balance for which level of “Idea Iteration Articulations” gives the most value and roll with that

    • Value of articulation = 1. Explanation power of model * 2. Your ability to ‘flex’ (change) the model and understand what is going on [see what I did here? Make a model about models! Meta model mania]

      • 1. Explanation power of model = 1.1 Number of variables * 1.2 How the variables interact * 1.3 quality of MECE

    • You’ll find the balance point where the model has the most value. 

    • Paradoxically the ‘greater’ the absolute explanation power of a model the harder it is to understand / use. So the model might be pretty but it’s useless...  pretty useless! 

    • I normally try to do ‘Occam’s Razor’, ie the simplest solution is the best… but so simple that it can’t explain what is going on isn’t useful either. Ie you solve complexity through simplicity.

    • “Don’t be simply stupid, or pretty useless, be beautifully balanced ;)” TM DA 2019!

Example 1: “L2: equation” at Edrolo (a company I co-founded) one product we are now getting into are textbooks for Years 11&12. 

  • We were trying to figure out what the line of minimum sufficiency is for saying yes to building a textbook. 

  • The generations: 

    • Generation 1: a product that is better than existing textbook offerings (seriously, I know right!)

    • Generation 2: a product that was 50% better as measured by the growth it provided to a student

    • Generation 3: 1. Everything an existing textbook does + 2. 2x+ Dealmakers

    • Generation 4: 1. Everything an existing textbook does + 2. 2x+ Dealmakers - 3. No dealbreakers

    • Generation 5: 1. Everything an existing textbook does + 2. 2x+ Instantly recognisable dealmakers - 3. No dealbreakers

    • Generation 5: 1. Everything an existing textbook does + 2. 2x+ Instantly recognisable irrefutable dealmakers - 3. No dealbreakers

  • Comment: 

    • We have a lot more detail in eg ‘how to find dealmakers’. That has been systemised into a “L5: machine with flowchart and speed and quality metrics for each node”. But that is not for today :)

    • With each new generation of the articulation of ‘minimum sufficiency to greenlight building a textbook’ we’ve found it increasingly easy to (hopefully) make great products. 

    • “The quality of your ability to articulate the problem affects the quality of your solution!”

    • In case it’s not clear, god I love this stuff! 

Example 2: “L3: equation + Taxonomy” there some VCs on this email list so I’ll make up something for you :). Should I invest in a startup or not? 

  • *note: I’m just making this up on the spot, so hopefully useful

  • V1: an unstructured verbal description of whether you like a startup or not

  • V2: Invest in startup = 1. Team * 2. Product * 3. Market 

  • V3: V2 + taxonomy for each of the of the variables

    • Taxonomy suggestion: 

      • Dealmaker

      • Nice to have

      • Not nice to have

      • Dealbreaker

    • Eg it doesn’t matter if there are lots of good things about a startup, 1x dealbreaker and it’s game over for investing in them! 

  • V4: I’m going to zoom in on the team variable and break it into it’s on mini equation

    • 1. Team = 1.1 Founder market fit * 1.2 Learning machine (ie eats information like no tomorrow) * 1.3 Hectic work ethic * 1.4 Grit

  • Comment: 

    • I’m going to stop here making equations for investing in a startup. 

    • What I’ve found is that if you have a base ‘equation’ for which to try explain a ‘problem space’ you can then have a much more useful conversation with people. 

    • If you can hang off your thoughts onto a model you can start to calibrate where things fit. Instead of ‘two ships sailing past each other in the night’ you can see ‘ok person 1 is talking about this part of the equation, person 2 is talking about another part, how do we try and put this altogether?’

    • Example: 

      • Let’s say you are discussing a startup and you are talking about the founders 

        • Person 1: I don’t like the team as I don’t think they have ‘1.1 Founder market fit’ 

        • Person 2: I like the founders because they are ‘1.2 Learning machines’. 

      • How can you use an equation + taxonomy to help here? 

        • You discuss what level each person’s point about the founders are on the taxonomy. 

        • Person 1’s ‘1.1 Founder market fit’ => Not nice to have. 

        • Person 2’s ‘1.2 Learning machines’ => Dealmaker. 

          • ‘1.2 Learning machine’ level is a ‘Dealmaker’ because over time the founders will improve “1.1 Founder market fit” as they learn about the market (they are learning machines after all) and as such you are good to move ahead. 

Example 3: “L4: visual representation of equation (aka model)”

Example 4: “L5: machine with flowchart and speed and quality metrics for each node”

Example 5: “L6: L5 + having numbers that flow through each node”

  • You’ll need to come and work at either Edrolo or OwlTail to see these. 

  • We are always hiring at both!