Mental models.
Mental models
Demartini: seeing both sides
The promotion has an unseen cost
Demartini’s perspective method looks for the overlooked side of a judgement. Leila sees a missed promotion as entirely bad. Comparing the money, time and duties begins to reveal several consequences within the same event.
Higher pay makes the new role shine
Leila imagines the promoted job mainly through its salary. Her current position appears inferior because its pay is lower. That narrow comparison gives the desired role a larger emotional pull than its complete conditions might support.
The work schedule changes the comparison
The promoted role also requires weekend duties. Leila’s existing job preserves free evenings and familiar tasks she enjoys. Those details reveal benefits in staying and drawbacks in the attractive alternative, using the same concrete situation.
Both sides can alter the reaction
Demartini teaches that seeing benefits in the painful setback and costs in the desired role can bring perception into balance. He calls the intended softening of resentment and idealisation a reduction in emotional charge. Actual effects vary.
The weight of each cost belongs to Leila
Free evenings may matter greatly to Leila, while another person might prefer the extra pay. The picture supplies consequences to consider. Their personal weight remains open, alongside any question about fairness in the promotion process.
Demartini proposes balancing perception by seeing benefits and drawbacks together, aiming to soften emotional charge.
His account connects balanced perceptions with gratitude. He describes a felt shift in the relationship to an experience. Saying the charge cancels out refers to that proposed emotional shift.
The full Demartini Method includes other questions about people, traits and values. These cards explain its central benefits-and-drawbacks idea through an original everyday illustration.
A useful detail needs support from the actual situation. Inventing a benefit or ignoring an important cost can distort the picture. Clearer understanding can lead to boundaries, a complaint or a different choice.
The example offers several consequences to consider. Their importance varies with the person and situation. Demartini's universal claim of equal balance requires separate evaluation.
- John Demartini: Turning Setbacks Into Comebacks, part one, examining setbacks and imagined alternatives
- John Demartini: Black and White Thinking, his account of positive and negative judgements
- John Demartini: Love and Wisdom, benefits within difficulty and drawbacks within attraction
- John Demartini: the Demartini Method, his description of balancing perceptions and emotional charge
When learning changes what we do
An idea reaches the shopping basket
Practical learning changes how someone acts or decides. Amir learns that checking the fridge before shopping can reduce duplicate purchases. The idea becomes useful when it changes the contents of his next shopping basket.
The old routine produces duplicates
Amir arrives at the shop without remembering what is at home. He buys another pot of yoghurt, then finds two unopened pots in the fridge. The repeated purchase reveals a gap in his routine.
A check changes the next purchase
Before his next trip Amir looks in the fridge. He sees the yoghurt and leaves it off the list. The practical consequence follows through a clear chain: better information changes what he buys.
A missed item improves the routine
Amir still forgets one item while shopping. On the following trip he carries a photograph of the shelf. That small adjustment responds to a remaining weakness instead of treating one successful trip as the finished system.
Learning and feedback form a loop
Amir’s idea led to a check, the check changed a purchase, and the result improved the next attempt. Behaviour is one sign of practical learning. Understanding can also exist before circumstances allow somebody to act.
Check before shopping changes routine → Amir
Amir changes purchases → Shopping basket
Shopping basket reveals a gap → Shelf photograph
Shelf photograph refines the method → Check before shopping
Practical knowledge becomes visible when an idea changes a choice and the result improves the next attempt.
A person can know a fact before using it. Someone may know where a country is while having no reason to travel there. Philosophers discuss knowledge of facts and knowledge of how to do things. This lesson focuses on the practical value of using learning.
The gap between understanding an idea and acting on it is often called the theory-practice gap. John Dewey's account of reflective thinking connects a difficulty with possible explanations, reasoning and further observation. Our three steps are a simplified teaching picture of learning through inquiry.
Action also depends on circumstances. The COM-B model describes capability, opportunity and motivation: what someone can do, what their situation allows, and what moves them to act. Time, tools, habits or competing demands can affect whether knowledge reaches behaviour.
One successful attempt gives limited evidence. Chance or a change in the situation may have helped. Repeated observations and reflection make it easier to judge when an idea is useful and when it needs adjustment.
Behaviour is one practical sign of learning. Facts can be understood before they are used, and circumstances can prevent a knowledgeable person from acting.
What can I do next?
A cancellation leaves several possible moves
George Mack uses high agency for the habit of looking for useful action within a situation. Leah’s drawing class closes. The class is gone, while her interest in learning and several other possible routes remain.
The usual path disappears
Leah had relied on a weekly class for instruction and company. Its closure removes both. Waiting for it to reopen would leave her learning paused, so the problem includes finding another source of guidance and practice.
One message reveals another route
Leah contacts the former tutor, who mentions a library group. The useful new option came from an action she could take herself. The tutor’s reply and the group’s availability still depend on other people.
A friend supplies a second option
The library group meets monthly, so Leah also arranges an hour of drawing with a friend. This creates a practical weekly routine from available resources. It may offer less instruction than the original class.
Action expands possibilities within real limits
Mack’s agency idea appears in Leah’s messages and experiments with alternatives. Her actions increase the options she can see. Outcomes also depend on opening hours, money, access and whether other people can help.
Former tutor offers information → Leah
Library group offers instruction → Leah
Drawing friend offers practice → Leah
A useful action can reveal a new route while real constraints remain part of the situation.
Ideas from George Mack · High Agency
Mack develops and popularises this tool; his essay also credits Eric Weinstein. Examples here are original.
Use high and low agency to describe an approach in a moment. People can act differently in different situations.
A useful next action may be asking for support, resting or stopping a harmful plan.
Agency increases possible actions. Outcomes still depend on constraints and other people’s choices.
How do I make a vague worry useful?
A broad worry becomes one visible problem
George Mack encourages turning vague concerns into specific questions. Ben thinks his mornings are a mess. Looking at one repeated delay reveals something concrete: he spends several minutes searching for keys left in different places.
The keys move around the house
On Monday Ben leaves his keys in a coat. On Tuesday they stay on a table. Each morning begins with uncertainty because their location changes, adding a search to the steps before leaving.
A bowl gives the keys a home
Ben puts a bowl beside the front door and leaves the keys there on arrival. The change targets the location problem directly. It needs little explanation because the object and its intended place are visible.
The next morning contains less searching
When Ben prepares to leave, the keys are in the bowl. One delay disappears from that morning. Traffic or preparing breakfast may still create pressure, so the result concerns a specific part of the larger problem.
Specificity makes the result assessable
Mack’s principle connects a broad concern to an identifiable cause and a visible change. Ben can now see whether the bowl reduces key searches. Other morning difficulties can be understood separately as further evidence appears.
A specific description makes it easier to see which change affected which result.
Ideas from George Mack · High Agency Habit: Specific > General
Mack presents this practical habit. Examples and explanations here are original.
A clear description helps another person understand what help you need.
Keep checking whether your chosen measure represents the problem you actually care about.
Some situations contain several causes. One clear question is a starting point.
What should I let into my attention?
Learning depends partly on what arrives
An information diet is the mix of things someone reads, watches and hears. George Mack discusses choosing that mix deliberately. Mina wants to grow tomatoes, and her attention is split across an endless stream of gardening videos.
Different videos offer different advice
Mina hears watering advice for several climates and types of container. Each video may describe a real situation, while their differences make her own plant harder to understand. More incoming material has added several unlabelled conditions.
A local guide narrows the comparison
Mina chooses a guide for her climate and the kind of pot she uses. Its advice now connects to identifiable features of her plant. The narrower information source gives her a clearer starting point for observation.
The plant adds its own evidence
Mina records watering and notices how the plant changes. The guide supplies an explanation; the plant supplies observations under her actual conditions. Together they can reveal whether the advice fits this particular situation.
Local tomato guide guides care → Tomato plant
Tomato plant produces observations → Watering notes
Watering notes checks the advice → Local tomato guide
A useful filter remains adjustable
Mack’s information-diet idea concerns what receives attention. Mina’s local guide and notes create a clearer learning loop. They still cover only part of gardening, so new pests or weather can make further sources useful.
Local tomato guide provides explanation → Mina
Watering notes records experience → Mina
Tomato plant presents new conditions → Mina
Relevant information becomes more valuable when it connects with observations from the situation being understood.
Ideas from George Mack · Information Diet: High Agency vs Low Agency
Mack presents a way to choose inputs. The gardening example and review questions are original.
Keep room for an unfamiliar viewpoint. A comfortable feed can hide useful disagreement.
Reliability and enjoyment are different questions. A funny post may serve a break; a factual claim needs checking.
Filtering information always leaves gaps. Review your choices as your needs change.
What else could this attention go towards?
One repeated thought occupies other time
George Mack’s thinking-cost idea concerns what attention given to one matter displaces. Leo repeatedly checks a parcel’s tracking page. Each check uses a small part of the evening he also wanted to spend drawing.
The tracking page keeps returning
The parcel is expected this evening. Leo opens the same page several times, often finding no change. The information stays similar while the repeated switching interrupts his involvement in the drawing beside him.
Leo checks again → Tracking page
Tracking page shows same status → Expected parcel
Expected parcel keeps concern active → Leo
A delivery window sets a boundary
Leo sees that the parcel is due between six and eight. This narrows the relevant period. A reminder near the end of that window gives him a way to return to the question later.
Drawing receives a longer stretch
Leo spends the next part of the evening on the drawing. The delivery remains uncertain, yet a specific follow-up has already been arranged. Fewer switches leave more continuous attention available for the activity in front of him.
The cost includes what attention displaces
The parcel example makes thinking cost visible through time and interrupted focus. Mack’s framing can clarify a choice about attention. Some worries are persistent or involuntary, so the amount a person can redirect will vary.
The cost of a repeated concern includes the attention it takes from other valued activities.
Ideas from George Mack · Thinking cost
Mack applies the established idea of opportunity cost to attention. This card uses an original example.
Writing a next step can make a practical question easier to manage.
Some thoughts carry important feelings. Listening to them and seeking support can also be useful ways to spend attention.
Attention is only partly under voluntary control. This tool describes a choice to try.
What is first-principles thinking?
Keeping tea warm begins with heat
First-principles thinking builds an explanation from basic facts. Hot tea loses heat to its cooler surroundings. Understanding the routes by which heat leaves can explain why changing the cup or adding a lid affects cooling.
The tea has several routes to cool
Heat can pass through the cup, and the exposed surface allows heat transfer and evaporation. A lid affects the surface route. This connects a familiar object with a specific part of the cooling process.
Hot tea loses heat through cup → Cooler room
Hot tea loses heat at surface → Exposed surface
Cup lid changes that route → Exposed surface
Two matching cups isolate one change
An illustrative comparison uses matching cups with equally hot water in the same room. One receives a lid. Keeping the other conditions similar makes a later difference more informative about the effect of covering the surface.
The readings connect explanation with evidence
After the same interval, temperature readings show how much each cup has cooled. If the covered cup stays warmer, the observation supports the explanation under those conditions. A different result would call for further investigation.
Hot tea supplies one measurement → Temperature readings
Basic facts still need checking
The reasoning travels from heat flow to the exposed surface, then to a lid and a comparison. First principles organise an explanation around underlying relationships. Assumptions and measurements determine how much confidence the conclusion deserves.
First-principles thinking connects a practical change to the basic relationships that could explain its effect.
Every explanation needs a starting point. A computer lesson can start with switches. If switches still feel mysterious, open that box and learn how electrical charge lets one part control another. Continue until each starting idea makes sense.
Basic assumptions still need evidence, and rebuilding every established result can consume substantial time.
Map and territory
A map keeps some details and leaves others
A model is a simplified representation used to understand something. Mina’s street map shows two equally nearby cafés. It describes distance clearly while leaving out a detail that matters for her visit: the entrances.
Both cafés look equally convenient
Mina compares the map’s distances and sees no obvious difference. That conclusion fits the information supplied. The map contains roads and locations, while doorway height and access arrangements are outside its selected level of detail.
The entrance reveals a missing difference
At the first café Mina finds steps. The second has a level entrance. Her friend uses a wheelchair, so this previously omitted feature changes which venue will work for their meeting.
Another layer answers the access question
Mina adds entrance information to her comparison. Distance still matters, and the extra layer now makes the relevant difference visible. The more useful representation is the one containing details needed for this particular decision.
Street map shows distance → Mina
Entrance steps reveals a barrier → Mina
Level entrance reveals access → Mina
The representation has a purpose
The street map helped with location. Access required another kind of information. Map-and-territory thinking keeps the actual café separate from any one description, making a model’s purpose and omissions part of its interpretation.
A model’s usefulness depends on which real features it includes for the decision being made.
A score on a dashboard is also a map. If a queue is summarised by its average waiting time, some long waits disappear inside that average. Open the distribution of waits and inspect individual journeys to recover the missing detail.
A detailed model can still omit the particular feature that matters for your decision.
Circle of competence
Expertise has a boundary
A circle of competence covers tasks someone understands reliably through learning and practice. A bicycle mechanic knows familiar brakes well. A new electronic gear system introduces a mechanism the mechanic has yet to understand.
Repeated repairs make the brakes familiar
The mechanic has inspected many worn pads and seen how replacement changes braking. Similar cases connect observations with outcomes. This repeated experience supports a confident judgement about the next bicycle with the same fault.
New practice builds reliable knowledge → Bicycle mechanic
Electronic gears introduce a different mechanism
The new gear system uses sensors and control software. Familiar mechanical experience offers some background, while the electronic fault could have several unfamiliar causes. Confidence based on brake repairs would exceed the evidence for this task.
Reading and practice expand the boundary
The mechanic studies the service manual and practises the diagnostic process. New observations gradually connect the electronic system’s parts with its behaviour. Expertise can expand as these connections become more accurate and repeatable.
Reliable confidence follows the task
The mechanic’s confidence now distinguishes familiar repairs from newer work. The circle of competence is useful because knowledge varies by task and can become outdated. Recognising its edge helps connect decisions with evidence and suitable help.
Expertise is most useful when its boundary and its supporting experience are visible.
Imagine colouring a map by evidence strength. A skill you have practised successfully several times gets a stronger colour than one you have only watched online. The exercise makes confidence depend on experience with the actual task and conditions.
Past expertise can become less reliable when the environment or task changes.
Inversion
Imagining failure can reveal an obstacle
Inversion means examining a problem from the opposite direction. An exhibition organiser wants visitors to enjoy the work. Imagining how the visit could fail reveals a simple possibility: people arrive while the entrance remains locked.
The display alone cannot admit visitors
The organiser has prepared the pictures and opening time. Those achievements support the exhibition, while visitors still need a clear address, an unlocked door and a display that arrives before they do.
Unclear directions guides arrival → Expected visitors
Locked entrance allows entry → Expected visitors
Late display provides something to see → Expected visitors
A failure story exposes the missing key
The imagined locked entrance leads the organiser to discover that only a neighbour holds the key. That small detail had escaped attention because most planning focused on the pictures inside the building.
Available key controls opening → Locked entrance
Locked entrance controls access → Weekend exhibition
Arranging access removes one avoidable failure
The organiser arranges the key, checks the address and confirms delivery. Each action addresses an identified obstacle. Visitors still may arrive late or dislike the display, so these changes improve conditions without guaranteeing the whole outcome.
The opposite view adds useful detail
Inversion complemented the attractive picture of a finished exhibition with a picture of blocked access. Its value came from a plausible failure and a checkable dependency. Imagined risks still need judgement about likelihood and importance.
Looking at how an outcome could fail can reveal conditions that success depends on.
Use a second pass to look for success requirements. A room can open on time and still contain little of interest. Inversion helps remove identifiable obstacles; combining it with a positive description of the desired experience gives you a fuller plan.
Imagining failure can miss unfamiliar risks and can also exaggerate very unlikely possibilities.
Second-order thinking
A decision has consequences after its first result
Second-order thinking follows what happens after an immediate effect. A library reduces opening hours to save money. The saving matters, and the changed hours also affect when people can visit and borrow books.
Shorter opening hours changes access → Evening visitors
Closing earlier reduces some costs
The library shortens its evening opening. Fewer staffed hours can reduce part of its running costs. This immediate effect is clear enough to attract attention, especially when the library faces a constrained budget.
People working late lose a visiting window
Some residents previously arrived after work. The new closing time removes that opportunity. Their reduced access follows from the same decision that created the saving, although it appears in a different part of the system.
Borrowing and support may then change
If those residents cannot find another time, they may borrow fewer books. Over time this could affect the library’s usage figures or local support. The later links are plausible possibilities whose strength needs observation.
The full decision contains several timescales
The library faces immediate costs, changed access and possible longer-term effects. Second-order thinking places these in one picture. More distant effects deserve cautious estimates because people may adapt and other conditions may also change.
Shorter opening hours changes access → Evening visitors
A useful decision picture includes the first effect and the consequences that may follow it.
Responses can reverse an initial effect. A faster road can attract additional trips, changing congestion later. Whenever the action changes people’s incentives or choices, ask how their response feeds back into the original situation and what evidence could test that prediction.
Longer chains contain more uncertainty, so distant effects need cautious estimates and review.
Thinking in probabilities
A chance describes possible outcomes
Probability describes how likely an outcome is under stated conditions. A bag contains three blue counters and seven yellow ones. A random draw can produce either colour, with blue accounting for three of the ten possibilities.
The draw treats each counter equally
The counters are mixed and selected without looking. With an equal chance for each counter, the three blue ones give blue a three-in-ten chance. The seven yellow ones account for the remaining seven-in-ten chance.
A yellow counter comes out
The first selected counter is yellow. That outcome fits the stated probabilities because yellow was always possible and more likely. One draw gives one result; the probability described the possibilities before that result became known.
Returning the counter preserves the conditions
The yellow counter goes back before another draw. The bag again contains the same three blue and seven yellow counters. Repeated draws under these conditions allow the original probability to be compared with a growing record.
Bag of counters provides a random pick → One random draw
One random draw adds a result → Repeated draws
Repeated draws counter is returned → Bag of counters
The percentage belongs to the model
Three in ten can also be written as thirty per cent. It describes the random draw from this known bag. Real estimates often depend on uncertain evidence, so the precision of a number deserves its own assessment.
A probability describes possible results under stated conditions while each attempt produces only one result.
Suppose you label ten events as having an 80% chance. If your estimates are well calibrated across many such groups, about eight in ten occur. A single unexpected outcome supplies limited evidence about your forecasting skill; a record of predictions supplies much more.
Precise percentages can hide weak evidence or changing conditions.
Thought experiments
An imagined scene can expose a relationship
A thought experiment uses an imagined situation to explore an idea. Amina and Ben have one umbrella in the rain. Changing how they walk makes the relationship between sharing, distance and shelter easier to see.
One umbrella can provide shelter → Amina
One umbrella can provide shelter → Ben
Falling rain creates the need → Amina
They begin beneath one umbrella
Amina and Ben walk close together. The umbrella can shelter both to some extent because its covered area overlaps their positions. Their shared arrangement matters alongside the umbrella’s size and the direction of the rain.
One umbrella covers one side → Amina
They take separate paths
The imagined scene changes only their arrangement: Amina and Ben walk apart. The single umbrella can now accompany one of them. Distance has changed what sharing can achieve even though the amount of equipment stays the same.
The resource limit becomes clear
With one umbrella and two separated paths, both people cannot receive that umbrella’s direct coverage at once. The result follows from the chosen assumptions. Another umbrella or a shared route would create a different situation.
A useful thought experiment shows its assumptions
The umbrella story isolates a connection between sharing and physical arrangement. Real shelter also depends on size, wind and movement. The imagined result helps organise reasoning, while claims about actual rain require observations of those conditions.
One umbrella offers limited coverage → Amina
One umbrella offers limited coverage → Ben
Falling rain changes actual protection → One umbrella
A thought experiment reveals what follows from its assumptions and identifies what reality would still need to establish.
Thought experiments can test consistency. Suppose a rule claims everyone should always go first through a narrow doorway. Imagine two people following it simultaneously. The conflict exposes a problem in the rule before you attempt to apply it in a real crowd.
A coherent imagined result still needs empirical evidence when it makes claims about the real world.
Occam’s razor
A simple explanation can be checked first
Occam’s razor favours explanations requiring fewer unnecessary assumptions when they fit the evidence. A speaker falls silent. A loose cable is one possible explanation, while several simultaneous internal failures would require additional events to have occurred.
Silence leaves several causes possible
The speaker worked earlier and has now stopped. Silence by itself reveals little about the cause. Power, connections and internal components all remain possible sources, so an explanation needs more evidence than the symptom alone.
Loose cable could interrupt sound → Silent speaker
Internal failures could stop operation → Silent speaker
The cable provides a direct test
The cable is visibly loose. Reconnecting it tests an explanation with few added assumptions and little effort. The value lies in how well that explanation fits the observation and how directly its consequence can be checked.
Music returns after reconnection
When the cable is secured, the music returns. The result supports the cable explanation for this episode. If silence had continued, the simple first explanation would have lost support and further causes would remain open.
Simplicity works alongside evidence
The speaker story shows why a simpler fitting explanation can be a useful starting point. Its success came from a checkable relationship and the result. A simple account that fails to explain observations would need revision.
A simple explanation earns confidence through its fit with observations and the tests it survives.
Simplicity depends on the question and the evidence. A two-part explanation may fit observations much better than a one-part explanation. Keep a visible balance between explanatory fit and extra assumptions, then seek fresh evidence where the explanations make different predictions.
The simplest explanation can be wrong, especially when it fails to account for important observations.
Hanlon’s razor
Silence can have an ordinary explanation
Hanlon’s razor encourages considering mistakes or limitations before assuming hostile intent. Mina’s friend has left an invitation unanswered. A crowded inbox, confusion about the date or an intentional refusal could all produce the same initial silence.
Unanswered invitation awaits a reply → Her friend
Crowded inbox may hide it → Unanswered invitation
Her friend has yet to answer → Mina
Mina first feels rejected
Mina notices the unanswered message and feels hurt. That feeling makes deliberate rejection seem plausible. The evidence currently consists of an absent reply, which leaves the friend’s actual circumstances and intentions largely unknown.
A missed message becomes possible
Mina remembers that her friend receives many messages while travelling. This supplies another explanation for the silence. Considering it broadens the picture without establishing which explanation is correct in this particular case.
A clear follow-up produces information
Mina sends the date again, and her friend replies that the first message was missed. That reply provides additional evidence. The initial interpretation can now change because the explanation is based on more than the silence alone.
Mina sends a clear follow-up → Her friend
Her friend explains the missed message → Later reply
A useful assumption remains revisable
Hanlon’s razor helped Mina keep an ordinary mistake in view. A repeated pattern or conflicting evidence could justify a different judgement. The principle provides a starting possibility while the actual relationship supplies the evidence.
Considering an ordinary mistake can prevent a premature judgement while leaving room for evidence about intent.
A repeated event provides new evidence. If the same problem continues after clear communication and several chances to respond, revise the explanation. This heuristic guides an initial search among possible causes; continuing judgement depends on the evidence and the consequences.
This heuristic offers no guarantee about anyone’s intentions and should remain responsive to evidence.
Systems thinking
A kitchen works through connected parts
A system is a set of parts that affect one another. In a kitchen, orders, preparation, cooking and serving connect. Meals leave only after these stages work together, so speed at one stage can influence every stage after it.
Orders arrive faster than meals leave
The kitchen receives more requests than its cooking stage can finish. Prepared food and waiting customers accumulate. The delay emerges from the relationship between arrival and completion, even when every worker is busy.
Faster ordering increases the pressure
Another person takes orders more quickly. That improves the front counter while sending extra work into the same cooking stage. The local improvement can enlarge the queue because the full process still has the same completion limit.
The cooking stage receives attention
Rearranging preparation around the oven reduces gaps between batches. If cooking had been the limiting stage, more completed meals may now leave. The useful change depends on identifying the actual relationship that held the process back.
The connections explain the outcome
The kitchen’s performance depends on flow across all stages. Systems thinking keeps those connections visible so a local improvement can be assessed against completed meals and waiting. Changing demand or equipment can shift which relationship matters most.
A system’s result depends on how its parts interact, including where work accumulates and what limits completion.
Suppose faster cooking produces meals before servers can collect them. A local improvement creates a new queue elsewhere. Record the overall outcome, such as meals served warm, alongside each component’s speed so you can see how the connected system behaves.
A system diagram simplifies reality and can miss important relationships or changing conditions.
Feedback loops
A result changes the next action
Feedback occurs when a process’s result influences what happens next in that same process. Cold shower water makes someone turn the control warmer. The resulting warmth then influences whether they keep adjusting or stop.
Water temperature is sensed → Felt warmth
Felt warmth guides adjustment → Shower control
Shower control changes temperature → Water temperature
Cold water prompts an adjustment
The person feels cold and turns the shower control towards hot. This action changes the mixture entering the pipe. The warmer mixture takes time to reach them, creating a gap between action and sensation.
More adjustments happen before the result
The water still feels cold during that delay, so the person turns the control further. Several adjustments now accumulate before the first change is felt. The control has moved beyond what one informed adjustment might require.
Still feels cold prompts another turn → Turned again
Turned again adds more heat → Effects still travelling
Effects still travelling postpones new information → Still feels cold
The arriving water becomes too hot
When the warmer mixture finally arrives, the person feels excessive heat and turns the control back. The loop has overshot a comfortable temperature. The sequence explains how reasonable responses to old information can produce oscillation.
Timing changes how the loop behaves
The shower loop connects sensation, action and a delayed physical result. Allowing the changed water to arrive before another adjustment can reduce overshoot. Multiple delays or controls would create a more complicated feedback system.
Water temperature provides feedback → Felt warmth
Felt warmth guides a change → Shower control
Shower control starts a response → Pipe delay
Pipe delay delivers the effect → Water temperature
Feedback changes future action, and delays determine how accurately that action matches the current situation.
The words positive and negative describe the direction of feedback in technical usage. Positive feedback reinforces a change; negative feedback counteracts it. Use the labels reinforcing and balancing on a beginner diagram to make their effect easier to see.
Multiple loops and delays can make the combined behaviour difficult to predict from a single loop.
Bottlenecks
The slowest required stage limits the line
A bottleneck is the part of a process limiting its completed output. In an illustrative sandwich line, fillings can be prepared faster than sandwiches are assembled. Faster preparation alone therefore cannot make finished sandwiches leave faster.
Fillings accumulate before assembly
The preparation team can supply twelve sets of fillings per minute, while assembly handles six. Extra prepared ingredients wait. Their accumulation makes the mismatch visible even though the preparation workers are performing quickly.
Faster wrapping leaves the limit unchanged
Wrapping already handles ten sandwiches per minute but receives only six. Increasing wrapping speed would leave assembly unchanged. The improvement creates spare ability at a stage that was already waiting for work.
More assembly capacity changes the output
Suppose assembly improves from six to nine sandwiches per minute. Preparation can supply twelve and wrapping can handle ten, so the simplified line can now complete nine, assuming steady supplies and enough demand.
The constraint can move after improvement
If assembly later exceeded ten, wrapping could become the next limit. Bottleneck thinking follows the whole flow as conditions change. The illustrated rates describe a simplified line; interruptions and variable demand can alter real performance.
Improving the stage that currently limits completion can change the output of the whole process.
Constraints can move during the day. Preparation may limit the morning while the till limits lunchtime. Observe several periods and check interruptions. Averages can conceal short bursts when demand exceeds capacity and queues grow even though daily capacity seems sufficient.
Real processes can have several interacting constraints, variable demand and interruptions.
Compounding
An increase can enlarge the next increase
Compounding means each change builds on the result of earlier changes. An idealised collection begins with one hundred seeds and grows by one tenth per season. The second season starts with the enlarged collection.
- 100 seeds
- 110 seeds
- 121 seeds
The first season adds ten seeds
One tenth of one hundred is ten. Adding those ten produces a collection of one hundred and ten seeds. The added seeds now belong to the amount from which the next increase will be calculated.
The next season starts from more
The second increase is one tenth of one hundred and ten, which is eleven. The rate stays the same in this illustration, while the larger starting amount makes the number added slightly larger.
Growth begins to build on growth
The collection has gained twenty-one seeds over the two seasons. One of those seeds comes from applying the second increase to the first increase. That small extra contribution is the mechanism compounding describes.
The repeated rate is an assumption
The example makes compounding easy to see by holding growth at one tenth each season. Real seed production encounters changing conditions, losses and limits. The arithmetic describes how repeated percentage changes interact when its assumptions hold.
- 100
- 110
- 121
When an increase joins the base, later percentage increases act on the earlier gains as well.
Knowledge can sometimes help you learn further ideas because earlier concepts provide useful connections. That is an analogy to compounding. Learning has forgetting, plateaus and changing difficulty, so a smooth exponential curve serves as a simplified illustration with explicit assumptions.
Constant percentage growth is a modelling assumption and usually encounters changing rates or limits.
Opportunity cost
Choosing one use removes another use
Opportunity cost is the value of the best available alternative given up through a choice. Leah has one free hour. A language lesson, a walk and a film all compete for that same period.
One evening hour could support → A walk
One evening hour could support → Language lesson
One evening hour could support → A film
Leah ranks her possible evening
She most wants the lesson, followed by the walk, then the film. These preferences describe her actual alternatives tonight. Another evening could have a different order because her energy, needs or available opportunities have changed.
The lesson uses the available hour
Leah chooses the language lesson. The same hour can now serve that activity alone. Its cost includes the money or effort involved and the valued walk that can no longer happen during that period.
The forgone walk has its own value
Leah would have enjoyed fresh air and movement on the walk. Those benefits describe the opportunity she gave up. Adding the film’s value as well would count an option she could never have fitted into that hour together.
The cost depends on the real alternatives
Opportunity cost connects Leah’s choice to what the same time could have achieved. Its size depends on her priorities and feasible options. A rainy evening or an unavailable lesson would change the comparison.
Leah values learning → Language lesson
Leah values the alternative → A walk
One evening hour limits the choice → Leah
A choice uses resources that could have supported the next best available alternative.
Use comparable units when possible, but leave meaningful differences visible. An hour with a friend and an hour learning a skill may serve different goals. A small comparison table can show what each choice offers without pretending every value has an exact price.
The best alternative depends on your objectives and the options genuinely available to you.
Incentives
What is rewarded can change what people do
An incentive makes an action more or less attractive. A help desk praises staff for closing many calls. That reward can encourage speed, even when some callers still need help after the recorded call ends.
The dashboard celebrates a high count
Staff see recognition going to the people who close the most calls. The visible count signals which result matters to managers. Staff may reasonably give that count more attention when choosing how long to spend with each caller.
Ending the call becomes attractive
A complicated customer issue takes time. Closing that call quickly helps the recorded count, even if the underlying problem remains. The reward has created a reason to favour one visible action over a less visible outcome.
Resolution information changes the picture
The help desk also records whether the customer’s issue was resolved. Staff and managers can now see a consequence that the closure count missed. Recognition can be connected with useful outcomes as well as speed.
Rewards operate within a wider setting
The help-desk story shows one route from reward to behaviour. Staff also respond to values, workload, skills and management expectations. Incentives become easier to understand when the valued outcome and the actual human response are both visible.
Recognition for speed influences priorities → Help-desk staff
Help-desk staff chooses actions → Calls closed
Calls closed affects service → Customers still stuck
A reward changes behaviour through what people expect to gain from an action.
People care about different consequences. A public leaderboard may encourage one person and discourage another; a task may also be satisfying in itself. List several motives and observe the response so the incentive explanation remains connected to evidence about the people involved.
Incentives interact with values, norms, ability and circumstances, so behaviour remains context dependent.
Base rates
A relevant group gives a starting rate
A base rate describes how often something happens in a relevant group. In an illustrative record, ten of one hundred comparable deliveries arrived late. That history gives today’s delivery an initial context before its particular details are considered.
Comparable journeys make the history useful
The record concerns deliveries with similar routes and service conditions. That similarity matters because lateness rates can differ across settings. A history of unrelated deliveries would give a weaker starting point for understanding today’s parcel.
The starting estimate is ten in a hundred
Before further information, the recorded rate supplies a ten-per-cent baseline. It leaves ninety in a hundred comparable deliveries arriving on time. The number describes a group pattern while today’s parcel will have one actual outcome.
A weather alert adds specific evidence
Today’s route has a weather alert. That fact may change the estimate, depending on how similar alerts affected previous deliveries. The alert belongs alongside the baseline, with its importance determined by relevant evidence.
History and present details work together
The base rate prevents today’s vivid weather alert from becoming the whole picture. The current route also prevents an old average from becoming the whole answer. Their combination supports a more grounded judgement about the delivery.
Comparable deliveries supplies baseline → Estimated lateness chance
Route weather alert adds specific evidence → Estimated lateness chance
Today’s parcel defines the case → Estimated lateness chance
A relevant group frequency supplies a starting point that the particular case’s evidence can update.
Reference classes need care. Comparing local bicycle deliveries with international freight may give a misleading baseline because routes, processes and delays differ. Try a narrower group, check its sample size, and explain how uncertainty grows when fewer comparable cases are available.
A poorly chosen or outdated comparison group can give a misleading baseline.
Bayesian updating
New evidence can change a belief’s strength
Bayesian updating means changing confidence in explanations when evidence arrives. One of two bags is chosen equally at random. Seeing a red counter makes the bag containing more red counters a stronger explanation of that result.
The two bags begin equally likely
Bag A has eight red and two blue counters. Bag B has two red and eight blue. A random equal choice makes either bag equally likely before a counter is drawn, despite their different contents.
The selected counter is red
A red counter is drawn from the hidden chosen bag. Red was possible from either bag, so both explanations remain available. It was much easier to obtain red from Bag A because red fills most of it.
Bag A provides eight red possibilities → A red counter
The evidence favours Bag A
With equal starting chances, eight of the ten red possibilities across the bags come from A. The updated chance of A is therefore eight in ten under these assumptions. The observation changes confidence without guaranteeing the answer.
An update depends on the starting setup
The red counter shifted confidence because the bags had known contents and equal selection chances. Different contents or an unequal choice would change the answer. Bayesian reasoning connects a starting belief with how expected the evidence is.
Bag A explains red more often → Which bag was chosen
Bag B explains red less often → Which bag was chosen
A red counter updates the comparison → Which bag was chosen
Evidence changes confidence according to how expected it would be under each possible explanation.
With equal starting chances for two bags, one containing 80% red counters and the other 20%, a red counter makes the first bag 80% likely in this example. Different starting chances would produce a different answer even with the same observation.
An update is only as reliable as its starting assumptions and estimates of the evidence.
Regression towards the mean
An exceptional result can be followed by an ordinary one
Regression towards the mean describes a pattern in which an unusually extreme result is often followed by a less extreme one. A player who usually scores around six gets ten on an especially lucky throw.
- Usually around 6
- One score of 10
- Another throw follows
The high score includes some luck
The player’s skill helps the throw, and small variations also affect where it lands. On this occasion those variations are unusually favourable. Selecting the ten-point throw singles out an event with an exceptional combination of conditions.
Target-game player has established ability → Usual performance
Usual performance combines with favourable variation → Unusually high score
The next throw has fresh variation
The player throws again with similar ability, while the small accidental influences change. There is little reason for that unusually favourable combination to repeat. A score nearer the player’s usual range becomes a reasonable expectation.
Praise can be blamed for the lower score
Someone praises the ten, then the next score is lower. The sequence can create an impression that praise caused the decline. Regression provides another explanation because the first score was selected for being unusually high.
Extreme selection changes the expectation
The lesson concerns observations chosen for being exceptional. A later result can move either way, and skill can change. Regression matters because returning towards ordinary performance can happen without any intervention causing that return.
An unusually extreme result can include temporary influences that are less extreme next time.
Imagine each score as two stacked blocks: usual skill and temporary variation. Selecting the highest scores tends to select positive temporary blocks too. On another attempt those temporary influences change, so the selected group often moves closer to its typical level.
Regression is about expected patterns across selected cases. A particular excellent score can be followed by another excellent score. Check many comparable repetitions and consider genuine changes in skill or conditions before deciding how much of the change temporary variation can explain.
Regression patterns depend on how observations are selected and related, and individual cases can move either way.
Survivorship bias
The plants still visible tell an incomplete story
Survivorship bias occurs when visible successes stand in for a larger group whose missing cases matter. A gardener displays five thriving plants and praises a watering routine. Twenty plants using that routine have already been removed after dying.
The display looks like strong evidence
All five plants on the table look healthy. Someone seeing only that table might conclude that the routine works very well. The display supplies information about the survivors while hiding how they were selected.
The removed plants enter the record
The gardener explains that twenty other plants died and were discarded. Those cases used the same routine. Their disappearance from the display had removed important evidence about how the original collection actually fared.
The denominator changes the result
The original collection contained twenty-five plants, of which five thrived. That is five in twenty-five, or twenty per cent. Including the missing cases changes the apparent success without changing the five plants still on display.
Visibility is part of the explanation
The watering routine may interact with light, soil and other conditions. Survivorship bias identifies a more basic problem: the selected display hid part of the outcome. Understanding how cases became visible helps interpret any apparent success pattern.
Original twenty-five retained five plants → Five thriving plants
Visible successes need the missing failures and the original group to form a meaningful picture.
Selection can create associations. If only exceptionally tall or exceptionally skilled players join a team, shorter selected players may be unusually skilled. A relationship within the selected team can therefore differ from the relationship in everyone who originally tried out.
Recovering missing cases can be difficult, and incomplete records leave uncertainty about the original group.
Sampling
A smaller group stands in for a larger one
Sampling means studying part of a group to learn about the whole. A park survey asks ten runners about a longer running path. Their answers describe those runners, while other park users may have different needs.
The first survey finds enthusiastic runners
The survey takes place beside the running route early in the morning. Most people approached are there to run. Their support for a longer route makes sense given how the time and place selected the participants.
Other users enter at different times
Families visit the play area later, and some walkers use another entrance. Their experiences barely appear in the first survey. The gap comes from the way people were selected, even if every recorded answer was honest.
A wider selection changes the picture
The survey now includes different entrances, times and activities. Runners remain part of the evidence alongside other users. This broader selection offers more information about the park’s range of needs and competing priorities.
Ten runners share running needs → Park-use survey
Families and walkers share other needs → Park-use survey
Park-use survey informs the proposal → Proposed running path
More answers help only within a sound design
Asking more runners would enlarge the first sample while preserving its central gap. Sampling quality depends on who can be selected and who actually responds. The useful group is defined by the question the survey is meant to answer.
A sample’s value depends on how its members represent the larger group the question concerns.
A large sample can still be biased when its selection favours particular members of the population. Random samples also vary by chance. Larger suitable samples tend to reduce random sampling error, while selection bias requires attention to how the sample was obtained.
A large sample can remain systematically unrepresentative of the population you want to understand.
Correlation and causation
Two things can appear together
Correlation means two things tend to vary together. Causation concerns how one brings about another. Umbrellas and puddles often appear together on a street, creating a pattern whose explanation requires another fact: rain.
Wet days contain both observations
On rainy days more people open umbrellas and more puddles form. On dry days both are less common. Recording the two observations would reveal a relationship even before the underlying mechanism was understood.
Rain produces two separate effects
Rain falling on the street supplies water for puddles. The same rain gives people a reason to open umbrellas. These two pathways explain why umbrellas and puddles track each other without requiring an umbrella to create water.
Rain prompts protection → Open umbrellas
Rain supplies water → Street puddles
An indoor umbrella changes the picture
Someone opens an umbrella in a dry room and no puddle appears. The action separates umbrella opening from rain. This everyday comparison illustrates why changing one correlated feature may leave the other unchanged.
The mechanism gives the pattern meaning
The relationship between umbrellas and puddles becomes understandable through rain and its two effects. Correlation supplies a clue. Establishing causation requires evidence about a mechanism and the assumptions under which a comparison identifies that effect.
Rain changes behaviour → Open umbrellas
Rain changes the ground → Street puddles
A pattern can come from one thing causing another, a shared cause or other relationships that need investigation.
Imagine a reading club whose members score well on a vocabulary test. Reading might improve vocabulary; people with strong vocabulary might enjoy joining; education might influence both. Each arrow suggests information needed before estimating the club’s effect on a new member.
Random assignment can help create comparable groups when it is feasible and ethical. If participants are assigned to two teaching methods by chance, systematic starting differences are reduced on average. Measurement quality, sample size and later differences still require attention.
Even well designed studies support causal conclusions within stated assumptions and conditions.
Reversibility
Some choices are easier to undo
Reversibility concerns how easily a decision’s effects can be reversed. A shop can move shelves back after a weekend trial. Removing newly built walls would require more work, cost and perhaps permission.
The shop tries a new arrangement
Staff move shelves to create a wider browsing route. They can observe whether customers move comfortably and find the products. The experiment changes the shop while keeping its furniture available for another arrangement.
The old arrangement stays within reach
After the weekend the shelves can return to their previous positions. Some time and effort have been spent, yet the main physical change is easy to reverse. That lowers the cost of learning from the trial.
Earlier layout provides the starting plan → Movable shelves
Movable shelves tests a new layout → Weekend trial
Weekend trial can return to it → Earlier layout
Building walls creates a different commitment
Permanent walls would change the building itself. Undoing them could require skilled work and produce waste or disruption. The same aim of improving layout now carries a larger cost if the first arrangement proves unsuitable.
Reversal cost affects how much to learn first
The shelves permit learning through a modest trial. The walls justify more preparation because reversal would be harder. Reversibility comes in degrees: even an easy change can leave costs, lost time or effects on other people.
The cost of changing course helps determine how much evidence a decision needs before commitment.
Make a three-column table for cost, time and affected people. A choice can be reversible in one column and lasting in another. Returning a hired projector recovers the equipment cost, but the time spent setting it up has already been used.
A trial can turn an uncertain commitment into a learning sequence. State what you want to learn, how long the trial lasts and what result would justify continuing. Keeping the exit conditions visible makes the return path part of the decision design.
Reversal may recover only part of what a decision changes.
Asymmetric bets
The size of gains and losses can differ
An asymmetric bet has differently shaped possible losses and gains. In a fictional portfolio, each investment can lose its stake, while one successful investment could return many times that amount. The chances of success still matter.
Ten stakes create the starting exposure
The example contains ten investments of one hundred thousand pounds each. Together they cost one million pounds. That total establishes what this simplified portfolio can lose before any outcome has occurred.
Nine failures remove most individual stakes
Nine investments fail completely. Their losses add to nine hundred thousand pounds. The remaining investment must do enough to outweigh those losses before the portfolio as a whole can show a profit.
Ten fictional bets leaves one outcome → One returns £10m
One large result changes the whole portfolio
Suppose the remaining investment returns ten million pounds. Against the total one-million-pound starting cost, that leaves nine million pounds of profit before costs and tax. This is a worked outcome with an unusually successful winner.
One returns £10m returns ten million → £9m gross profit
A large possible payoff leaves the odds open
The arithmetic explains this particular fictional result. It supplies no chance of finding that winner. An attractive upside can coexist with poor expected value, unaffordable losses or correlated failures across the entire portfolio.
An asymmetric payoff can cap a loss while allowing a larger gain, but value also depends on odds and affordability.
The arithmetic depends on a 100× gross return, including the winning stake. The £9 million profit accounts for all ten original stakes. If all ten fail, the full £1 million is lost.
The chance of success, hidden obligations and shared risks affect the decision. A cap on the initial investment only caps the loss when the agreement creates no further liability.
A favourable looking payoff shape can still have poor expected value or an unaffordable loss.
The 1-3-1 framework
A problem arrives with considered options
The 1-3-1 framework describes one problem, three possible solutions and one recommended choice. A team member reports late customer orders and brings possible changes, making the discussion about the problem and the reasoning behind a response.
The delay is described concretely
Orders wait after picking because packing falls behind near closing time. Naming that stage narrows the problem. A general statement that deliveries are bad would leave the location and cause of the delay unclear.
Closing-time shortage creates a delay → Orders wait for packing
Orders wait for packing defines the decision → Manager
Three options expose different consequences
The team considers stocking more ready-packed items, adding a short packing shift and moving the daily order deadline earlier. Each changes a different part of the process, with different costs for staff, inventory and customers.
A short packing shift becomes the recommendation
The team member recommends a limited extra shift because the backlog occurs during one predictable hour. A small trial could reveal whether that capacity resolves the delay before the arrangement becomes a continuing commitment.
The framework structures shared reasoning
One problem, several options and a reasoned choice allow the manager to examine the thinking. The number three is a convention. The practical value comes from realistic alternatives, evidence about the cause and an affordable way to learn.
A clear problem and genuine alternatives make the reasons behind a recommendation easier to assess.
A decision includes the chosen action and the reasons supporting it. Explaining both lets other people question assumptions and learn from the outcome.
The coaching question “What are you going to do about it?” stands separately. The 1-3-1 framework adds a specific structure for preparing a decision.
The number three is a practical convention, and the quality of the options determines the value of the exercise.
Expected value
The average can be useful before a result
Expected value combines possible outcomes with their chances. A fictional fair-coin game gains three tokens on heads and loses one on tails. Its expected gain is one token per play, while an individual play has only one outcome.
The coin gives equal chances
A fair coin makes heads and tails equally likely in the example. Each possible outcome therefore receives the same weight in the calculation. Different chances would change the expected result even if the gains and losses stayed fixed.
Fair coin half the chance → Tails loses 1
The equally weighted outcomes combine
One heads outcome contributes a gain of three tokens. One tails outcome contributes a loss of one. Combining these equally weighted possibilities gives a total gain of two across two cases, averaging one token per case.
Heads gains 3 adds three → Average gain 1
One real play can still lose
The next toss may land tails and cost a token. That result fits the game’s stated chances. Expected value describes an average across possibilities or many comparable repeats; it does not promise a gain on each attempt.
The calculation inherits its assumptions
The fair coin and the three-token gain create the positive expected value. Unreliable probabilities, fees or changing rules would alter it. The average also leaves open whether someone can afford the losses occurring along the way.
Fair coin sets probabilities → Average gain 1
Heads gains 3 sets a gain → Average gain 1
Tails loses 1 sets a loss → Average gain 1
Expected value is a probability-weighted average, with individual outcomes and affordability requiring separate attention.
An average can conceal an unacceptable downside. A game with a rare enormous loss might have a positive expected value under estimated probabilities. Add the worst plausible loss and the number of attempts available to see whether the average answers your actual decision question.
Expected value depends on estimated probabilities and does not describe the result of every individual attempt.
Margin of safety
Spare room absorbs some surprises
A margin of safety is space for events to differ from a plan. A traveller’s station journey usually takes twenty minutes. Leaving extra time gives a queue or slow crossing room to happen without immediately missing the train.
The average journey leaves no spare time
The traveller leaves exactly twenty minutes before departure. If the trip matches its usual duration, arrival is just in time. Any additional delay now reaches the deadline because the plan has no room between expectation and requirement.
A small queue consumes the remaining room
A queue at the crossing adds several minutes. The journey still resembles an ordinary trip, yet those minutes can cause a missed departure. The consequence comes from combining variable travel with a fixed deadline.
Leaving earlier creates a buffer
On another trip the traveller allows additional time. The same small queue can now occur within the allowance. The extra time changes how much variation the plan can tolerate before the fixed departure becomes unreachable.
The buffer has a cost and a limit
Leaving earlier uses time that could serve something else, and a sufficiently large delay can still exceed the allowance. A useful margin depends on variation in the journey and the consequence of missing this particular train.
A margin of safety adds room between the expected demand and the point where a plan fails.
A buffer provides room for variation in demand or performance. Spare time, capacity and supplies can all serve this purpose. Backups sharing the same vulnerable dependency may fail together.
A buffer costs resources and can still be exceeded by sufficiently large or unexpected demands.
Trade-offs
Getting more can require giving something up
A trade-off occurs when improving one feature costs another valued feature. A small day bag is light but holds less. A larger bag carries extra clothing and adds weight for the person walking.
The small bag makes carrying easy
The walker can move comfortably with a small bag. It holds food and water for a short outing, while extra warm clothing may no longer fit. Its advantage and limitation come from the same compact size.
Small light bag limits spare capacity → Extra warm clothing
The larger bag creates additional capacity
The larger bag makes room for warm layers and more supplies. Filling it also increases the load the walker carries. The choice changes two linked consequences rather than improving a single feature in isolation.
The route changes which balance fits
A long exposed route makes spare clothing more valuable than it would be on a short sheltered walk. The physical bags stay the same while the circumstances change how much their different features matter.
The available options set the trade-off
The walker’s decision links weight, capacity and the actual route. A lighter material or a different route could alter the available choices. Trade-offs describe constraints within a situation that can itself change.
Small light bag offers lower weight → Day walker
Larger heavier bag offers capacity → Day walker
Long exposed route sets practical needs → Day walker
A trade-off becomes clear when the benefits and costs of the same choice appear together.
A design change can move the available boundary. A stronger lightweight material might improve capacity and weight together, while adding cost. Revisit all relevant dimensions after the change so an apparent free improvement does not conceal a new constraint elsewhere.
Trade-offs depend on the available options, and innovation can change the set of feasible choices.
Diminishing returns
Another helper can add less than the first
Diminishing returns means additional input eventually creates smaller additional gains while other conditions stay fixed. A second gardener can help water a small bed. A sixth may mostly wait because the available space has become crowded.
One gardener does all the work
One person waters the bed with a single can. There is room for someone else and another can. At this starting point, extra help could share the work without much interference between the gardeners.
Working space allows free movement → One gardener
A second gardener shares the bed
The second person waters another section using another can. Both can work at once. The added helper creates a noticeable improvement because the arrangement still has room and enough equipment for two active workers.
Working space allows parallel work → Two gardeners
Two gardeners cover separate sections → Small flowerbed
Six gardeners compete for the same space
The bed’s size stays fixed while more gardeners arrive. They begin blocking one another and waiting for access. Each extra person can add less useful work because space, rather than willingness, now constrains simultaneous activity.
The conditions explain the declining gain
Diminishing returns appears here because the bed and working area remain unchanged. Expanding the bed or reorganising the task could alter the result. The point of diminishing gain depends on which resources are held fixed.
An additional input can help less once another required resource becomes increasingly constrained.
Write total output as 4, 10, 14 and 16 pots finished with successive helpers in a fictional example. The extra outputs are 4, 6, 4 and 2. Diminishing returns describe the later falling additions, even while total output keeps increasing.
The fixed condition matters. Giving the group a larger workspace and more equipment changes the question. Compare adding labour in the original space with redesigning the whole process. This separates short-run congestion from the effects of growing an entire operation.
The point where extra returns diminish varies with the task and the conditions held fixed.
Scale
A bigger task changes its relationships
Scale concerns how a process changes as its size changes. Designing a community poster takes time whether ten copies or a hundred are printed. Printing more spreads that design effort while increasing paper use and distribution work.
The design takes one hour
The organiser prepares a poster with the event’s date and place. That initial work creates a file ready for printing. The same completed design can support a small batch or a much larger batch.
Poster design can supply this batch too → Hundred posters
More copies share the initial effort
With a hundred posters, the same design hour is spread over more copies than with ten. The design effort per copy therefore falls. This benefit arises from reusing work whose amount stayed roughly fixed.
Printing and delivery still grow
A hundred posters need more paper and more places to display them. The printer may become busy, and distributing the batch takes extra work. Larger scale introduces these demands alongside the saving in design effort per copy.
Scale changes which part matters most
At a small scale, design effort dominates the job. At a larger scale, printing capacity and distribution may become more important. Understanding scale means following which relationships change as the quantity grows.
As a process grows, shared costs, repeated costs and capacity limits can change at different rates.
Growth can create additional coordination. Four people can talk directly with relatively few links; a large group has many potential connections. Common instructions, roles and interfaces may become useful. Check whether the new coordination effort offsets the benefits of producing more.
Benefits and costs of scale differ across processes and can change at different sizes.
Emergence
A queue grows from individual actions
Emergence means a larger pattern forms through smaller interactions. At a serving counter, each arriving person joins behind the person already waiting. Those individual actions together create an ordered queue stretching away from the counter.
One person waits at the counter
The first customer arrives while the server is occupied. They stand in one place to wait. There is only one customer so far, with no long line and no need for anyone to direct a larger crowd.
The next person joins behind
A second customer sees the first and stands behind them. A third copies that relationship. Each person needs to notice who is already waiting; those nearby interactions are enough to extend the line in an orderly direction.
Many small choices create the larger pattern
As more customers arrive, each follows the same rule. A queue now has a length, an order and a waiting time. These describe the collection, produced by the way its individual members interact.
Changing the arrangement changes the pattern
The first customer is served and everyone advances. A second counter or a different joining rule could produce a different arrangement. Emergence names the larger pattern, while the local rule and service process explain how it forms.
A larger pattern can arise because many individuals repeatedly respond to nearby people and shared conditions.
Emergence connects levels of description. A single neuron has electrical and chemical activity; a network supports patterns involving many cells. Linking those levels requires evidence and models. The word emergence identifies the explanatory task while the detailed mechanism still needs investigation.
Labelling a pattern emergent leaves the work of explaining its particular mechanisms and limits.
Game theory
One choice depends on another person’s choice
Game theory studies situations where each person’s outcome depends partly on others’ decisions. Amina and Ben want to meet on a picnic walk. Choosing the same path matters more than either person’s preferred scenery.
Amina can choose → River path
Amina can choose → Hill path
Ben can choose → River path
Ben can choose → Hill path
Two attractive paths create a coordination problem
Amina prefers the river and Ben prefers the hill. Either route could lead to an enjoyable walk together. If they decide separately and follow their own preferences, they may spend the afternoon on different paths.
Different routes mean they miss each other
Amina takes the river path while Ben climbs the hill. Each chose a reasonable route individually. Their joint result is disappointing because the value of either choice depended on coordinating with the other person.
A message aligns the decisions
For the next walk they agree on the river path beforehand. The message changes each person’s expectation about the other. Their choices now fit together, making the shared picnic possible through coordinated action.
Agreed route sets the same plan → Ben
The outcome belongs to the combination
The picnic illustrates a coordination game: matching decisions benefits both people. Other situations involve different incentives, information or conflicting interests. Game theory makes the dependence between choices visible through a simplified account of the participants and outcomes.
Agreed route coordinates choice → Amina
Agreed route coordinates choice → Ben
Amina joins the meeting → River path
Ben joins the meeting → River path
Some decisions work only when their consequences are understood together with other people’s choices.
A repeated interaction can change incentives. Two people sharing equipment may care about tomorrow’s cooperation as well as today’s convenience. Add future consequences to the table and observe how the attractive choices can change under the assumed relationship and rules.
Real people have limited information, varied motives and changing behaviour that simplified games may omit.
Network effects
Another member can make the group more useful
A network effect occurs when additional participants change the value for existing users. A tool-sharing group becomes more useful to Amina when Ben joins with a drill. Each member brings possible resources and needs.
Amina with ladder offers a ladder → Tool-sharing group
Ben with drill offers a drill → Tool-sharing group
Amina alone has few exchanges available
Amina owns a ladder but sometimes needs a drill. An empty sharing group offers nobody to borrow from or lend to. The service’s usefulness depends partly on other people who can make complementary contributions.
Ben creates the first useful match
Ben joins with a drill and needs a ladder for painting. Amina can lend what he needs and borrow what he offers. The additional member makes a useful exchange possible for both existing and incoming participants.
Amina with ladder lends the ladder → Ben with drill
Ben with drill lends the drill → Amina with ladder
Borrowing opportunities meets a need → Amina with ladder
Leah adds further possible matches
Leah joins with a saw and her own borrowing needs. Amina and Ben now have another person to exchange with. New connections can add usefulness, provided the tools are available and the arrangements are reliable.
Amina with ladder shares a ladder → Tool-sharing group
Ben with drill shares a drill → Tool-sharing group
Leah with saw shares a saw → Tool-sharing group
The quality of participation matters
More members can also create waiting, poor information or unreliable returns. The network effect depends on useful connections, not merely a rising count. Availability and coordination determine whether the growing tool group actually serves its members.
Amina with ladder needs availability → Borrowing opportunities
Ben with drill needs reliable returns → Borrowing opportunities
Leah with saw needs clear arrangements → Borrowing opportunities
A new participant can change a service’s value for others through the connections and resources they add.
A new participant adds value when they enable useful exchanges. Some networks connect two sides, such as buyers and sellers, where growth on either side can benefit the other. Quality, congestion and search effort also affect the experience.
More participants can also create congestion, low-quality activity or additional costs.
Goodhart’s law
A useful count can change when it becomes a target
Goodhart’s law describes how a measure can lose usefulness when people concentrate on hitting it. A reading challenge rewards books finished. Choosing very short books raises the count while revealing little about how much participants understand.
Finished books originally supply a rough clue
The organiser uses books completed as a rough sign of reading activity. More completed books may often accompany more time spent reading. The count remains an indirect clue because book lengths and readers’ understanding differ.
A prize changes the reason for choosing books
Participants learn that the largest completion count wins. Someone selects several tiny books because they are quick to finish. The target changes book selection, which changes the relationship between the count and the original learning aim.
The higher number leaves learning uncertain
The participant finishes more titles, but the organiser still lacks evidence about understanding. The count has improved under the prize system while its meaning has become less clear. A description of what was learned would add different evidence.
The target and the real aim need comparing
The reading challenge makes the mechanism visible: a reward changes behaviour around a proxy measure. Some measures remain useful with careful design and review. Their value depends on whether the targeted behaviour still serves the underlying aim.
A measure can become less informative when people change their behaviour specifically to maximise it.
Measurement remains useful when its limits stay visible. A library might track visits, borrowing and user feedback together, then inspect unusual changes. Several imperfect indicators can reveal a fuller picture, although each can also influence behaviour once it becomes a target.
The effect is context dependent, and a target can remain useful with thoughtful design and review.
Testable claims
A claim becomes testable through what could happen
A testable claim predicts an observable difference that could fail to appear. One claim says a brighter lamp will help a reader recognise smaller print. Comparing two settings on the same page gives that prediction a concrete check.
The prediction identifies a particular change
The reader uses the same page at the same distance. The lamp’s brightness changes between settings. Holding the page and distance similar helps focus the comparison on the proposed influence of the light.
The brighter setting is tried
Under the brighter setting the reader attempts the smaller print again. Whether the letters can be read supplies a result. The procedure has meaning because a lack of improvement would conflict with the simple prediction.
The outcome can challenge the first explanation
Suppose the reader gains no improvement. That observation challenges this prediction under the tested conditions. Glare, measurement difficulty or an already sufficient starting brightness could require a more detailed explanation of what happened.
Adjustable lamp requires conditions to be considered → Brighter lamp helps
A test has a defined reach
The lamp comparison concerns one reader, one page and selected conditions. Its result contributes evidence while leaving broader claims open. Testability means a claim risks disagreement with observations; interpreting that disagreement still requires careful reasoning.
Adjustable lamp defines tested settings → Brighter lamp helps
Same printed page defines the material → Brighter lamp helps
Reader at fixed distance defines the tested case → Brighter lamp helps
A testable claim identifies observations that could count against it under stated conditions.
Make the prediction specific enough to inspect. “This helps somehow” leaves many escape routes. “Under these conditions, the average number of correctly read letters increases” identifies an outcome. You can then decide how much evidence would count as a meaningful change.
A failed prediction can arise from the main explanation or from supporting assumptions, such as a faulty instrument. Check both. Scientific testing uses repeated evidence, competing explanations and careful measurement to work out which part of the account needs revision.
A single test rarely settles a broad theory, and tests depend on measurement and supporting assumptions.
Optionality
A choice can be valuable before it is used
Optionality is the value of keeping a useful future choice available. A community group expects twenty or forty guests. An upgrade agreement lets it reserve a small room now and choose a larger one after bookings become clearer.
The guest count is still uncertain
The group needs to arrange a venue before everyone replies. A small room risks crowding if many attend, while a large room could cost more than necessary. The uncertainty affects both available choices.
A fee keeps the upgrade available
The venue offers a limited upgrade option for a fee. Paying it gives the group a right to move to the larger room under agreed conditions. The group buys flexibility during a period of uncertainty.
Upgrade option fee keeps upgrade available → Forty-person room
More replies make one option clearer
Bookings rise towards forty before the deadline. The preserved upgrade can now be used when its benefit is easier to judge. If replies had stayed low, the group could have retained the smaller room.
Bookings near forty makes extra space useful → Forty-person room
The option has conditions and a price
The fee would be spent even if the group never upgraded, and the choice expires at its deadline. Optionality is useful when later information can improve a consequential decision enough to justify the cost of keeping choices available.
Upgrade option fee buys flexibility → Forty-person room
Upgrade deadline limits availability → Forty-person room
Bookings become clearer informs use → Forty-person room
A useful option preserves a decision until better information arrives, subject to its price and conditions.
Keeping every option open can consume attention and delay worthwhile action. Give each option a purpose and a review date. When the relevant information arrives, exercising or releasing the option can free resources for the next decision and reduce unnecessary complexity.
Options can expire, cost resources and distract from taking useful action.
Assessing people and expectations
An expectation can outrun its evidence
Shahzad describes disappointment after hiring as sometimes beginning with an incomplete assessment. In this fictional example, a founder mistakes a confident interview for evidence of job performance. The later mismatch exposes a gap in the original judgement.
The interview reveals one kind of ability
The candidate speaks clearly about past work. That provides useful information about communication and experience. The founder has yet to observe the particular planning and delivery tasks that will occupy most of this role.
The job exposes an untested requirement
After joining, the employee struggles to organise several simultaneous deadlines. The founder feels disappointed, but the difficulty concerns a skill the selection process barely examined. Expectations had extended beyond what the interview actually demonstrated.
A work sample makes a future judgement more specific
For a later vacancy, a relevant sample can show how a candidate plans competing tasks. That evidence supports a narrower, more grounded judgement. Clear instructions, reasonable support and the conditions of the task also affect what performance reveals.
Acceptance and boundaries can coexist
Shahzad’s wider reflection concerns seeing people more accurately and setting expectations accordingly. An assessment remains provisional and specific to a role. People can learn, and working conditions can change, so evidence and responsibility belong on both sides of the relationship.
Actual role tasks sets a demand → Job candidate
Company founder provides conditions → Job candidate
Job candidate supplies performance evidence → Role expectation
A fair expectation is specific to the role and proportionate to the evidence supporting it.
Ideas from Shahzad Ali · My first assessment
Personal reflection supplied by Shahzad on 18 September 2026, edited for clarity in the linked Journal entry. The visual examples and connections are editorial explanations.
Shahzad’s hiring reflection places responsibility on his own initial assessment. A later problem can have several contributing causes, including unclear expectations, limited resources, unsuitable work and someone’s choices. A specific case needs specific evidence.
Work samples can show how someone performs a representative task. Assessment should distinguish abilities needed on entry from abilities a role will teach. Structured interviews use comparable questions and scoring criteria.
Acceptance means recognising the evidence available about a person. Boundaries can protect time, safety and trust. People remain responsible for their actions, and an assessment can be revised when their behaviour changes.
The height example in the Journal illustrates complaining about an already-visible characteristic. Job assessment here concerns the abilities and behaviours relevant to the work.
An assessment is provisional and specific to a role or relationship. People can develop, and the conditions around their performance matter.
How a setback can contribute to success
A failed attempt can influence the next one
A setback can contribute to later success when it reveals information that changes action. A designer loses a project after presenting an unclear proposal. Feedback helps her explain the next proposal more clearly, creating one possible contribution to its success.
The first proposal leaves the client confused
The designer describes many features without making the proposed result clear. The client chooses another supplier. Losing this project has a real cost, while the decision alone gives only limited information about exactly why she lost.
Feedback identifies a changeable weakness
The client says the proposal was difficult to follow. This points to a specific feature the designer can change. The useful information comes from the feedback, adding detail that the bare fact of rejection lacked.
The next proposal explains the result clearly
The designer reorganises a later proposal around the client’s problem, proposed outcome and costs. A new client accepts it. The revision may have contributed, alongside price, fit, timing and competing offers.
The benefit travels through what changed
The earlier failure did not guarantee the later success. The plausible chain includes feedback, interpretation and revision, with other causes still present. Learning can coexist with lasting losses, and successful stories omit many people whose setbacks ended differently.
A setback can contribute to success through information and changed action, alongside other causes and continuing costs.
A 2019 study examined scientists whose grant applications fell just above or below a funding threshold. Near misses increased the chance of leaving the funding system; among continuing researchers, later research impact was higher. The finding concerns that particular setting and study design.
Experiments by Eskreis-Winkler and Fishbach found that people sometimes learned less from failure feedback than success feedback. A setback can make information harder to attend to. Learning depends on how people process and use what happened.
An alternative-history question considers how the outcome might have differed without the earlier event. That can clarify a proposed cause. Evidence and assumptions are needed to assess the imagined alternative.
Success after a setback can have several causes. A useful lesson can coexist with lasting costs. The story of one successful person gives an incomplete picture of everyone who faced the same difficulty.
Why do I keep putting off something I want to do?
Delay can appear around valued creative work
Steven Pressfield calls the inner pushback against important creative work Resistance. Amina wants to draw a birthday card but keeps searching for ideal materials. His framework interprets that expanding preparation as one possible way of avoiding the actual drawing.
The paper search replaces the first mark
Amina spends her available drawing time comparing paper. Material quality can matter, yet the search keeps expanding while the card remains blank. The visible sequence is preparation followed by more preparation, with the intended work repeatedly postponed.
A new condition appears when paper is chosen
After choosing paper, Amina decides the pen must also be perfect. The threshold for starting has moved. Pressfield would interpret this recurring pattern as Resistance when it protects the person from the discomfort of making the work.
An imperfect drawing creates something to improve
Amina draws a simple cat on scrap paper. The result has uneven lines, but it now exists and can be revised. The action changes the problem from choosing ideal conditions to responding to an actual first attempt.
The author’s frame has a defined scope
Pressfield’s Resistance language highlights avoidance around valued work. Amina’s scene is one illustration. Fatigue, illness, caring duties and missing resources can also prevent action, so the label needs consideration alongside the person’s actual circumstances.
Pressfield’s Resistance describes a pull to postpone valued work, illustrated by preparation that continually delays a first attempt.
Ideas from Steven Pressfield · The War of Art
Resistance is Pressfield's concept. The birthday-card examples were created for Learn.
The Resistance label can help you notice a pattern. Keep the specific facts beside it: how much time is available, how rested you feel, what support you need and whether the task still matters. A label becomes useful when it helps you choose what to do next.
Resistance is the author's framing. Illness, fatigue, distress, caring duties and missing resources deserve attention in their own right.
How do I begin when the whole task feels huge?
A small beginning creates a concrete next problem
Pressfield’s work encourages beginning the work despite discomfort. A learning library can feel too large to start. Leah’s first explanation of a keyboard turns the broad ambition into a small piece that can be inspected and improved.
The imagined library contains too many decisions
Leah thinks about subjects, diagrams, navigation and hundreds of lessons at once. The whole project offers many possible starting points. With no actual explanation on the page, the next useful decision remains difficult to identify.
Large learning library contains many demands → Leah
One keyboard explanation offers one bounded part → Large learning library
One sentence establishes a first piece
Leah writes that a keyboard turns key presses into signals a computer can use. The sentence gives the lesson a starting claim. Its limitations are now visible in words instead of existing only as a vague imagined project.
The draft reveals a missing connection
Leah notices that the sentence leaves out how the signal becomes a letter on screen. That specific gap identifies another useful piece. The first attempt has produced information about what the explanation needs next.
Beginning and finishing require different amounts of work
The first sentence does little of the whole library, yet it creates a usable learning cycle. A small start can expose the next decision. Completing the project still requires sustained work, resources, skill and choices about its scope.
First clear sentence reveals missing detail → Key-to-screen connection
Key-to-screen connection guides further work → Next explanation piece
Next explanation piece adds one useful part → Large learning library
Large learning library contains more subjects → First clear sentence
A small first version changes a vague ambition into something that reveals its next useful improvement.
Ideas from Steven Pressfield · The War of Art and essays on starting
Inspired by Pressfield's discussion of beginning. The small-action example is Learn's application.
A small completed action exposes information that planning alone may leave hidden. A rough explanation can reveal an unclear word or a missing connection. That feedback helps define the next session.
Small starts also let you check whether the obstacle was unclear scope. If the first action still feels impossible, inspect the situation again. Perhaps you need information, rest, help or a different task. Keep the action small enough to learn from.
A small start can reveal the next step. Finishing a large project also requires time, skill, resources and repeated decisions.
How can I keep going when enthusiasm comes and goes?
Regular practice gives work a recurring place
Pressfield emphasises showing up for creative work. Ben wants to learn drawing and sketches the same kitchen cup twice a week. A recurring session makes another attempt more likely even when initial enthusiasm fluctuates.
The cup supplies a stable subject
Ben draws the cup’s outline and handle on the first page. Keeping the subject familiar reduces one decision at the next session. The main changing element becomes how he observes and represents the same object.
Dated pages make changes visible
Over several sessions Ben dates each drawing. Comparing the pages shows how the handle’s shape has changed and which difficulties persist. The record connects repeated effort with evidence about what practice is actually improving.
- First drawing
- Later dated pages
- Same observed cup
A new responsibility changes the available time
Ben takes on an evening caring duty and moves the drawing sessions. Keeping the original timetable would now compete with a real responsibility. Adjusting the arrangement allows practice to fit the circumstances that support or constrain it.
Consistency includes observing and adjusting
Ben’s repeated sessions provide opportunities to learn, while comparison makes the learning more specific. Pressfield’s practice theme values continued engagement. The useful arrangement also includes rest, feedback and changes when health or responsibilities affect what is feasible.
Twice-weekly session creates another attempt → Sketchbook
Sketchbook builds a record → Dated drawings
Dated drawings reveals improvement → Ben
Ben adjusts the practice → Twice-weekly session
Regular practice creates repeated opportunities to learn when its timing, feedback and demands fit the person’s life.
Ideas from Steven Pressfield · The War of Art; Turning Pro; essays on practice
Pressfield supplies the creative-practice framing. The weekly sketchbook example was created for Learn.
A recurring time and prepared materials reduce the decisions needed to begin. Comparing successive attempts supplies feedback about what is improving and what needs more practice. The routine can adapt when circumstances change.
Sustainable practice includes rest and changing circumstances. A schedule can be adjusted when health, responsibilities or priorities change.
How do I finish when I keep making tiny changes?
A version becomes finishable when its needs are clear
Finishing means bringing a version to a defined useful state. Leah keeps changing an album’s font while its essential names and dates await checking. Pressfield’s creative-work approach highlights the importance of completing and sharing work at an appropriate stage.
Changing the font keeps the album moving sideways
Leah has tried six fonts without resolving the album’s outstanding information. Each change feels like work, while the same completion issues remain. The visible activity leaves the version’s readiness almost unchanged.
The required checks narrow the remaining task
The family wants readable captions, accurate names and correct dates. Leah compares the draft with those needs and finds two dates to correct. The end point becomes clearer when the required result is explicit.
A private draft brings useful outside information
Leah sends the checked draft to her brother, who notices a missing name. His response creates a specific final edit. Sharing an appropriate version reveals something repeated font changes were unlikely to uncover.
Completion can lead into the next revision
Leah adds the name and finishes that version. The album may still improve later, while its current purpose has a clear stopping point. The checks required before sharing depend on the work’s privacy, audience and consequences.
A clear purpose and concrete completion requirements make the final checks more useful than endless small changes.
Ideas from Steven Pressfield · The War of Art and essays on finishing
Inspired by Pressfield's discussion of finishing. The album, review stages and checklist are Learn's examples.
In this example, finishing has two levels. A draft is ready for a family check; a later version is ready to print. Writing those stages down prevents the first review from carrying every expectation of the final object.
Before sharing, ask what you need to learn from the reader. You might ask whether an explanation is clear, whether facts are accurate or whether something important is missing. A focused question makes their response easier to use.
The checking needed depends on the stakes. Health, safety, privacy and financial claims need suitable review before publication or use.
What should I understand before making a plan?
A plan depends on the actual situation
Sun Tzu emphasised understanding conditions before acting in warfare. Our everyday analogy concerns a picnic. A field looks ideal in a photograph, while a visit reveals entrance steps and missing facilities that change whether it suits the group.
Visit to the site reveals practical conditions → Proposed field
The photograph highlights the attractive features
The organisers see open grass, trees and an inviting view. Those features support the idea of a pleasant picnic. The photograph leaves the entrance, toilets and routes from public transport largely outside the decision picture.
The visit reveals a barrier
At the site, the organisers find steps at the entrance and no toilets. Some guests would struggle with access. These conditions change the plan because they connect the location with the actual people expected to use it.
Another park fits the gathering better
A nearby park provides an accessible entrance and suitable facilities. The organisers change the location while keeping the shared aim of spending time outdoors. Better information changes the route towards the aim while preserving the desire to gather outdoors.
The lesson travels as an explicitly labelled analogy
Sun Tzu’s original context was warfare; the picnic has cooperative goals and different values. The transferable idea here is narrower: people, place and resources shape whether a plan works. The visit supplies information the first impression omitted.
A plan becomes more useful when it fits the actual people, place, resources and conditions involved.
Ideas from Sun Tzu · The Art of War, Chapter I
Translation consulted: Lionel Giles (1910), via Project Gutenberg. Everyday applications are Learn's analogies.
Some questions deserve attention earlier because the answer could change the whole plan. Access and permission may determine the location. Napkin colour can wait. Order the checks by how much they could change your decision.
Everyday cooperation has its own values and goals. This picnic is our analogy; the source book addresses warfare.
Is this disagreement worth the cost?
A disagreement can hide compatible needs
An objective is the result someone wants to achieve. In this everyday analogy to strategic thinking, two neighbours argue over a shared driveway. Their practical needs concern a delivery and a repair that occur at different times.
Neighbour Amina needs delivery access → Shared driveway
Neighbour Ben needs repair space → Shared driveway
Both neighbours initially ask for the driveway
Amina says the driveway must stay clear, while Ben says he needs to use it. Stated this broadly, the requests appear incompatible. The language leaves out the timing that determines whether both activities can actually fit.
The required times make the needs specific
Amina’s delivery arrives at nine, while Ben’s repair begins at eleven. Revealing those times changes the available arrangements. The driveway is still shared, yet its uses no longer require the same space at the same moment.
- Delivery at nine
- Different required times
- Repair at eleven
Both activities become possible under one arrangement
The neighbours keep the driveway clear for the delivery and use it for repairs later. The outcome serves their practical needs while avoiding a prolonged argument. Communication changes the arrangement by exposing information that the initial positions concealed.
The desired outcome gives conflict a useful test
This analogy draws attention to aims and the costs of pursuing them. It leaves serious conflicts over rights or safety with their own demands. The neighbours’ solution works because their specific objectives are compatible once the relevant details are visible.
Clarifying the desired result can reveal agreements that broad opposing positions conceal.
Ideas from Sun Tzu · The Art of War, Chapter III
Translation consulted: Lionel Giles (1910), via Project Gutenberg. Everyday applications are Learn's analogies.
A clear outcome makes alternatives comparable. Getting a delivery safely to the door, for example, may be possible through several arrangements. Each has different consequences for time, access and the people involved.
Also ask who carries each cost. An arrangement that works for you may block someone else's access or use their unpaid time. Making those effects visible can reveal a fairer agreement and a more durable result.
Some disputes involve rights, safety or serious unfairness and deserve firm action. The appropriate response depends on the situation.
What will this plan need to keep working?
An outcome depends on the resources behind it
A plan requires people, equipment, time and materials. In this everyday analogy to Sun Tzu’s attention to resources, an organiser promises twenty roast lunches, then discovers that the available oven cannot cook all the trays together.
The menu creates a specific equipment demand
The organiser lists roast meals and begins planning ingredients. Each meal requires oven space at the same serving time. The number of guests therefore creates a capacity requirement that a shopping list alone does not resolve.
Six trays reveal the gap
The oven holds six trays, leaving a much larger demand than one batch can meet. Repeated batches would take extra time and need a workable serving arrangement. The plan’s constraint becomes clear before ingredients are purchased.
A different menu uses available equipment
The organiser chooses dishes that can be prepared in larger pots. The guests can still receive lunch, while the process now fits the equipment more closely. The original desired outcome remains, and the production method changes.
Resources matter throughout the plan
The lunch also needs preparation, serving and cleaning after cooking. The strategic analogy highlights the full chain of requirements. Estimates can change with availability and timing, so a workable plan includes the stages beyond the appealing initial promise.
A feasible plan connects the desired outcome with the capacity and resources needed at every stage.
Ideas from Sun Tzu · The Art of War, Chapters II and VII
Translation consulted: Lionel Giles (1910), via Project Gutenberg. Everyday applications are Learn's analogies.
Then test one awkward possibility: a helper cancels or food takes longer to cook. Decide which adjustments are affordable, such as a simpler menu or more preparation time. Keep a small reserve that fits the actual uncertainty.
Planning uses estimates. Prices, availability and people's circumstances can change, so review the assumptions as the event approaches.
When should I change a plan that worked before?
A successful plan depends on its conditions
Adaptation means changing an approach when relevant conditions change. In an everyday analogy to Sun Tzu’s situational strategy, Leah’s familiar fifteen-minute journey becomes a twenty-five-minute diversion after a road closes. The old departure time now produces a different result.
The old route supports a reliable routine
Leah usually leaves fifteen minutes before she needs to arrive. The routine has worked under familiar traffic and route conditions. Its apparent reliability depends partly on those conditions continuing, even when they have become easy to overlook.
The closure adds a longer journey
A diversion now takes roughly twenty-five minutes. Leaving at the usual time would make Leah late. The relevant change is concrete: the journey needs ten additional minutes under the newly observed route conditions.
An earlier departure fits the new route
Leah leaves ten minutes earlier and observes the next journeys. The adjustment responds to the measured difference, while the result shows whether the new timing is sufficient. Extra variability could require another change to the allowance.
Reopening creates another reason to review
When the road reopens, the diversion-based routine may no longer be useful. The analogy highlights responsiveness to current conditions. Changing plans also uses attention and effort, so a relevant change and enough observation help distinguish adaptation from needless switching.
Road closed changes the route → Twenty-five-minute diversion
Twenty-five-minute diversion changes the timing → Leaving ten minutes earlier
Leaving ten minutes earlier continues until conditions change → Road reopens
Road reopens ends the original constraint → Road closed
A plan remains useful when its assumptions are checked against the conditions that now exist.
Ideas from Sun Tzu · The Art of War, Chapter VI
Translation consulted: Lionel Giles (1910), via Project Gutenberg. Everyday applications are Learn's analogies.
The aim is reaching the station in time. The method includes a route and a departure time. A road closure changes the available methods while the aim can remain useful.
Choose a simple signal for reviewing the plan. In this example, it is a road closure or a repeated late arrival. A visible signal helps you explain why you changed course and gives you something concrete to check afterwards.
Changing course also costs time and attention. Use relevant information and allow enough observation to judge whether an adjustment helped.