# What Physics Actually Is > Physics is the craft of predicting the world with simple models: measurement, units, and the idea that a few rules explain enormous amounts. --- # What Physics Actually Is If physics ever felt like a wall of formulas someone expected you to memorize, you were handed the wrong end of it. Formulas are the *receipts*. The actual thing - the part that's been quietly running for four hundred years - is building the simplest possible model of a piece of the world and then checking whether reality agrees. That's it. That's the whole craft, and you can learn to think that way even if equations make you tense. ## How to read this Read the three phases in order. Each opens with something you already know in your body - a thrown ball, a checked answer, a coffee cooling - and builds the idea from there. There's no math you have to pre-load. When numbers show up, they're worked out in front of you, slowly, with the units kept visible so you can see *why* the answer is the answer. If equations make you nervous, there's a whole separate guide for that: [/guides/why-math-isnt-your-enemy](/guides/why-math-isnt-your-enemy). ## The phases 1. [What a model actually is](01-what-a-model-is.md) - physics as model-building, not memorizing; why a frictionless plane is a *deliberate* lie that tells the truth. 2. [Measurement and units (your built-in mistake detector)](02-measurement-and-units.md) - how to measure, why units travel with every number, and how they quietly catch your errors before anyone else does. 3. [The deepest idea: things that never change](03-conservation-and-the-loop.md) - conservation laws, and the scientific loop that turns a guess into knowledge. --- # What a Model Actually Is ## You already predict the world Toss your keys across the room to someone. You didn't solve an equation. But something in you *predicted* - the arc the keys would take, roughly where they'd land, how hard to throw so they'd reach without overshooting. You ran a model of the world, fast and wordless, and it was good enough. Physics is that same instinct, slowed down and written out so you can trust it past the range your gut covers. Your gut is great at catching keys. It is terrible at predicting where a satellite will be in six months, or how much your bridge will sag under traffic. The gut doesn't scale. A written-down model does. So before any formula, hold onto this: **physics is the practice of building a small description of part of the world that lets you predict what happens next.** Not memorizing what happened. *Predicting* what will. ## A model is a deliberate simplification Here's the part school usually skips. A physics model is not a perfect copy of reality. It's a *cartoon* of reality - and the cartooning is on purpose. The real world is hopelessly tangled. A rolling ball touches air, which pushes back. It touches the floor, which is slightly rough, slightly warm, very slightly soft. The ball itself flexes a hair. Track every one of those and you'd never finish a single calculation. So physicists do something that sounds like cheating and is actually the whole skill: **they throw away everything that doesn't matter much.** The famous "frictionless plane" you may have heard mocked? That's the move in its purest form. Nobody thinks friction is zero. The physicist is saying: *for this question, friction is small enough that pretending it's zero gives me an answer close enough to be useful - and far simpler to get.* ```text REAL situation MODEL (deliberate cartoon) ----------------------- -------------------------- ball: spinning, flexing, --> point with a mass warming, dimpled floor: rough, soft, sloped --> flat, frictionless surface air: pushing, swirling --> ignored result: messy, near-true result: simple, close enough ``` *What just happened:* we traded a perfectly accurate description we can't compute for a slightly wrong one we *can*. That trade is the engine of physics. The art is knowing what's safe to throw away. ## "Wrong but useful" is the goal, not the failure This reframes everything: a model being *wrong* is not a scandal. Every model is wrong, because every model leaves things out - that's what makes it a model and not the universe itself. The only question that matters is: **wrong by how much, and does that matter for what I'm doing?** Treating a thrown ball as a point with no air resistance predicts its landing spot to within a step or two. For playing catch, perfect. For a long-range artillery shell, that same model is dangerously off, because over a long flight the air you ignored adds up. Same model, same physics - *useful in one context, useless in another.* Knowing the difference is judgment, and judgment is the thing you're actually building when you learn physics. > A model isn't true or false. It's *appropriate* or *inappropriate* for a question. The skill is matching the cartoon to the job. ## A worked first cartoon Let's predict something. You drop a stone off a bridge and want to know roughly how long until it hits the water. The complete answer involves air resistance, the stone's shape, and the wind. The *useful* answer throws all of that away and keeps one fact: near Earth's surface, things speed up by about 9.8 metres per second, every second they fall. ```text Model: ignore air. Falling speed grows 9.8 m/s each second. Rule: distance fallen = 1/2 x 9.8 x (time in seconds)^2 Bridge height measured: about 20 metres. Try time = 2 seconds: 1/2 x 9.8 x (2 x 2) = 4.9 x 4 = 19.6 metres. 19.6 is almost exactly 20. So the stone hits at roughly 2 seconds. ``` *What just happened:* with one stripped-down rule and no air, we predicted a real event before it happened. The prediction is a little optimistic - real air would slow the stone, so it'd take a touch longer - but for "is it 2 seconds or 20 seconds?" the cartoon nails it. That gap between our 2 seconds and the true answer is real information, not a mistake to hide. **For builders:** this is exactly what a good first version of anything is. You don't model every edge case before shipping - you build the simplest thing that predicts the common path, measure where it's wrong, and add detail only where the wrongness costs you. A frictionless plane and a happy-path prototype are the same instinct. ## Why a few rules cover so much The reason this craft is worth your time: the cartoons compose. A handful of rules - how things fall, how they push on each other, how energy moves - recombine to describe a staggering range of the world. The rule that drops a stone in 2 seconds is *the same rule* that holds the Moon in orbit. You don't learn a new physics for every situation. You learn a few deep rules and a skill for cartooning, and between them they reach almost everywhere. That's the promise of this whole pillar. Not a pile of disconnected formulas to cram. A small toolkit, and the judgment to point it at the world. ```quiz [ { "q": "In physics, what is a 'model'?", "choices": [ "A perfect, complete description of a real situation", "A deliberate simplification that predicts well enough to be useful", "A formula you memorize without understanding", "A physical object built to scale" ], "answer": 1, "explain": "A model intentionally throws away unimportant details so the situation becomes simple enough to compute, while staying close enough to reality to be useful." }, { "q": "Why do physicists talk about a 'frictionless plane' when no such thing exists?", "choices": [ "They believe friction is genuinely zero", "It's a mistake that gets corrected in advanced courses", "Ignoring small friction makes the problem far simpler while staying close enough to true", "Frictionless surfaces are common in laboratories" ], "answer": 2, "explain": "It's a deliberate cartoon: when friction is small for the question at hand, pretending it's zero gives an answer that's close enough and much easier to get." }, { "q": "A model treating a thrown ball as a point with no air resistance predicts a backyard toss perfectly but fails for a long-range shell. What does this show?", "choices": [ "The model is broken and should never be used", "Models are appropriate or inappropriate for a question, not true or false", "Air resistance can always be safely ignored", "Longer distances require fewer assumptions" ], "answer": 1, "explain": "The same model is useful in one context and useless in another. Matching the simplification to the question is the core judgment of physics." } ] ``` --- # Measurement and Units (Your Built-In Mistake Detector) ## A number alone is a lie Someone tells you the trip is "5." Five what? Five minutes is a coffee run. Five hours is an afternoon gone. Five days is a holiday. The number `5` by itself carries almost no information - and worse, it *feels* like it does, which is how it fools you. This is the first hard rule of measurement, and it sounds too obvious to matter until it saves you: **a physical quantity is a number *and* a unit, always together, never apart.** "5 metres." "20 seconds." "9.8 metres per second per second." The unit isn't decoration tacked onto the number. It's half the meaning. Drop it and you've thrown away the half that tells you what kind of thing you're even talking about. A famous spacecraft was lost because one team worked in one set of units and another team assumed a different set, and nobody reconciled the two. The numbers matched. The units didn't. The number alone was a lie, and it cost a mission. ## Why we agree on a small set of units You could measure distance in your own footsteps and time in heartbeats, and for personal use that's fine. But the moment you want to *share* a prediction - hand your model to someone across the world and have them get the same answer - everyone needs the same rulers. So science settled on a shared set, the SI units. You don't need to memorize the whole table. A handful carries most of what you'll meet early: ```text Quantity SI unit Symbol "It's roughly..." -------- ------- ------ ---------------- length metre m a long stride mass kilogram kg a full water bottle time second s one heartbeat-ish ``` *What just happened:* we picked fixed, agreed reference sizes so that "3 metres" means the same thing in Lagos, Lima, and London. Everything else - speed, force, energy - gets *built* from these few. Speed is metres per second. That "per" is doing real work: it's a division, baked right into the unit. ## Units are math you can do on the labels Here's the quietly powerful part. Units obey arithmetic. You can multiply, divide, and cancel them exactly like numbers - and when you do, they tell you whether your calculation even makes sense *before* you trust the answer. Say you drive 120 kilometres in 2 hours and want your speed. ```text speed = distance / time = 120 km / 2 h = 60 km/h <- the units divided right alongside the numbers ``` *What just happened:* the `km` and `h` didn't vanish - they combined into `km/h`, which is exactly what a speed *should* be measured in. The units came out describing the right kind of thing, which is your first signal the calculation is sound. Now watch them catch a mistake. Suppose you fumble and *multiply* distance by time instead of dividing: ```text 120 km x 2 h = 240 km·h ``` *What just happened:* `km·h` - kilometre-hours - is not a unit of anything you'd ever want. Speed is never measured in kilometre-hours. The units came out nonsensical, and that nonsense is a flashing warning light: you used the wrong operation. You caught the error without knowing the right answer, only by reading the labels. ## The trick that checks your work for free This is called dimensional analysis, and it's the cheapest insurance in physics. Before you trust any result, ask: *do the units of my answer match the units the answer should have?* Remember the falling stone from phase 1. The rule was distance equals one-half times 9.8 times time-squared. Let's check it on the labels alone, ignoring the actual numbers: ```text The "9.8" is an acceleration: metres per second, per second -> m/s^2 Time squared is: s x s -> s^2 m/s^2 x s^2 = m x (s^2 / s^2) = m x 1 = m ^^^^^^^^^ the seconds cancel completely ``` *What just happened:* we multiplied the units and the seconds-squared cancelled the per-seconds-squared, leaving plain metres. A distance *should* come out in metres - and it did. We've now confirmed the formula is built correctly without computing a single digit. If those units had come out as `m/s` or `s`, we'd know the formula was wrong before wasting time plugging in numbers. > Make this a reflex: glance at the units of your answer before you believe it. Right units don't prove you're correct, but wrong units *prove* you're wrong - and that catch is free. **For builders:** units are types. `5` is an untyped number waiting to be misused; `5 metres` is a typed value that refuses to be added to `5 seconds`. Dimensional analysis is the compiler check that runs in your head - same reason a typed function signature catches a category of bugs before the code ever runs. Mixing metres and feet is the physics version of passing a string where an integer was expected. ## Precision is a measurement too When you measure something, you don't get infinite digits - you get as many as your instrument can resolve. A tape measure gives you millimetres, not nanometres. So when you write "the bridge is 20 metres," you're really saying "somewhere close to 20, give or take." Reporting a falling time as "2.0000001 seconds" from a rough height estimate is claiming a precision you never had. Good physics keeps your answer's confidence matched to your measurement's confidence. Carrying ten decimal places from a one-decimal-place ruler is its own kind of lie. ```quiz [ { "q": "Why is a physical quantity always a number paired with a unit?", "choices": [ "Units make equations look more professional", "The unit carries half the meaning; the number alone doesn't say what kind of thing it is", "It's a convention with no practical effect", "Only large numbers need units" ], "answer": 1, "explain": "A bare '5' could be minutes, metres, or kilograms. The unit tells you what the number measures, which is half the information." }, { "q": "You divide a distance in km by a time in h and the answer comes out in km/h. What has this told you?", "choices": [ "Nothing - units are only labels", "The answer must be exactly correct", "The units came out as a valid speed, a first signal the calculation is sound", "You should convert everything to metres first" ], "answer": 2, "explain": "Units follow the same arithmetic as numbers. Getting sensible units (km/h for a speed) is evidence your operation was right; nonsense units flag an error." }, { "q": "In dimensional analysis, what does it mean if your answer's units come out wrong?", "choices": [ "Nothing definitive - units can't reveal errors", "Your answer is definitely correct", "It proves the calculation is wrong, even before you compute the number", "You only need more decimal places" ], "answer": 2, "explain": "Right units don't prove correctness, but wrong units prove the calculation is broken - a free error check before you trust any digits." } ] ``` --- # The Deepest Idea: Things That Never Change ## The accountant's trick Imagine watching a chaotic scene - a break in pool, balls scattering everywhere, spinning off cushions, colliding, impossible to track. Now imagine someone hands you a single number that *doesn't change* through the whole mess. No matter how the balls fly, you add up that number before the break and after, and it's identical. You'd cling to that number. In a storm of moving parts, the one thing that stays put is gold. That's a conservation law, and it's the deepest idea in physics. Underneath the changing, swirling world, certain quantities **never get created or destroyed - only moved around or swapped between forms.** Find one, and you've found a fixed point in the chaos. You can predict the end of a process you couldn't possibly trace step by step, because whatever else happens, *this number must come out the same.* ## Energy is the famous one The headline conservation law: energy is never created or destroyed. It only changes costume. Lift a book onto a shelf and you've stored energy in its raised position - call it stored-up energy. Let it fall and that stored energy turns into motion energy, faster and faster. It smacks the floor and the motion energy doesn't vanish - it becomes a tiny bit of sound, a tiny bit of heat in the floor and the book. Add it all up at every instant and the total is constant. The energy was never gone. It was only ever moving between forms. ```text On the shelf: all stored-up (height) energy, no motion Halfway down: half stored, half motion (still adds to the same total) Just before floor: almost all motion, little stored After impact: motion energy -> sound + heat (total unchanged) ``` *What just happened:* we tracked a falling book without solving its motion second by second. We only insisted the *total* energy stay fixed and watched it change disguise. That's the power: conservation lets you skip the messy middle and still know how the story ends. ## A prediction you can almost feel Conservation laws make predictions that sound like magic until you see the bookkeeping. A swinging pendulum, pulled back and released, will climb the other side to *almost exactly* the height it started from - never higher, because climbing higher would mean ending with more energy than you put in, and there's nowhere for extra energy to come from. The law forbids it before you do any calculation. ```text Start: pulled to height H, let go. Bottom of swing: all that height-energy is now motion (fastest point). Other side: motion converts back to height -> rises to about H again. It can't exceed H. That would create energy from nothing. (Real pendulums creep slightly lower each swing - a little energy leaks to air and friction as heat. Still conserved, it only left the pendulum's account.) ``` *What just happened:* the law told us the *limit* of the swing with zero arithmetic. And the small real-world droop isn't a violation - the missing energy is conserved too, it's been quietly transferred to heat in the air and the pivot. The books always balance; you sometimes have to look in another account. ## How an idea becomes knowledge: the loop So where do laws like this come from, and how do we trust them? Not from authority. From a loop that physics runs over and over, and you can run it yourself: ```text observe something | v guess a simple rule (a model) | v predict what the rule says happens next | v test it against reality (measure!) | +--> matches? -> trust it more, push it harder | +--> doesn't? -> the model is wrong. fix it or drop it. (loop back to the top) ``` *What just happened:* this is the whole machine of science in one diagram. The non-negotiable arrow is the test. A guess that hasn't met reality is a story, not knowledge. And reality always wins - when a beautiful model disagrees with a careful measurement, the model is what changes. Conservation of energy earns its place not because it's elegant but because in two centuries of relentless testing, nobody has caught it breaking. > A model that *can't* be tested isn't wrong - it's worse than wrong, it's outside the loop entirely. The test is what separates physics from a nice-sounding story. **For builders:** you already run this loop. A bug report is an observation. Your hunch about the cause is a model. "If I'm right, changing this line fixes it" is a prediction. Running it is the test. When the bug survives your fix, you don't argue with the program - you update your model of what's happening. Debugging *is* the scientific method, pointed at a codebase instead of a pendulum. ## Where this pillar goes You now have the three load-bearing ideas. Physics is **model-building** - the deliberate, useful cartoon. Every quantity is a **number with a unit**, and the units quietly guard your work. And underneath the motion sit **conserved quantities** that never change, tested and re-tested by **the loop**. Everything else in this pillar - motion, forces, energy in depth, heat, light, the strange quantum rules - is these three instincts applied harder and to stranger corners of the world. You're not memorizing a wall of formulas. You're learning to think like the world's most patient, most careful accountant. And if the math still makes you flinch, go meet it on friendlier terms first: [/guides/why-math-isnt-your-enemy](/guides/why-math-isnt-your-enemy). ```quiz [ { "q": "What does it mean for a quantity to be 'conserved' in physics?", "choices": [ "It slowly decreases over time", "It is never created or destroyed, only moved around or changed in form", "It only applies to objects at rest", "It can be measured but not predicted" ], "answer": 1, "explain": "A conserved quantity stays constant overall - its total doesn't change, even as it swaps between forms or locations, giving you a fixed point in a changing system." }, { "q": "A real pendulum rises slightly lower on each swing. How does this fit conservation of energy?", "choices": [ "It violates the law, which only holds in theory", "Energy is destroyed by the swinging", "The missing energy leaked to heat in air and the pivot - still conserved, only transferred", "The pendulum gains energy from the air" ], "answer": 2, "explain": "Energy isn't lost, it's transferred out of the pendulum into heat. The total is still conserved; you only have to look in the other 'account'." }, { "q": "In the scientific loop, what happens when a model's prediction disagrees with a careful measurement?", "choices": [ "You trust the model and assume the measurement was wrong", "The model is wrong and must be fixed or dropped", "You ignore the result and keep the model", "It proves measurement is unreliable" ], "answer": 1, "explain": "Reality is the arbiter. When a model conflicts with a careful test, the model changes - that test is what turns a guess into knowledge." } ] ```