I saw the charging curve for the first time in high school physics. The teacher had us wire a battery, a resistor, and a capacitor in series, close the switch, and read the voltage off a multimeter every two seconds. I plotted the points by hand on graph paper. They traced a clean exponential approach toward the supply voltage, fast at first, slowing as the capacitor filled. I labeled the axes, drew the asymptote, handed it in, and moved on to whatever came next that week.

I saw it again during my Elektroniker apprenticeship. The same curve, in a different context, attached to actual hardware that someone was paying me to build. Capacitors in power supplies, in filters, in timing circuits, all of them charging and discharging through that same shape. I stopped drawing it by hand and started recognizing it without trying.

I saw it a third time in the first semester of electrical engineering at university, in 2014. This time we derived it from Kirchhoff's voltage law. $\tau$ equals $RC$. After one $\tau$ you have crossed 63% of the gap. After five $\tau$, for practical purposes, you are full. The instructor said something I have never forgotten, which was that the boring experiment in front of us was one we would see again, for the rest of our careers, in more places than we expected. I assumed at the time he meant other circuits.

A year after that, I started trading. I am still trading. Somewhere in between learning how an order book worked and learning what a 1.5 risk-reward setup actually felt like in the chest, I noticed something quiet. The curves I was watching on charts were curves I had seen before. A breakout decelerating into resistance looked like a charging capacitor. The bleed-off of a panicked stock looked like a capacitor discharging. Range-bound chop looked like an LC circuit my apprenticeship instructor would have drawn on a whiteboard without thinking. Volatility after a news event rang and faded in exactly the shape of an RLC system going through its impulse response.

This is not a story about a single moment of recognition. It is a story about a slow accumulation, spread across about a decade, of seeing the same handful of shapes show up in domains that had nothing structurally to do with each other. By the time I had seen each curve enough times in each domain, the pattern was undeniable and a little bit obvious. The article you are reading is what happens when I finally sit down to write the pattern out instead of just noticing it again.

The thesis is the same in any case. The universe does not have an infinite alphabet for "how things change." It has maybe five letters. They keep getting reused, in domains that look nothing alike, because the underlying mathematics is short and the underlying mathematics is common. This article is a field guide to the five letters, and to what a chart looks like when read in their language.

What to notice: three completely different systems drawing the same shape at the same time. The point of the article in one image.

The grammar of change

Almost every smooth change in the natural world is produced by a rule about how fast something changes relative to its current state. Things fill faster when the gap to the target is large. Things drain slower when the level is low. Things oscillate when a restoring force pushes them back toward a center. These rules are short, usually a single line of arithmetic, and they happen to be so common that the curves they produce show up across electronics, biology, chemistry, economics, and (for our purposes) the price action of a market.

The curves I'll walk through are:

1. Exponential growth, the filling-up shape.
2. Exponential decay, the draining shape.
3. Bounded growth, the S-curve.
4. Oscillation, the sine wave.
5. Damped oscillation, the ringing-bell shape.

Plus a brief stop on linear, which is rarer in nature than you would think.

Every one of these has at least one electrical engineering archetype, one biological one, one physical one, and at least one trading one. The fact that the same handful of curves shows up everywhere is not coincidence. It is the consequence of a small set of underlying differential equations that nature seems to particularly like.

1. Exponential growth: the charge-up

In an electronics lab, the simplest possible experiment with a capacitor goes like this. You take a battery, a resistor, and a capacitor, wire them in series, and flip a switch. Before the switch closes, the capacitor holds no charge. After the switch closes, current flows, and the capacitor begins to fill.

It does not fill at a constant rate. It fills fast at the start, when the difference between its voltage and the supply voltage is large, and it fills slower as it approaches the supply, because the driving force is proportional to the gap that remains. The voltage on the capacitor as a function of time is:

$V(t) = V_0\,(1 - e^{-t/\tau})$

The parameter τ, called the time constant, is equal to the product of resistance and capacitance: $\tau = RC$. It sets the personality of the system. After one τ the capacitor has reached about 63% of its final voltage. After three τ it has reached about 95%. After five τ, for practical purposes, it is full.

What to play with: the τ slider. Watch how the same shape stretches or compresses in time, but never changes its essential character. The shape is in the equation; τ is just the clock.

The shape is greedy at the start because the system is being pulled by a large gap, and slows down as the gap shrinks. This single rule, speed of change proportional to distance from target, produces this curve every time, no matter what the underlying mechanism is.

You have seen this curve in places that have nothing to do with electronics:

• Bacteria in an unlimited nutrient bath grow this way during their early phase. Each cell divides at a roughly constant rate, the population doubles, doubles again, and the upward slope looks unmistakably like a charging cap viewed in log space.
• Viral adoption of a new app, before it saturates the addressable market, traces the same shape. Each new user introduces the product to friends, who become users themselves at a rate proportional to the size of the existing user base.
• A bull breakout in its first leg often looks like this. The earliest buyers move price the fastest, because there is the most "gap" left to close between the current price and where the new equilibrium lies. As price approaches that new equilibrium, the marginal next buyer faces less margin of return, and the slope decelerates.
• Filling your lungs. Airflow is fastest when they are empty and slows as you near full capacity, for the same reason as the capacitor: the gap pulling air in shrinks as the lungs fill.
• Compound interest in its early phase, before any meaningful base effect, is the same equation viewed discretely. Each interval grows the principal by a fraction proportional to itself.

The trader habit this curve should kill is the impulse to extrapolate the steepest part of a move forward in a straight line. Exponential growth toward a target looks explosive in its first third and then politely decelerates. The novice draws a ruler along the steep early part and projects a moonshot. The cap is indifferent to the ruler. The cap responds to the gap.

There is also a subtler habit it should create. When you see a sharp move that begins to decelerate, ask whether the deceleration is evidence the move is dying or evidence the move is approaching its target. These look identical for a few bars. The difference is what happens next. A move that is dying gives back some of its gain. A move that is approaching its target sits at the target and stops. The capacitor in the lab does the second thing.

2. Exponential decay: the drain

Now run the experiment backwards. The capacitor is fully charged. Disconnect it from the battery, short its terminals through the resistor, and watch the voltage fall:

$V(t) = V_0\,e^{-t/\tau}$

This is the mirror image of the first curve. Fast drop at the start, because the cap has the most energy to lose, tapering into a long, slow tail that mathematically never quite reaches zero. The same τ governs the shape, except now it controls how fast the system empties instead of how fast it fills.

What to notice: three completely different physical systems, all collapsing on the same curve, all governed by the same τ. The radioactive sample, the cooling coffee, and the price column are running the same math.

This curve is doing far more work in the universe than the growth curve. A non-exhaustive list of things that decay this way:

• Radioactive isotopes. Carbon-14, uranium-238, the americium-241 in your old smoke detector. The "half-life" you learned in school is just $\tau \cdot \ln 2$, about 0.69 times the time constant. Every atom in the sample has the same independent probability of decaying in the next second. The aggregate behavior of a large population of those atoms is exactly the discharging capacitor.
• Hot coffee cooling on the counter. Newton's law of cooling: the rate of heat loss is proportional to the temperature difference between the cup and the room. Same equation, different units.
• Drugs in your bloodstream. Caffeine's half-life is roughly five hours. The dose at 8am has lost half its punch by 1pm, three-quarters by 6pm, seven-eighths by 11pm. Which is exactly why a 4pm espresso, however small it feels at the time, is still doing 60% of its work at midnight.
• Memory. The Ebbinghaus forgetting curve, established in the 1880s with one of the loneliest research programs in psychology history (Ebbinghaus memorized lists of nonsense syllables on himself for years), found that the recall of newly learned material decays approximately exponentially. The fact you read at the start of this article is fading right now, on roughly that kind of curve.
• A struck bell, a plucked guitar string, the reverberation of a hand-clap in an empty room. All of them are mechanical energy in a resonant system bleeding out through friction with the surrounding air. All of them decay exponentially.
• Light through a medium. The Beer-Lambert law: each layer of a translucent material absorbs the same fraction of the light passing through it, so total intensity falls exponentially with depth. Same shape, again.
• Volatility after a news event. The VIX spikes on a shock and bleeds back toward its baseline. The decay looks remarkably like an RC discharge with a τ of roughly two to five trading days for first-order effects.
• Selling pressure after capitulation. The price chart of a stock blowing off the bottom of a panic move is a discharging capacitor on its side. The largest sellers exit first, when the chart is in its steepest section, because they have the most position to unload and the lowest patience. The long, slow tail is composed of late-stage holders giving up one by one.

For trading, this matters in one specific way. After a major shock (a Fed surprise, an earnings miss, a regime change in macro), the temptation is to treat the second day as "the day it's over." It is not. The volatility you see on day two is roughly half of what you saw on day one. The volatility on day five is roughly an eighth. Acting as if a shock has passed because the chart looks calmer is the same as acting as if a drug has worn off because you can't taste it anymore.

3. Bounded growth: the S-curve

What if growth has a ceiling?

In real biology, bacteria do not keep doubling forever. They eat the sugar in the dish, they fill the available space, their waste products poison the medium, and growth slows. The differential equation gets an extra term, and the curve gets a new shape.

$\frac{dx}{dt} = r\,x\,\left(1 - \frac{x}{K}\right)$

This is the logistic equation. $K$ is the carrying capacity, the ceiling. When $x$ is far below $K$, the bracket is close to 1 and growth looks exponential, indistinguishable from the curve in section 1. When $x$ approaches $K$, the bracket goes to zero and growth politely shuts off. The solution traces an S-curve: a slow start, an explosive middle, and a graceful settle into the ceiling.

In German-language physics class this is called beschränktes Wachstum, "bounded growth", and once you have a name for it, you start seeing it everywhere.

What to play with: the r slider. Watch how a faster growth rate doesn't change the final population, only the speed of getting there. The ceiling K is set by the system; the rate is set by the conditions inside it.

• Yeast in a sealed bottle of beer. They eat the sugar fast, then run out, then stop. Brewing is logistic growth that smells nice. The fact that beer has the alcohol content it does is largely because yeast becomes its own limiting factor; ethanol above twelve or thirteen percent kills the organism producing it. The S-curve is built into the chemistry.
• A new technology spreading across a population. First the early adopters (the steep middle of the S), then a long deceleration as the people who are going to adopt have already adopted. Television, smartphones, electric vehicles. Even social media platforms, with the wrinkle that the K for a network technology shifts upward over time as the network's value grows.
• A trader scaling into a position over multiple days. The first share is the easiest to buy. Each subsequent share runs into less of the trader's available capital and more of their available risk-appetite, so the position grows on an S-curve, even though the individual buying decisions feel unbounded at the start.
• An epidemic in its early stages, before interventions. Exponential at the start, S-shaped overall, as the pool of susceptible hosts is consumed. The SIR model is the logistic equation with one extra coupling.
• A stock that becomes "overowned." Every fund that wants to be in it is in it. The remaining buyers are scarcer, so the slope flattens even before any actual selling shows up. The chart of a crowded trade has a logistic flavor whether or not the analyst notices.

The reason this matters for charts is the inflection point. The S-curve has a precise moment where it stops accelerating and starts decelerating, mathematically at exactly half of K. Before the inflection, the chart looks like a moonshot. After it, the chart looks like a topping pattern. They are the same chart photographed at different times.

If you only learn one thing from this section: a parabolic move never resolves into the sky. It resolves into an S. The question is where the ceiling is, and the question is almost always answerable in retrospect and almost never in advance, which is one of the reasons trading is hard.

4. Linear: the boring honorable mention

Linear growth is, in nature, surprisingly rare in pure form. The universe seems to prefer feedback loops over indifference. A linear curve corresponds to a system whose rate of change has nothing to do with where it currently is, which is the absence of feedback, and feedback is what most physical and biological processes are made of.

You still see linear in a few places:

• An inductor with a constant voltage across it builds current linearly. $V = L\,dI/dt$ rearranges into $I = (V/L)\,t$, which is a straight line until something gives. Inductors do not have a "full" state the way capacitors do; they store energy in a magnetic field whose growth, in this idealized setup, is unbounded.
• Theta decay on a stock option close to expiration, on a quiet day, is close to linear over short windows. The textbook formula actually involves $\sqrt{t}$, which curves slowly, but at the scale a trader notices, the option premium just bleeds out steadily.
• A trader dollar-cost-averaging into an index. The position size grows by exactly the same amount each pay period regardless of price. Pure line, by deliberate design.

If you see a stretch of chart that looks truly linear, suspect a non-natural process. Algorithmic buying with a fixed shares-per-minute, scheduled accumulation by a corporate buyback program, a stop-loss being walked. Nature curves. Lines are usually a fingerprint of someone with a clipboard.

5. Oscillation: the inductor joins the party

Now add the second great player from intro electronics: the inductor.

Where a capacitor stores energy in an electric field between two metal plates, an inductor stores energy in a magnetic field around a coil of wire. The two devices are duals of each other. A capacitor resists changes in voltage; an inductor resists changes in current. Put them together in the same loop, with no resistor, and something beautiful happens.

The capacitor dumps its stored charge into the coil. As the current rises, the inductor builds a magnetic field. When the capacitor has fully discharged, the magnetic field carries all of the energy. The collapsing magnetic field then pushes the charge back into the capacitor, but in the opposite direction. The capacitor fills up the other way around. And then the whole thing happens again in reverse.

The system has two stores of energy that endlessly trade with each other. The math is:

$\frac{d^2 q}{dt^2} = -\omega^2\,q \quad\Rightarrow\quad q(t) = A\,\sin(\omega t + \phi)$

A sine wave. Energy sloshing back and forth between two storage modes, in principle forever, until something interrupts.

What to notice: the capacitor glow and the inductor glow are 90° out of phase. When one is full, the other is empty. The energy never disappears; it only changes form. The sine wave below is just the projection of that exchange onto one axis.

This is the most copied pattern in the universe. The same math applies to:

• A pendulum, where kinetic energy and gravitational potential energy trade places at every swing.
• A spring with a mass, where kinetic and elastic potential energy trade.
• A vibrating molecule, where atomic motion and bond-stretching energy trade.
• A heartbeat (approximately, as one harmonic of a more complex shape).
• Predator-prey populations. Lynx eat hares, lynx population grows, hares crash, lynx starve, hares recover. The Lotka-Volterra equations produce closed oscillations that, in phase space, look exactly like an LC circuit's voltage-current diagram. Two stores of biological energy, trading.
• Tidal patterns on a coastline, driven by the orbital coupling between Earth and Moon.
• Range-bound markets, where price ping-pongs between support and resistance until something breaks the symmetry.

The deep reason oscillation is so common is short. Any system with a restoring force, a force that pushes it back toward a center, and without enough friction to kill the motion, will oscillate. The center is whatever "equilibrium" means for that system. A pendulum's equilibrium is straight down. A market's equilibrium is, roughly, fair value as perceived by the marginal participants. The math is indifferent to the choice.

A trader looking at range-bound chop is looking at an LC circuit. The energy is just trading between two stores. The breakout, when it comes, requires a forcing function: news, a level break, a sudden change in conditions. Without one, the oscillation continues. People who try to predict a breakout direction while the system is still in pure oscillation are calling something that isn't structurally present yet.

One nuance worth flagging. Real oscillations almost never run forever, because real systems always have some friction. The pure LC circuit, with zero resistance, is a mathematical idealization. In the real world, every oscillator is at least slightly damped. Which brings us to the next curve.

6. Damped oscillation: friction enters the room

Now put the resistor back into the LC circuit. It becomes an RLC circuit, and the sine wave starts to lose amplitude with every swing.

$x(t) = A\,e^{-t/\tau}\,\sin(\omega t + \phi)$

Look closely at that equation. The exponential decay from section 2 is wrapping the sine wave from section 5. The envelope is a discharging capacitor. The wiggle inside is the LC oscillation. Two curves we already know, multiplied together.

What to play with: the two sliders, ω and τ. Notice that frequency and damping are independent. A system can ring fast and die fast, or ring slow and die slow, or any combination. The product ω·τ (roughly the quality factor Q) is the number of audible swings before the ring is gone.

The shape describes:

• An earthquake and its aftershocks. The biggest jolt first, then progressively smaller tremors over hours and days as the fault system bleeds off the elastic strain energy that was stored when the rocks first slipped. The Gutenberg-Richter law and the Omori law of aftershocks are quantitative versions of exactly this idea.
• A struck bell or a plucked guitar string. The amplitude dies away exponentially as friction in the medium converts kinetic energy to heat. A bell with very low internal friction (a high-quality bronze, say) rings for a long time, with a small decay constant. A wooden block dropped on a tile floor barely rings at all.
• A car suspension going over a pothole, ideally damped enough that you get one wobble, not a sustained bounce. Bad shocks let the car ring like a bell.
• Volatility after a major news event. Price oscillates around its new perceived value, with each swing slightly smaller than the last, as the market converges on a new equilibrium. The same envelope law applies to the VIX itself in the days following a regime change.
• A stock that just IPO'd or just had a binary catalyst resolved. Wild swings the first week, calmer the second, "normal" by the third. Each swing is a fraction of the previous one, set by how much friction the market has against the new information.
• Recovery from a market crash. Bounces, retests, lower lows, lower highs, until something settles.

The damping ratio $\zeta$ controls how the oscillation dies. Underdamped (small $\zeta$): rings for a long time, like a bell. Overdamped (large $\zeta$): no oscillation at all, just a slow approach to equilibrium, like a thick door closing on a hydraulic. Critically damped: returns to equilibrium as fast as possible without overshooting, which is what you want in a car shock absorber.

Markets can be in any of these regimes depending on the conditions. A liquid index after a small surprise tends to be critically damped: it absorbs the shock and re-anchors with minimal ringing. A thin small-cap after an earnings beat is often badly underdamped, ringing for weeks as participants disagree about the new fair value and the lack of liquidity prevents fast averaging. Knowing which damping regime the chart is in tells you, qualitatively, whether you should expect aftershocks.

Combinations and transitions

Real charts are almost never any one of these curves in pure form. They are combinations and transitions. The skill, once you have internalized the five letters, is reading the transitions.

The most common transition is exponential growth into bounded growth. A bull breakout starts as an exponential approach to a target (section 1). At some point, the system runs into its ceiling: an obvious resistance level, the limit of buying interest, the price at which selling becomes structurally heavy. At that moment the curve stops being a charging capacitor and becomes a logistic S. The first half of the chart and the second half are governed by different equations, even though the chart itself is continuous.

The interesting question is when. Sometimes the ceiling is visible from the start. Other times, the system reveals its ceiling only by hitting it. A trader watching the early-phase exponential cannot, in general, distinguish between an unbounded charge-up and an early-stage S-curve. The two shapes are identical until the inflection.

The second most common transition is from oscillation to damped oscillation. An LC circuit with even a tiny leak loses energy over time, and the pure sine wave becomes a slowly decaying sinusoid. Range-bound markets do the same thing. A symmetric oscillation between two levels is metastable. Eventually some resistance enters the system, the amplitude shrinks, and the market resolves into one direction.

A third pattern is decay following a spike, the impulse response. A market that absorbs a sharp news event behaves like a system that was suddenly displaced from equilibrium. The displacement decays back toward baseline. If the market is overdamped (a calm, liquid market) the price simply slides back, slowly. If it is underdamped (a thin, illiquid market) it rings around its new equilibrium for days, oscillating with decreasing amplitude. The same physics describes a tuning fork dropped on a table.

A fourth transition is less common but worth knowing: from one S-curve to another. A technology that saturates one market may jump into a new domain and start a fresh logistic curve from a new base. The trade-paper plot is the celebrated double-S, sometimes called innovation diffusion at the firm or industry level. When you see what looks like a stalling logistic that suddenly re-accelerates, suspect that a new K has been unlocked.

These transitions are where the analogy gets interesting. The pure curves of sections 1 through 6 are first-year-physics-class clean. Real systems combine them, transition between them, layer them. A market in a sustained trend with periodic volatility shocks is an exponential drift plus a damped oscillation, mathematically. A bull market in an emerging sector is a logistic with a slowly rising K. A capitulation panic recovering through smaller and smaller swings is a damped oscillation laid over an exponential decay.

The good news is that you do not need to write the equations. You only need to recognize which curves are in play in the current section of the chart, and notice when the regime changes. A regime change in the underlying equation is, in trading, almost always the most important thing on the screen.

The half-life lens

Every one of the curves we have looked at has a personality parameter. The capacitor has $\tau = RC$. The radioactive isotope has its half-life. The oscillator has its angular frequency. The damped system has both. These parameters set the clock of the system, the timescale at which it does its thing.

Asking "what's the half-life of this?" is the single most useful question I have stolen from physics and started applying to everything else.

• Caffeine: about 5 hours.
• An argument with someone you love, if you don't talk about it: about 6 hours of residual tension, longer if it goes unprocessed.
• A market shock from a Fed decision: 1–3 days for first-order effects, weeks for second-order repricing.
• Carbon-14: 5,730 years.
• Cesium-137 (the famous smoke-detector isotope, also the worry isotope of nuclear accidents): 30 years.
• A bear market in collective memory: about 18 months before retail starts forgetting why they were scared.
• A great song stuck in your head: about 30 minutes after distraction.
• The dopamine hit from checking your portfolio: roughly a minute, which is why you keep refreshing.
• A new piece of vocabulary in your head: about a day, then a week, then a month. Spaced repetition is engineered specifically to interrupt this decay.
• A first impression of a stranger: long. Months, sometimes years.

Half-life is a way to put very different systems on the same ruler. A trading-floor adrenaline spike and a uranium decay can be compared, not because they are physically similar, but because both decay exponentially with a characteristic time, and that characteristic time is the system's clock.

The strange and counterintuitive feature of exponential decay is that it is memoryless. The atom that has been sitting around for ten thousand years has the same probability of decaying in the next second as the atom that was created yesterday. The system has no history. Its future depends only on its present state.

Markets, by contrast, are not memoryless, and that is one of the places where the analogy starts to leak.

Where the analogy ends

I want to be honest about the limits of this whole framing, because the analogy is too useful to overstate.

A capacitor does not know it is a capacitor. It cannot decide to charge faster or slower depending on what other capacitors are doing. It has no narrative, no expectation, no opinion about its own voltage. Its behavior is determined entirely by the physical constants of its construction and the voltage across its terminals.

A market is structurally different. The participants in a market are agents with beliefs, narratives, expectations about what other participants will do. They watch each other and remember what they saw. They tell each other stories about what is happening, and they become afraid of patterns they have seen before, even when the underlying conditions are different. They become greedy in conditions that resemble past rallies, even when the conditions are not really the same.

This produces three behaviors that no electrical circuit exhibits.

First, reflexivity, in George Soros's sense. The participants' beliefs about the system change the system. If everyone expects a stock to break out, their expectation becomes part of the cause of the breakout. A capacitor's expectation about its own voltage is not part of the wiring.

Second, narrative dependence. The same chart, presented with two different stories about why the price is where it is, will produce two different next moves, because participants will trade according to the story. The same RC discharge does not change its time constant depending on what story you tell about the capacitor.

Third, non-stationarity. The "physical constants" of a market change over time. The volatility regime of 2008 is not the volatility regime of 2023. The participants of 1995 are not the participants of 2025. A τ that fits one decade does not fit the next. A capacitor's RC, by contrast, is a property of its physical construction and does not change unless the device is rebuilt.

So the analogy is not "markets are circuits." It is "markets, like circuits, exhibit certain canonical curves because the underlying processes producing those curves are mathematically simple, and the simple processes are common." The shapes carry over. The mechanism differs. The robustness of the curves is what makes the framing useful. The difference in mechanism is what stops it from being literally predictive.

There is a second, smaller honesty. Real charts are noisy. The capacitor charging curve in the lab is silky-clean. The bull breakout curve on a five-minute timeframe is fuzzed with hundreds of micro-shocks: large orders, sentiment shifts, news ticks. The underlying shape is present. The signal is wrapped in a thick layer of stochastic detail. If you stare at it too closely, you lose the shape. If you stare at it too zoomed-out, you lose the inflection points. The right scale of attention is part of the skill.

With those limits in mind, the framing still pays. The five curves are real. The fact that they show up across domains is real. The diagnostic value of recognizing which curve is in play, and what that says about the underlying forces, is real. The pieces of the analogy that work are the pieces I want.

A diagnostic

Here is the practical version of all this. When you look at any curve on a chart, run three questions.

The first question: which of the five letters is this? Look at the gross shape. Is it approaching a target? Falling toward zero? Settling into a ceiling? Oscillating around a center? Ringing with decreasing amplitude? Most charts will fit one of the five at any given moment. If none fits, the chart is probably in a transition between two of them, which is itself information.

The second question: what does that imply about the underlying force? This is the part most people skip. Each of the five curves is produced by a specific kind of rule:

• Exponential growth implies a force proportional to a gap.
• Exponential decay implies a force proportional to the current amount, draining it.
• Logistic growth implies a force proportional to a gap, in a system with a hard ceiling.
• Oscillation implies a restoring force pulling toward a center.
• Damped oscillation implies a restoring force in a system with friction.

Identifying the curve identifies, by deduction, what the underlying mechanism must look like. The mathematical shape of the mechanism is most of the useful information, even when the specifics remain unknown.

The third question: what is the time constant? What is τ for this particular system? How fast is it doing the thing? If it is exponential decay, how long until it is half gone? If it is damped oscillation, how many cycles before the ringing is inaudible? If it is logistic growth, where is the inflection?

Answering this third question is mostly a matter of eyeballing the curve and reading off the timescale. You do not need to fit a regression. You need to know roughly what the clock looks like.

If you can answer those three questions for the chart in front of you, you have done something most chart-readers do not do. You have stopped looking at the chart as a picture and started looking at it as the output of a process. The first orientation gets you head-and-shoulders and cup-and-handle. The second orientation gets you a piece of the mechanism.

The picture orientation is sometimes right and frequently wrong, with no reliable way to tell the difference. The mechanism orientation has more legitimate uncertainty, but it is more honest about what it knows. I would rather be honestly uncertain than confidently wrong. The five curves are a tool for being honestly uncertain.

Coda

The instructor in my first semester said the boring experiment in front of us was one we would see again, for the rest of our careers, in more places than we expected. I have come back to that sentence many times. He was right, and I think he was being modest about how far it went. He probably meant other circuits, or other physical systems. I do not think he meant trading charts, or a kettle on a counter, or a conversation between friends that wound down past midnight in a slow damped oscillation toward "we should call it a night."

The five curves do not stop at the edge of a textbook. They sit underneath most of the moving things in a life. The kettle reaches its boil on a logistic. The dopamine cycle of a notification climbs exponentially on the open and falls exponentially on the scroll. A career arc, looked at honestly, is usually a logistic with a slowly rising ceiling, sometimes a damped oscillation, occasionally an LC circuit between two long-running passions that trade energy across decades.

The point I would leave you with, after all of this, is small. Five curves. Five letters in the alphabet of how things change. They partition the world of moving things into a small number of recognizable shapes, and reading that world becomes faster and more humble at the same time. Faster because there are only five shapes. More humble because you can see how often you used to misread them.

I find that astonishing every time I think about it. The S&P 500, a yeast colony in a beer bottle, and the capacitor in the wall-charger next to me are running, at different speeds and with very different details, the same equations.

The universe, somehow, only knows a few curves.