A logarithmic vs linear chart shows exactly the same price observations but assigns vertical distance differently. A linear axis gives equal space to equal absolute price changes; a logarithmic axis gives equal space to equal ratios, which are equivalent to equal percentage changes. Neither version changes the data, and neither provides predictive evidence. The appropriate choice depends on whether the research question concerns currency-unit moves, proportional performance, or both.
This distinction matters when a crypto price history spans several orders of magnitude. On a linear chart, high-price periods dominate the available height and early fluctuations may appear nearly flat. A log chart expands those low-price periods and compresses later absolute moves. That can improve proportional comparisons, but it can also make a dramatic increase look visually restrained. Responsible research therefore treats scale as a declared analytical setting, not a cosmetic preference.
How a logarithmic vs linear chart encodes price
Suppose price is represented by P. A linear axis places the point according to P itself. Moving from 10 to 20 occupies the same vertical distance as moving from 110 to 120 because both changes equal 10 currency units.
A logarithmic axis instead places the point according to log(P). The vertical distance between two positive prices is therefore determined by log(P2/P1). Moving from 10 to 20 has the same height as moving from 100 to 200 because both represent a multiplication by two, or a 100% gain. The chosen logarithm base affects tick labels, not the underlying relative positioning.
| Move | Absolute change | Percentage change | Emphasized by |
|---|---|---|---|
| 10 to 20 | +10 | +100% | Log scale relative to 100 to 200 |
| 110 to 120 | +10 | +9.09% | Linear scale relative to 10 to 20 |
| 100 to 200 | +100 | +100% | Log scale relative to 10 to 20 |
The linear view answers, “How many quote-currency units did price gain or lose?” The log view answers, “By what proportion did price change?” Calling one scale accurate and the other distorted misses the central point: each preserves one kind of visual equality while sacrificing another.
Current TradingView chart documentation, checked on 15 August 2026, distinguishes regular, percentage, indexed-to-100 and logarithmic modes. These are not interchangeable. Regular and logarithmic describe different axis mappings. A percentage view is generally a linear transformation relative to a selected starting value, so equal vertical distances represent equal percentage-point differences from that baseline—not necessarily equal returns between every adjacent pair of dates. Indexed-to-100 similarly rebases the first displayed value to 100, which can help compare series with different price units.
Worked example: one history, two visual stories
Consider a hypothetical token with six observations. The numbers are illustrative, use no real asset, and imply no forecast.
| Period | Price | Change | Period return |
|---|---|---|---|
| 0 | 10 | — | — |
| 1 | 20 | +10 | +100% |
| 2 | 30 | +10 | +50% |
| 3 | 60 | +30 | +100% |
| 4 | 120 | +60 | +100% |
| 5 | 60 | −60 | −50% |
On a linear price scale, the first two upward segments—from 10 to 20 and from 20 to 30—have equal height because each adds 10. The rise from 60 to 120 is six times as tall because it adds 60. Periods 0 to 2 may look relatively quiet beside the later movement, even though the first interval doubled.
On a log scale, the moves from 10 to 20, 30 to 60 and 60 to 120 have equal height: every one is a doubling. The 20-to-30 move is shorter because its ratio is 1.5. The fall from 120 to 60 has the same magnitude, in the opposite direction, as a doubling because halving and doubling are reciprocal proportional moves.
This also exposes a common return misconception. A 50% fall requires a 100% subsequent gain to regain the starting price. The apparent symmetry on a log axis reflects reciprocal ratios—120/60 equals 2 and 60/120 equals 0.5—not equal signed arithmetic percentages.
Before interpreting either shape, calculate the relevant return directly: (Pend/Pstart − 1) × 100%. For the full example, 60/10 − 1 equals 500%. That numerical result is unchanged by axis selection, chart height, zoom or screen size.
Which crypto chart scale fits the question?
Use a linear scale when absolute price levels and currency-unit moves are the subject. It can be appropriate for examining how far price sits from a fixed currency-denominated threshold, checking an execution range, or reading the absolute magnitude of a move over a limited price interval. It is also generally more immediately intelligible to broad audiences. The UK Office for National Statistics notes in its guidance on axes and gridlines that linear scales are more intuitive than non-linear ones, while allowing cropped axes for line charts when the crop is clearly presented and helps reveal a trend.
Use a log scale when ratios are the object of comparison. Typical questions include whether two growth phases had similar compounded rates, whether volatility was comparable in percentage rather than currency terms, or how an early phase relates to a later phase in a long-term crypto chart covering a very large price range. A straight line on a log-price chart corresponds mathematically to a constant compounded growth rate over evenly spaced time observations. It does not establish that the rate will continue.
Use both when the research has mixed objectives or will be communicated to readers with different levels of chart literacy. Side-by-side views can reveal that a late move is dominant in absolute units while an early move was equally or more important proportionally. Keep the asset, venue, pair, instrument, time range, interval, aggregation and chart dimensions identical so scale is the only changing variable.
For cross-asset comparisons, a percentage or indexed view may be clearer than placing unrelated nominal prices on one axis. The starting date then becomes a major assumption: changing it changes the baseline and the displayed relative path. Disclose that date and check total and subperiod returns numerically.
Why slopes and trend lines change
A visual slope is not an intrinsic property of a price series. It depends on the vertical transformation, horizontal time span, axis limits and chart aspect ratio. Switching from linear price to log price changes the variable being plotted, so a straight line fitted or drawn in one view represents a different relationship from a straight line in the other.
- A straight trend in linear price models a roughly constant absolute change per unit of time.
- A straight trend in log price models a roughly constant proportional or continuously compounded change per unit of time.
- A manually drawn support or resistance line may touch different observations after the scale changes because the geometry has changed.
- Two screenshots with different axis bounds can make identical fluctuations appear calmer or more violent.
Consequently, a trend line should be labelled with its scale and should not be silently transferred between scales. If researchers compare assets or periods, they should use consistent scale modes, axis policies and chart dimensions. The ONS guidance explicitly warns that different scales in comparable charts can mislead and recommends consistent axes.
This is also why “the curve looks exponential” is an incomplete conclusion. A linear chart can make sustained proportional growth curve sharply upward, while a log chart can make the same history look nearly straight. The transformation has not discovered a law of price behaviour; it has changed the coordinates used to display it.
What each view can reveal—and hide
A linear chart reveals absolute economic distance clearly. It shows that a move from 40,000 to 50,000 is much larger in quote-currency units than a move from 1,000 to 2,000. Yet it can hide the fact that the first move is 25% while the second is 100%. Over a history with a large range, early percentage volatility can be compressed into a nearly unreadable strip.
A log chart reveals proportional structure. Equal percentage moves are comparable across low- and high-price regimes, and early observations receive more visual room. Yet this compression can hide the practical size of late absolute changes. Readers who intuitively interpret height as currency units may also underestimate growth or risk.
Experimental evidence confirms that presentation affects judgment. In a peer-reviewed Experimental Economics study, Christoph Huber and Jürgen Huber varied the width of vertical scales while participants assessed hypothetical assets. Narrower axes increased perceived risk even where underlying volatility was the same, and inconsistent scaling made comparisons between assets harder. This study concerned wide versus narrow presentation rather than linear versus logarithmic transformation, but it demonstrates the broader point: axis design can influence financial judgments without changing the observations.
A separate series of experiments reported in Behavioral Science & Policy showed participants identical pandemic data on linear or logarithmic axes. Those shown logarithmic graphs made less accurate growth predictions and judged the threat as lower; explanatory education reduced but did not eliminate the effect. The subject was public-health data, not crypto, so applying the exact measured effects to markets would be an inference. The relevant lesson is narrower: unfamiliar axis transformations can materially alter how non-specialists read the same growth path.
The zero and non-positive-value limit
A standard positive logarithmic price axis cannot contain zero or a negative value. Log(0) is not finite, and the logarithm of a negative real number is not defined on the ordinary real-valued scale. The D3 logarithmic-scale documentation therefore requires a log domain to remain strictly positive or strictly negative and not cross zero.
Ordinary traded token prices are non-negative, but zero can still appear in faulty, missing, rounded or placeholder data. Negative values can arise in other financial series, such as profit, cash flow or a spread. Software may reject, omit, clip or transform such observations. A missing point on a log chart must not be interpreted as a genuine market gap until the raw series and the platform’s handling rules are checked.
Adding an arbitrary constant to every value so that zero becomes plottable changes all ratios and can distort the very proportional relationships the log scale is intended to show. Alternatives such as a symmetric-log transformation exist for signed data, but they are not standard log-price axes and require explicit explanation.
A practical scale-disclosure checklist
Before publishing, comparing or relying on a chart, record the following:
- Name the scale. State linear, logarithmic, percentage, indexed-to-100 or another transformation in the caption—not only in a hidden platform control.
- State the units and baseline. Identify the quote currency or unit. For percentage and indexed charts, give the baseline date and value.
- Preserve comparison settings. Use the same time window, interval, dimensions, axis mode and axis-limit policy when comparing panels.
- Inspect axis bounds. Note whether the range is automatic, fixed or cropped. A narrow range can amplify apparent variation.
- Check data eligibility. Look for zeros, negative values, missing observations and software-specific log handling.
- Recalculate key moves. Report starting price, ending price, absolute change and percentage return for every conclusion that depends on magnitude.
- Test the other view. If the history spans a large range, inspect both linear and log versions and explain any conclusion that survives only one presentation.
- Treat shape as description. A straight line, steep curve or apparent channel describes transformed historical coordinates; it is not evidence that the next observation will follow the same path.
The durable rule is simple: choose the scale that matches the question, disclose it prominently, and inspect numerical returns before drawing a conclusion from shape. A chart can organize historical evidence. Its axis cannot turn that history into a forecast.
Sources and further reading
- How to configure your Supercharts — TradingView, Accessed 2026-08-15
- Chart details: Axes and gridlines — Office for National Statistics, Accessed 2026-08-15
- Scale matters: risk perception, return expectations, and investment propensity under different scalings — Experimental Economics, Springer Nature, Published 2018-12-01; issue date 2019-03-15
- Graphs with logarithmic axes distort lay judgments — Behavioral Science & Policy, SAGE Journals, Published online 2020-10-01
- Logarithmic scales — D3 by Observable, Accessed 2026-08-15
Editorial note: This article is general educational information, not personalized financial, accounting, tax, or legal advice. Product capabilities and obligations can change; verify current facts and consult a qualified professional where needed.
