Granthara03.02.03Aspect ratio and the illusion of steepness

03.02.03 · Aspect ratio and the illusion of steepness

PreviouslyA logarithmic axis makes slope mean rate, which is why long charts need it.

The same move, at three different angles.

Two people look at the same share on the same day, over the same twelve sessions, with the same correct axis labels. One sees a calm drift. The other sees a climb. Neither chart is wrong, and neither person is careless.

An angle on a chart has no units. It is the data plus the shape of the window.

Updated

+8.0%
The move, in every frame on this page
Part 03 · Axis, Scale, and Visual Deception
  1. 03.02.01 Linear vs logarithmic scale
  2. 03.02.02 Why log scale is mandatory for long-term charts
  3. 03.02.03 Aspect ratio and the illusion of steepness
  4. 03.02.04 Auto-scaling traps
0 of 4 read
Learn

Steepness is not in the data

A chart converts price into height and time into width. How steep a line looks is the first divided by the second: rupees per centimetre of height, against sessions per centimetre of width.

Neither number belongs to the market. Both are decided by the size and shape of the box the chart is drawn inside — the aspect ratio. Make the box shorter and wider and every line flattens. Make it narrower and taller and every line rears up.

So a line's angle is not a measurement. It is a measurement wearing a costume. Two charts can disagree completely about how dramatic a move looks while agreeing exactly on every price in it.

See it

Twelve sessions, three frames

An illustrative share from ₹250 to ₹270. The prices and the axis labels are identical in all three. Only the box changes.

The caption under each frame carries the actual move. Check it against what the shape made you expect.

Learn

Where you meet this without anyone lying

Your phone and your laptop. A phone chart is tall and narrow; the same chart on a wide screen is short and long. The identical series looks more violent in your hand than it does at your desk, all day, every day.

A screenshot. Cropping a chart to fit a post changes its aspect ratio. So does resizing it into a report column or a slide. A person forwarding a chart to make a point does not have to alter a single price to make the point look stronger.

Zooming. Dragging the price axis on most platforms stretches the vertical scale without touching the horizontal one. That is a deliberate feature — it is how you inspect a small range — and it also means the shape you end up staring at is one you built.

The same twelve sessions, measured rather than looked at
First price ₹250, last price ₹270
+₹20 · +8.0%
Highest ₹270, lowest ₹249
range ₹21
Wide frame
the reference angle
Tall frame, same data
about 7.5 times steeper on screen
+8.0%in every frame, measured from the axis
0prices changed by reshaping the box
Try it

An illustrative share moved from ₹250 to ₹270 across those twelve sessions. Drag to the percentage change.

20%

Learn

What this costs anything built on angles

If steepness comes partly from the window, then any rule stated as an angle describes your screen rather than the market. A line 'at 45 degrees' is 45 degrees only at one particular price-to-time ratio, and becomes something else the moment the chart is resized.

Platforms know this. Several offer a setting that locks the price scale to the bar spacing, so zooming and resizing no longer reshape the chart. If an angle matters to how you read something, that setting is what makes the angle mean anything twice.

The same caution applies to comparing two charts side by side. Two series can only be compared by shape if they share a scale type, a time span and a frame shape. Otherwise compare the numbers — the percentage change, the range, the high and the low — and let the pictures be illustrations.

None of this makes a chart untrustworthy. A chart is a drawing, and a drawing has a size. The error is treating the drawing's proportions as though they were evidence.

Try it

Someone sends you a chart with no axis labels visible, showing a line climbing steeply from the bottom left to the top right. What can you conclude about the move?

The three seconds that fix this

Before reacting to a chart's shape, read the first and last axis labels and the length of the time axis. Those two facts convert a shape into a measurement: a range and a span. A move you can state as "8% over twelve sessions" cannot be made bigger by a narrower window, and a move nobody can state that way was never evidence of anything.

The frame is one presentation choice. The other is the range of prices the axis covers — and on most platforms, that one is not a choice you made. It is made for you, silently, every time you scroll.

Remember

How steep a line looks is the data divided by the shape of the window, so the same +8% can read as a drift or a climb — which makes the axis labels and the time span, not the angle, the only things worth reacting to.

Everything this chapter has told you, in order

  1. 03.02.01A linear axis measures rupees and a logarithmic axis measures percentages, so the scale decides which of two moves looks bigger.
  2. 03.02.02On a logarithmic axis the slope of a line is the rate of return, which is why a long chart needs it — and why it cannot draw a zero.
  3. 03.02.03How steep a line looks is the data divided by the shape of the window, so an angle is a measurement wearing a costume.
Next
03.02.04 — Auto-scaling traps

Next: the axis that redraws itself every time you scroll.