Examples · Scatter plot

Scatter plot examples

A scatter plot refuses to summarize: every reading is its own dot, and patterns — clusters, spread, outliers — emerge from the crowd. These examples plot values across ordered positions, the same label-and-value rows every other chart here uses.

When to use it

  • Seeing spread and outliers that an average would bury
  • Many individual readings where each one deserves its own mark
  • A first look at raw data, before deciding what to summarize

Not the right shape?

Exact category-to-category comparison — points floating in space are hard to rank. Bars do that job.

Real-world examples

The same readings, joined up

Connect the dots and the noise resolves into a direction — same rows, one chart type away (the chart maker can morph between the two).

Open in maker with this data

View the spec 
{
  "type": "line",
  "theme": "editorial",
  "title": "Daily support tickets, as a trend",
  "source": "Helpdesk, three weeks",
  "data": {
    "labels": [
      "D1",
      "D2",
      "D3",
      "D4",
      "D5",
      "D6",
      "D7",
      "D8",
      "D9",
      "D10",
      "D11",
      "D12",
      "D13",
      "D14",
      "D15"
    ],
    "series": [
      {
        "name": "Tickets",
        "values": [
          24,
          31,
          18,
          42,
          27,
          12,
          9,
          35,
          29,
          44,
          21,
          16,
          38,
          26,
          31
        ]
      }
    ]
  }
}

Two runs side by side

Two point series in one frame: where the clouds separate, the difference is real; where they mix, it probably isn't.

Open in maker with this data

View the spec 
{
  "type": "scatter",
  "theme": "mint",
  "title": "Page load time by build (ms)",
  "source": "Synthetic monitoring",
  "data": {
    "labels": [
      "R1",
      "R2",
      "R3",
      "R4",
      "R5",
      "R6",
      "R7",
      "R8",
      "R9",
      "R10"
    ],
    "series": [
      {
        "name": "Old build",
        "values": [
          420,
          390,
          445,
          410,
          460,
          430,
          415,
          470,
          425,
          440
        ]
      },
      {
        "name": "New build",
        "values": [
          310,
          340,
          295,
          330,
          305,
          350,
          315,
          300,
          345,
          320
        ]
      }
    ]
  }
}
Open this chart in a maker Scatter plot maker from scratch

Questions people ask

What is a scatter plot used for?

Showing every reading on its own, so you can judge spread, spot outliers, and see clusters. Averages and totals compress information; a scatter plot is the chart you use before compressing.

Can it draw a trend line through the points?

Not a fitted regression line — that's beyond this engine's honest scope today. You can draw a fixed reference with options.target (say, a threshold every point should stay under), or morph the points into a line chart to read the sequence directly.