Bar charts excel at comparing quantities across discrete categories. The height/length of each bar makes differences immediately visible, unlike pie charts where comparing slice sizes is harder.

Line charts are ideal for time-series data. They connect data points to show continuity and progression over time, making trends, patterns, and anomalies easy to spot.

Pie charts become hard to read with many slices, and humans struggle to compare angles precisely. For many categories or when exact comparisons matter, bar charts are better.

When points trend upward left-to-right, it shows positive correlation - as one variable increases, the other tends to increase too. For example, delivery distance and delivery time.

In heatmaps, color intensity encodes the numeric value - typically darker colors mean higher values. This lets you quickly spot patterns across two dimensions, like order volume by hour and day.

Dual y-axes allow comparing metrics with different scales (e.g., orders in hundreds vs revenue in thousands) without one line appearing flat. Each metric gets its own scale.

400-800ms is the sweet spot - fast enough to feel responsive, slow enough for users to track changes. Shorter feels abrupt, longer feels sluggish and delays access to information.

Frequent animations become distracting and can obscure the data. For real-time dashboards, use chart.update('none') to skip animations, keeping the display calm and readable.