Study 10
Follower growth dynamics
An exploratory analysis of historical follower growth using simple growth models.
Data source: Wayback Machine follower count snapshots. Irregular time resolution.
Enter a handle and click Fetch data to load Wayback snapshots.
How to read this chart
Raw data
- Points show follower counts; the line connects them in time order.
- Linear: constant growth per day. Exponential: percentage growth. Logistic: S-curve with saturation.
- Models are descriptive aids, not predictions or causal explanations.
Measurement choices & limitations
- Points are follower counts at the snapshot dates returned by the archive for the entered handle. Spacing between points is irregular and depends on archive coverage.
- Models are fit to the loaded points only; they are not extrapolated beyond the last date.
- Wayback coverage is sparse; gaps and missing values are expected.
Pitfalls
- Do not infer causality. Different models may fit similarly; overfitting is a risk.
What this chart shows (in this dataset)
- In this dataset, the raw follower count rises over the period shown.
- Growth appears to slow toward the end of the series in many cases.
- The fitted curves diverge from each other in the later part of the chart; which sits closest to the points depends on the loaded data.
- The line connecting points often shows irregular steps; spacing between points varies over the period.
In simple terms
Growth often slows over time—early phases can look exponential, then level off as limits are approached. We test different models to see which shape best describes the observed pattern, not to predict the future.
This study does not claim that any model is “correct,” that growth will continue, or that any factor caused the observed pattern. It is exploratory and descriptive only.