Preferential Attachment, explained.
Preferential attachment is a growth mechanism where existing connectivity increases the probability of receiving a new connection.
Why it happens
Begin with two connected nodes. Each additional node attaches with one undirected edge, choosing an existing node in proportion to degree or uniformly. A fixed seed preserves each growth prefix. Inspect hub degree, total edges and the resulting graph up to forty nodes.
Preferential attachment is a growth mechanism where existing connectivity increases the probability of receiving a new connection.
Read the result
The growing prefix is seeded, and every new node adds exactly one edge. Degree advantages can accumulate without a built-in intrinsic quality difference.
A worked example
A network where connections attract connections
Grow the preferential network from two connected nodes to twenty and inspect the highest-degree node.
Switch to uniform attachment using the same node count, then compare the degree concentration rather than only the drawing's shape.
Increase to forty nodes, recognising that one small network cannot establish a distributional power-law claim.
OPTIONAL DEEPER DETAILGo deeper: inside the model
Inside this model
Begin with two connected nodes. Each additional node attaches with one undirected edge, choosing an existing node in proportion to degree or uniformly. A fixed seed preserves each growth prefix. Inspect hub degree, total edges and the resulting graph up to forty nodes.
Where this idea is useful
A practical use
Cumulative visibility can create unequal connection opportunities even without differences in intrinsic node quality.
A common misconception
“A hub must have been the most intrinsically valuable node.”
Early connections can increase later opportunities under preferential attachment. Real networks can also include quality, capacity and other mechanisms this model excludes.
What this explanation leaves out
- Forty nodes cannot establish a power-law fit. The model omits deletion, capacity, fitness and multiple links per new node. Preferential attachment is one possible mechanism, not a complete account of every unequal network.
Is every heavy-tailed network generated this way?
No. Different growth, fitness, copying and selection mechanisms can produce unequal networks. Observed concentration alone does not identify the generating process.
What advantage in a system compounds merely because it already exists?
Associated thinkers
Associations marked provisional are awaiting source review.
Further reading
Explore the original research or the teaching reference behind this experiment.