Simulation Results
Time for infection to spread globally in different network structures:
- Fully clustered network: 73 days
- 10% shortcuts (small-world): 26 days (almost 3x faster!)
- Fully random network: 25 days (barely faster than small-world)
The Critical Finding
With less than 1% of links being shortcuts, diseases spread almost as fast as if everyone were randomly connected.
Why This Matters
Global pandemics spread quickly because:
- Most transmission is local (clustered)
- But just a few long-distance connections (shortcuts) dramatically accelerate spread
- You don’t need many international travelers to create global pandemic This also explains:
- Why information goes viral rapidly on social media
- How ideas permeate seemingly separated communities quickly
- Why targeting hubs is so effective for control
The Exponential Effect
In the simulation, introducing shortcuts:
- Created exponential growth at the beginning
- Then nearly linear spread (can’t go faster than moving to each node)
- Final result: 3x faster than without shortcuts The ramp-up is what’s dangerous—by the time you notice widespread infection, it’s already too late to prevent global spread.
Public Health Implications
Prevention strategies:
- Target hubs for maximum impact (Thailand’s HIV success)
- Expect rapid spread in interconnected modern world
- Plan assuming small-world properties
- Few restrictions on shortcuts can have large effects Why blanket approaches often fail:
- Treating all connections equally ignores network structure
- Hubs and shortcuts have disproportionate impact
- Network-aware interventions are much more efficient
index We simulated if you can really reach anyone in 6 steps Six Degrees of Separation - Key Takeaways Network Shortcuts Network Hubs and Preferential Attachment