Network Analysis Comparison

Al Karama, Dubai - Regular Grid Pattern

Select Analysis Method

Metric Betweenness HIGH for grids

Counts how often each street lies on shortest distance paths between all pairs of locations. High values = major through-routes.

Best for: Identifying main corridors, traffic flow prediction

Angular Choice (NACH) HIGH for grids

Like betweenness, but paths minimize turn angles instead of distance. Based on how people actually navigate (prefer fewer turns).

Best for: Pedestrian movement, wayfinding routes

Correlation with Metric: 0.46 (moderate - reveals different patterns)

Global Integration LOW for grids

How few turns needed to reach all other streets. In regular grids, all streets have similar global access.

Best for: Organic/irregular street patterns, historic centers

Local Integration (r=3) LOW for grids

How few turns to reach nearby streets (within 3 steps). Even more uniform in regular grids.

Best for: Areas with cul-de-sacs, dead-ends, irregular blocks

Degree Centrality LOW for grids

Simply counts number of connections at each intersection. Most intersections in a grid have 4 connections.

Best for: Quick connectivity overview, not detailed analysis

Effectiveness Summary

Method CV% Grid? Organic?
Metric Betweenness 216% ✓ Good ✓ Good
Angular Choice 278% ✓ Good ✓ Excellent
Global Integration 15% ✗ Poor ✓ Excellent
Local Integration 5% ✗ Poor ✓ Good
Degree Centrality 15% ✗ Poor ~ Medium
Key Finding: For Al Karama's regular grid, Betweenness metrics (both Metric and Angular) effectively differentiate streets, while Integration metrics show uniform values. This is expected - grids have uniform accessibility but variable through-traffic.

Legend

Low → High

📖 Glossary

CV (Coefficient of Variation): Std ÷ Mean × 100%. Higher = better street differentiation.

OD (Origin-Destination): Where trips start (hotels, offices) and end (shops, restaurants).

NACH: Normalized Angular Choice - space syntax term for angular betweenness.

Betweenness: How often a street lies on shortest paths. High = major through-route.

Integration: How easy to reach from everywhere. Uniform in regular grids.

Angular Distance: Path cost in turn angles, not meters. People prefer fewer turns.

Grid Pattern: Regular layout (like Al Karama). Integration metrics fail here.

Organic Pattern: Irregular historic layout. Integration excels here.

🎯 When to Use Each

OD-Weighted: Real pedestrian flows (needs POI data)

Angular Choice: Wayfinding routes (people minimize turns)

Metric Betweenness: Vehicle traffic, shortest paths

Integration: Only for organic/historic street patterns