Riot and Protest Intensity Metrics
Insights for public safety and community awareness.
Riot and Protest Intensity Metrics
Insights for public safety and community awareness.
Understanding Riot Intensity
What is RPI-100?
The RPI-100, or Riot/Protest Intensity Index, is our innovative framework designed to evaluate the intensity of protests and riots. Utilizing a composite score from various metrics, it provides a real-time assessment that can guide public safety responses effectively, subverting bloviation.
Welcome to Not So Peaceful
Mission Statement
Not So Peaceful — Our structured Methodology for Evaluating Riot or Protest Intensity, using specified metrics and presented in a framework that allows for quantitative analysis, time-based tracking, and comparative evaluation.
Our RPI Framework
Framework Name: RPI-100 (Riot/Protest Intensity Index – Scale 0–100)
The RPI-100 scale is a composite score representing the intensity of a protest or riot, updated hourly based on real-time inputs across five weighted categories: Crowd Presence, Law Enforcement Engagement, Damage Level, Political Response, and Media Influence.
Metrics Categorized by Dimension
- Crowd Presence & Behavior (Weight: 25%)
- Law Enforcement Engagement (Weight: 25%)
- Property Damage (Weight: 25%)
- Political Response (Weight: 15%)
- Media & Public Influence (Weight: 10%)
Timeline Tracking
Use hourly intervals to track:
• Changes in crowd size
• Law enforcement escalation
• Property damage reports
• Political statements issued
• Social/media momentum
This provides a dynamic “riot curve” showing acceleration, peak, and de-escalation phases. Viewing the riot--through a Machine Learning lens--as a single, large scale rolling fight.
Composite Log Output
riot_data_log = {
"eve*****": "RPI_EVAL_0611",
"loc*****": "Downtown Sector 3",
"met***********": "normalized_scale_0_10",
"dat*********": "hourly",
"rec*****": [
{
"tim*******": "13:00",
"met*****": {
"cro************": 2,
"max*******": 10,
"nor*******************": "RPI-100"
}
.
Detailed Open Methodology
Metrics Categorized by Dimension - ML-driven Image Scoring
A. Crowd Presence & Behavior (Weight: 25%)
- Number of protesters (log scale)
- Number of protest hours (total and peak concurrent)
- Number of Block-Side hours denied to pedestrians (a square block has 4 block-sides)
- Number of hard objects thrown at law enforcement (stones, bricks, concrete, frozen water bottles)
- Number of looters (e.g., individuals exiting shops with objects in hand)
- Number of civilian injuries (wounds reported, trauma cases, ER admits, critical injuries, head trauma, non-lethal hits)
B. Law Enforcement Engagement (Weight: 25%)
- Number of law-enforcement responders deployed
- Foam rounds fired
- Real bullets fired
- Tear gas canisters deployed
- Stun grenades used
- Baton strikes recorded
- Number of arrests made
- Mounted Unit Dashboard Score
- Number of law enforcement injuries (wounds reported, trauma cases, ER admits, critical injuries, head trauma, non-lethal hits)
- Machine Vison: Before and After Riot Shield Strike Imagery (cracks, impact marks, scorches, warping, scratches, edge damage)
C. Property Damage (Weight: 25%)
- Civilian vehicles damaged
- Official/government vehicles damaged
- Civilian business damage ($ value)
- Minority business damage ($ value)
- Minority damage ratio (%)
- Small business damage ratio (%)
- By Name small business bankruptcies (projected)
- Government property damage ($ value)
- Number of broken windows
- Linear inches of damaged curbs
- Days of civilian business closure
- Days of government facility closure
- Number of graffiti tags
D. Political Response (Weight: 15%)
- Time delay (in hours) until political leader aligned with protesters calls for cessation
- Whether condemnation is explicit or ambiguous (binary scale: 0 or 1)
- Hours to Police Riot Recall-Alert / Rollout / On Scene
- Hours to National Guard Recall-Alert / Rollout / On Scene
- Hours to Federal Troops Recall-Alert / Rollout / On Scene
- Hours to 1807 Insurrection Act invocation
- Hours to first 1878 Posse Comitatus violation
E. Media & Public Influence (Weight: 10%)
- Social media mentions (supportive vs. critical)
- Number of unique media reports
- Rate of hourly increase or decrease in coverage
- Media Metics Mentions of the phrase "Mostly Peaceful"
F. Timeline Tracking
Use hourly intervals to track:
- Changes in crowd size
- Law enforcement escalation
- Property damage reports
- Political statements issued
- Social/media momentum
This provides a dynamic "riot curve" showing acceleration, peak, and de-escalation phases.
G. Scoring System (per metric)
Each metric gets normalized to a 0–10 scale, and is then weighted according to the category. Example:
- 10,000 protesters = 10
- 1,000 protesters = 6
- 100 protesters = 3
Property damage of:
- $10M = 10
- $1M–10M = 7
- <$100k = 3
Normalize all metrics in a scoring matrix, then apply weights to compute a final score out of 100.
H. Visual Input Methodology (Machine Vision Video Analysis)
- Real-time Object & Action Detection: Identifies protesters, law enforcement, vehicles, thrown objects, and specific actions (e.g., vandalism, clashes) using frame-by-frame analysis.
- Escalation Pattern Recognition: Tracks visual cues such as crowd density, movement speed, fire/smoke presence, and property damage to classify threat levels.
- Facial & Gesture Mapping: Detects aggression indicators via body language and facial expressions where visible, adjusting intensity scores accordingly.
- Temporal Layering: Synchronizes video input with timeline data to correlate visual events with RPI-100 spikes for forensic review and live forecasting.
I. Output Visualization
- Heat map: geographic spread of activity
- Hourly timeline chart: showing RPI score evolution
- Bubble chart: X-axis = Crowd size, Y-axis = Property damage, bubble size = Law enforcement use of force.
J. TDA - Tower Data Analytics for Riot Sector Intelligence
- IMEI-Based Presence Detection
Every device connects via an IMEI (International Mobile Equipment Identity). By aggregating tower handshake logs, analysts can determine:
- Time on Site: total minutes per IMEI within a protest zone
- Sector Residency: which cell sectors saw the most persistent device presence
- Ingress/Egress Patterns: timestamped entry and exit sequences for movement mapping
- Triangulated Geocoding
Using signal strength and timing from three or more towers, analysts triangulate approximate geolocations of each device per minute. This enables:
- Heatmaps of crowd density and dispersal
- Path tracking of high-mobility actors
- Identification of loitering or pattern anomalies
K. Synchronized Video Mapping
By integrating triangulated locations with citywide camera feeds and our Retail Store Subscribers:
- Estimated device positions can be overlaid onto live or archived video
- Visual IDs (clothing, body language, object possession) can be matched to anonymous IMEIs
- Real-time movement paths can be traced in a multi-angle video timeline replay
L. Cross-Riot Comparative IMEI Matching Synchroni
Devices that appear at multiple unrelated riot sites across time and geography are flagged for deeper analysis:
- Repeat presence suggests potential organization or professionalism
- Coupling IMEI logs with facial recognition and social media metadata improves attribution accuracy
- This method supports classification into tiers: civilian, agitator, organized actor, or outside agitator
M. Integrated Mobility Data Intake & Synchronization Hub
- Aggregates data from diverse sources: traffic cams, LPRs, rideshare APIs, and MetroCard swipes
- Supports real-time and batch ingestion with automated preprocessing pipelines
- Standardizes and tags input data by timestamp and location for cross-system alignment
- Enables timeline-based synchronization across multiple video camera angles
- Prepares unified datasets for downstream analysis, replay, and forensic mapping
- Facilitates identity correlation and mobility pattern reconstruction with privacy-aware redaction options
- Modular architecture supports easy integration of additional transport data feeds
N. Sample Py
riot_data_log = {
"eve*****": "RPI_EVAL_0611",
"loc*****": "Downtown Sector 3",
"met***********": "normalized_scale_0_10",
"dat*********": "hourly",
"rec*****": [
{
"tim*******": "13:00",
"met*****": {
"cro************": 2,
"max*******": 10,
"nor*******************": "RPI-100"
}
O. Mounted Unit Metrics Dashboard
- # of Protesters Within 10 Feet of Horse
- # of Protester-Horse Physical Contacts
- # of Aggressive Gestures Toward Horse (e.g., object thrown, shouting, waving signs)
- # of Fireworks Detonations Within 50 Feet of Horse
- # of De-Escalation Maneuvers by Mounted Unit
- # of Controlled Area Reclaims by Mounted Units (square footage)
- Horse Stress Indicators (e.g., elevated heart rate, erratic movement, documented distress signs)
- Duration of Continuous Deployment (minutes per horse)
- # of Officer-Horse Separation Incidents (loss of control, dismount)
- # of Medical Interventions (human or horse) from Mounted Interaction
- # of Video-Confirmed Trampling Incidents
- # of Mounted Warnings Issued (verbal or by loudspeaker)
- # of Horses Deployed vs Protester Density Ratio
- Protester Compliance Rate After Horse Advance (within 30 seconds)
- # of Mounted Unit Withdrawals Under Duress
- # of Riot Control Formations Involving Horses (e.g., wedge, line, sweep)
P. Interior Damage Assessment (IDA)
Retail Store Impact During Civil Unrest
- Inside graffiti tags: [number]
- Broken windows and doors: [number]
- Number of items removed: [number]
- Dollar amount of insurance claim: $[amount]
- Square feet with water damage: [number] sq ft
- Number of door entry team members involved: [number]
- Number of arrests on block-sides with a store entrance: [number]
- Internal smoke damage: 0 (No) / 1 (Yes)
- Internal water damage: 0 (No) / 1 (Yes)
- Internal graffiti damage: 0 (No) / 1 (Yes)
- Number of display cases and racks smashed: [number]
- Business license expired (checked quarterly): 0 (No) / 1 (Yes)
- Minority or Protected Class owner: 0 (No) / 1 (Yes)
- Willing to Interview: 0 (No) / 1 (Yes)
Q. Backtested Riot Metrics for Predictive Benchmarking
- Historical riot data across multiple U.S. cities and years
- Metrics normalized for comparison against event type, scale, and response intensity
- Includes National Guard deployment timing and force size
- Tracks use of federal troops under Insurrection Act conditions
- Flags Insurrection Act invocation risk indicators
- Measures distance to milestone events (e.g., elections, verdicts, public funerals)
- Enables average-based comparisons for early warning systems
- Supports regression analysis and anomaly detection across unrest phases
- Suitable for tactical planning, claims modeling, and civil stability forecasting
R. Riot Monitoring: Turn Chaos into Evidence, Insight & Protection
Subscriber Benefits for Retailers in Riot Risk Areas:
- 📹 Securely Upload Your Security Footage
Contribute to a protected, access-controlled video repository that syncs across time and location markers - 🧠 Leverage Our Riot Data Engine
Match your footage against known riot timelines, protester trajectories, and damage vectors—built from nationwide unrest events - 🛡 Strengthen Your Insurance Claims
Use our certified impact reports and synchronized event logs to bolster insurance documentation with verified third-party data - 🚨 Get Real-Time Alerts from Nearby Subscribers
Be notified immediately if a fellow retailer or entity in your area flags violence, looting, or suspicious group activity - 🛰 Geopositioned Offender Mapping
Match footage to external scenes and trace the movement of repeat actors across different retailers and days - 🔍 Cross-Retailer Offender Recognition
Identify known bad actors with AI-powered facial and clothing matching (with privacy controls and legal compliance) - 🧾 Access Prebuilt Insurance & Law Enforcement Reports
Instantly generate exportable PDFs summarizing damage events, entry paths, timestamped footage links, and synced testimony - 🔗 Integrate with Your In-House or Private Security Team
Feed alerts, offender IDs, and location data into your existing dashboard or dispatch system - 🧭 Peace-of-Mind Interface
No technical skills needed. Just log in, upload footage, and activate alerts—it’s built for retail teams, not engineers.
R. Volunteer Opportunities
We offer a range of volunteer opportunities for community members to get involved and make a difference. Join us and help us build a safer community.
S. Education and Training
Our staff receives ongoing training and education to ensure they are equipped with the knowledge and skills necessary to provide the highest level of service to our community.
