ClarusC64/cascade-f1-teamstrategy-coupling-rival-response-tyre-delta-gap-v0.1

F1 TeamStrategy–Coupling–RivalResponse–TyreDelta–Gap Cascade

A five-node coupling model for team-level strategic collapse in Formula 1.

This repository models how two-car strategy alignment, rival response speed, tyre delta, and track position gap interact to produce non-linear outcome cascades.

It shifts analysis from single-car optimisation to multi-agent interaction geometry.

---

What This Repo Demonstrates

You can:

• Score a two-car strategic state for cascade risk
• Identify interaction drivers of team-level instability
• Compare coordinated vs split strategies
• Estimate distance to strategic collapse boundary
• Export structured stability reports

The dataset is synthetic.
It demonstrates the geometry of multi-agent instability.

---

Core Five-Node Set

• car1_strategy
• car2_strategy
• rival_response_time
• tyre_delta
• track_position_gap

This is a coupled decision system.

The model captures how:

Aligned aggressive strategies → increased coupling risk
Fast rival response → compressed decision windows
Tyre delta mismatch → pace asymmetry
Small track position gaps → amplified timing consequences

Collapse emerges from interaction across both cars and rivals.

---

Prediction Target

label_cascade

• 0 = Stable team strategy state
• 1 = Team-level cascade region reached

A cascade represents:

Loss of both cars’ track position
Undercut chain amplification
Intra-team interference
Strategic dead-end state

---

Row Structure

Each row is a normalized race snapshot (0.0–1.0 scale).

car1_strategy
Higher values indicate aggressive or delayed commitment

car2_strategy
Higher values indicate similar commitment increasing coupling

rival_response_time
Higher values indicate faster competitive reaction

tyre_delta
Higher values indicate larger tyre performance mismatch

track_position_gap
Lower values indicate tighter pack compression

---

Use Cases

Two-Car Strategy Design

Evaluate whether to split or align pit windows.

Rival Reaction Analysis

Quantify risk under rapid competitor response.

Multi-Agent Boundary Mapping

Identify when team coupling creates instability.

What-If Strategic Exploration

Test small adjustments in timing or tyre choice.
Measure impact on cascade probability.

---

What Makes This Different

vs Single-Car Strategy Models

This models two-car interaction explicitly.

vs Deterministic Decision Trees

It evaluates instability surfaces rather than scripted branches.

vs Outcome Simulation Only

Simulation predicts finishing position.
This predicts collapse risk.

---

Example Output

Input Strategic State

{
  "car1_strategy": 0.44,
  "car2_strategy": 0.48,
  "rival_response_time": 0.52,
  "tyre_delta": 0.54,
  "track_position_gap": 0.46
}
Risk Assessment
{
  "cascade_probability": 0.41,
  "risk_band": "AMBER"
}
Boundary Interpretation

If rival_response_time decreases (faster response) and track_position_gap narrows:

Cascade probability can exceed 0.75.

Distance-to-RED can be estimated via L1 / L2 perturbation norms.

Batch Testing Capability

Compare multiple team strategies:

def batch_test(strategies: list) -> list:
    results = []
    for s in strategies:
        risk = score(s)
        results.append((s, risk))
    results.sort(key=lambda x: x[1]["cascade_probability"], reverse=True)
    return results

Applications:

• Compare split vs aligned pit windows
• Evaluate aggressive vs conservative team play
• Rank coordinated race plans by stability margin

Exportable Stability Reports

Structured reporting supports:

• Strategy room reviews
• Pre-race documentation
• Post-race analysis
• Governance and audit trail

Example concept:

def export_report(state, risk, boundary_configs, mitigations):
    report = {
        "configuration": state,
        "risk_assessment": risk,
        "boundary_configs": boundary_configs,
        "mitigations": mitigations
    }
    return report
Files

data/train.csv
Synthetic training data

data/tester.csv
Evaluation dataset

scorer.py
Outputs:

• accuracy
• precision
• recall
• f1
• confusion matrix

Evaluation

Run:

python scorer.py

Scope

This repository demonstrates five-node coupling geometry using synthetic data.

It does not represent calibrated team telemetry.

Small samples reveal structure.
Production-scale data determines operational exposure.

Production Direction

Production deployment enables:

• 50K–1M row telemetry-calibrated datasets
• Real-time team-level stability scoring
• Multi-car coordination monitoring
• Early warning before strategic cascade
• Integration into race engineering dashboards

License

MIT

Structural Note
This dataset identifies a measurable coupling pattern associated with systemic instability.
The sample demonstrates the geometry.
Production-scale data determines operational exposure.

Enterprise & Research Collaboration
Clarus develops production-scale coherence monitoring infrastructure for motorsport, healthcare, finance, infrastructure, and AI systems.

team@clarusinvariant.com

Instability is detectable.
Boundaries are measurable.