Private Calibration Protocol (PCP)
A multi-stage calibration protocol for identifying alignment between two parties without premature disclosure of private thresholds.
1. Definitions
Parties: Initiator (A) and Responder (B).
Threshold Range: A range [min, max] representing a party's "No Regret" zone.
Calibration Flexibility (σ): A percentage (e.g., 20%) defining the maximum allowed expansion of the range.
Target Value (T): The ideal value at the center of your range.
2. Protocol Lifecycle
Phase I: Initialization & Commitment
Initiator A defines a target value TA and selects a calibration flexibility (σ).
The system generates an initial range: [TA - σ, TA + σ].
This range is committed to the protocol.
Phase II: Secure Response
Responder B provides their target value TB.
The system generates their range using the same flexibility σ selected by the Initiator to ensure fairness.
Phase III: Overlap Verification (Round 1)
The protocol checks if the two ranges overlap.
Overlap Max = Minimum of both Maximums
If an overlap exists, the calibration is SUCCESSFUL. The final agreement value is the center point of the overlapping section.
Phase IV: Feasibility & Directional Disclosure
If the ranges do not overlap, the protocol checks for Bounded Feasibility. This determines if a deal is possible if both parties were to use their maximum flexibility.
Case: B is below A
Feasible if B's maximum possible range reaches A's minimum possible range.
Case: B is above A
Feasible if B's minimum possible range reaches A's maximum possible range.
If feasible, the protocol reveals a Directional Hint(e.g., "Party B is Above") to both parties. If not feasible, the protocol concludes as NOT ALIGNED and terminates without revealing any directional data.
3. Security Properties
Internal Comparison
Calculations are performed privately within the system. Ranges are never shared between parties. In a non-aligned state, no information (including the direction of the gap) is revealed.
Anti-Fishing Constraint
By locking the calibration flexibility (σ) globally, a Responder cannot use a disproportionately large range to "fish" for the Initiator's target.
Equitable Surplus Extraction
The use of overlap midpoints removes the first-mover disadvantage. Neither party can "win" the calibration through aggressive anchoring; the protocol extracts the shared surplus equally.
4. Reference Implementation
The complete, open-source implementation of the calibration engine and this application is available on GitHub.
View Source on GitHub