The mockingjay contribution score, in numbers

On standard peer-to-peer platforms, reputation is an inflationary asset. An Airbnb host with 400 five-star reviews from 2019 can let a property decay for three years and still present a 4.98 rating to an unsuspecting guest. The metric measures historical volume, not current reliability. WEVONE’s internal reputation engine—code-named the Mockingjay Contribution Score—operates on an entirely different calculus: trust as a decaying, multi-category velocity vector.

Instead of averaging scalar ratings from 1 to 5, the Mockingjay algorithm synthesizes raw transactional data across WEVONE's active universes (Tutus for second-hand items, Nest for short-term stays, Mission for local services, Pilote for co-transport). The output is a dynamic integer bounded between 0 and 1,000, updated asynchronously upon every escrow release or dispute resolution.

The Vector Architecture

The primary formula hinges on four distinct inputs: transactional escrow volume ($V_e$), category multiplier ($M_c$), time-decay factor ($D_t$), and the dispute coefficient ($C_d$).

The decay equation follows a strict half-life parameter where unmaintained score drops over time:

$$D_t = e^{-\lambda t} \quad \text{where } \lambda = \frac{\ln(2)}{90 \text{ days}}$$

If a user accumulates 100 points through successful services in January and performs zero actions through April, their active weight drops to precisely 50 points. Reputation is not a trophy room; it is a live electric current that requires continuous maintenance.

| Input Variable | Metric / Weight | Platform Function | | :--- | :--- | :--- | | Escrow Volume ($V_e$) | $\log_{10}(1 + \text{EUR settled})$ | Sub-linear scaling to stop whale bias | | Category Multiplier ($M_c$) | 1.0x to 1.4x | Adjusts for transaction complexity | | Half-life ($D_t$) | 90 days ($t_{1/2}$) | Prevents dormant accounts from banking old trust | | Dispute Penalty ($C_d$) | -4.2x multiplier | Asymmetric drag applied on lost arbitrations |

Worked Example: Calculating Elena’s Rolling Score

To understand how these numbers interact, consider a concrete 30-day window for Elena, an active participant across three universes on WEVONE.

In March, Elena completed three distinct transactions:

1. She delivered a €120 freelance copy-editing task on Mission. The escrow released without contest after the standard 48-hour dispute window.
2. She provided a ride from Lyon to Geneva on Pilote for €25.
3. She rented out a hammer drill on Tools for €15.

The base score increment ($S_i$) for a single settled transaction is derived via:

$$S_i = \log_{10}(1 + V_e) \times M_c$$

For her Mission job (€120 volume, category multiplier $M_c = 1.2$ due to verified manual labor): $$S_1 = \log_{10}(121) \times 1.2 = 2.0827 \times 1.2 = 2.499 \text{ units}$$

For her Pilote route (€25 volume, $M_c = 1.0$): $$S_2 = \log_{10}(26) \times 1.0 = 1.415 \text{ units}$$

For her Tools rental (€15 volume, $M_c = 1.1$): $$S_3 = \log_{10}(16) \times 1.1 = 1.324 \text{ units}$$

When activity spans three or more distinct universes within a rolling 30-day window, Mia’s context engine triggers a Cross-Universe Diversity Multiplier ($\Omega = 1.35$). Elena’s raw monthly increment becomes:

$$\text{Total Increment} = (2.499 + 1.415 + 1.324) \times 1.35 = 7.07 \text{ points}$$

Now consider the reverse scenario: if the hammer drill on Tools had been returned damaged, leading to an arbitrated dispute where Elena was found at fault for failing to provide proper usage instructions, the penalty is non-linear. A lost dispute applies an asymmetric factor ($C_d = -4.2$) against her total rolling 90-day score, burning down 4.2 times the transaction's weighted yield across her global account profile.

The Economics of Sybil Defense

A persistent vulnerability in marketplace reputation architectures is wash trading: colluding users creating artificial micro-transactions to manufacture fake trust.

In the Mockingjay architecture, wash trading is rendered economically non-viable through platform escrow fees. Because raw score points scale logarithmically with volume, cycling €10 between two accounts ten times yields significantly fewer points than moving €100 once, while incurring ten separate baseline transaction costs.

To manufacture 100 baseline Mockingjay points purely through circular transactions, colluding accounts would need to pay approximately €340 in non-refundable fee allocations to the WEVONE escrow protocol. The average economic yield of holding a 100-point score (such as reduced collateral requirements in Nest or top-tier placement in Mission) delivers roughly €45 in value per quarter. The cost of attack exceeds the extracted utility by a ratio of 7.5 to 1.

Honest Limitations and Edge Cases

WEVONE is early, and the current implementation of the Mockingjay engine contains explicit structural limitations currently being monitored in live production.

1. High-Value, Low-Frequency Asymmetry

Universes like Nest (short-term property rentals) inherently penalize hosts who do not participate in everyday universes like Tutus or Pilote. A host renting a high-end apartment once a month accumulates baseline reputation significantly slower than a commuter driving daily on Pilote, despite managing thousands of euros in physical assets. WEVONE is currently testing a capital-commitment offset in beta—allowing users to stake WEVAR tokens to buffer their decay floor—but this risks introducing wealth bias into a system designed to measure active contribution.

2. Regional Dispute Variance

Human arbitration data from Mission indicates a 12% variance in dispute initiation rates between Northern and Southern European markets, driven primarily by differing cultural expectations regarding service delivery times. While Mia’s context memory normalizes language sentiment during dispute reviews, the raw mathematical penalty remains uniform, creating slight geographical bias in net score velocity.

Platform Infrastructure

The Mockingjay Contribution Score does not rely on a static database column. It is computed asynchronously from state proofs logged across universe-level ledgers.

When a buyer confirms receipt on Tutus, the transactional escrow releases funds and emits a cryptographically signed completion event. Mia reads the event, calculates the logarithmic score delta, applies the exponential decay vector relative to the account's last timestamp, and updates the public contribution ledger.

By anchoring trust to verifiable economic velocity rather than passive accumulation, WEVONE eliminates legacy review monopolies. On this marketplace, trust isn't a badge earned five years ago—it is a measurable metric validated by today's transactions.