Case Study · Optimism Seasons 4 & 5

1.46M OP Tokens Gone
Before the Market Could Blink

Optimism distributed 6.63M tokens to 11,024 wallets. 22% were moved to exchanges within a single day — and 59% of those senders had done the exact same thing before. A behavioral filter could have prevented 73% of the leakage.

22%Tokens moved within 24 hours1.46M of 6.63M OP distributed
74.3%Wallets in lowest score band8,187 of 11,024 scored below 100
59%Repeat movers1,491 of 2,529 prior offenders struck again

How value leaked

1

Optimism

Protocol distributes 6.63M OP tokens

2

S4 & S5 Airdrop

11,024 wallets receive tokens

3

19.3% Exit

2,124 wallets sell within 24 hours

4

DEX / CEX

1.46M OP liquidated instantly

The Problem

Governance Tokens for Ecosystem Builders
Funneled to Fast-Exit Wallets

Optimism Seasons 4 and 5 were designed to reward wallets that actively participated in the Superchain ecosystem — voting, deploying contracts, bridging, and providing liquidity over sustained periods.

What actually happened: 74.3% of all recipients (8,187 of 11,024) scored below 100 — wallets with shallow histories and zero depth of protocol engagement. Among those who moved tokens quickly, nearly 59% were repeat movers who had exited previous distributions the same way — median exit time: within 1 day.

The wallets that claimed the most were, by behavioral measure, the least likely to stay.

In the Age of AI Agents, This Gets 100× Worse

  • $20/month AI agents can simulate governance participation — casting votes, bridging assets, and interacting with Superchain contracts across hundreds of wallets at once.
  • They mimic multi-season Optimism farming patterns, exploiting sequential airdrop criteria, then liquidate within hours — leaving behind a history of fabricated participation.
  • Without behavioral intelligence that reads cross-protocol exit history, the next wave of leakage will reach hundreds of millions.
The Evidence

Three Behavioral Patterns
Drove 22% Instant Token Exit

Analysis of 11,024 wallet records reveals structural, repeatable patterns that separate opportunistic claimers from genuine ecosystem participants.

74.3%Score-Band Crowding8,187 of 11,024 wallets

of all recipients concentrated in the lowest score band (0–100). These wallets had zero deposit activity in 46% of cases — accounts with no sustained protocol commitment, present only for the claim.

19.3%Same-Day Exits2,124 wallets within 1 day

of all wallets moved tokens within 24 hours of receiving them. Almost entirely concentrated in the 0–200 score band, where average holding horizon is measured in hours — not months.

59%Habitual Repeat Offenders1,491 of 2,529 prior movers

of wallets that had previously exited other airdrops quickly did exactly the same here. This is not coincidence — it is a repeatable playbook applied sequentially across season after season.

Wallet Outcomes by Score Range

Higher scores correlate sharply with holding behavior

Repeat Behavior Signal by zScore

Prior repeat movers vs Optimism S4–S5 quick exits — wallet counts

The Solution

If zScore Was Used

zScore was not live during Optimism Seasons 4 and 5. The analysis below simulates outcomes if a behavioral filter (zScore ≥ 200) had been applied to the observed dataset — a conservative threshold that excludes only the lowest-quality behavioral profiles while preserving all established ecosystem participants.

The Bottom Line

73%Leakage PreventedSame-day token exit reduction
1.07MOP Tokens PreservedKept from same-day exits
9,813Low-Quality Wallets FilteredOpportunistic claimers removed

One behavioral filter — applied before Season distribution — would have stopped nearly three-quarters of all immediate token leakage, preserving 1.07 million OP from wallets that had no intention of staying.

MetricActual (No Filter)With zScore ≥ 200Impact
Eligible wallets11,0241,211−89%
Bad wallets (<1 day)2,124 (19.3%)289 (23.9%)−86%
OP moved within 1d1,460,987392,064−73%
OP tokens preserved1,068,923
01

Score Before You Snapshot

Apply a zScore ≥ 200 baseline before publishing any eligibility list. Wallets below this threshold have demonstrated shallow activity patterns — zero deposits, single-event claims — that raw onchain metrics cannot expose.

02

Track Seasonal Repeat Patterns

Run cross-season behavioral analysis to identify wallets that exit consistently. A wallet that exited Seasons 1, 2, and 3 will exit Season 4. History is the most reliable predictor available.

03

Shadow → Enforce → Iterate

Run the behavioral filter in shadow mode for one distribution cycle. Compare actual vs. simulated outcomes. Enforce in the next cycle. Iterate thresholds based on real false-positive rates and ecosystem feedback.

What Comes Next

For Protocols

  • 01

    Score before you distribute. Cross-protocol behavioral history reveals exit-pattern wallets that simple activity metrics — transactions, contracts touched, governance votes — cannot detect.

  • 02

    Treat seasonal repeat behavior as a first-class disqualifier. A wallet that exited three consecutive seasons is not a participant. It is an extractor operating a playbook.

  • 03

    Stage token unlocks by behavioral tier. High-score wallets earn faster access. Low-score wallets earn the right to participate by demonstrating holding behavior over time.

What ZeruAI Provides

zScore — a universal behavioral reputation score (0–1,000) for every EVM and non-EVM wallet, derived from onchain activity across 40+ chains. Queryable via API. Mintable as an onchain credential.

Protocols integrate zScore to filter airdrop recipients, identify seasonal repeat extractors, underwrite credit risk, assess agent reliability, and allocate incentives based on real behavioral contribution — not just eligibility checklists.

Every number in this report is derived from public onchain data. Download the raw dataset to verify the findings independently.