Open Datasets

CashPop publishes anonymized aggregate datasets under CC-BY-4.0 license at datasets.cashpop.meme.

Release cadence

• Daily: rolling 30-day metrics endpoint via API.
• Quarterly: full quarterly snapshot, including Round-level aggregates.
• Annually: annotated research dataset with methodology paper.

Data products

1. cashpop-rounds-Qx-YYYY.parquet

Per-Round aggregates. Schema:

round_id          int64
question_id       int64
start_ts          int64 (UTC ms)
participant_count int32
commit_count_by_second  array<int32>  // 30 elements
reveal_count_by_second  array<int32>  // 15 elements
majority_outcome  int8           // 0 or 1
vote_ratio        float64        // share of majority
country_buckets   map<string, struct{count: int32, ratio: float64}>

2. cashpop-reservoir-Qx-YYYY.parquet

Question reservoir snapshot. Schema:

question_id       int64
question_text     string
category          string
language_origin   string
calibration_distribution  array<float64>  // synthetic-population estimate
actual_distribution       array<float64>  // observed
focal_point_index float64       // Schelling divergence

3. cashpop-timeseries-Qx-YYYY.parquet

Daily aggregates. Schema:

date              date
dau               int32
mau               int32
total_rounds      int32
ad_revenue_usd    float64
prize_pool_distributed_usd  float64
unique_countries  int32
median_round_size int32

4. cashpop-focal-matrix-Qx-YYYY.parquet

Cross-country focal-point similarity. Schema:

question_category string
country_a         string  // ISO-3166
country_b         string
similarity        float64  // cosine on response distributions
sample_size_a     int32
sample_size_b     int32

Methodology paper

Each quarterly release is accompanied by a methodology paper covering:

• Sampling decisions and known biases
• Anonymization parameters (k-anonymity threshold, ε for differential privacy)
• Calibration analysis of question reservoir
• Replication instructions

How to access

# Via API
curl https://datasets.cashpop.meme/api/v1/Q2-2027/rounds.parquet

# Via direct download
wget https://datasets.cashpop.meme/releases/Q2-2027/

# Via Python
import requests, pandas as pd
df = pd.read_parquet('https://datasets.cashpop.meme/releases/Q2-2027/cashpop-rounds.parquet')

Replication code

Open-source pipeline at github.com/cashpop-protocol/data-pipeline. Anyone can:

1. Re-run anonymization with different parameters.
2. Re-verify k-anonymity and DP guarantees.
3. Reproduce published analyses.

License

Creative Commons Attribution 4.0 International (CC-BY-4.0). You may use, redistribute, and adapt the data for any purpose, including commercial, with attribution.

Academic partnerships

Active or being negotiated:

• Behavioral economics group at a research university (target announcement Q3 2026).
• Computational social science consortium (target announcement Q4 2026).

External researchers seeking access to non-aggregate data (with appropriate IRB review and confidentiality protocol) may contact: support@cashpop.meme.

Citation

If you use CashPop data in published research, please cite:

CashPop Protocol Team. (2027). CashPop Round Aggregates Qx-YYYY [Dataset]. https://datasets.cashpop.meme

A formal BibTeX entry will be provided per release.