Methodology & downloads

Documentation of data selection, timing windows, account classification, analytical limitations and the underlying downloadable files.

Reporting boundary

Report run date: October 19, 2026. Include only bid records with timestamps before 2026-10-19 00:00:00, interpreted as America/Toronto local wall time. The source export does not supply explicit timezone offsets, so this interpretation should be checked against original page timestamps. Data collected later is retained separately, never counted in this reporting period.

Detection rules

For each account, sort bids by recorded timestamp and bid ID. Calculate rolling windows of 60 seconds. An account is designated high-frequency if it reaches at least 15 bids in one window on at least 10 distinct listing IDs in that same window. Compute cross-listing transitions from consecutive bids by that account on different listing IDs. Count gaps of 0–3 seconds or 0–5 seconds inclusive. These definitions alone cannot determine whether activity is human or automated.

Category rules

Each listing title is matched against ordered keyword groups; the first match determines category. Items with no match are labeled Other / unclassified. These categories are approximate and may not agree with KOTN's site taxonomy.

Important limitations

The first observed date (October 13) starts part-way through the day. Timestamp accuracy is one second. Listings and histories may be captured at different times. Account names are public pseudonyms; account ownership, seller identities, automation usage, final wins and real spending are not established by these exports.

Financial and pickup methodology

Current catalogue leaders are provisional. The October 19 listing snapshot is not valid as an October 18 end-of-day position. Without confirmed closed results there is no verified win rate, spending, collection volume or calculated price inflation. Queue examples are workload scenarios only.

Downloadable evidence

Reproducibility

All dashboard metrics are generated from included CSV records and serialized in dashboard.json. The catalogue CSV is explicitly dated to distinguish its collection time from the bid-analysis cutoff.