A Polymarket whale tracker is a system that monitors the live trade feed across the platform and surfaces positions large enough to indicate directional conviction. The Rivo whale tracker extends the same logic to Kalshi, producing a unified view of meaningful trade activity across the two largest prediction-market venues. This page describes how the tracker works, what it surfaces, and how copy-traders use it to identify positions worth following.
What counts as a whale on Polymarket and Kalshi
Rivo defines a whale trade as a single large opening position on a Polymarket or Kalshi event contract. The definition has three deliberate components.
The size floor is large enough to filter out retail-scale activity, market-maker rebalancing, and small-scale arbitrage flow. Below the threshold, trade activity is dominated by behavior that does not carry directional signal. The threshold is calibrated to produce a manageable daily volume of qualifying trades across both platforms.
The opening-position requirement excludes closes and partial exits. A $50,000 sell on a position previously entered at $30,000 is a $20,000 paper gain being realized, not a $50,000 directional view. Including exits in the whale feed would conflate two fundamentally different kinds of activity.
The per-contract framing focuses the definition on single-ticket conviction. A trader who spreads $50,000 across ten small positions is operating differently from a trader who places a single $50,000 ticket on one contract. The whale definition captures the second behavior. For a more detailed treatment, see our article on what is a whale on Polymarket.
How the Rivo whale tracker works
The tracker is built on a four-stage ingestion pipeline.
Stage one is platform polling. Rivo polls the Polymarket Gamma and CLOB APIs and the Kalshi REST API every few seconds, retrieving the most recent trade activity across all active markets. The polling cadence is calibrated to balance freshness against API rate limits and is typically between three and ten seconds per cycle per platform.
Stage two is normalization. Trades from each platform are normalized into a unified event schema with consistent field names, currency denominations, and timestamps. The normalized schema allows downstream filters to operate identically on Polymarket and Kalshi trades, which is important for the cross-venue copy-trading workflow.
Stage three is filtering. Trades below the size threshold are dropped. Trades identified as closes or partial exits are tagged but excluded from the directional whale feed. Trades on contracts where the YES price is above 0.75 are excluded as effectively decided. Trades on books too thin to support the executed size at the moment of entry are filtered out as fake-whale activity.
Stage four is enrichment. Each qualifying trade is enriched with the market title, category, current price, and wallet history where available. The enriched event is written to the public feed, persisted in the whale-events table, and passed to the alert pipeline for downstream delivery.
Position-delta classification
A critical part of whale tracking is distinguishing opening positions from closes. Without this distinction, the whale feed is contaminated by exit flow that looks superficially similar to fresh directional bets.
Rivo classifies each trade into one of six position deltas: open, add, flip, trim, close, or unknown. Opens are fresh positions in a market where the wallet had no prior exposure. Adds are increases to an existing position on the same side. Flips are reversals from one side to the other. Trims are partial reductions of an existing position. Closes are full exits. Unknown applies to cases where the trade history is insufficient to classify reliably.
Only opens, adds, and flips are surfaced in the directional whale feed and the alert pipeline. Trims and closes are tagged but excluded, because they represent profit-taking and risk management rather than new directional views. The classification logic is one of the more substantive engineering features of the tracker.
Wallet history and resolved PnL
Each whale trade is associated with the underlying wallet address (on Polymarket) or trader identifier (on Kalshi). The tracker maintains a per-wallet history of all recorded trades, the markets they were placed on, and the resolved outcomes where the underlying markets have settled.
The wallet history supports a per-wallet skill evaluation, which is the most important filter on the whale tracker for copy-trading purposes. A wallet with twenty or more resolved positions and positive category-specific PnL is a meaningfully more reliable copy-trading target than a wallet appearing in the feed for the first time. For the underlying analysis of whale profitability, see our article on whether Polymarket whales are actually profitable.
Public surfaces of the whale tracker
Several surfaces of the whale tracker are publicly available without authentication.
- The biggest wins page, which ranks resolved whale trades by paper PnL on the winning side.
- The biggest losses page, which ranks resolved whale trades by paper PnL on the losing side.
- The all-time wins leaderboard and all-time losses leaderboard, which provide full ranked lists across the resolved sample.
- Daily date-archive pages under /wins and /losses, which surface the resolved trades from any given trading day.
The live trade feed itself is gated behind a Rivo account. The free trial includes feed access; ongoing access requires a subscription.
Comparison to alternative whale trackers
Several alternative services offer some form of Polymarket whale tracking. The differentiators of the Rivo tracker relative to alternatives are the cross-venue coverage (Polymarket and Kalshi in one feed), the position-delta classification (which most alternatives do not perform), the AI quality filter on alerts, and the public resolved- PnL leaderboards that allow users to verify the tracker's signal value before subscribing.
Free-tier alternatives often surface raw trade activity without filtering or classification, producing a feed that is technically complete but practically unusable for copy-trading. Paid alternatives typically focus on a single platform and lack the resolved-trade scoring that makes wallet skill comparable across the population. The Rivo tracker is engineered to fill the middle: complete cross-venue coverage with the filtering and classification required for actionable copy-trading.
What to read next
For the broader copy-trading framework, see our copy-trading prediction markets guide. For the Telegram alert delivery channel that surfaces whale activity in real time, see our Polymarket Telegram bot page. For the full alerts infrastructure, see our alerts page.
Frequently asked questions
What is a Polymarket whale tracker?
A Polymarket whale tracker is a system that monitors the live trade feed on Polymarket and surfaces positions large enough to indicate directional conviction. Rivo applies a minimum opening position threshold and additional quality filters to identify whale activity in real time across Polymarket and Kalshi.
How does the Rivo whale tracker work?
Rivo polls the Polymarket and Kalshi APIs every few seconds, normalizes the trade data into a unified event stream, and filters for large opening positions that clear our size threshold. Each qualifying trade is enriched with market context, wallet history where available, and a position-delta classification, then made available through the public feed and the alert pipeline.
What is the minimum trade size for whale detection?
Rivo applies an internal minimum notional value for opening positions as the default whale threshold. The threshold filters out retail-scale activity, market-maker rebalancing, and small-scale arbitrage flow. Users can configure a higher threshold for their personal alerts if they prefer a more selective feed.
Does the whale tracker cover both Polymarket and Kalshi?
Yes. Both platforms are tracked in the same pipeline with the same threshold and the same filtering logic. The unified feed allows for direct comparison of whale activity across the two largest prediction-market venues.
How accurate is the whale identification?
The size filter has zero false negatives, by construction; any opening trade above our internal threshold is captured. The classification of position deltas, separating opens from closes, is approximately 95 percent accurate based on internal validation against subsequent trade activity from the same wallet.