Prediction Market Leaderboard Data: What It's For Beyond Copy Trading

Every major prediction market already has a leaderboard. Polymarket ranks traders by profit and volume. Limitless ranks by points and volume. A whole layer of consumer apps, Polycopy, PolyScalping, Polycool, exists just to make those rankings easier to browse and copy. What's missing from all of it is the same data treated as something to run a real research question against, not just a live scoreboard to glance at.

What this data actually is

A trader leaderboard is a ranked snapshot: who's currently ahead, by how much, on which venue. A per-trader trade history is the record of exactly what that trader did to get there, every position, every entry and exit, over time. Together, that's something structurally different from the resolved-market data most prediction-market research already uses. Resolved-market data tells you what happened. Trader-level data tells you who called it, how early, and how consistently, across many markets, not just one.

What it's actually for

Testing whether skilled-trader positioning adds information beyond the market price itself. A market's price already aggregates everyone's positioning. Whether a specific top-ranked trader's position, isolated from that aggregate, predicts the eventual outcome any better is a genuinely open, testable research question, not something to assume either way without checking.

Category-specialist analysis. The pattern researchers and copy-trading discussion both point to is that top traders tend to concentrate in one category, politics, sports, a specific kind of crypto market, rather than trading everything. Per-trader trade history lets you actually test that: does a trader's win rate hold up when you isolate it to their specialty, or does it regress once you look past their best category?

Building your own logic instead of trusting a black-box app. The existing copy-trading tools make the follow/don't-follow decision for you, based on whatever ranking method they've chosen. Raw leaderboard and trade-history data lets you define your own criteria, weighting recency, category, position size, drawdown, however your own research says it should work, instead of inheriting someone else's.

The current landscape, and where it stops

Polycopy, PolyScalping, and Polycool are real, actively used tools, ranking traders by P&L and ROI, tracking whale-sized moves, surfacing category specialists. They're built for a specific job: helping a retail user decide who to follow, right now, inside one app. That job is done well. It's also a different job from pulling the underlying data to test a hypothesis, build a model, or run a backtest against trader positioning over time.

The scope limit on all of them is the same: Polymarket only. None extend to Limitless or Predict.fun, and none are built as a programmatic data source, an API or bulk export a researcher would actually integrate into a pipeline, rather than a dashboard meant to be read directly.

Where broader access fits

Convex Lake's leaderboard and per-trader trade-history data covers Polymarket, Limitless, and Predict.fun, available via the same API as the rest of the catalog, on Research tier and above. That's the multi-venue, research-oriented version of what the consumer copy-trading apps already do for a single venue, aimed at someone testing a hypothesis about trader performance rather than deciding who to follow this afternoon. See the API docs for the current endpoint reference.

FAQ