Polymarket, a decentralised prediction market platform, isn’t as busy as the numbers about its volume show. Storm, a researcher at the venture capital firm Paradigm, found that big analytics platforms have been overestimating Polymarket’s trading volume. The problem comes from the way blockchain records every transaction: in each transaction, the platform creates distinct “OrderFilled” events for both buyers and sellers. Dashboards like DeFiLlama, Alium, and Blockworks have been adding these trades separately and displaying them as separate transactions, even though they are actually the same trade from different perspectives.
For example, instead of seeing a $100 trade as one, analytics tools treat it as generating $200 in volume, since their systems record it on both the buyer’s and seller’s sides. Storm said that this is causing over-reporting about Polymarket’s activity, and this has nothing to do with any manipulation. He also said that the glitch makes important numbers like notional volume (the total value of contracts traded) and cashflow volume (the actual amount of money moved) look bigger, which could mean that reported monthly volumes drop from $3.7 billion to about $1.85 billion.
Counter Narrative Emerges
Deployer, who helped start Bankr, a commercial AI trading agent and crypto platform, disagreed with Storm’s criticisms, saying that third-party platforms messed up the data instead of Polymarket doing anything wrong. He says that Paradigm’s criticism is biased because they invested in Kalshi, a competitor of Polymarket.
The controversy could have a big effect because prediction markets are becoming more popular as tools for making predictions. To keep this new field credible, we need to have accurate metrics about these people.