Accuracy in Prediction Markets: A Game Changer for Polling and Forecasting

Well, prediction markets may be facing some legal challenges, but that's not slowing their growth.

We've seen PolyMarket and Kalshi's valuations nearly double in a few months, and the industry's all-time notional volume has also grown 13 times over. six months.

And with prediction markets being a feature of DeFi, we're also seeing them grow in blockchain ecosystems like BASE and BNB chain.

Well, joining us to weigh in this morning is DJ Hà Trang, the Head of Research at Birdeye.

DJ, great to have you on.

Thank you so much for joining us.

So all time notional volume of prediction markets now exceeds $150 billion.

So take us through what's driving this growth.

Yeah, for sure.

So, prediction markets have grown rapidly in the past year.

As you mentioned, all-time notional volume exceeding $150 billion, growing 13 times within six months, while open interest increased six times over the past year.

I think several factors are driving this growth.

One of which is the institution adoption of prediction markets.

And the second one is more and more focus is being placed on how prediction markets accuracy, it's outperforming traditional polling and it's becoming a major important factor in the institutional risk management.

Yeah, and names such as PolyMarket and Calshi have become household names.

And I understand that your research shows PolyMarket and Kalshi accounting for 79% of total volume.

But can you tell us why and how we're seeing activity emerge in ecosystems such as BASE and BNB Chain?

For sure.

So apart from PolyMarket and Kalshi which are the big names in the industry we're seeing more and more activities on base such as limitless exchange and BNB chains such as opinions predict that fund.

And even though these Exchanges are small compared to PolyMarket and Kalshi.

There has been actually a moment where opinions on BNB chain had a big share in weekly prediction market volume at the beginning, at the end of 2025 and going into 2026.

And I think With more and more LP records incentives, I think there will be more players in the field and users are going to try to explore more new entrants within the field.

Yeah, and DJ, I want to get your take on infrastructure here.

So for the layperson who may not be as familiar, can you take us through DeFi oracles and also Chainlink's role?

For sure.

So for prediction markets, the structure of it is that it has a prediction event.

It has a market structure that are cloud-based.

It's trading.

And also resolution mechanism is very important in prediction markets.

Resolution decides whether the predict the events.

It's how the event is resolved.

And decentralized oracles such as Chainlink empower empower the resolution and chain-link using decentralized nodes, give information to the prediction markets itself, and so that it doesn't have to rely on a single point of failure.

And now that chain-link has kind of perfected its decentralized Oracle nodes, it has also evolved into AI Oracles that also will power the resolution mechanism under prediction events, so that the events can resolve quickly and also correctly.

Yeah, and DJ, I have a question about accuracy.

So your research shows PolyMarket and Kalshi perform traditional benchmarks such as polling futures and expert forecasts.

But in reality, how accurate are the predictions?

Yeah actually prediction markets are fairly accurate compared to traditional polling and an average of 90.8 percent of the markets reaching right resolution four hours before resolution.

So that is a very high accuracy rate compared to average polling.

And they have an average Bria score of point zero point zero six eight six two which is fairly low.

Well which is very like high accuracy rate.

And I think what's more interesting is that the higher the deeper the liquidity is the higher the accuracy of prediction markets are.

And we're seeing like Bria's score as low as 0.0247 in markets that has liquidity of higher than $1 million.

And so this means that as liquidity flows into prediction markets and as infrastructure matures, we will see prediction markets having more and more higher accuracy.

And that will mean that it will be an important factor in institutional risk management and outcomes-based prediction.