On September 2, KTX hosted a special Twitter Space focused on prediction markets, bringing together experienced traders and industry participants to discuss where the opportunities are today, what beginners often get wrong, how strategy changes as capital grows, and where AI may fit into the next phase of the market.
- KTX LIVE Date: September 2, 2026
- Format: Twitter Space
-
Guests:
- Martin, Senior Prediction Market Trader at ForeGate
- Iverson, Head of Prediction Markets at KTX
- Levi, Full-Time U.S. Equities & Crypto Trader
- Media Partner: MyToken
Martin has experience across sports, event, and price prediction markets, including managing accounts from a few hundred to tens of thousands of USDT. Iverson currently leads KTX Prediction Markets, with a focus on making prediction markets easier to access through an exchange-based experience. Levi focuses mainly on trader education and systematic prediction market strategies.
After the World Cup Hype, Where Are the Opportunities Now?
Activity naturally cooled after major events such as the World Cup, but the guests agreed that this does not mean prediction markets are losing users. Instead, activity is becoming more regular and more diversified.
Martin noted that prediction markets now go well beyond major sports events. Football, tennis, basketball, esports, BTC and ETH price markets, Fed decisions, entertainment, and other real-world events can all create recurring opportunities.
His view is that traders should not depend on one blockbuster event. Once the World Cup ends, the better question is: which categories do you actually understand well enough to have an edge?
Iverson made a similar point. He noted that users have been moving toward smaller and faster-settling markets such as MLB and tennis. Sports, crypto, and politics are still major categories, but long-tail markets around equities, technology, weather, gaming, and other topics continue to expand.
For him, prediction markets are not really about trading one World Cup or one election. They are about trading changes in information. As new information appears, probabilities move—and those movements create opportunities.
Levi grouped the most interesting areas into three broad categories: macro and market events such as Fed rate expectations, CPI, nonfarm payrolls, and short-term crypto moves; event-driven markets around AI, technology, and geopolitics; and high-frequency markets such as intraday crypto price predictions and regular-season sports.
If the Space Is Already Competitive, Why Do New Platforms Keep Appearing?
The prediction market sector may look crowded, but the guests do not see it as fully saturated.
Martin pointed out that the number of people who have actually used prediction markets is still relatively small. At the same time, different categories attract very different users. Politics, sports, esports, and entertainment all have their own trading patterns and product needs.
That makes it difficult for one platform to serve every type of market equally well.
He also compared the development of prediction markets with the evolution of crypto exchanges. Exchanges have gradually added access to assets such as equities and precious metals. Prediction markets go in another direction: they turn information itself—economic events, sports, entertainment, and social developments—into something that can be priced and traded.
Iverson emphasized the sector’s potential to reach beyond existing crypto users.
For many people, questions such as “Will BTC reach a certain price?”, “Will the Fed cut rates?”, or “Who will win this match?” are much easier to understand than wallets, Gas, cross-chain transfers, or DeFi.
That means the real competition may not simply be about taking users from another prediction market. It may be about bringing in people who have never participated in Web3 products before.
Where Do Beginners Most Often Go Wrong?
One of the biggest mistakes is assuming that being right about the real-world event automatically means being right in the market.
Martin believes many new traders lose discipline because they become too confident in their original view. In fast-moving markets, prices can change quickly, and traders may start chasing short-term moves instead of sticking to a clear thesis.
Iverson and Levi both stressed another issue: settlement rules.
A market headline can look simple, but the actual outcome depends on the exact deadline, data source, and Yes/No criteria defined by the platform.
Levi gave the example of a market asking whether a proposal would pass by Wednesday. News outlets may report that it has passed, but if the settlement rules require an officially signed document, the market may not resolve the way a trader expects.
That leads to one of the most important lessons for beginners:
You are not simply trading your interpretation of an event. You are trading the event as defined by the market’s settlement rules.
Iverson also pointed out that prediction markets are not necessarily “enter a position and wait for final settlement.”
If a Yes position moves from 0.40 to 0.70 before the event ends, a participant may be able to exit and realize the change in value, provided there is enough liquidity.
So prediction markets can involve trading both the final outcome and the changing probability along the way.
🔔 Explore sports, crypto price, and other prediction markets on KTX Prediction Markets.
If You Only Have 100 USDT, How Should You Approach It?
For smaller accounts, the guests generally favored spreading risk rather than concentrating everything in one market.
Iverson said that with 100 USDT, he would rather divide the capital across many small positions and focus on shorter-duration markets that settle within a day, a few days, or a week.
The goal is not necessarily to maximize profit immediately.
With smaller capital, one of the biggest advantages is being able to test more ideas and learn where you actually have an edge—whether that is crypto, sports, macro, or another category.
Levi takes a more aggressive approach. He believes smaller accounts can look at short-duration markets with asymmetric risk-reward, including very short-term BTC price prediction markets.
Martin also noted that smaller traders sometimes have more flexibility in thinner markets because their position sizes have less impact on liquidity.
In short, small capital can be useful for experimentation. The priority is learning where your judgment is strongest before scaling up.
What Changes When the Position Size Gets Bigger?
This was one of the clearest points of agreement during the discussion: once capital grows, the strategy has to change.
Martin said larger traders need to stop asking only, “How much can I make?” and start asking, “If I am wrong, can I absorb the loss?”
That means stricter market selection, clearer position management, and more disciplined entries and exits.
Iverson focused on liquidity.
A 10,000 USDT position may be easy to execute in a liquid market. A 1 million USDT position is different. If it is entered too quickly, the trader may push the price against themselves.
At larger size, being right about the event is no longer enough. Traders also have to think about market depth, slippage, entry execution, and whether they can exit efficiently.
Levi described this as a shift in both liquidity constraints and probability-based thinking.
Small accounts can pursue more flexible, higher risk-reward opportunities. Larger accounts need to care more about deep liquidity, probability edges, arbitrage opportunities, and drawdown control.
How Would You Allocate 10,000 USDT?
None of the guests recommended putting the entire amount into one market.
Martin would divide the capital into three parts: a smaller allocation for short-duration, higher-frequency trades; a core allocation for markets he understands well and where liquidity is stronger; and a reserve for opportunities that may appear later.
Iverson uses a 5:3:2 framework.
He would put 50% toward markets where he believes he has a stronger probability edge, 30% toward areas he understands particularly well, and 20% toward smaller, short-duration opportunities.
He also stressed that high probability does not mean guaranteed return. Even events that look very unlikely can happen, and larger positions can still take meaningful losses.
Levi’s approach is more arbitrage-focused.
He would use a smaller portion of the capital for higher risk-reward trades, while using most of the rest to monitor price differences across prediction market platforms, or between prediction markets and derivatives such as perpetual futures and options.
For larger capital, that kind of relative-value strategy may be more scalable than repeatedly trying to predict individual outcomes.
Can AI Actually Improve Prediction Market Trading?
AI was another major theme of the Space.
Martin sees AI’s biggest strengths in information aggregation, probability analysis, and disciplined execution.
AI can process large amounts of information quickly and compare a market’s implied probability with a trader’s own estimate. When the difference is large enough, that gap may point to a potential opportunity.
This is especially useful in short-duration markets, where information can move quickly. For more complex, longer-term events, Martin believes human research combined with AI support may work better.
Iverson sees AI slightly differently.
For him, AI does not need to “predict the answer” for the user. Its first job can simply be helping the user understand the market.
If someone is unfamiliar with finance, sports, crypto, or esports, AI can summarize the background, key data, and recent developments. From there, it can also help users think through entry prices, position sizes, and risk management.
KTX Prediction Markets also plans to move further toward AI Skills.
The longer-term idea is straightforward: a user tells AI how much capital they have, how much risk they can tolerate, and what areas they are interested in. AI then helps surface relevant markets, explain the context, and provide strategy support.
Levi also sees strong value in opportunity screening. AI can monitor many markets at once, review historical performance, and help traders narrow a large number of events down to those that better fit their strategy.
The value of AI, then, may not be that it “guesses better.” It may be that it helps traders process information faster, compare opportunities more systematically, and execute with more discipline.
What Will Actually Keep Users Around?
As more prediction market platforms launch, simply offering more markets may not be enough.
Martin believes platforms need to make the experience more engaging and social, especially for smaller users. Short-duration markets, broader topics, and more accessible content can all help people stay interested.
Iverson believes the core challenge is reducing three types of friction: entry friction, understanding friction, and trading friction.
Users should not have to learn wallets, bridges, and Gas just to get started. They should be able to quickly understand what a market is asking, how it will settle, and what the current price represents.
He also expects prediction markets to become more closely connected with content.
A user may first discover a real-world event through content, then see the probability implied by the market, and from there decide whether to take a position.
Levi believes access, content, and trader education all play different roles in the same journey.
Easy access brings users in. Good content helps them understand what they are looking at. Education helps them develop a more sustainable approach over time.
If users never learn probability, risk, position sizing, or capital management, even a smooth product experience may struggle to create long-term retention.
From Predicting Outcomes to Trading Probabilities
The broader takeaway from the discussion is that prediction markets are becoming less dependent on a handful of major events and more like continuous markets for information and probability.
For smaller accounts, the focus may be on testing more ideas and finding where an edge exists.
As capital grows, liquidity, slippage, drawdown, and execution matter much more.
And as AI becomes more involved, it may lower the barrier to research, market discovery, and systematic decision-making.
The next stage of prediction market competition may therefore come down to more than who offers the most markets.
It may come down to which platforms can help users understand probability more clearly, manage capital more systematically, and develop a trading approach that fits them.
🔔 Explore sports, crypto price, and other prediction markets on KTX Prediction Markets.