Layer 1: Market Data
The foundation is raw market data: current prices, volumes, and order book depth across all platforms. This tells you what the market currently believes and how much capital backs that belief. Without this layer, you are operating blind.
But most traders underestimate how much information lives in this basic layer. A contract trading at 52 cents with $50,000 in daily volume tells a different story than one at 52 cents with $500 in volume. The first suggests genuine price discovery with meaningful capital backing the probability assessment. The second might just be noise or a few small traders moving the needle.
Order book depth matters even more. I've seen contracts where the bid-ask spread widens dramatically beyond the top few orders. A market showing 51-52 cents on the surface might actually be 48-55 cents if you want to trade more than $100. This depth data reveals liquidity constraints that can make or break your execution strategy.
Real-time price feeds also catch momentum shifts before they show up in slower aggregators. When a major news event breaks, watching individual platform feeds can give you 30-60 seconds of edge before the information propagates everywhere. Those seconds matter when you're trying to get positioned before the crowd reacts.
Layer 2: Cross-Platform Analysis
Comparing prices across Polymarket, Kalshi, Manifold, PredictIt, Metaculus, and Opinion Trade reveals consensus, divergence, and potential arbitrage. This layer transforms six independent data points into relational intelligence.
The cross-platform view shows you where smart money is actually flowing. When Polymarket shows 45 cents and Kalshi shows 52 cents on the same outcome, that 7-cent spread is rarely the free money it looks like. It's information about user bases, liquidity constraints, and sometimes regulatory differences affecting each platform.
Polymarket tends to move fastest on crypto-adjacent events because of its user base. Kalshi often leads on traditional finance and economic indicators due to its regulated status attracting institutional flow. PredictIt, despite lower limits, sometimes shows early signals on political events because of its engaged political trader community.
I track these platform personalities because they create predictable patterns. When a Federal Reserve announcement is coming, I watch Kalshi first. For election developments, PredictIt often moves before others catch up. Metaculus provides longer-term perspective that helps calibrate whether short-term price movements represent genuine information or just noise.
The Blockcircle aggregation dashboard makes this cross-platform analysis systematic rather than manual. Instead of checking six different websites, you can spot divergences and track convergence patterns in real-time.
Arbitrage Opportunities in Practice
Cross-platform arbitrage works, but it's not as simple as buying low and selling high. Withdrawal times, deposit requirements, and position limits create friction that eats into theoretical profits. A 5-cent spread might only be worth 2-3 cents after accounting for these costs.
The bigger value comes from using cross-platform data to validate your analysis. When your independent probability estimate disagrees with one platform but aligns with others, that's useful information about where the mispricing might exist.
Layer 3: Information Gathering
For each contract of interest, gathering relevant news, data releases, expert commentary, and historical analogues provides the raw material for probability estimation. This layer is where domain knowledge matters most.
Effective information gathering means knowing which sources matter for different types of events. For economic indicators, I prioritize Federal Reserve communications, economic data calendars, and analyst consensus forecasts. For political events, polling aggregators, campaign finance reports, and local news sources often provide better signal than national media coverage.
Historical analogues deserve special attention because prediction markets are still relatively new. Most contracts don't have deep historical data, so you need to find comparable situations from traditional markets or past events. When analyzing an election contract, looking at polling accuracy in similar races provides better context than just the current polling numbers.
Social media sentiment can be valuable, but it requires careful filtering. Twitter/X trending topics might indicate public attention, but they don't necessarily predict outcomes. Reddit discussions in specialized communities often provide more substantive analysis than general social media buzz.
The key is building information pipelines that update automatically rather than relying on manual research for each trade. RSS feeds, news alerts, and data subscriptions create systematic information flow that scales across multiple positions.
Domain Expertise vs. General Analysis
Some contracts reward deep domain knowledge, while others favor general analytical skills. A pharmaceutical approval contract might require understanding clinical trial processes and FDA procedures. A sports betting market might be more about statistical modeling and injury reports.
Knowing your knowledge boundaries helps you focus on contracts where you have genuine advantages. I avoid biotech contracts because I lack the expertise to properly evaluate clinical trial data. But I'm comfortable with economic indicators because I understand the data release processes and historical patterns.
Layer 4: Probability Estimation
Using the gathered information, you produce an independent probability estimate for each contract. This estimate, formed before looking at the market price, is your analytical anchor. Comparing it to the market price identifies potential mispricings.
Independent probability estimation is harder than it sounds because market prices create anchoring bias. Seeing a contract at 60 cents makes it difficult to genuinely believe the true probability is 40%. The solution is developing your probability estimate first, then comparing it to market prices.
I use a structured approach: base rate analysis, specific factor adjustments, and confidence intervals. For an election contract, I start with historical base rates for similar races, adjust for specific factors like polling, fundraising, and campaign quality, then estimate a range rather than a point estimate.
Base rates provide the foundation because they prevent overconfidence in specific factors. If incumbent governors historically win reelection 75% of the time, that's your starting point regardless of current polling. Specific factors can move you away from the base rate, but they need strong evidence to justify large adjustments.
Confidence intervals matter because they determine position sizing. A 60% probability estimate with high confidence justifies larger positions than a 60% estimate with significant uncertainty. The width of your confidence interval directly feeds into the Kelly criterion calculations in Layer 6.
Layer 5: Whale and Smart Money Analysis
Checking whether whale positioning confirms or contradicts your estimate adds a layer of validation. Whale cluster detection, track record filtering, and position change monitoring provide smart-money context for your analytical conclusions.
Whale analysis isn't about blindly following large positions. It's about understanding when whale activity provides additional information versus when it might be noise. A whale with a strong track record taking a large position opposite to your analysis should make you reconsider your reasoning.
The Blockcircle Whale Finder tracks position sizes and timing across platforms. When multiple whales cluster around similar positions, that's meaningful information. When a single whale makes a large bet, it might just be someone with more capital than sense.
Track record analysis requires looking beyond simple win rates. Some whales might be profitable overall but weak in specific categories. A whale who's great at sports betting might not have edge in political markets. Context matters more than overall performance.
Position timing also provides information. Whales who position early before news breaks might have information advantages. Those who pile in after price movements might just be momentum followers. Understanding these patterns helps you interpret whale activity correctly.
Smart Money vs. Dumb Money
Not all large positions represent smart money. Some whales are just wealthy individuals with strong opinions and weak analytical skills. Others might be hedging positions in other markets rather than making directional bets.
The key is identifying whales who consistently demonstrate analytical edge rather than just capital. This requires tracking performance over time and understanding their typical position sizing and timing patterns.
Layer 6: Risk and Sizing
Kelly criterion sizing based on your estimated edge, fractional adjustments for model uncertainty, portfolio-level correlation checks, and maximum position limits translate your analytical conclusion into a specific, risk-managed position.
Kelly criterion provides the mathematical foundation for position sizing, but raw Kelly often suggests positions that are too large for practical trading. If you estimate a 70% probability on a contract trading at 50 cents, Kelly suggests betting 40% of your bankroll. That's rarely wise in practice.
Fractional Kelly reduces position sizes to account for model uncertainty and psychological comfort. Most successful prediction market traders use 25-50% of full Kelly sizing. This provides meaningful exposure to positive expected value opportunities while maintaining manageable risk levels.
Portfolio correlation checks prevent concentration risk across related contracts. Betting heavily on multiple economic indicators that all depend on the same underlying data creates hidden correlation that increases overall portfolio risk. Diversification across event types and time horizons provides better risk-adjusted returns.
Maximum position limits create hard stops regardless of Kelly calculations. Even with high confidence and attractive odds, limiting individual positions to 10-15% of total capital prevents single-contract disasters from destroying long-term performance.
Layer 7: Execution and Monitoring
Placing the order with appropriate limit prices, setting alerts for position monitoring, and scheduling review when new information arrives or the contract approaches resolution completes the cycle.
Execution timing can significantly impact returns. Market orders in thin prediction markets often face substantial slippage. Limit orders provide price protection but might miss fast-moving opportunities. The choice depends on urgency and market conditions.
Post-trade monitoring requires systematic alerts rather than constant manual checking. Price alerts notify you of significant movements that might require position adjustments. News alerts help you stay informed about developments affecting your positions. Resolution date reminders ensure you don't miss important deadlines.
Position reviews should happen when new information arrives, not on arbitrary schedules. A weekly review might miss important developments, while daily checking creates noise and overtrading. Event-driven reviews based on news alerts and price movements provide better timing.
The Complete Information Stack in Practice
Each layer builds on the previous ones. Skipping a layer produces weaker decisions. Sizing without probability estimation, or estimating probability without information gathering, creates systematic gaps in your analytical process.
The full stack, applied consistently, produces decisions that are systematically better than partial analysis. It also creates a feedback loop where tracking results helps you identify which layers need improvement in your personal process.
Most traders are strong in some layers and weak in others. Identifying your weak points and systematically improving them provides more edge than trying to perfect your strongest areas. The information stack framework makes these gaps visible and actionable.
Explore these tools on Blockcircle: Prediction Markets Mispricing Engine | Whale Finder