Liquidity and Spreads: The Hidden Cost of a Prediction
A market's headline probability leaves out the price, quantity, and timing assumptions that determine what an order could actually receive.
Read the articleThe prediction market conversation
Big questions.Better signal.
The future isn’t a fact. Let’s talk about the odds. Explore prediction markets, Polymarket, Kalshi, and the AI tools changing how we think about what comes next.
Curiosity first. Evidence always. Certainty optional.
Find your starting point
Follow your curiosity.
From your first market question to your next research workflow. There’s a way in.
01 / READING PATH
Understand the markets
Read the question behind the price. Learn contracts, probability, liquidity, and resolution.
Explore market fundamentals02 / READING PATH
Follow the conversation
Explore the podcast’s reading room, conversation tracks, and questions worth asking.
Discover the podcast03 / READING PATH
Meet the new toolkit
Go inside market APIs, language models, research bots, and agents with an audit trail.
Explore AI & agentsTwo platforms. Better questions.
Same headline.
Different details.
A meaningful Polymarket vs. Kalshi comparison starts with the contract: what counts, when it closes, and how it resolves. Get beyond the headline percentage.
A reader’s field guide01 / Platforms
Polymarket
Explore event structure, outcome tokens, and the path from a market question to resolution.
Explore PolymarketKalshi
Explore event contracts, the market’s stated rules, and the details behind a Yes or No outcome.
Explore KalshiThe Signal Journal
Worth a closer look.
Agentic Prediction Markets: Design a Research Agent You Can Audit
Design a research agent that preserves evidence, separates observations from inference, and lets a reviewer reconstruct every published forecast.
Read the article
AI Prediction Bots: Paper Trading, Permissions, and Risk Controls
A practical framework for evaluating prediction bots without confusing an impressive simulation with a reliable trading system.
Read the articleInside the research toolkit
A model is only
part of the story.
Trace the path from raw market data to an evidence-based brief. Ask what the model saw, how its forecast was scored, and which decisions stayed with a person.
Explore AI & agents010203
Start with dependable data
Identifiers, timestamps, price units, and market rules give an API response context.
Measure the forecast
Test calibration and scoring before drawing conclusions from a convincing answer.
Keep the research inspectable
Preserve the sources, tool calls, and review decisions behind an agent’s work.
A few good questions
Let’s clear
a few things up.
What is a prediction market?
A prediction market lets participants take positions on a defined event. Prices can be read as market-implied signals about that event, with caveats about liquidity, costs, incentives, and the contract’s rules. Start with our prediction market guide.
Is a market price the same as a probability?
A price may be interpreted as an implied probability under simplifying assumptions. It is also a transaction price shaped by the available order book. Neither interpretation makes the outcome certain. Our liquidity guide explains the distinction.
Do you cover both Polymarket and Kalshi?
Yes. Our Polymarket and Kalshi reading guides focus on market definitions, resolution, and how to compare questions. PredictionPodcast.com is independent and is not affiliated with either platform.
Can an AI model reliably predict a market?
A model’s performance needs to be measured on questions it could not already know the answers to. Calibration, question selection, and comparison baselines matter. We discuss how to evaluate forecasts without promising an edge or a return.
Where can I listen to episodes?
Audio episodes are not available on this site yet. The podcast page introduces the conversation tracks, and the Signal Journal provides the complete written guides. Share a topic with us at [email protected].
Is this financial advice or a trading service?
No. PredictionPodcast.com publishes educational explanations and research methods. It does not operate a market, hold funds, execute orders, or make individualized trading recommendations.
What question should we explore next?
Send a topic, a correction, or a perspective worth discussing.