I have taught and led Business, Economics and Computing in different parts of the world, and one lesson travels well across all three subjects: complicated vocabulary can hide a simple idea. Prediction markets are a good example. At their core, they start with a question about whether something will happen. People trade positions linked to the possible answers, and the market price changes as information and opinions change.
If a YES contract that pays one dollar when an event happens is trading around 63 cents, people often call that a 63% implied probability. That is useful shorthand, but it is not a scientific statement that the event has been proved to have a 63% chance. It is the current market price attached to one specific contract, under one specific set of rules.
Read the question first, read the rules second, read the price third.
Why the question comes before the percentage
A large percentage on a screen is visually persuasive. Before I care whether it says 42%, 63% or 91%, I want to know what exactly has to happen, by when, and according to whose definition. A football market asking whether a team will win a match is not automatically the same as one asking whether that team will qualify for the next round. Extra time, penalties and competition rules can make those questions resolve differently.
The same issue appears in politics, economics, technology and entertainment. “Will rates be cut at the September meeting?” is not identical to “Will the policy rate be lower by the end of September?” A product being announced is not the same event as it being available to buy. A candidate winning a primary is not the same event as winning the general election. Similar words are useful for discovery; they are not enough for identity.
Why different platforms can disagree
Imagine asking two rooms of people to estimate how many sweets are in the same jar. They may produce different answers because the groups contain different people and different information. Prediction-market platforms can behave similarly: participants, liquidity, fees, access and information flow differ. Two materially compatible contracts can therefore trade at different prices.
That disagreement can be interesting, but only after we have established that both rooms are looking at the same jar. This is why I built Prediction Market Radar around compatibility rather than title similarity. The price comparison is the visible part; the identity work underneath is what makes the comparison defensible.
Settlement is where the wording becomes real
“Will Candidate A win?” sounds simple until we ask which election, which jurisdiction, what counts as winning, what happens in an unusual edge case, and which source decides the final result. Settlement rules are not legal fine print that can safely be ignored. They are part of the market itself.
That is also why Prediction Market Radar keeps original source links and wording visible. I do not want a reader to have to trust a normalized percentage without being able to inspect what sits underneath it. If a comparison is challenged, there should be a path back to the source evidence.
Why refusing a match can be the right answer
One of the habits I value in both teaching and system building is being willing to say, “There is not enough evidence yet.” A matching system that always produces an answer can look productive while creating false certainty. Prediction Market Radar is deliberately willing to withhold a possible relationship when the event, outcome, timing or settlement evidence is not strong enough.
Coverage matters, but trust matters more. A missing match is visible. A false match can quietly corrupt every percentage, chart or downstream decision built on top of it.
Prediction markets are information systems, not crystal balls
A market priced at 90% can still resolve NO. A poll and a prediction market can disagree. A highly liquid market can move sharply when new information arrives. None of those facts makes prediction markets useless. They are systems for expressing and updating uncertainty, and they become most useful when we are precise about what each number belongs to.
That is the plain-English idea behind Prediction Market Radar. I am not trying to tell people what to trade. I am trying to make fragmented prediction-market information easier to understand, compare and audit—whether the reader is a trader, journalist, researcher, developer, platform or simply someone who wants to know what the percentages actually mean.