The six-step research workflow behind Sikt Intelligence, from reading the rules to finding the real gap with the market, done by hand in 15 minutes.
By Kristian FagerlieFounder & CEO, Sikt Intelligence
Most prediction market traders lose money, and most of them skip the research. Sikt Intelligence was built to do that research on every question: read the market, check the evidence, form an independent probability, and put it next to the market's price. This guide walks through the same six steps, so you can do them by hand in about 15 minutes, and see exactly what Sikt does for you.
A price is the market's probability, and it is usually a decent one. To do better than the price, you need a reason to believe the true probability is different, and a reason is something you can write down and check. In October 2026, Galaxy Research found that 69% of retail Polymarket accounts finished below break-even (why most prediction market traders lose money). Instinct is free; it is also what the majority is trading on.
The title is a summary. The rules decide who gets paid. Polymarket's own documentation puts it plainly: "Always read the resolution rules before trading. The market title describes the question, but the rules define how it resolves" (Polymarket).
Check three things:
What Sikt does: before any research starts, Sikt reads the market itself: the question, its outcomes, its rules and its current prices. Every forecast is about the market as written, not the headline.
Before looking at today's news, ask how often events like this happen. How often do incumbent senators win re-election? How often does a central bank move at a meeting where it signalled a hold? This "outside view" anchors you to history instead of to the loudest story. We explain how in base rates: the first step in every good forecast.
What Sikt does: every forecast starts from what usually happens in situations like this, and only then adjusts for what is different this time.
Look for what has changed since the base rate: polls, official statements, data releases, schedules, court filings. Three rules:
What Sikt does: Sikt researches each question from scratch in today's news and data, checks every fact against its source, and throws out what it cannot verify. It does this for every question, every time, without getting bored or tired.
This is the step almost everyone skips. Once you have seen that a contract trades at 63¢, your own estimate tends to drift toward 63%. Psychologists call this anchoring (Tversky and Kahneman, 1974). Write your probability down first, with one sentence on why.
What Sikt does: Sikt forms its probability from the evidence, independently of the market's price. That independence is the whole point: a second number is only useful if it was not copied from the first.
Now look at the price, and at more than the price:
Then ask the most important question: why might the market see something you don't? Someone may know more, or the rules may differ from your reading. Only a gap that survives these checks is worth anything.
What Sikt does: Sikt shows its probability next to the market's price, outcome by outcome, so the gap is visible at a glance, together with the reasons behind it and an honest note on how much to trust it.
Record your probability, the price and the outcome when it resolves. After 20 or 30 forecasts you will know whether your 70% calls happen about 70% of the time (calibration) and how your accuracy compares with the market's (the Brier score). Without a record, you cannot tell skill from luck.
What Sikt does: every Sikt forecast is scored against what actually happens, misses included. Our first public test is the Sikt Midterm Bench, on the key 2026 Senate races.
| Step | You, by hand | Sikt |
|---|---|---|
| 1. Rules | Read the resolution source, deadline and edge cases | Reads the market and its rules first |
| 2. Base rate | Ask how often this usually happens | Starts every forecast from the outside view |
| 3. Evidence | Primary sources, dated, verified | Researches from scratch, checks every fact |
| 4. Your number | Write it down before seeing the price | Forms its probability independently of the market |
| 5. The gap | Spread, fees, other venue, long shots | Shows its number next to the price, with reasons |
| 6. Score | Keep a record, check calibration | Scored on every forecast, in public |
Doing this well takes discipline, and doing it on dozens of markets takes a team. Professional trading desks build exactly this kind of independent probability in-house (see how hedge funds and quant firms use prediction markets). Sikt Intelligence is building it for everyone: an AI superforecaster that does steps 1 to 6 on every question, consistently, with its sources showing, and puts an honest second number next to the market's.
Watch the demo above to see it in action, follow the live odds on our 2026 midterms page, and join the waitlist for early access. Nothing here is financial or investment advice; research lowers the odds of a bad trade, but nothing removes the risk of losing money.
Read the resolution rules, start from the base rate, gather dated and verified evidence, write down your own probability before looking at the price, compare it with the market after fees and spreads, and keep a record of every forecast.
Because seeing the price pulls your estimate toward it, a bias called anchoring. A probability formed independently is the only kind that can tell you whether the market is wrong.
Sikt Intelligence is building an AI superforecaster that researches each prediction-market question, checks its sources, forms its own probability independently of the market and shows it next to the market's price. It is in research; you can join the waitlist for early access.
69% of retail Polymarket accounts lose money, and 0.1% of accounts take 67% of the profits. Why most traders lose, and five habits of the ones who don't.
Read the article
Quant firms make markets on Kalshi, Wall Street buys Polymarket data as signals, and funds start hedging with event contracts. What institutions do in 2026.
Read the article