Learn · Glossary
Prediction market glossary
New to prediction markets? Here are the words you'll keep running into, in plain English — what a YES contract is, why a price is really a probability, how a market settles, and the handful of TinyCorp-specific terms (a signal, an anchor, the Evolution Lab) you'll see around the site. Each term links to where you can see it live. Everything here is paper trading — real markets and real prices, simulated money, no card, ever.
Markets & contracts
- Prediction market
- A market where you trade contracts on the outcome of a real-world event — an election, a temperature, where an index closes. The price moves with the crowd's collective probability that the event happens. New here? The how-it-works walkthrough covers the mechanics.
- YES / NO contract
- The two sides of the question. Each contract settles at $1 if its side is right and $0 if it isn't, so its price between 0¢ and 100¢ reads straight off as a probability. Buy YES at 40¢ and you collect $1 if it happens (a 60¢ profit) or lose your 40¢ if it doesn't. (Cents here are illustrative — check the live book.)
- Settlement
- When the event finishes and contracts pay out $1 or $0. What decides it isn't a vibe — it's an official source named in the contract: the National Weather Service daily climate report for temperature markets, the official closing level for an index, the published figure for an economic release.
- Lock vs settle
- The outcome is often effectively decided (locked) hours before the market officially settles and pays out. Knowing the source and the lock timing is most of the game for weather bots — see WX Lookup.
- Bracket / range market
- A ladder of yes/no questions over ranges of the same number for one event (the S&P closing in 5,475–5,500, 5,500–5,525, and so on). The prices across the ladder add up to a probability distribution over where the number will land. The stock & index signals board shows live brackets next to the real level.
- Governing number
- The single official figure a contract resolves on — the NWS high, the index close, the CPI print. Find the governing number and you know exactly what to watch.
- Kalshi
- A US, CFTC-regulated prediction-market exchange that settles in cash. It's one of the two books TinyCorp Signal bots paper-trade. See how it stacks up in Kalshi vs Polymarket.
- Polymarket
- A prediction market that settles in stablecoin and resolves outcomes through the UMA oracle. The other book our bots paper-trade. (We also read prices from Manifold for data, but our bots trade Kalshi and Polymarket only.)
- Oracle (UMA)
- The decentralized dispute-and-resolution process Polymarket uses to decide how a market settles when the answer isn't perfectly obvious.
Probability & pricing
- Implied probability
- A contract's price read as a probability: 40¢ ≈ a 40% chance. It's the single most useful habit — stop seeing a price and start seeing the odds the crowd is quoting. More in what is a signal.
- Edge
- The gap between your own estimate and the market's implied probability. If a market implies 40% and your well-grounded number is 55%, that 15-point gap is your edge — provided your number is genuinely better calibrated than the market's.
- Calibration
- Whether your stated probabilities match reality over many bets: if the things you call 70% likely happen about 70% of the time, you're calibrated. Calibration is the only thing that turns an "edge" into something real rather than luck.
- Spread
- The gap between the best price to buy and the best price to sell. A wide spread is a real cost that can quietly eat a thin edge before the outcome is even known.
- Liquidity
- How much can trade without moving the price much. Deep, liquid markets are harder to beat with a hunch because more people are already pricing them carefully.
Trading & strategy
- Paper trading
- Trading with simulated money against real markets, real live prices and real outcomes. The whole site is paper trading — no card, no real-money mode, nothing to lose. Why we do it this way, and why every bet is flat: paper trading vs real money.
- Backtest
- Running a strategy over past data to see how it would have done before you trust it forward. Honest only when it's tested on data the strategy didn't get to peek at. Try it on the backtester.
- Out-of-sample / walk-forward
- Testing a strategy on data it never saw while it was being designed — the cure for ideas that look brilliant only because they were quietly fitted to the past (overfitting).
- Flat sizing
- Betting the same fixed amount on every trade. Here every bot bets a flat $5, which keeps the leaderboard an honest read on the strategy itself rather than a clever bet-sizing trick.
- Return per unit of risk taken. The public leaderboard ranks bots by Sharpe, so a steady performer outranks a bot that got lucky with one big win.
- Drawdown
- The worst peak-to-trough drop in a strategy's running profit. A good average win means little if the path there includes a drawdown you couldn't have stomached.
- Win rate
- The share of settled bets that won. On its own it can mislead — a bot can win often and still lose money if its losses are bigger than its wins.
- Arbitrage / line-shopping
- When the same event trades at two different prices on two books, buying the cheaper side is a mechanical edge that needs no forecast. The gap is a real disagreement, not guaranteed risk-free profit — fees, spreads and settlement differences are real. Our free arbitrage scanner finds the gaps; this guide explains them.
TinyCorp Signal terms
- Signal
- A real-world number plus a condition that tells a bot when to act — "only when the yield curve is inverted", "only when the VIX is above 20". Signals are the heart of the site; browse the free signal library.
- Anchor
- The builder block that wires a real data feed (weather, finance, crypto, and so on) into a bot's rule, so a plain-English condition is actually backed by live numbers.
- Bot / strategy
- A saved set of plain-English rules that paper-trades prediction markets automatically once you turn it on. Build one in the builder — no code.
- Fork
- Making your own copy of an existing bot so you can tweak its rules (or have an AI improve it) without touching the original.
- Evolution Lab
- The public lab where an AI proposes changes (mutations) to real bots, backtests each one head-to-head against its parent, and only promotes genuine winners to live paper trading. Watch it run on the Lab.
- Magic link
- Passwordless sign-in — you get a one-tap link by email instead of choosing a password. It's how free accounts work here.
Free account · no card
Now put the words to work
TinyCorp Signal is a free, paper-trading sandbox for prediction markets. A one-tap magic-link account (no password, no card, all simulated) unlocks the whole thing:
- Build & save a bot from these signals in plain English, then turn it on and watch it paper-trade live markets.
- Track it on the public leaderboard — plus the full signal library and every tool on the site, all in one account.
- Prefer to just play first? Start a $10,000 paper bank and see if you can beat the market.
Ready to go deeper on one market? See the weather, crypto, sports, macro, politics, energy, stocks and tropical primers — or learn how cross-platform arbitrage works.