September 17, 2026
Most days, the interesting thing on Polymarket isn't a single price — it's the way three unrelated books resolve into three completely different shapes of uncertainty.
One is a barrier question that lives or dies on a single threshold being touched. One is a disclosure tail, where the headline probability sits close to zero and almost all of the information is in the resolution language rather than the price. And one is a policy line, where the clock is a scheduled meeting and the market mostly re-prices around a handful of data releases. Same screen, same order book mechanics, three entirely different reasons a number moves.
Live numbers: Bitcoin price prediction from Polymarket odds: the median price and 50%/80% ranges implied by each day's strike ladder for the next week, plus reach and dip odds, refreshed twice a day. Also for Ethereum, Solana and XRP.
Barrier markets are the easiest to misread because they look like price forecasts and behave like volatility forecasts. A question about whether Bitcoin reaches $45K inside a defined window is not asking where the market settles — it is asking whether the path ever touches the level, even briefly. That distinction shows up in how the book reacts: a fast move toward the barrier can lift the price far more than a slow grind of the same magnitude, because the same distance travelled with more energy implies more chance of a touch before the clock runs out.
On my watchlist, the things I track here are mechanical rather than directional. Which price source resolves the market, and does it use spot, an index, or a time-weighted reference? Does an intraday wick count, or does the rule require a close above the level? How much time remains relative to the distance to the barrier — because barrier probability decays non-linearly as the window narrows, and the last stretch of a window can drain value even when the underlying has not moved much at all. None of that tells you where Bitcoin goes. It tells you whether the quoted probability is answering the question you think it is.
A market quoted around 0.5% is a different animal entirely. At that level, the price is not really an estimate of likelihood — it is a floor set by the cost of tying up capital against a remote outcome, plus whatever premium sellers demand for headline risk. Tails like this are where resolution language does almost all the work, because the gap between "something newsworthy happened" and "the specific documented condition in the rules was met" is usually where disputes live.
My checklist on disclosure-style questions is short and boring on purpose: who or what counts as the source of record, what format of confirmation satisfies the rule, whether reporting by outlets is sufficient or an official filing is required, and what the deadline stamp actually says. A rumour cycle can move a tail market noticeably in percentage terms while changing nothing about whether the resolution criteria can be satisfied in time. That asymmetry is the whole reason I keep tails on the watchlist as a study of rule-reading, not as a signal.
The policy market is the most structured of the three, because the calendar is known in advance. A hike question resolves off a scheduled decision, which means the uncertainty is concentrated into a handful of scheduled inputs beforehand and then collapses almost instantly at the announcement. Between now and the October meeting, the shape I expect to see on the chart is stair-step rather than drift: quiet stretches punctuated by re-pricings around inflation and labour prints, then a hard snap to near-certainty once the statement lands.
What I watch is the definition, not the vibe. Does the market resolve on any increase, a specific increment, or the published target range? Which document is authoritative, and what happens in the case of an unscheduled action before the meeting? Policy markets are unusually clean to study precisely because the resolution source is unambiguous — which makes them a good calibration exercise against your own read of the macro data.
Three clocks, three shapes: a path question, a rules question, and a calendar question. Putting them side by side is a research exercise in matching your reasoning to the structure of the market, not a set of picks. Nothing here is a trade recommendation, and every one of these questions can resolve against a perfectly sensible thesis because of a definition you skimmed. If you want to compare notes on the mechanics, fellow traders discuss this kind of structure in our Telegram channel: https://t.me/PolymarketView.
Because a barrier question only requires the level to be touched at some point in the window, while an above/below question requires the level to hold at settlement. Touch-style questions are therefore more sensitive to volatility and to how much time remains, and less sensitive to where the market is expected to finish.
No, and that framing is exactly what I try to avoid. Deep tails carry headline risk, capital is committed until the deadline, and the payoff profile is heavily skewed. I treat these as case studies in resolution-rule reading rather than as anything actionable.
Read the resolution text first: which decision it references, what counts as a hike under the rules, which official publication is authoritative, and how an unscheduled action would be handled. The calendar makes the timing predictable, but the definitions decide the payout.