99market vs Predikt

What 99market does

Free scanner for Polymarket outcomes priced between 98¢ and 99.5¢, ranked by annualized return. It reads Polymarket's public market data, keeps markets with at least $1,000 of volume, and shows the price, simple APR, time left, volume, liquidity and gross profit per $100. An automated score rates how clear each market's rules are. Around it sit an APR calculator, strategy guides, a directory of 231 Polymarket products and a jobs board. You trade by hand on Polymarket, and there are no accounts.

• Feed of outcomes at 98–99.5¢ with at least $1,000 of volume, ranked by APR and then end time
• Views for Today, Ending Soon, 15-minute crypto, Weather, Sports and Crypto
• Rule-quality score from 1 to 100: resolution clarity 45%, market realism 55%
• Gross profit per $100 and reported liquidity on each row
• APR calculator

• Buying at 98¢ makes about 2¢ if you are right and loses 98¢ if you are wrong, so one loss takes back some 49 wins and the strategy needs better than 98% just to break even. The scanner does not tell you whether a market clears that bar. It has no win-rate history for these prices; the methodology says the score does not use a complete database of past results and does not predict the winner. That number is the one a user of this strategy most needs, and they will have to work it out themselves.
• The APR column looks better than it is. It divides a small gain by a short time, so on Nov 7, 2026 the top of the feed was same-day weather markets at >1000%, each worth about $2 per $100. Those rows showed liquidity of about $400 to $1,600, so the figure cannot be scaled up or repeated at will, and it leaves out fees, slippage and any delay in settlement. The site says all of this itself in its methodology, and that APR can pass 1,000% for exactly this reason, but it still sorts by that column and marks it with a flame.
• The score is the more original part. It is a fixed text heuristic that looks for official sources and measurable thresholds and marks down subjective wording, and price or volume cannot raise it. That suits this strategy, where losses tend to come from disputed or badly worded rules. It is still a reading of text: a “Good” label is not evidence that the market will resolve as priced, and the site tells you to read the original rules.