. Stop orders are effectivly limit orders but pointing away from the current market probability rather than towards. For example, if a market is at 50% and I place a stop order for M$100 at 60%, nothing happens until the market hits 60%, at which point I automatically buy M$100 shares of YES, sending the market even higher. (Or sell shares of NO; buying YES and selling NO are equivalent.)

The purpose of stop orders is for situations where you think that a market is likely to undercorrect in response to new information; you place a stop order to automatically do the remaining correction yourself. But a stop order is a standing offer to buy shares at a price that's worse than the current market probability, which is inherently dangerous. In my previous example, someone else could buy the market up to 60%, triggering my stop order, then immediately sell all their shares at a profit, leaving me holding the bag.

You can get away with stop orders if the amount of the order is smaller than the amount it would cost a manipulator to manipulate the price, and people do use them in large markets like the stock market. But on Manifold there are far fewer users, fewer bots, and fewer standing limit orders, meaning the cost of manipulation is often very close to 0.

In the event that Manifold (or a popular third-party interface for Manifold) adds stop orders, I will wait 1 month, then distribute a poll among everyone who had both asked for stop orders and used at least one stop order. This market resolves to the percentage of them that indicate they regret using them, this was an unhelpful feature, or any other sentiment that would retrocausally indicate that Manifold should not add this feature.