The last batch of sports questions didn't have bots trading on them via the API, though Yuna was still trading by hand, so I'm not sure if you could really tell the difference. Did you like them better than when Yuna was trading via the API?
Because we're going to be implementing this new plan for liquidity, I'm reenabling taker orders on sports questions. You can read about the new plan here: https://manifoldmarkets.notion.site/The-Yunaplan-for-Liquidity-18254492ea7a80e0984add56470a9770
market:
1,000@ian What's the difference between the new Yunaplan and going back to the old version with bot taking?
The reason there are so few orders in the order book is that the AMM subsidy is high relative to the amount of (organic) trading activity. The AMM is like a limit order at every level that always gets priority. So new orders have to compete with the AMM, and there isn't enough trading activity to make it profitable. When the organic trading activity is high relative to the depth of the order book (and AMM), you get a spiky graph and it is a magnet for people to fill up the order book with more orders.
Right now, most of the trading activity on sports sweeps markets is people trying to make profitable trades against the AMM. For robust markets, you need traders that will go against the consensus price to take a specific position (like betting on their favorite team).