i wonder if aggregators should just aggressively down-weight sim and instead build a statistical model of what liquidity will be around when the trade lands. so much happens on chain between sim and execution
benedict
benedict1.8. klo 23.05
the main drawback of dark amms is that they push a lot more complexity to the routing layer i would expect to see meaningful differentiation between aggregators as a result
you also need a centralized system to track recent txs you’ve sent that haven’t landed and a model of every dark amms liquidity refresh rate so that you don’t take out a full side of the book if a lot of users swap in a short window
1,4K