I did this for a while and ran into the same problem. There is no perfect way to paper trade. Anyone who tries to switch from paper trading to actual trading soon realizes how difficult it becomes when they move the bid\ask on you if you attempt limits or run you up on a big market order. You have to accept the limitation of a system like that.
Best practices for a paper trading system is to trade highly liquid stocks, or indexed ones and accepting market orders for the open or close price for the next day. So to open a position 5/18s market open you would have to post the share size before the market closes today. To get tomorrows close you would have to have the order in prior to tomorrows market open. It kept me from cheating myself and fudging the numbers at the sake of intraday shifts in the trend. It may be one of the reasons I don't like day trading, you really have to have a head for it. Even though it is trading it is definitely not the same as someone who simply trades daily or weekly charts.
In the end the only think that mades a good trader is actual risk and protection of capital. With a paper system it's very easy to just scrap the whole thing when its not going one's way. When actual money is risked one has to follow through to the end.
Market Data 
Markets 
AI
