Max Cost only works reliably for percentages if it is configured to a fixed value.
If it is configured to unlimited (aka -1), Max Cost always equals Current Cost.
This doesn’t give us a base to compare our bots against.
So I would suggest using the Max Parallel Cost instead, that shows how many funds were really used at the same time in the past. That way even bots with a fix value of Max Cost could be valued based on that.
This would give us a view on our bots that is more
realistic and finally also more comparable
Of course, for this to work, the bot would have to know what the maximum parallel cost used by the bot in the past was, since it can be different from both Max Cost and Current Cost.