I've had to do this when training research monkeys in a former life, particularly when there is only one or two parameters being varied. They would learn the 'patterns' far too quickly. After seven heads in a row, they'd stop paying attention and always choose heads for some time afterwards.

The solution we settled on was block randomization. But that's probably not what you'd want for a human game as they could potentially figure out the end of each block (and what value remains). But it worked wonderfully for us.