This is certainly possible. I've relaxed and loosened up on 0% beer as though it had alcohol when I knew that's what it was too.

I strongly suggest with little evidence that if you took 10 people who had previously taken MDMA and gave them placebos at least 9 would know it if not all 10. A single digit percentage of the control who don't know they got a placebo probably renders it pretty useless.

The BBC did a doco [1] on MDMA as a symptom mitigator for parkinson's disease with a stuntman who had it bad but liked to party at music festivals and saw it completely alleivated the symptoms. They did the two trials thing with him, take a pill (once mdma, another time placebo) and try to perform specific tasks in a gym and grade how well he could do it. The cameras got him saying something like this to the boffins in white coats: "I know this is the real thing because I can do it all easily and THE MUSIC ISN'T LOUD ENOUGH!!!"

[1]https://www.bbc.co.uk/science/horizon/2000/ecstasyagony.shtm...