Ganzfeld experiments, controls

I was watching this
and he talked about ganzfeld. A very important point he hasn't mentioned: did the experiments have controls where a computer picked the images, or a person picked but didn't try to mentally send the image, and are the guess results in this way consistently 25%? If they're getting 25% consistently in this way, and 32% when the image is being mentally sent, there's something really up.