> If the second group solves it consistently faster than the first one They can also transmit cooperation information non-verbally, by posture for example, so

> If the second group solves it consistently faster than the first one They can also transmit cooperation information non-verbally, by posture for example, so your experiment testing for speed would test two things, possibly cancelling each other. It's possible that the verbal communication exists but lowers the efficiency of the cooperation (maybe one of them is yelling at the other how stupid he is and that they should use his superior technique).