you can study patterns of other poker players or even watch tape of them to look for tells

I wouldn't call that "science," any more than a lion sizing up a herd of antelopes to identify which one it's most likely to catch is practicing science. Say you identify an individual person's tell, are you then going to publish a paper about it, send it out for peer review? If the player in question reads that paper, don't you think they'll try to fix it? And even if not, then when they die, all your work becomes useless, you leave nothing for future scientists to build on, because you were only studying the tells of that individual.

I don't think it's really a controversial claim to say that some questions fall outside the domain of science, I believe the majority of scientists would agree with that.