[QUOTE=YogiAdam;48983]I mean “is it a car” is pretty random in my opinion, and that could easily be a guess, so I’m not impressed… but I’m not that stubborn… the whole poles turning into tires on a racing car thing… now that’s a tad freaky lol 
Now, I’m a scientist and skeptic… so I’m not going to start just believing in stuff just cause I’m impressed by a lucky guess… maybe we should try it again to minimize the possibility of random chance :)[/QUOTE]
Nice one Kareng! I was waiting for conclusion of this “game”.
Adam, if you want to approach it from more scientific point of view, I guess we should assess probability of a random guess. Repeating whole test will not give any new information - it can always be happy guess. To assess probability, we would need a number of images that average person (or Kareng in our case) can generate. Is it 100, 1000 or 10000? Also, are there images that are often repeated? (e.g. dress, house, dog). But its not very feasible…
Now I recall that people testing such things use numbers. Its very easy to assess probability of a random guess. And if probability is below given threshold it is considered as not random (e.g. in clinical sciences is probability lower than 0.05 - but here it should be much lower I guess). Sad thing is that we can’t be 100% sure because it can always be a random guess.