There is a concept of “degree of being wrong”. In this area of thinking, you can have two wrong statements, but one is “more wrong than the other”. For instance, if I am in London, and I say two things: “I am standing in Brussles” and “I am standing on Saturn”, both can easily be classified as wrong. However, when measured, the first statement is *drastically* less wrong than the second. In this example, instead of measuring wrongness, we are attempting to measure unknown-ness. By virtue of the structure of the claim, we can impart some level of likelihood of correctness – even without considering the specific claim itself. “I have seen a purple polar bear” : the speaking individual claims to have seen 1 extraordinary version of a known real animal. “every Polar bear in Juno Alaska is purple” : the speaking individual claims that every instance of a known real animal within the borders of an entire city is extraordinary. With out even looking at the original claim, it should be fairly easy to determine that the first statement has a MUCH better chance of being correct, simply because it requires less to be true. ONE instance of ONE oddity seen by ONE person is enough. The second statement most likely requires much more for it to be true – be it many instances of an extraordinary event spread over a fairly large area, or a trickery in the wording that changes the suggested meaning of the original statement. So without *KNOWING* anything more than the structure of the unknown claims, one *should* be placed closer to correct than the other. It should never be claimed as true without further evidence, but it should also not be considered equal to the other until such time as both are determined true. To end this logical exersize, here is the purple polar bear that I have seen. It got a fungal infection in the hollow space of its transparent hairs, which turned it green. The medication to kill the fungus turned it purple for a while. http://www.nwbotanicals.org/mediawatch/purplebear.htm This idea of "improbable but guaranteed" is handled nicely in video #2. Picking out 8 coin flips in a row is very improbable. However, if you have 100 people, and 50 pick heads, the other 50 tails, the half will be right. Divide up those people for round two into 25 heads and 25 tails; again, half will be right....so on and so forth. eventually, you'll end up with one person who has picked the "random" event multiple times in a row correctly, defying logical chance. But not because they are psychic. Instead, because you have enlarged the sample size to the point where the unlikely becomes a given. As I said, I don't always agree with Dawkins - to me, the non-existence of God is just as improvable as the existence. As such, the theory he supports for abiogenesis, to me, didn't *have* to happen, but it's a hell of a lot more likely than the existence of an omnipotent self-aware all-pervasive, but yet invisible creator being. Especially once you start dealing with billions upon billions upon billions of molecules in billions upon billions of tiny pools of water on billions upon billions of hills of semi-dry land on billions of planets around billions of stars in billions of galaxies.