PZ's place is still brimming with activity. I'm generally peeking at samples, lately, rather than keeping all the way up. One comment I bumped into on a non-cracker thread recently makes a point I used to quite often when I ran into anti-science trolls: How do you know if you're wrong?
One of the consistent problems with woo of any form is that they have never stopped to consider the impact of that question. They rarely put their beliefs to any sort of test. Most of the time I've seen them try, they have ready-made interpretations for failure, often shrug it off as a one-time snafu, and generally disregard the test.
The scientific method is about falsification. If a theory is wrong, experiments or new observations can prove it wrong. Every hypothesis has to meet that criteria: Falsifiability. If we're wrong, we have escape hatches we can use to move towards more accurate hypotheses that explain what the old theory couldn't.
That's where the null hypothesis comes in. Let's say we're dealing with the null hypothesis of "Psychic clairvoyance does not exist." That can be proven wrong with experiment, which is the point of various skeptical challenges. Let's take the card reading example as an experiment. Get a magician who knows all sorts of card tricks, and knows all the ways to cheat. Set up an experiment to negate those tricks. If the psychic succeeds at a statistically significant margin, that's a big point in their favor. If they do it repeatedly under tighter and tighter controls, we can be confident. Of course, psychics don't have a very good success rate at this. That's one reason why you don't see Zener cards very often. Having solid, empirical measurements isn't their bag.