I don't think one would simply ignore the dice, and what data is the frequentist drawing upon in the comic which specifies the null?
How about "the probability of our sun going nova is zero and 36 times zero is still zero"?
Although... continuing with the XKCD theme if you divide by zero perhaps that would increase the odds. ;)
http://xkcd.com/1132/
Is this a fair representation of frequentists versus bayesians? I feel like every time the topic comes up, 'Bayesian statistics' is an applause light for me, and I'm not sure why I'm supposed to be applauding.