Assume a coin is fair or completely biased towards heads or tails.

Let F = the coin is fair
Let N = the coin lands on n heads in a row

P(F|N) = P(F) 1/2^n / (P(F) 1/2^n + (1-P(F))/2)

P(F|N) < 1/2 when P(F) < 2^n/(2^n+2)

So we should only think the coin is biased if our initial credence that the coin is fair was less than 2^n/(2^n+2). Thus one cannot simply conclude that any coin which lands on some large number of heads is biased, it depends on our initial belief that the coin is fair.

So the creationist argument regarding the supposed improbability of abiogenesis is incomplete.