If you are familiar with the evangelical style of Ray Comfort and Kurt Cameron, then you will be familiar with the question “what if you’re wrong?”. For those of you who are not, it is a tactic that attempts to highlight in the minds of atheists the repercussions if they are wrong about the existence of the god that Comfort and Cameron believe in (i.e. that the atheist will go to Hell). I have seen many refutations of this argument (which is essentially Pascal’s Wager), however, I am going to actually do the opposite; I think it is a valid question in certain contexts and should be answerable by any person about any belief that they hold.
To demonstrate why I think this is the case, I would like to modify a scenario used by Richard Carrier in his book, Sense and Goodness Without God;
Suppose a friend told you they had purchased a new car, would you believe them? As this is a fairly unremarkable claim (many people own cars), it would require very little evidence for you to believe them, perhaps even just their word alone. However, suppose now that you were relying on this friend to drive you to a very important meeting. Would you be willing to rely on just their word or would you require more evidence now that the claim has the potential to impact upon your life? If you believe them, and they are wrong (either by lying or just being misinformed; say they thought the car would be ready for their use on that day, but it was delayed), you are now stuck without a way to get your meeting.
The point that I am trying to drive at is that the amount of evidence needed to support a claim is not simply just how ordinary or extraordinary the claim is, but also how much of an impact the claim’s truth or falseness will have. Claims that will have very little effects require less evidence than claims that will have profound effect, all other things being equal. The way in which Comfort and Cameron use this question is still wrong; in that, they are essentially throwing in a possibility, Hell, which has such a low probability of actually existing that it isn’t worth considering. As such, the question is only valid when used in the context of known negative outcomes. However, when used in this way, it is very useful at highlighting how effects can impact upon our evidential standards.