Another approch,
You could also consider Pascal's wager. You try to be good, whether there is an after life or not. In either case, you win, because being good gives a better life, over all, in the present. So if you reject the Null Hypothesis and are wrong (Type II error) it does't matter because you win in that case as well.
Pascal was a statistician, Decarte was the co-discoverer of analytical geometry (as in Cartesian coordinates). Therefore, the concept of Type I and Type II error is more applicable to Pascal.
Type I error - Accepting the null hypotheses when it is false.
Type II error - Rejecting the null hypotheses when it true.
By the way, statistics began when a gambler asked the mathimatician, Pascal, to help him improve his odds of winning. From that shady beginining, many important scientific contributions have occurred by the sophisticated use of statistics. For an example, look at 1999 volume of Psychological Bulleton for the Meta Analysis by Hunter and Schmidt regarding the most important predictors of job success. Simple topic, complex stats. Hunter and Schmidt's work on validity generalization in psychological testing won them a major scientific award.
Another interesting study is the work on Black Jack by a PhD. mathimatican, Thorpe, that turns the game into an even or better bet for the player. Using applied statistics, means that you are not gambling when you play this game properly you will in the long run - win. I have emopirical evidence to verify that.