Simple version

  1. The probability of Reliability on Naturalism is low.
  2. The probability of Reliability on Theism is high.
  3. So, Reliability supports Theism over Naturalism.

 

Defeater version

  1. The probability of Reliability on Naturalism is low or unknowable.
  2. If 1, then we have an undercutting defeater for all our beliefs if Naturalism is true.
  3. We do not have an undercutting defeater for all our beliefs.
  4. So, Naturalism is not true.

 

Self-defeat version

  1. The probability of Reliability about Highly Theoretical Matters on Naturalism is low or unknowable.
  2. If 1, then we have an undercutting defeater for all our highly theoretical beliefs if Naturalism is true.
  3. Naturalism is a highly theoretical.
  4. So, if naturalism is true, we have an undercutting defeater for believing it.