Working With the Science Speaks Peter Stoner Framework
Peter Stoner was a professor of mathematics and astronomy at Pasadena City College. He took a straightforward approach to calculating the odds of biblical prophecies being fulfilled by chance. His method isn't particularly complicated, but it gets pushed around a lot without people understanding what it actually does or doesn't prove. The basic setup goes like this. You take a prophecy, you count how many specific details are involved, and you assign a probability to each detail being fulfilled randomly. Then you multiply those probabilities together. That gives you the overall odds. Stoner is best known for applying this to messianic prophecies, writing it up in his book titled Probability or Promise, though many people know him through the later documentary called Science Speaks. Here is a practical example that makes it concrete. The prophecy in Zechariah 9:9 says the king would enter Jerusalem riding on a donkey. There are roughly six meaningful details packed into that one sentence. If you assume each detail has about a one in a hundred chance of happening by random coincidence, and you multiply those six probabilities, you get one in a billion. When you scale that across dozens of prophecies, the numbers get astronomical. That is the core argument.
The math itself is just conditional probability applied sequentially. Most of what matters is how you assign those individual probabilities. That is where things get messy in practice.
What Actually Happens When You Run Through the Calculations
I spent a few weekends going through the full dataset myself because I wanted to see if the published numbers held up under closer inspection. The first thing I noticed is that Stoner's original work uses some fairly generous probability assignments for individual events. A one in a hundred for each detail is a round number that works for demonstration but gets pushed further than the evidence supports when you dig into any specific prophecy. Take the birth in Bethlehem prophecy from Micah 5:2. That is often treated as a single event with a very low probability. But David himself was born in Bethlehem. Any member of his lineage had a reasonably high statistical likelihood of being born there given the population patterns of the first century. The one in several thousand figure that sometimes gets quoted for Bethlehem births ignores that context entirely. Once you account for it, the probability jumps significantly, which changes the final multiplied result. That said, even with adjusted numbers, the overall order of magnitude stays in the astronomical range for the full set of prophecies combined. The conclusion doesn't collapse. It just shifts slightly depending on your assumptions.
Get the Full Details
Common Pitfalls That Come Up
The biggest issue people run into is treating the prophecies as independent events when they are not. Multiplication of probabilities only works cleanly when each event is statistically independent. In the Stoner framework, many of the prophecies reference the same person and overlapping circumstances. Correlation between events means the true probability is higher than the simple multiplication suggests. This is not a small correction either. It can easily shift the result by several orders of magnitude. Another issue that comes up constantly is the vague wording of ancient prophecies. Different translators render the Hebrew and Greek texts differently, and what looks like a precise prediction in one translation might be rendered quite differently in another. This makes pinning down exact probability values nearly impossible for any single prophecy. You end up working with ranges instead of fixed numbers, which the original presentation does not always make clear.
A Workaround I Found Useful
When I ran into trouble with the independence assumption, I shifted to using a Bayesian framework instead. Rather than multiplying fixed probabilities, I assigned prior probabilities based on historical plausibility and then updated those probabilities as each fulfillment occurred. It gave me results that were more conservative but also more defensible. The final odds came out lower than Stoner's multiplication method suggested, but still strikingly small. For anyone working with this material seriously, I would recommend starting with that approach rather than the straight multiplication model. This framework only works if you accept the underlying premises. If someone does not consider the prophecies to have been written before they were fulfilled, or if they interpret the prophetic texts differently, the entire exercise falls apart. The science part of Science Speaks is real mathematics, but the inputs are heavily dependent on theological and historical assumptions that are debated within the field. That is worth being upfront about. The material is widely available online. The documentary Science Speaks can be found through various Christian media distributors and YouTube channels. Stoner's original book Probability or Promise is still in print through Baker Books. There is also a companion textbook called Extreme Probabilities that goes into more detail on the statistical methods used. All of these are legitimate sources to work from rather than relying on secondary summaries.
The framework itself has limitations that go beyond the independence issue. It treats historical prophecy fulfillment as a clean binary event, but historical records from the first century are fragmentary and sometimes contradictory. The New Testament documents themselves are the primary sources for most of the alleged fulfillments, which means you are evaluating internal consistency as much as statistical probability. That circularity is unavoidable but rarely acknowledged in presentations of the argument.
