The chart on the right illustrates the power of STR profiling. The upper line in red shows theoretical match probabilities for a fictional dog having the most common allele at each locus. The match probability is calculated with the simple product rule and uses no correction factor for population substructure, or the allele dropout common in sub-optimal DNA samples. With the standard 10-marker panel used in forensic cases, the Match Probability exceeds 1 in one million for all but the most inbred dog breeds.

In forensic cases it is more correct to express these statistics as likelihood ratios. A likelihood ratio is a measure of how much more likely it is that the reference animal is the source of the evidence than some other random animal. In most forensic casework, that can be considered the "prosecution's hypothesis". In the calculation, the actual allele frequencies found in the case are used as well as an adjustment for the population substructure found in dog breeds and further adjustments for missing information (such as a partial profile). A common scale of interpretation is that a likelihood ratio from 10 -100 "supports" the prosecution's hypothesis, from 100 -1000 "strongly supports" it while a ratio exceeding 1000 "very strongly supports" it. In the chart above, the line in blue shows the likelihood ratio in an actual example typical of successful casework with dog hairs. In the case of a match, each marker that is successfully typed adds additional weight to the likelihood ratio. In this example. markers that did not amplify are indicated by the lack of increase in the likelihood ratio (the blue line remains flat). While the conservative numerical value of the likelihood ratio is far less dramatic than the match probability, it has been well accepted in court cases (see Closed Casework).