In order to define best criteria that must be satisfied to establish inference for fingerprint exclusion, research was performed to seek out and find samples of largest and best amounts of non-corresponding ridge features, absent clear distortion markers, in two impressions from the same source.  The number of samples bearing non-corresponding Level I ridge features, e.g., different pattern classification, and/or non-corresponding Level II ridge features, e.g., ending ridge features, bifurcating ridge features, and so on, were collected and counted.  In addition, the number of samples bearing non-corresponding Level I and/or Level II ridge features absent clear distortion markers and an amount of corresponding ridge features sufficient to infer positive identification (based on T-Model theory) were collected and counted.  The results were as follows:

The number of samples bearing non-corresponding Level I ridge features absent clear distortion markers was 1.  The number of samples bearing non-corresponding Level I ridge features absent clear distortion markers and an amount of corresponding ridge features sufficient to infer positive identification was 1 (see below Miller Match).

The number of samples bearing non-corresponding Level II ridge features absent clear distortion markers was more than 1.  The number of samples bearing were non-corresponding Level II ridge features absent clear distortion markers in two impressions from the same source and an amount of corresponding ridge features sufficient to infer positive identification was more than 1 (see below 5 samples).

The number of samples found in which there were non-corresponding Level I and /or Level II ridge features absent clear distortion markers and contained an amount of corresponding ridge features insufficient to infer positive identification was 0.

The following images display 5 samples of largest and best amounts of non-corresponding ridge features found in a two fingerprints from the same source:

No. 1:  Miller Match

The following "Miller Match" displays non-corresponding ridge features absent clear distortion markers from the same source:

Miller Match  (Source:  Judith Miller / www.clpex.com)

No. 2:  Ashbaugh Match

The following "Ashbaugh Match" displays non-corresponding ridge features absent clear distortion markers from the same source (Exemplar A):

Exemplar A

The red arrows point to continuous ridges absent clear distortion markers in Exemplar A that fail to appear in Exemplar B (same source - see below).

Exemplar B

Exemplar A

Ashbaugh Match (Source:  David Ashbaugh / Ridgeology Course, Portland, OR 2007)

No. 3:  Triplett Match

The following "Triplett Match" displays non-corresponding ridge features absent clear distortion markers from the same source:

Triplett Match (Source:  Michele Triplett / www.clpex.com)

No. 4:  Templeman Match

The following "Templeman Match" displays non-corresponding ridge features absent clear distortion markers from the same source:

The red arrow in Exemplar B shows a visually clear and apparently reliable non-corresponding bifurcation not present in Exemplar A (see below close-ups).

Exemplar B (close-up)

 Exemplar A (close-up)

No. 5:  Maciel Match

The below "Maciel Match" displays non-corresponding ridge features absent clear distortion markers from the same source.

Note:  During the analysis phase of Exemplar A, ten (10) latent print examiners (4 CLPEs) all deemed the non-corresponding ridge feature (red arrow) appeared reliable enough such that they would expect and predict it to be present in other exemplars from the same finger.

Maciel Match  (Source:  Dawn Maciel, Latent Print Examiner II, San Jose PD Central ID Unit)

Exemplar B

 Exemplar A

Close-up of non-corresponding ridge feature (marked with a red dot and red arrow).

 Exemplar B

 Close-up of continuous ridges in between two ridge features marked with red dots (not present in Exemplar A).

Conclusion

The likelihood of finding an amount of non-corresponding ridge features absent clear distortion markers and an amount of corresponding ridge features insufficient to infer positive identification in two impressions form the same source is so remote as to be considered a practical impossibility. 

As a result the following criteria is used to define the T-Model's standard to infer fingerprint exclusion with a degree of probability that borders on certainty:

• An amount of non-corresponding ridge features absent clear distortion markers, and
• Insufficient corresponding ridge features to infer positive identification, and
• Applied to all comparable anatomical areas, and
  
• Reproducible conclusion.

Note:  The "non-corresponding" ridge feature is defined as a Level I or Level II ridge feature absent clear distortion markers present in one fingerprint impression and not in another which fails to lends itself to reasonable explanation.

Example No. 1

If the two fingerprint impressions in the Miller Match (see above) were visually faint and blurred to the extent the corresponding Level II ridge detail were not visible, and the non-corresponding Level I ridge feature, e.g., pattern class, were visible, then the conclusion would be “the two impressions can be excluded as being from the same source with a degree of probability that borders on certainty”.

Example No. 2

If the Maciel Match (see above) displayed the one clear, apparently reliable non-corresponding ridge feature and only a small amount of corresponding Level II ridge features, e.g., insufficient to infer positive identification, then the conclusion would be “the two impressions can be excluded as being from the same source with a degree of probability that borders on certainty”.

Note

The T-Model standard for fingerprint exclusion has a zero error rate to make an erroneous exclusion (so far) for naturally occurring, e.g., not cropped or digitally manipulated, fingerprints.  Athough the T-Model can theoretically produce false negatives, based on it’s current zero error rate to make false negatives, the T-Model’s standard to establish inference for fingerprint exclusion is considered more accurate and therefore the preferred standard to use in criminal casework.

Inference for fingerprint exclusion depends on the relevant population for the case at hand, because the larger the population group the more same-source fingerprint impressions displaying non-corresponding ridge features will exist. 

Like all scientific theories, the following T-Model demarcation for sufficiency to infer fingerprint exclusion  is fixed and uncertain.  Like all scientific theories, it is refutable, testable and falsifiable by experiment [93].  

The mathematical equations that define the T-Model's demarcation for sufficiency to infer fingerprint exclusion is two-fold as follows:

NRF ≥ 1

T < RP

where,

NRF = The number of Non-Corresponding Ridge Features present in two impressions absent clear distortion markers which do not lend to reasonable explanation, and

T = T-Value, e.g., the aggregate quantitative-qualitative weight of the amount of corresponding ridge feature present in two impressions), and

RP = Relevant Population for the case at hand, e.g., the number of parts per fingerprint in the plausible number of fingers that can be the source of the latent fingerprint impression for the case at hand.

Example

Let a latent v. exemplar display 1 clear, apparently reliable non-corresponding ridge feature and an amount of corresponding ridge features having a T-Value equivalent to the minimum number of least weighted level II ridge features needed to infer identification based on a world fingerprint population of 66 billion, e.g., 12 ending ridge in a funnel in excellent agreement or 10^12. 

Let the relevant fingerprint population group for a case at hand is defined as 3 billion (e.g., 300 million people (roughly the total United States population and equivalent to the population group used by the FBI to match routine DNA profiles) times 10 (conservative upper-bound number of fingers per person).

Based on T-Model formula, the number of fingerprint "parts" (P) is calculated as follows:

T^P = 10^120

(10^12)^P = 10^120

P= 10

and,

the number of look-alikes is calculated as follows:

L = RP/T

L = (10)(3 billion) / 10^12

L = 30 billion / 10^12

L = 0.03

The number of predicted look-alikes is less than 1.  As a result there is a sufficient amount of corresponding ridge features to infer identification.  Therefore based on the presence of the single discrepancy, there is no scientific basis to establish inference for exclusion.

"The one-dissimilarity doctrinbe is bunk"

David Ashbaugh

Fingerprint matches displaying disagreements in ridge formations due to distortion are not uncommon, while as a result of decades of experience by fingerprint examiners around the world, matches displaying unusually clear amounts of non-corresponding ridge formations absent distortion markers or clear “red flags” have been recorded and presented during examiner training. The above fingerprints from the same source represent samples of such events.  The fact that such events have occurred establishes empirical basis that exclusion cannot be based on the presence of a single non-corresponding ridge formation. As a result, the "One-Dissimilarity Doctrine" is a fallacy.