Content uploaded by Siobhan Parish
Author content
All content in this area was uploaded by Siobhan Parish on Mar 30, 2015
Content may be subject to copyright.
Journal of Interdisciplinary Science Topics
Predicting the First Recorded Set of Identical Fingerprints
David Evans & Siobhan Parish
The Centre for Interdisciplinary Science, University of Leicester
24/03/2015
Abstract
Fingerprints have been used to identify criminals in the UK since the beginning of the 20th century, with 1901
marking the initial development of Scotland Yard’s fingerprint database. Since this time the UK database has
continued to grow and now has approximately 7 million sets of fingerprints on record. Sir Francis Galton’s 1982
calculations stated that there is a 1 in 64 billion chance that two fingerprint sets are identical. Using these match
probability calculations and the average yearly growth of the database, this paper shows that it will be at least
1,042,277 years before the British database will contain two sets of identical fingerprints.
Introduction
A fingerprint is an impression left by a special type
of skin found on the tips of a person’s fingers. This
skin, known as friction ridge skin, forms in the womb
during pregnancy and leaves a distinctive ridge
pattern that is unique to an individual. Due to this
individuality and the fact that they remain constant
throughout a person’s life [1] - providing the
individual’s fingertips are not subjected to deep
scarring [2] - fingerprints can be used as a means of
identification [1].
Fingerprints are classed as identical if a significant
number of the ridge patterns minute details match
up. The evaluation of these details and the
determination of fingerprint matches is left to highly
trained fingerprint experts who are the only ones
with the power to declare a fingerprint match within
a criminal trial [1]. While this is common global
practise, it is important to note that different
countries have different standards of what
constitutes a fingerprint match. Until 2001 the UK
used a 16 point match standard (now a match is left
to the discretion of the expert) where as other
countries have different standards, e.g. Australia
uses a 12 point standard [3].
Fingerprints started to be become a recognised
means of criminal identification towards the end of
the 19th century. While many historical figures
contributed to their eventual worldwide use, Sir
Francis Galton was the first person to publish
mathematical evidence that fingerprints were
unique to an individual. In his 1982 publication
“Finger Prints” Galton’s calculations showed that
there was a 1 in 64 billion chance of two fingerprints
sets being identical [1]. His work gave mathematical
proof that fingerprints were unique to individuals
and lead Sir Edward Henry establishing the Henry
Classification system in 1901. Henry presented this
system to Scotland Yard in 1901 and this lead to the
establishment of a British fingerprint database [4].
Since this time the database has continued to grow
and is currently stored on the IDENT1 computer
system, which to this date contains approximately 7
million sets of fingerprints [5]. This paper models the
average rate at which the British fingerprint
database has grown each year since 1901 and uses it
to establish how long it would take for the British
database to contain two sets of identical fingerprints
based on Galton’s original calculations.
Identifying an Identical Match
As previously stated there is a 1 in 64 billion chance
of identical set of fingerprints according to Galton’s
probability (P).
Fingerprints have been collected and classified by
the Henry classification system by Scotland Yard
since 1901. If it is assumed that there have been 7
million prints collected for the British database (as
Predicting the First Recorded Set of Identical Fingerprints, March 24th 2015
stated in 2014) [5], that no prints have been deleted
from the records, and that there has been a steady
collection of prints from the day they first started
the records, the number of prints collected per year
(N) can be determined for the 114 years.
prints per year
It is necessary to note here that that there is not an
even number of prints that would have been
collected per year. Since the time that print
collection began many new techniques for
visualising and lifting fingerprints has been found, as
well as more efficient methods for print collection. It
is therefore presumed that there will have been a
greater number collected in the past 50 years
compared to the beginning few years.
Once the value for the number of sets collected per
year was defined, it could then be applied to
Galton’s value for the probability of finding an
identical set of prints. From the probability of 1 in 64
billion it is assumed that the 64th billion set of prints
will be exactly identical to one set collected prior to
it. Using 61404 prints per year and the probability, it
was calculated what year the first identical set of
prints will appear in the database (T).
As prints were not collected until 1901 this is added
to the year in order to determine the final date for
the first set of identical prints: 1044178 AD.
If fingerprints were collected from the moment that
anatomically modern humans were present and
roaming the earth – 200,000 years ago [6] – then
this year would be reduced to 842277 AD.
Conclusion
The validity of using fingerprints in court trials has
been questioned due to the similarities between
different prints, and the difficulty with distinguishing
between certain prints.
However, from our calculations here, it can be seen
that using Galton’s value for probability, an identical
print will not be found in the British database for
over a million years (1,044,178 AD).
The rates of print collection in other countries have
not been accounted for in our calculation; however,
it is assumed that although some countries – such as
the USA – may have collected more prints due to
larger populations, the raw number of prints will not
differ by any order of magnitude. It can therefore
still be assumed that even if all the databases were
collated, it would still be many hundreds of
thousands of years before an identical fingerprint
set is identified.
References
[1] Jackson A.R.W. & Jackson J.M. (2011) Forensic Science, 3rd Edition. Pearson.
[2] Yoon, S. (2012) Altered Fingerprints: Analysis and Detection. Pattern Analysis and Machine Intelligence.
34, pp 451.
[3] Lynch, M. (2003) God’s signature: DNA profiling, the new gold standard in forensic science. Endeavour,
27, 2, pp 93-97.
[4] Hawthorne, M.R. (2008) Fingerprints: analysis and understanding. Boca Raton: CRC Press
[5] The Scottish Government (2014) Fingerprint Database - IDENT1. Available:
http://www.gov.scot/Topics/Justice/law/dna-forensics/scottishdnadatabase/ident1 [Accessed
16/03/2015].
[6] Barton, N., Briggs, D., Eisen, J., Goldstein, D. & Patel, N. (2007) Evolution, 1st Edition. Cold Spring Harbor
Laboratory Press, Cold Spring Harbor, New York.
