Ballistic Imaging: Not Ready for Prime Time

Policy Backgrounders | Crime

No. 160
Wednesday, April 30, 2003
by David B. Kopel, J.D., & H. Sterling Burnett, Ph.D.


  1. Frederic A. Tulleners, Laboratory Director, Bureau of Forensic Services, California Dept. of Justice, Technical Evaluation: Feasibility of a Ballistics Imaging Database for All New Handgun Sales (Oct. 5, 2001), pages 2-2 to 2-3, percent20Evaluation.pdf.
  2. Tulleners, page 2-4.
  3. Tulleners, pages 2-2 to 2-3 and 3-1 to 3-2. For automated imaging, only the firing pin impressions, breech face marks, and ejector marks are used. Tulleners, page 3-2.
  4. Some bolt-action rifles do show ejector markings. Shotgun shells, like .22 cartridge cases, are often of very little value for ballistic identification. Tulleners, page 2-4. See also Jan De Kinder, Review AB1717 Report. Technical Evaluation: Feasibility of a Ballistics Imaging Database of All New Handgun Sales, page 17, "Shotgun shells do not carry many marks..."
  5. De Kinder, page 17.
  6. Tulleners, pages 2-4, 6-3; De Kinder, page 17.
  7. De Kinder, page 17.
  8. Peter Weiss, "A Shot in the Light," Science News, Jan. 11, 2003.
  9. Tulleners, page 5-2.
  10. The Glock has a quite distinctive breech face and other marks. Tulleners, pages 3-3 to 3-4. A test conducted by FTI, the manufacturer of IBIS, reported a correlation rate of 83 percent to 85 percent for Glock 9mm pistols - far higher than the 38 percent to 62 percent correlation for Smith & Wesson pistols in the California study discussed below. De Kinder, page 14. Real-world evaluation of ballistic imaging must take into account the fact that there are many varieties of handguns other than Glocks. Although it is debatable how accurately BATFE gun traces reflect actual patterns of criminal gun use, it should be noted that a BATFE 2000 report found Smith & Wesson guns in the top 10 of nearly every major category of handgun trace requests, while Glock pistols never appeared in the top 10 in any category. Department of the Treasury, Bureau of Alcohol, Tobacco and Firearms, Crime Gun Trace Reports (2000), pages 13-22, Tables 5-6, 8. On page 18 of the same report, BATFE did state that Glocks were "frequently recovered" in some cities, albeit not often enough to put Glock in the top 10 for juveniles, youths or adults from whom guns were seized,
  11. Tulleners, pages.1-1 and 2-4. Bullets from .22 handguns are especially difficult to identify. Being small, .22 bullets deform more when they strike a target.
  12. NIBIN Web site,
  13. Tulleners, page 1-1.
  14. "Success Stories,"
  15. The authors thank Larry Keane of the National Shooting Sports Foundation for bringing this data to our attention.
  16. De Kinder, page 12.
  17. De Kinder, page 12, citing Tulleners.
  18. Revolvers don't typically leave shell casings at the scene. The shells remain in the cylinder, to be replaced manually when the shooter chooses.
  19. For some inexpensive guns, the stamped breech faces produce identical ballistic images, so that a bullet from one gun may be mistakenly identified with a different gun, even by an expert examiner. Tulleners, page 3-4 to 3-5. These inexpensive guns are especially likely to produce false hits when included in an automated system. Tulleners, page 3-5.
  20. Trained firearms examiners can usually confirm a match between a gun's early ballistic signature and a later ballistic signature. This does not mean that a computer-matching program can succeed in matching a late-fired bullet to a database image from the gun's youth. Tulleners, page 1-1.
  21. Tulleners, page 2-4.
  22. Also, dirt and lead buildup can partially obscure the breech face impression. Tulleners, pages 3-2 - 3-3.
  23. Tulleners, page 1-2. The BATFE disputes this point, arguing that the marks do not change, although they may be shallower or deeper on different brands. The evaluator commissioned by the California Attorney General explained: "Both opinions can be easily brought together by noting that as the depth of mark decreases, it will be invisible even for microscopic observations." De Kinder, page 11.
  24. Tulleners, Appendix F, reprinting Bill Twist, "Erasing Ballistic Fingerprints," (downloaded Aug. 10, 2000).
  25. Tulleners, pages 1-5, 8-9 to 8-10.
  26. De Kinder reported that about 10 percent of the guns examined by his laboratory in Belgium had erased serial numbers and estimated that the rate of ballistic image alteration would be about the same. He added: "Whereas the BATFE sees altering a firearm as a non-issue, it is a real problem: Any reduction in the proportion of 'hits' caused by such an alteration to a firearm is of concern when evaluating the usefulness from a technical point of view of a 'gun sales database.'" De Kinder, page 17.
  27. Anthony A. Braga, Harvard University, "Reducing Gun Violence in Boston: Intervening in Illegal Markets," presentation to the annual meeting of the Academy of Criminal Justice Sciences, Anaheim, Calif., Mar. 8, 2002. As criminal awareness of ballistic imaging grows, some criminals may attach "brass catchers" to their semiautomatics. (Brass catchers are inexpensive nets which catch the cartridge case being ejected from a semi-automatic.) Criminals might also retrieve cartridge cases belonging to innocent people (hundreds or thousands of empty cases can be found at shooting ranges) and drop them at a crime scene. Soft lead bullets (as opposed to lead bullets with copper jackets) are more likely to deform greatly on impact, reducing or eliminating their forensic utility. Sophisticated criminals could use ammunition with old-fashioned black powder, which burns much dirtier than modern smokeless powder and would obliterate most of a ballistic image. Tulleners, Appendix F.
  28. Gail Gibson and Dennis O'Brien, "Ballistic 'fingerprint' database isn't foolproof tool, experts say. System unable to account for stolen firearms, wear," Baltimore Sun, October 15, 2002.
  29. Tulleners, p. 1-1.
  30. Other participants included firearms examiners from the Los Angeles Police Department, Orange County Sheriff's Department, Oakland Police Department and the Sacramento County District Attorney's Office.
  31. Tulleners, page 1-3. De Kinder estimated that an actual California database would include at least 3,300 guns of this model. De Kinder, page 8.
  32. Tulleners, pages 1-4, 7-2 and 8-1 to 8-4.
  33. Tulleners, pages 1-4, 7-3 and 8-7 to 8-9.
  34. Tulleners, page 1-5.
  35. Tulleners, page 8-12.
  36. Tulleners, page 1-2.
  37. Tulleners, pages 1-2 to 1-3.
  38. Tulleners, page 1-2 and page 6-2.
  39. Robert M. Thompson, Jerry Miller, Martin G. Ols and Jennifer C. Budden, National Integrated Ballistics Information Network; Bureau of Alcohol, Tobacco and Firearms, U.S. Dept. of the Treasury, Ballistic Imaging and Comparison of Crime Gun Evidence by the Bureau of Alcohol, Tobacco and Firearms (May 13, 2002).
  40. Thompson et al., page 15.
  41. De Kinder, Review AB1717 Report. Technical Evaluation: Feasibility of a Ballistics Imaging Database of All New Handgun Sales.
  42. De Kinder, pages 14 and 15.
  43. De Kinder, page 3.
  44. De Kinder, page 3.
  45. De Kinder, page 3.
  46. De Kinder, pages 3-4, 8-12. The Federal primer had a hardness (measured on the Vickers Hardness scale) of 108. Other brands measured 157 (Remington), 114, 159, 186, and 166, De Kinder, page 9. Details about the Vickers Hardness measurement standard can be found at the Web site of the United Kingdom's National Physical Laboratory, De Kinder explained that other factors besides primer hardness also affect how well a cartridge case will accept breech face markings. These factors include primer seating and buildup of gas pressure. Overall, a 1997 test by the Forensic Institute in the Netherlands showed Federal cartridges to be medium in their acceptance of breech face markings. De Kinder, page10. Federal cartridges accept distinctive breech face marks and firing pin marks far better than do Remington cartridges. De Kinder, page 11.
  47. De Kinder, page 9, note 6.
  48. De Kinder, pages 4 and 14.
  49. De Kinder, page 12; Tulleners, page 1-2.
  50. John Lott, "Bullets and Bunkum," National Review, Nov. 11, 2002, page 30.
  51. The Maryland and New York programs use the IBIS software program, but they are not part of the BATFE's NIBIN database for crime guns, since federal law requires that NIBIN compile only images for crime guns. Thompson et al., page 4. (Of course Maryland and New York police can still use the NIBIN database.) Even without the express limitation for NIBIN, the federal statute forbidding the compilation of federal registration of (law-abiding) gun owners may, arguably, make it illegal for the Maryland and New York databases to be integrated into NIBIN. 18 U.S. Code sect. 922(t)(2)(C).
  52. Leslie Koren, "Matching a Firearm to a Crime," The Record (Bergen County, N.J.), Nov. 4, 2002.
  53. Grand Lodge, Fraternal Order of Police, F.O. PAGE Viewpoint: Ballistics Imaging and Comparison Technology (Washington: 2002),
  54. "New Device May Offer Alternative to Ballistic Imaging," The Firearms & Outdoor Trade, Dec. 1, 2002, page 4.

Read Article as PDF