Chapter 15: Firearms and Ballistics: What a Bullet Tells You (and the Limits of "Matching")

"The fired bullet and the fired cartridge case are the two most eloquent witnesses a firearm leaves behind — but like any witness, they can be misread, and the examiner who forgets that is the most dangerous person in the room." — constructed teaching epigraph, in the practitioner voice of this book

Overview

A gun is, mechanically, a machine for doing violence to a small piece of metal at very high speed, and like any machine it leaves its marks on the work it does. A bullet scraped down a barrel comes out scored with fine parallel lines. A cartridge case slammed against the breech and struck by a firing pin comes out stamped, dented, and scratched. The promise of firearms examination — the part that is real — is that those marks carry information: about the kind of gun, sometimes about the individual gun, and about what happened in the few thousandths of a second the weapon was in use. The peril of firearms examination — the part this chapter spends as much time on — is that the field has, for most of a century, claimed to read those marks with a precision the science has never actually demonstrated, and has put the word "match" in front of juries as though it meant what it means on television.

So we ask the investigative question first, before any technique: what is a bullet, or a casing, actually being asked to tell us, and how much of that can it honestly deliver? Sometimes the answer is a great deal — the cartridge cases at a scene came from a gun later found in a suspect's car, and the class characteristics line up so well, with so many corresponding fine marks, that the association is strong. Sometimes the answer is almost nothing — a deformed bullet recovered from a wall carries class characteristics shared by millions of weapons and too little detail to say more. And sometimes, as in our cold case, the most important thing a firearm-related item teaches is that it has nothing to do with the crime at all, and the discipline that matters most is the discipline to say so.

This is a comparison field, like the fingerprints of the last chapter, and it sits in the same uneasy place: real, useful, used every day in real investigations — and flagged by the 2016 PCAST report as a method whose foundational validity has not been established to the standard the courtroom pretends it has. We will neither dismiss it nor oversell it. We will learn exactly what a bullet tells you, exactly where examiner judgment enters, exactly what the databases can and can't do, and exactly what an honest examiner can say on the stand without stepping over the line into the certainty the science has not earned.

In this chapter, you will learn to:

• Tell internal, external, and terminal ballistics apart, and match each to the real questions it can answer.
• Explain how rifling marks a bullet with lands and grooves and striations, and sort those marks into class and individual characteristics.
• Read what a cartridge case records — the breech-face, firing-pin, extractor, and ejector marks — and say why a casing is often a better witness than a bullet.
• Describe how comparison microscopy yields a firearms "identification," and locate the examiner's subjective judgment and the missing error rate inside that process.
• Explain why NIBIN/IBIS produces leads, not matches, and why confusing the two is dangerous.
• Place firearms identification on the validity spectrum, and state the difference between what an examiner may say and what overstates the method — including the moment in our cold case where a firearm item turns out to be a red herring.

Learning Paths

🔎 Investigator/CSI: Your work is at the scene — finding, marking, photographing, and collecting bullets and casings without destroying the very marks that matter, and reconstructing trajectory. Weight §15.1 (terminal ballistics and trajectory) and §15.3 (casings as a map of where the shooter stood). 🧪 Lab analyst: You will sit at the comparison microscope. §15.2–§15.4 are your bench; pay special attention to subclass characteristics in §15.2 and to the honest description of the "identification" decision in §15.4. ⚖️ Law/courtroom: The action for you is §15.4 and §15.6 — what an examiner may testify to, how PCAST and a string of court rulings have narrowed that, and how the "match" is attacked on cross. 👥 General reader/juror: §15.4 and §15.6 are the antidote to the television "ballistics match." Learn the difference between "this casing was fired by this gun, to the exclusion of all others" and what the science can actually support.

15.1 Internal, external, and terminal ballistics

Start with the word, because it is the source of a great deal of confusion. Ballistics is, strictly, the physics of projectiles in motion — the science of how things fly when launched. In ordinary forensic usage "ballistics" has come to mean the whole discipline of firearms examination, including the comparison of marks on bullets and casings, which is not really ballistics at all. We will use the loose courtroom sense because everyone does, but keep the distinction in your head: the part of this chapter that compares striations under a microscope is firearms identification or firearms and toolmark examination; the part that concerns how the bullet moved is ballistics proper. Conflating them is the first small inaccuracy television teaches, and small inaccuracies are how this field gets into trouble.

Ballistics proper divides into three phases, defined by where the projectile is.

Internal (or interior) ballistics is everything that happens inside the firearm, from the instant the firing pin strikes the primer until the bullet leaves the muzzle. The primer ignites, the powder burns, pressure builds in the chamber to thousands of pounds per square inch, the bullet is driven down the barrel, and — critically for us — the barrel's interior shape grips and marks the bullet on its way out. Internal ballistics is where the forensically useful marks are created. It is also where malfunctions live: a failure to feed, a light primer strike, a squib load that lodges a bullet in the barrel. An investigator who understands internal ballistics can read a malfunction as a clue.

External (or exterior) ballistics is the flight of the bullet through the air, from muzzle to target: its trajectory, the drop under gravity, the drift from wind and spin, the loss of velocity to drag. This is the phase that lets an investigator, sometimes, reconstruct where the shooter was standing by working a trajectory backward from where a bullet ended up. It is genuine physics, but in real scenes it is bedeviled by the things physics textbooks leave out: a bullet that ricocheted, deflected off glass, or tumbled after passing through a body does not travel a clean parabola, and a trajectory reconstruction built on the assumption that it did can be confidently, precisely wrong.

Terminal ballistics is what happens when the bullet strikes — the transfer of energy into a target, the wound it makes in tissue, the way it deforms or fragments, the depth and shape of the wound track. In a death investigation, terminal ballistics is shared territory with the forensic pathologist (Chapter 11), who reads the wound to estimate range, direction, and the sequence of shots, and with the anthropologist (Chapter 12) when the wound is in bone. The crucial humility here: a wound rarely tells you the specific weapon. It can suggest a caliber range, a contact versus distant shot, an entrance versus exit — but "what gun" is a question for the marks on the recovered bullet, not the shape of the hole.

🔬 At the Bench Trajectory reconstruction at a scene uses simple, honest tools: rods and string (or, increasingly, laser lines) run through bullet defects — a hole in a wall, a graze in a door frame, a perforation in a car seat — to project the bullet's path backward in space and define a probable region the shot came from, often expressed as a cone, not a line, because the entry and exit points have measurement uncertainty. Done carefully, with two or more defects from the same shot, it can place a shooter within a defensible zone. Done carelessly — a single defect, an assumed straight path, a bullet that actually deflected — it produces a precise-looking line that means nothing. The output should always be a region with stated uncertainty, never a pinpoint. When you hear a trajectory described to three decimal places, ask what defects it was built from and whether deflection was ruled out.

What can the three phases establish, and what can't they? Internal ballistics can establish that a given firearm is mechanically capable of producing a given mark, and can explain a malfunction; it cannot, by itself, say which of two identical-model guns fired a round. External ballistics can constrain where a shot came from; it cannot survive an undetected ricochet. Terminal ballistics can characterize a wound — range, direction, energy — and is strong evidence about how someone was shot; it is weak evidence about which weapon and almost silent about who pulled the trigger. Hold those boundaries. Much of the overstatement in firearms cases comes from letting one phase answer a question that belongs to another.

This division also tells you where, in our running cold case, a firearm question can and cannot reach. The cold case is, the autopsy will eventually establish, a blunt-force homicide disguised by fire — there is no gunshot wound at all. Terminal ballistics, then, has nothing to read; there is no wound track, no recovered bullet from a body. And yet a single fired cartridge case was logged in the cabin's debris. Hold that tension: an item exists that belongs to this chapter's discipline, in a death where the discipline's central phase — terminal ballistics — finds nothing to do. We will return to it, because that gap is the whole lesson.

15.2 Rifling: lands, grooves, and the marks they leave

Here is the mechanism at the heart of bullet comparison, and it is worth getting exactly right.

The inside of most modern firearm barrels is not smooth. It is cut, or formed, with a series of spiral grooves running the length of the bore — rifling. The raised metal between the grooves are the lands; the cut-away spiral channels are the grooves. (The naming trips up nearly everyone at first: the lands are the high parts, the grooves the low parts, exactly as on a landscape — high land, low groove.) The purpose of rifling is aerodynamic, not forensic: the spiral forces the bullet to spin as it travels down the bore, and a spinning bullet is gyroscopically stabilized in flight, far more accurate than a tumbling one, the same way a thrown football spirals to fly true. Rifling is one of the great firearm improvements of the last few centuries, and its forensic usefulness is an accident of its design.

Because the bullet is slightly larger than the bore's land-to-land diameter, it is engraved as it is forced through: the lands bite into it, cutting grooves into the bullet's surface, and the bullet emerges carrying a negative impression of the barrel's interior. Two layers of information come off this engraving, and the entire validity question of bullet comparison turns on telling them apart.

The first layer is class characteristics — features shared by every barrel made to the same specification. These include the caliber (the bore diameter), the number of lands and grooves, their width, and the direction and rate of twist of the rifling (a right-hand or left-hand spiral, turning once in so many inches). A barrel is manufactured to a design, and every barrel of that design imparts the same class characteristics. Read off a bullet, these class features can narrow the field of possible weapons dramatically — "this bullet was fired from a barrel with six lands and grooves, right-hand twist, of roughly .38 caliber" can exclude whole categories of firearm. But class characteristics, by definition, are shared by every gun of that make and model; they cannot point to one weapon. They exclude, and they narrow, and that is a real and underrated power. If the recovered bullet has five lands and grooves with a left twist and the suspect's gun has six with a right twist, the suspect's gun did not fire it — a clean, defensible exclusion of exactly the kind §1.6 taught us to value.

🔬 Read the Evidence

text FIGURE 15.1 — "A fired bullet's rifling impressions" [constructed teaching example] THE ITEM A fired, slightly deformed copper-jacketed bullet recovered from a wooden wall stud, examined under a comparison microscope alongside a test-fired bullet from a suspect gun. THE CONTEXT Recovered intact enough to read; photographed in place, collected without forceps gripping the bearing surface (forceps can scratch in fresh "individual" marks that mimic real ones). WHAT IT SHOWS Six land impressions and six groove impressions, right-hand twist; caliber consistent with .38/9 mm class. Within the land impressions, fine parallel striations run lengthwise. WHAT IT DOESN'T The class features (6 lands/grooves, right twist, ~.38) are shared by very many firearm models — they cannot name one gun. Deformation has obliterated part of the bearing surface, so some striations are simply missing, not "non-matching." THE INFERENCE CLASS-consistent with the suspect weapon (not excluded on class grounds). Whether the striations correspond well enough to support an identification is a separate, harder, and more contestable judgment — §15.4. THE LESSON Class characteristics narrow and can exclude; they never individualize. The leap from "consistent class" to "this gun" is the leap the whole validity debate is about.

The second layer is individual characteristics — and this is where the marketing of the field outran its science. As a barrel is manufactured (cut, button-rifled, broached, or hammer-forged) and then used and worn and corroded, its interior surface acquires microscopic irregularities: tiny tool marks from the rifling process, pits from rust, wear from thousands of rounds. These irregularities, the theory goes, are random and unique to that one barrel, and they scrape a corresponding pattern of fine lines — striations — into every bullet the barrel fires. If the striations on a crime-scene bullet correspond, line for line, to those on a bullet test-fired from a suspect's gun, the argument runs, the two were fired from the same barrel "to the exclusion of all other firearms." That sentence — to the exclusion of all other firearms — is the claim the field made for a century, and it is the claim PCAST and the NAS found unproven. We will hold it up to the light in §15.4 and §15.6. For now, note only the structure: class characteristics are real and validated; individual characteristics are real as a phenomenon (barrels do differ, striations do form) but the examiner's ability to read them reliably enough to individualize is exactly what has never been adequately measured.

🧠 Cognitive-Bias Watch There is a third category, finer than class but coarser than individual, that quietly produces some of the worst firearms errors: subclass characteristics — marks shared by a batch of barrels because they were made in sequence with the same imperfect tool, before the tool was sharpened or replaced. A reamer or broach that is itself nicked imparts the same "random-looking" striations to every barrel it cuts that day. An examiner who does not recognize a subclass mark for what it is can declare two bullets a "match" when they were in fact fired by two different guns from the same manufacturing run. We met this category in Chapter 1 (§1.3); in firearms it is not a curiosity but a live failure mode, and the safeguard is consciously asking, of every persuasive-looking agreement, "could this be subclass — shared by a batch, not unique to one barrel?" The danger is precisely that subclass marks look like the individuality the examiner is hoping to find.

A complication worth naming: not every barrel is rifled, and not every bullet can be compared. Shotguns fire through smooth bores and shot pellets carry essentially no individualizing marks; some pistols use polygonal rifling (smooth, rounded lands and grooves rather than sharp-edged ones) that leaves shallower, harder-to-read impressions. And a bullet that strikes bone, glass, or steel may be so deformed that its bearing surface — the part that touched the barrel — is destroyed. None of these gives the examiner anything to overstate, which is the honest silver lining: a badly deformed bullet usually announces its own uselessness, and the temptation it presents is not false confidence but the urge to read marks that aren't really there.

15.3 Cartridge cases: breech, firing-pin, ejector marks

Now turn from the bullet to the other half of the fired round, and to a piece of evidence that is often better than the bullet: the cartridge case — the metal cylinder (usually brass) that holds the primer, the propellant powder, and the bullet seated in its mouth, and which is left behind when a self-loading firearm cycles. (In a revolver the cases stay in the cylinder; in a semi-automatic pistol, rifle, or many shotguns they are flung clear, which is why scenes involving those weapons are littered with brass.) The case never goes downrange, never deforms against a target, and is frequently recovered intact — and in the moment of firing it is gripped, struck, and thrown by several different parts of the firearm, each of which can leave its own mark. Where a bullet records essentially one structure (the rifled bore), a cartridge case can record four or five distinct contacts.

Work through them in the order the firing cycle creates them:

• Firing-pin impression. When the trigger releases the firing pin (or striker), it slams forward and dents the primer at the base of the case. The shape, depth, and fine internal texture of that dent reflect the firing pin's tip — its profile (round, elliptical, rectangular on some designs) is a class feature; microscopic irregularities on its face can be individual. A drag mark as the pin retracts (firing-pin drag) is itself sometimes characteristic.
• Breech-face marks. This is usually the richest source. At the instant of firing, chamber pressure drives the case violently backward against the breech face — the flat steel wall of the firearm that closes off the rear of the chamber. The case head is stamped against that wall hard enough to take an impression of its surface texture: the fine tool marks, machining patterns, and wear on the breech face are pressed into the soft primer and case head. Because the breech face is a relatively large, flat, heavily-machined surface, its impression on the case head often carries more detail than the striations on a bullet — which is why, given a choice, many examiners would rather have the casing.
• Extractor and ejector marks. In a self-loading firearm, after firing, the extractor (a hook) grips the rim of the case and pulls it from the chamber, and the ejector (a fixed lug or spring-loaded plunger) strikes the case to kick it clear. Each leaves a mark — a scrape or notch on the rim from the extractor, a dent from the ejector impact — whose location and shape are characteristic of that firearm's design and, sometimes, its individual wear. The position of the ejector mark relative to the firing-pin impression is itself a class clue to the type of weapon.
• Chamber marks. As the case expands against the chamber wall under pressure and is then drawn out, the chamber's interior can scratch its surface — useful occasionally, less commonly relied upon.

🔬 Read the Evidence

```text FIGURE 15.2 — "What a fired cartridge case records" [constructed teaching example] THE ITEM A fired 9 mm brass cartridge case recovered from a floor; examined base-on and side-on under a comparison microscope against a test-fired case from a recovered pistol. THE CONTEXT Collected with gloved fingers or a soft tool, not a magnet (magnets are fine for steel cases but the case must not be dragged across grit that adds fresh scratches).

     BASE OF CARTRIDGE CASE (schematic — not to scale)
        _____________________
       /        rim          \        ← extractor scrape here (3 o'clock)
      |    ___________        |
      |   /  primer    \   E  |       E = ejector mark (dent near rim, ~9 o'clock)
      |  |   ( ● )  ▒▒  |      |       ● = firing-pin impression (center of primer)
      |  |   pin    bf  |      |       bf = breech-face texture pressed across the head
      |   \___________/       |
      |        head           |
       \_____________________/

WHAT IT SHOWS A central firing-pin dent (round, class-consistent with a striker-fired pistol); breech-face texture across the primer/head; an extractor scrape on the rim; an ejector dent. Several of these correspond in position and fine detail to the test-fired case. WHAT IT DOESN'T Even strong correspondence is a similarity judgment, not a counted probability — there is no validated random-match statistic for breech-face marks the way there is for DNA. Position of marks confirms weapon type; only the fine texture argues for the individual gun. THE INFERENCE Strong CLASS agreement; the fine breech-face correspondence may SUPPORT same-source — stated at that strength, never as "identical, to the exclusion of all other firearms." THE LESSON A casing is often a richer witness than a bullet — four or five contacts instead of one — but richer evidence does not come with a probability attached, and "richer" is not "certain." ```

The schematic in Figure 15.2 is the mental map to carry to any scene with ejected brass: the base of a fired case is a small record of four or five separate handshakes between the round and the gun. Each handshake is a comparison opportunity — and each is, again, a class feature in its gross form and a possible individual feature only in its finest texture. The same honest boundary from §15.2 applies: the positions and shapes of these marks reliably tell you what kind of firearm and can exclude a candidate weapon cleanly; the claim that the fine texture identifies one gun to the exclusion of all others is the contested claim.

Cartridge cases carry one further investigative gift that bullets do not. Because semi-automatics eject cases in a roughly repeatable direction and distance, the distribution of fired cases on the floor can help reconstruct where the shooter was standing and how they moved — a scene-reconstruction use that is independent of any "match" and often more reliable than the matching itself. An investigator who maps and numbers every casing before anything is moved is preserving a kind of evidence the comparison microscope never touches.

⚖️ In the Courtroom Notice that the strongest, least contestable firearms testimony is frequently the most modest: "These five cases were all fired by the same firearm" (cartridge-case-to-cartridge-case comparison, which avoids needing the gun at all and avoids the deformation problems of bullets) or "This casing's class characteristics are consistent with — and not excluded as — having been fired by the recovered pistol." Those statements live within what the method can support. The sentence that draws a Daubert challenge (Chapter 5) and a hard cross is the leap to "this casing was fired by this gun and no other gun on Earth." The art of honest firearms testimony is staying on the near side of that leap — and the art of cross-examination is forcing the examiner to admit which side of it they are on.

15.4 Comparison microscopy and the "match"

Everything so far has been about the marks. Now we confront how an examiner turns marks into a conclusion — and where, exactly, the science becomes human judgment.

The instrument is the comparison microscope, the same dual-stage optical bridge used in trace and hair examination (Chapter 19): two microscopes joined by an optical bridge so the examiner sees two specimens — the questioned bullet and a test-fired known — side by side in a single split field of view, and can rotate and align them to see whether their striations or breech-face textures run continuously across the dividing line. When the fine lines on the left specimen flow without interruption into the fine lines on the right as the examiner rotates the known into register, the examiner judges that the two surfaces were marked by the same tool — the same barrel, the same breech face. This is the moment television compresses into a triumphant click. In reality it is a patient, iterative visual comparison that can take a skilled examiner a long time and end in "inconclusive" as often as in a conclusion.

To generate the known, the examiner test-fires the suspect weapon — usually into a water tank or a box of dense gel that stops the bullet without deforming it — recovering pristine bullets and cases that carry that gun's marks under controlled conditions. The comparison is then questioned-versus-known, and the possible conclusions, under the framework long used by the Association of Firearm and Tool Mark Examiners (AFTE), are essentially: identification (same source), elimination (different source), or inconclusive (insufficient agreement to decide either way). It is worth pausing on those words. "Identification" in AFTE usage has historically meant a same-source conclusion expressed with great confidence; "elimination" is the exclusion this book prizes; and "inconclusive" is the honest verdict when the marks simply do not resolve the question — common with deformed bullets, subclass-heavy surfaces, or poor-quality impressions.

Here is the crux, and it must be stated plainly because the field spent decades not stating it. The standard for declaring an identification has historically been "sufficient agreement" — defined, in the AFTE framework, as agreement that exceeds the best agreement the examiner has ever seen between marks known to come from different sources, and that the examiner judges could not reasonably be explained by anything but the same source. Read that definition again and notice what it rests on: the examiner's accumulated experience and the examiner's judgment. There is no fixed number of corresponding striations required, no objective threshold, no statistic. "Sufficient agreement" is, in the end, a trained expert's subjective determination that two things look enough alike. It may be a good subjective determination — experienced examiners are not guessing — but it is a subjective one, and for a century it was presented to juries dressed in the language of objective, near-infallible science.

⚠️ Junk-Science Alert The phrase to flag, in a report or on the stand, is some version of "identified to this firearm to the exclusion of all other firearms, to a reasonable degree of scientific certainty." Each part of that phrase claims more than the method has been shown to support: to the exclusion of all other firearms asserts a universal negative no one has tested against all firearms; reasonable degree of scientific certainty is a legal-sounding phrase with no agreed scientific meaning, which several authorities and standards bodies have urged examiners to abandon. The honest core that survives the deletion of those flourishes is real and worth keeping: "I found agreement in the class characteristics and in the individual marks, exceeding what I have seen between different-source samples, and in my opinion the recovered firearm is the source." Stated as an opinion with its basis exposed, it can be evaluated and cross-examined. Stated as a fact of identity to the exclusion of all others, it is the overstatement that PCAST, the NAS, and a growing line of courts have pushed back against.

So where do errors come from? Four well-documented sources, each a place where the human enters:

1. Subclass characteristics (§15.2) mistaken for individual ones — declaring same-source from marks actually shared across a manufacturing batch.
2. Confirmation and contextual bias. If the examiner knows the detective is confident in this suspect, or has already seen the "match" in a database lead, the threshold for perceiving "sufficient agreement" can drift downward. The fix — analyzing the questioned sample before, and shielded from, the known and the case context — is context management / blind verification, which Chapter 31 treats in full and which most firearms units have not adopted. (The independent re-examination by a second examiner who declares identifications is a partial safeguard, but if the second examiner knows the first one's conclusion, the "verification" can simply confirm a shared error.)
3. The closed-set fallacy. Validation studies that ask examiners to match within a known closed set — "one of these test guns fired this bullet; which?" — can overstate real-world accuracy, where the true source may not be in the set at all and "none of them" must be a permitted answer.
4. Inconclusive as an escape hatch. Some studies and critics note that examiners may classify hard comparisons as "inconclusive" rather than risk an error, which can make measured error rates look artificially low — an inconclusive is neither a hit nor a miss, but in the real world a wrongly-withheld elimination can keep an innocent person in jeopardy.

🧠 Cognitive-Bias Watch The structure of the firearms decision is almost a laboratory model of how bias contaminates a judgment. The examiner is asked a yes/no-ish question ("same source?") with no objective threshold, often while knowing the answer the investigation wants, frequently after a database has already suggested a candidate. Every one of those conditions — subjective criterion, known desired outcome, prior suggestion — is a documented amplifier of error in human judgment (we name the experiments in Chapter 31). This is theme three of the book in its purest form: the biggest threat to a firearms result is not the microscope but the analyst's mind, and the safeguard is not better optics but blind, sequential, context-managed examination.

None of this means firearms identification is worthless — that would be the opposite overstatement. Class-characteristic exclusions are solid. Same-gun grouping of multiple casings is strong. A well-documented identification by a competent examiner who states it as an opinion, exposes the corresponding marks photographically, and concedes the absence of a numerical error rate is evidence — evidence a jury can weigh against everything else. What is not legitimate is the old courtroom theater in which "match" meant "certainty" and the examiner's judgment was presented as a reading of nature. The method is real; the certainty was not.

15.5 NIBIN and database leads

Suppose you have a fired cartridge case from a scene and no suspect weapon to compare it to. For most of the field's history that case was a dead end until a gun turned up. Since the late 1990s, that has changed, and understanding how it changed — and how it is constantly misunderstood — is essential.

The National Integrated Ballistic Information Network (NIBIN), administered in the United States by the Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF), is a national database of digital images of the marks on fired cartridge cases (and, to a lesser extent, bullets) recovered from crime scenes and from test-fires of recovered guns. The imaging and correlation technology underlying it is IBIS (Integrated Ballistics Identification System): a station captures a high-resolution 2-D and 3-D image of a casing's breech-face, firing-pin, and ejector marks, and the software compares that image against the database, returning a ranked list of possible matches — cases whose marks the algorithm scores as most similar. Together, the network and its imaging engine are usually referenced as NIBIN/IBIS.

Now the single most important sentence in this section, the one that prevents the central error: NIBIN/IBIS does not declare matches. It generates investigative leads. The computer's ranked list is a triage tool — a way of saying "these stored casings are worth a human's time to look at." Every candidate the algorithm flags must still be confirmed (or rejected) by a firearms examiner doing the conventional comparison-microscope examination of §15.4. The database finds possible links across cases and across jurisdictions — connecting, say, a casing from a Tuesday shooting to a casing from an unsolved Friday shooting, suggesting the same gun was used in both — and those links can break open serial shootings, gun-trafficking patterns, and cold cases. But the link is a hypothesis for an examiner to test, not a conclusion, and certainly not something to put before a jury as "the database matched it."

🔬 At the Bench The forensic value of NIBIN is overwhelmingly investigative, and its payoff depends on speed and comprehensiveness. The agencies that get the most from it enter every recovered casing promptly — not just the cases they think are important — because the network's power is combinatorial: each new casing can light up links to many old ones. A "NIBIN hit" (more properly, a lead or correlation) typically triggers a notification to investigators that two crimes may share a firearm, often within days, while leads are still actionable. What goes into court later is never "the computer matched these," but the examiner's independent microscopic comparison of the two physical casings, with all the honesty and limits of §15.4. The database opens the door; the examiner — and the examiner's caveats — must still walk through it.

⚖️ In the Courtroom Two errors recur when database evidence reaches a courtroom, and both are worth a juror's and an attorney's vigilance. The first is presenting the algorithmic correlation as if it were the finding — "IBIS identified this gun." It did not; it nominated a candidate. The second is subtler and connects to Chapter 31: a database lead is itself a source of contextual bias for the confirming examiner, who now approaches the comparison already told by the machine that these two casings are probably the same. Best practice has the confirming examination done as blindly as practical, precisely so the human is testing the lead, not rubber-stamping it. A lead that has only been confirmed by an examiner who knew it was a lead is weaker than one confirmed blind — a distinction a good cross-examination will draw out.

NIBIN is, on balance, one of the genuinely valuable infrastructure advances in the field — not because it "solves" the validity problem of firearms matching (it inherits that problem entirely, since human confirmation is still the standard) but because it does something the old workflow could not: it lets unrelated scenes find each other across time and jurisdiction. That is a real, defensible power, as long as everyone remembers it ends at "go look," not "we have our man."

15.6 What firearms ID can and cannot establish (PCAST's verdict)

We can now do for this discipline what Chapter 1 promised to do for all of them: place it honestly on the validity spectrum, using the 2009 NAS report and the 2016 PCAST report as the yardstick.

Begin with what the discipline can establish, because it is real and is sometimes lost in the criticism:

• Class-characteristic exclusion and narrowing. That a recovered bullet or casing could not have been fired by a candidate weapon (different caliber, different rifling count or twist, incompatible marks) is a clean, defensible conclusion — exclusion, the book's prized power.
• Same-source grouping. That several casings, or several bullets, were fired by the same (unknown) firearm is often strongly supportable and is independent of having the gun.
• Functional and reconstruction facts. Caliber, whether a weapon was operable, trajectory regions, casing-distribution reconstruction, range of fire from terminal effects (with the pathologist) — much of this is solid physics and mechanics.
• A same-source opinion, stated as such, by a competent examiner who exposes the basis and concedes the limits — real evidence for a jury to weigh.

Now the hard part — the identification claim, the assertion that a specific bullet or casing was fired by a specific firearm to the exclusion of all others. Here is the documented record, stated carefully:

The 2009 NAS report examined the firearms and toolmark discipline and concluded, in substance, that the field had not demonstrated the scientific basis for its core individualization claim: the "sufficient agreement" standard was unquantified and subjective, and the studies needed to establish how often examiners err had largely not been done. The NAS did not say the method was useless; it said its central claim was scientifically unproven and its language overstated.

The 2016 PCAST report went further and more specifically. Reviewing firearms analysis as one of several feature-comparison methods, PCAST asked whether the discipline had established foundational validity — whether properly designed "black-box" studies (examiners given samples where ground truth is known, error rates measured) showed the method could do what it claimed. PCAST's conclusion was that firearms analysis fell short of demonstrated foundational validity at the time of the report: too few appropriately designed studies, and the existing ones suggesting error rates that, while not enormous, were non-trivial and meaningfully larger than the "essentially zero" the field had long implied. PCAST recommended both that more rigorous black-box studies be conducted and that, in the meantime, examiners refrain from claiming certainty or zero error and instead disclose the measured error rates that did exist.

⚖️ In the Courtroom The legal system has, unevenly, absorbed this. A growing number of courts — federal and state — have, in documented rulings, limited firearms examiners' testimony: barring the phrase "to the exclusion of all other firearms," barring "reasonable degree of scientific certainty," requiring that conclusions be framed as the examiner's opinion, and in some instances permitting only that a firearm "cannot be excluded as" or "is consistent with" the source rather than "is" the source. Other courts continue to admit the traditional testimony largely unrestricted. The result is a genuine patchwork: the same examination can yield different permissible testimony depending on the courtroom. For the law/courtroom reader, this is the live front line of the chapter — and the lesson is that admissibility here is contested and moving, not settled. (How courts gatekeep this kind of evidence is the whole subject of Chapter 5; how bias inflames it is Chapter 31.)

So where does firearms identification sit on the spectrum? Above the discredited methods we will meet — it is not bite-mark analysis (Chapter 16); barrels and breech faces really do differ, the class characteristics are real and validated, and the discipline has a coherent physical basis. But meaningfully below DNA and instrumental chemistry, and below even latent prints in the eyes of some reviewers, because its central individualization claim rests on a subjective "sufficient agreement" standard without a validated, quantified error rate. Contested is the fair word — the same word the Chapter 1 validity table used. The honest one-line verdict: firearms identification is a real comparison discipline with solid class-level and same-source-grouping power, whose individualization claim has not been shown to meet the standard of scientific validity the courtroom has long assumed, and must be stated as a qualified opinion, not a certainty.

🔍 Check Your Understanding 1. An examiner testifies that a casing was "fired by the defendant's pistol to the exclusion of every other firearm in the world." Name two specific things wrong with that sentence, using this chapter's terms. (Think: what has been tested against "every other firearm," and what is the actual decision standard?) 2. Why is "these four casings were all fired by the same gun" often a stronger and more defensible statement than "this casing was fired by the defendant's gun"? 3. NIBIN returns a high-ranked candidate linking your casing to a recovered pistol. What, exactly, has been established at that moment — and what has not?

What does all this mean for the bullet (or casing) in your hands at a scene? Read its class characteristics and use them to exclude — that is your surest move. Group same-source items when the marks support it. Test-fire and compare honestly, declaring "inconclusive" when the marks earn nothing better, and stating any identification as a documented opinion with its basis shown and its error rate disclosed. Treat every NIBIN lead as a question, not an answer. And never, ever let the word "match" do the work of the word "certainty." If you carry only one sentence out of this chapter, carry the one that fits on a card: firearms evidence excludes cleanly, groups well, and identifies only as a qualified opinion — never as proof.

🗂️ The Case File

The stray cartridge case. Back in Chapter 3, when the cabin's evidence was inventoried, one item sat slightly apart from the rest: a single fired cartridge case, recovered from the debris of the front room, near the threshold. At the time it was logged with everything else — gas can, tools, charred documents, the pry-marked door, the victim's phone — its meaning unknown. It is the kind of item that, on television, would now become the spine of the case: a casing means a gun, a gun means a shooting, run it through the database and find your killer. This chapter is where we resist exactly that story.

Run the honest workflow. The casing is photographed in place, mapped, and collected. It is a 9 mm case; its breech-face, firing-pin, and ejector marks are legible. It is entered into NIBIN/IBIS, which returns a couple of low-confidence candidate correlations — leads, nothing confirmed — to casings from unrelated incidents elsewhere in the state. An examiner could, in principle, pursue them.

But here the discipline of relevance does its quiet, decisive work, and it comes from another chapter entirely. The autopsy (Chapter 11) established the cause of death as blunt-force head trauma, with no gunshot wound anywhere on the body — and no bullet, no gunshot wound track, nothing for terminal ballistics (§15.1) to read. Whatever this casing is, it is not the instrument of Marcus Diallo's death. It may be old — fired weeks or years earlier by a previous occupant, a hunter, a target-shooter on the property; cabins in rural Carrow County are not strangers to spent brass. It may be entirely unconnected to the events of 18 October. What it is not is evidence that Diallo was shot, because the body says, unambiguously, that he was not.

The case-file entry for this chapter, then, is a discipline, not a discovery. Log the casing fully — photograph, NIBIN entry, class characteristics — because you do not yet know it is irrelevant when you collect it, and an item you fail to document cannot be re-examined later. Then, when the autopsy comes back, record honestly: this cartridge case is a probable red herring — present at the scene, but inconsistent with the established manner of death; it neither implicates nor excludes any person of interest, and pursuing it as a "ballistics match" would chase the case away from where the evidence actually points. This is theme one of the book wearing work clothes: the casing's most important property is not what it matches but that, against the autopsy, it is irrelevant — and recognizing irrelevant evidence, and refusing to be led by it, is as much a forensic skill as any comparison. The exclusion/inclusion status of Keller, Whitfield, Salas, and Renner is unchanged by this chapter. The casing changes nothing — and saying so clearly is the whole point.

Conclusion

A fired bullet and a fired cartridge case are eloquent, but only about certain things. They can tell you, reliably, the class of firearm — caliber, rifling, the type of action — and that is enough to exclude candidate weapons cleanly and to narrow the field. They can tell you, often, that several rounds came from the same gun. They can, through a competent examiner's careful comparison-microscope work, support the conclusion that a specific firearm was the source — but that support is a qualified expert opinion resting on a subjective "sufficient agreement" standard, not the quantified certainty DNA can offer, and the 2009 NAS and 2016 PCAST reports both found the discipline's individualization claim scientifically underbuilt. NIBIN/IBIS multiplies the investigative reach of casings across cases and jurisdictions, but it generates leads for human examiners to test, never courtroom matches. And in our cold case, the chapter's central item — a stray casing — teaches the hardest lesson of relevance: against an autopsy showing no gunshot wound, the most disciplined thing to do with a piece of firearm evidence is to recognize that it has nothing to do with the crime, and to say so.

We have advanced two of the book's themes head-on: that forensic evidence excludes (and recognizes irrelevance) far more reliably than it proves (the casing is a red herring; class characteristics exclude but rarely individualize), and that not all methods are equally valid — firearms identification is real and useful but sits in the contested middle of the spectrum, not at the quantified summit where television places it. The next chapter stays in the comparison disciplines and pushes to their honest extremes: toolmarks and impressions, where footwear and tire evidence can be solid class evidence, and where bite-mark "matching" — the most thoroughly debunked discipline in all of forensic science — shows what happens when a comparison method claims individualization with no validity at all.

Key Terms

• Ballistics — strictly, the physics of projectiles in motion (internal, external, terminal); in common forensic usage, the broader discipline of firearms examination, including the comparison of marks on bullets and cartridge cases.
• Rifling — the spiral lands and grooves cut or formed into the interior of most firearm barrels to spin and stabilize the bullet, which in the process engraves class and individual marks onto every bullet fired.
• Lands and grooves — the raised metal (lands) and cut spiral channels (grooves) of rifling; their number, width, and twist direction are class characteristics readable from a fired bullet.
• Striations — the fine parallel lines scraped onto a fired bullet (and onto toolmarks generally) by microscopic irregularities of the barrel; their correspondence is the basis — and the contested basis — of a same-source identification.
• NIBIN/IBIS — the National Integrated Ballistic Information Network (the database) and its Integrated Ballistics Identification System imaging engine; it images and correlates cartridge-case and bullet marks to generate investigative leads for examiners to confirm, not matches.
• Cartridge case — the metal cylinder holding primer, powder, and bullet, ejected when a self-loading firearm fires; it records firing-pin, breech-face, extractor, and ejector marks and is often a richer comparison witness than the bullet.

Spaced Review

1. A bullet recovered from a wall has six lands and grooves with a right-hand twist; the suspect's revolver has five with a left-hand twist. What can you conclude, and is it an exclusion, a consistency, or an identification? (§15.2, and recall the honest verbs from §1.4)
2. Where on the NAS/PCAST validity spectrum does firearms identification sit, and what specifically about the "sufficient agreement" standard keeps it out of the quantified-validity tier with DNA? (§15.6; the validity-spectrum habit from §1.5)
3. In Chapter 14 you met Brandon Mayfield, a fingerprint "identification" made with 100% confidence that was 100% wrong. What single safeguard — the one most firearms units still lack — would most reduce the analogous risk in a firearms identification, and why? (§15.4; previews Chapter 31)
4. From Chapter 3, recall how the stray cartridge case entered the cold-case inventory. After this chapter, what is its forensic status, and which of the book's four themes does that status most directly illustrate? (§15.1, The Case File)
5. A detective says, "NIBIN matched the casing to a gun we recovered — that's our murder weapon." Identify two distinct errors in that sentence. (§15.5, and the manner-of-death fact from Chapter 11)