How does ballistic forensics work?

June 12, 2013 7:51 AM

Can someone who understands forensics or firearms (or who just reads more mystery novels than I do) explain how ballistic forensics is not easily circumvented?

I don't own a gun, or go to a shooting range (or know anyone IRL who does), so I'm coming at this as a total layperson.

Premise:
1) Forensic experts on TV are shown matching bullets and casings found at crime scenes to a suspect based on different indicia left as marks by the suspect's weapon.
2) TV and movies have also taught me that gun owners take their weapons apart to clean and maintain them.
3) Presumably, the perp could replace the parts that leave the marks with other parts so the marks don't match (or, rather than replace them, deform them or sand them or something).
3a) I'm assuming that there is not the same registration and licensing for buying a part that there is for buying a firearm (to the extent the system works at all, anyway). I'm also assuming that these parts are not the single most expensive parts of a firearm, but maybe that's not true. Plus there's that business now with 3D printers making guns, which seems like a terrible idea.

Am I missing something? Are all the bad guys we see on TV and read about in books just too lazy to spend a few minutes with a Dremel and thus get caught by forensics? Or are people just counting on evading suspicion in the first place? Or is ballistic forensics not the, er, silver bullet that TV procedurals make it out to be?

posted by Admiral Haddock to Science & Nature (20 answers total)

Best answer: What you're saying is possible, but no one actually does it.
posted by Chocolate Pickle at 7:53 AM on June 12, 2013

Best answer: Bullets fired are rarely in the condition you see on TV. Once they hit something they are generally catastrophically deformed. The best you can usually hope for is enough to determine the size or caliber of the bullet. Matching the rifling on a bullet is not usually possible due to this.

You wouldn't even need to change out the barrel to disguise the rifling, you could just run a rat tailed file through it.
posted by sanka at 8:00 AM on June 12, 2013 [1 favorite]

Best answer: Same reason most people don't wear gloves when they're committing crimes -- most shootings are spur-of-the-moment, carried out by people with poor impulse control who are freaked the fuck out after they've shot someone.
posted by Etrigan at 8:02 AM on June 12, 2013

Best answer: It is not as accurate as they make out on TV. Shell casings can be used more accurately than bullets, but even then the success rate is not terribly high. See here for more details: http://en.wikipedia.org/wiki/Ballistic_fingerprinting
posted by Nothing at 8:02 AM on June 12, 2013

Best answer: Ballistic forensics is a pretty inexact science in general - in real life it almost never works like it does on TV and there's concern that juries put too much stock in this evidence.

Sounds like you're suggesting someone would do this after using the gun to commit a crime, but to do that they would have had to have planned ahead enough to have the extra gun parts already (since presumably going out and buying extra gun parts in between when the crime is committed and when the police take your gun would be a suspicious-looking kind of thing to do), or maybe they have some secret source of extra gun parts. In which case I would think it probably be easier just to get an illegal gun that's not registered to them and discard it afterwards.

It would be a good twist in a "creepy gaslighting killer who's thought of everything" kind of story, but probably wouldn't make sense in real life.
posted by mskyle at 8:10 AM on June 12, 2013

Best answer: It's fairly common for crime labs to be able to look at a bullet (and particularly a shell casing) and tell you what sort of weapon fired it, sometime narrowing it down to a particular manufacturer and model, and sometimes not. It's much rarer for a particular weapon to leave its own "fingerprint" on a bullet or casing -- most often its from a small manufacturing defect in the weapon that leaves a similar imprint on each casing. If you want to do more reading on the issue, "toolmark evidence" or "toolmark identification" describes the field more generally, as it sometimes overlaps to a particular crowbar that was used to pry open a door, etc.
posted by craven_morhead at 8:21 AM on June 12, 2013 [1 favorite]

Best answer: The FBI can do some pretty advanced stuff, but in most cases it's just not necessary.

If you want the skinny on real crime procedure, The Wire and Homicide Life on the Streets are pretty good dramas that show how real police work is done.

Forensic evidence is good to have, but most criminals do everything but leave a trail with a sign that says "This way to the criminal". The other part of the criminal element leaves no trail at all, and those crimes are never solved.

Also, if you watch Law and Order, and compare the old ones with the later ones, you can see how technology is tripping up more and more people. Internet search histories, cell phone chips, cameras EVERYWHERE.

I also watch a lot of The First 48, and this is super-true. In the cases that are solved, usually there are witnesses who come forward to ID the criminal. Or there's surveilance cameras. Or a blood trail from the crime scene to the house where the criminals were partying. In many cases they KNOW the guy did it, but there's just not enough evidence to hold the guy.

In one The First 48, the guy who got away with murder in one crime, was the victim of murder in a separate crime. That was weird, but I'll bet it happens all the time.
posted by Ruthless Bunny at 8:39 AM on June 12, 2013

Response by poster: Thanks, all! This was very informative--I'll put the depiction of ballistics on those shows in the same category as the "computer, enhance" trope--convenient and implausible.
posted by Admiral Haddock at 9:00 AM on June 12, 2013

Most modern handguns have a serial number stamped on both the barrel and the frame/slide/receiver. (Some have a barrel that is part of the frame, so in those cases, you can't "swap parts" anyway)

So if you swap out parts, it is easy to tell - serials don't match.

I thought Maryland had passed a law about "registering" a fired shell when you bought a gun, for just such tracking, but wiki says no

But, as suggested above, someone can easily deform the firing pin, ejector, rifling etc. to defeat it.

An interesting follow up would be: how often is such evidence actually presented and used in court ? And is it usually circumstantial (ie victim shot by a .45, defendant owns a .45) or specific (victim shot by a .45 and recovered bullet matches defendant's .45)
posted by k5.user at 9:02 AM on June 12, 2013 [1 favorite]

Or is ballistic forensics not the, er, silver bullet that TV procedurals make it out to be?

Yup. My sister works for the crime lab in the state of Massachusetts. I've told her that she should do a whole blog called "Why we are not CSI." I've gotten tours of the MA ballistics department and I've also toured the ballistics department of the Vermont crime lab. There is a lot going on there that is not forensics, including the gun library where they have versions of may sorts of guns to test fire and many test firing situations where they can compare bullets that came out of similar guns. The toolmark stuff is pretty rarefied air for analysis and usually they've got better leads on criminals without getting to that point. Makes for good tv to get to use all the fancy computers but most of the work they do is knowing about guns and how they work and fire and not that sort of thing.

While it's technically possible to do the things you've described, there is not a lot of overlap on the Venn diagram of "People who are ingenious enough to work out a long con crime down to obscuring the ballistics fingerprint of their weapon" and "People who are criminal enough to do things that attract the attention of law enforcement for their crimes" as much as it seems it's all we ever see on tv.
posted by jessamyn at 9:14 AM on June 12, 2013 [1 favorite]

Or is ballistic forensics not the, er, silver bullet that TV procedurals make it out to be?

You can read more than you ever wanted to know about ballistic forensics in the National Research Council's 2008 report Ballistic Imaging. Key quote: "Although they are subject to numerous sources of variability, firearms-related toolmarks are not completely random and volatile; one can find similar marks on bullets and cartridge cases from the same gun. [But a] significant amount of research would be needed to scientifically determine the degree to which firearms-related toolmarks are unique or even to quantitatively characterize the probability of uniqueness."

In other words, we have no reliable research as to whether a subjective match between a bullet and gun indicates, well, anything. In court one expert would say it does and the other would say it doesn't, and given the lack of scientific basis it hopefully wouldn't be the deciding factor.

As to deciding factors, bear in mind that by the time you get to test-firing a suspected murder weapon (say), you have already identified, for some other reason, a particular person in possession of a gun of the kind used in the murder. That person is already close to conviction, and shouldn't be patting themselves on the back because they kept the gun around but ran a file through the barrel. If the test firing doesn't match, in a way that could possibly be caused by messing with the barrel, it won't significantly change their picture. If the tampering's obvious it might even make it worse.

One scenario where tampering might be useful is if a gun was used in two different crimes, and the bullet in the second crime matched in the NIBIN database to the first crime. That might provide some leads to check out even if the matches aren't super reliable. But at that point, yeah, we're talking about an improbably sophisticated criminal.
posted by jhc at 9:17 AM on June 12, 2013 [1 favorite]

Most modern handguns have a serial number stamped on both the barrel and the frame/slide/receiver. (Some have a barrel that is part of the frame, so in those cases, you can't "swap parts" anyway)

So if you swap out parts, it is easy to tell - serials don't match.

Hmm, this one right in front of me doesn't have anything stamped on the barrel except "45 ACP SIG", and it's brand new. even if there were serial numbers on the barrels, they are commonly swapped for perfectly legal reasons, typically to make the gun more accurate for target shooting.

If you wanted to avoid any sort of ballistic fingerprinting and you're going to do it in a pre-meditated way, I think you'd rather make sure the police never find the weapon, rather than assuming they'll find the weapon but have trouble matching it to the bullet they recovered from the victim. Or you could buy a new barrel just for your crime, shoot someone using the after-market barrel, then put the original back in the gun and throw the aftermarket one in the ocean. Or just disassemble the whole gun and drop the parts off the golden gate bridge or whatever.
posted by tylerkaraszewski at 9:24 AM on June 12, 2013

Also, this is so common it's a trope on procedurals. The cops are sweating a suspect and they make up forensic evidence to get the guy to confess.

There's a hilarious scene in Homicide Life on the Street, where Munch and Bolander trick a suspect into believing a copier is a lie detector.
posted by Ruthless Bunny at 9:24 AM on June 12, 2013 [2 favorites]

Okay, well, don't shoot navy guys. Abs can take one look at a bullet recovered from a dead SEAL and tell you what the shooter's mother had for breakfast, and Gibbs can get you to take back things you didn't even steal.

One of the real-life problems with forensic evidence is that the defendant's lawyer gets to put his own expert's testimony to the jury. The jury then gets to believe the police experts or the defense experts. It ain't what you can prove, it's what you can get the jury to believe.

The lie detector scene: hilarious. I'd forgotten how cool that show was. Also, RIP, Lenny. I'm sorry to see these law and order versions run out of steam. (BTW...a friend mine had a yard service business, which he called "Lawn Order.")
posted by mule98J at 10:32 AM on June 12, 2013 [4 favorites]

Re: the lie-detector hoax test in "H:LOTS" is based on a bit from David Simon's book that inspired the show, "Homicide: A Year on the Killing Streets," about his year-long ride-along with the Baltimore Homicide Squad. The same gag later appears in "The Wire," also created by Simon.

In the book, the cops scare an ignorant suspect by reporting that they're going to "neutron" him, or something like that. The test for gunpowder residue on the skin is usually referred to, scientifically, as a neutron activation analysis, hence the strange connection between nuclear particles to crime-solving.
posted by Sunburnt at 10:46 AM on June 12, 2013 [1 favorite]

It's a bit of a side-issue, but when they're testing for gunshot residue on a suspect's hands, they aren't searching for gunpowder (more accurately, smokeless powder). When a gun is fired it fuses a number of chemicals together that are found in the primer and the powder; the hallmark concentrations in GSR are rarely found elsewhere (if memory serves, they're particularly concerned with concentrations of lead, antimony and barium).

One other note: if the victim was shot with, say, a .45 caliber handgun with a weird toolmarking from the barrel, and the suspect has a .45 caliber handgun without the matching toolmarking, but with evidence that a big 'ol file has been reamed through it, that's probably going to raise some eyebrows as well.
posted by craven_morhead at 12:21 PM on June 12, 2013 [2 favorites]

Spoiler alert: The omniscient Abby Sciuto on NCIS is an actress, not a rl firearms expert. When there is only 30 minutes left after the commercials to solve a crime, some magic must be used.
Also, real cops know the troublemakers in their area. They can start with the facts of the shooting to match the signature acts of a known criminal.
posted by Cranberry at 1:44 PM on June 12, 2013

I taught forensic science at university for a few years. I am in no way a ballistics or firearms expert. What I can tell you from working with them is this:

Don't believe anything you see on tv.

Ballistics experts do indeed look at the unique striations on bullets to individualize them. There are often stray bullets at a scene that haven't been deformed badly on impact. This comparison can be useful for determining if the same gun was fired at different scenes.

Ballistics work is also used for determining things like location (of the shooter, of objects and people struck by bullets).
posted by Sublimity at 3:01 PM on June 12, 2013 [1 favorite]

Since you seem open to some possibly fictional or semi-fictional ideas, one of the ways I've seen this addressed is for the shooter to treat the gun as disposable, kind of like tylerkaraszewski describes. The shooter gets hold of a gun in some way that can never be traced back to them (like stealing it from another criminal in another state), and then just drops the gun at the scene. The cops can run all the ballistic tests they want, but if the shooter can't be connected to the gun and/or the crime scene in the first place, it won't matter.

For added insurance, they take the gun with them and ASAP pitch it into the nearest body of water, or a dumpster a couple of miles away, or somewhere else the cops are unlikely to ever find it. No gun, no ballistics tests.

IOW, even in fiction it's more believable that a sophisticated criminal capable of planning some way of fooling a ballistics ID would just put greater effort into making sure they never get connected to the crime or the gun in the first place, rather than futzing around with swapping parts or filing the barrel or whatever.
posted by soundguy99 at 8:32 PM on June 12, 2013

There are also saboted rounds where a smaller-caliber bullet is encased in a plastic jacket to fit a firearm chambered for a larger caliber. A lot of shotgun slugs are like this, and even some high-powered rifle ammo. Remington had a .30-06 Express which was (I believe) a .223 bullet encased in a sabot to fit the larger .30-06 bore. Because the bullet (not the brass shell casing) was almost/entirely encased in the plastic sabot which flew off and was discarded upon exiting the muzzle, there would be little or no contact between the actual bullet and the internal bore of the rifle. That would make any kind of forensic comparison between a recovered bullet and the rifle in question extremely difficult, if not impossible, especially as unless it was a point-blank shooting, the discarded sabot would be nowhere near the point of impact.

A setup like that would leave few strings tying the bullet to the shooter, or his rifle.
posted by xedrik at 8:50 PM on June 12, 2013