www.latimes.com/news/local/la-me-firearms5may05,1,1100423.story?coll=la-home-headlines

COLUMN ONE

LAPD's Top Gun Is on the Case

Richard Smith has an eye for firearms analysis that few can match. Detectives
compete for help from the 'one-man weapon against crime.'
By Jill Leovy
Times Staff Writer

May 5, 2004

Maybe it's farsightedness ? the ability of Richard Smith to read freeway signs
before anyone else in the car. Or maybe it's his high tolerance for boredom.

No one, least of all Smith himself, is sure how he does it ? how this once
ordinary Los Angeles Police Department patrol officer sees what others miss, and
how he came to be, as one supervisor put it, "a one-man weapon against crime."

But whatever the basis of his gift, the LAPD is cashing in. Today, as they have
every Wednesday for several months, detectives will cut through what used to be
an all but impenetrable wall of red tape to get to Smith, the top technician in
the LAPD's firearms analysis unit.

Detectives jump for this chance each week, driving across town with bits of
crime scene debris in manila envelopes to a dark, low-ceiling room where Smith
sits before a computer terminal.

There they hover, waiting for him to work his magic, hoping the random bits of
metal left by gunfire will produce what firearms analysts call "a cold hit" ? a
match to some other crime or weapon ? and to detectives, the kind of link that
might break a case.

The program, called "walk-in Wednesdays," started as the LAPD's answer to a
crushing backlog of firearms crimes.

By allowing detectives to skirt a formal priority system one day a week for fast
answers on any case they want, lab officials hit on a serendipitous fusion. A
rapid increase in the number of cold hits has not only helped solve recent
crimes, it has yielded surprising revelations about the way crime guns move
around, and demonstrated ways of using new digital technology and a national
database on crime guns.

But perhaps the most important lesson is that the technology is useless without
the right craftsman ? a fact no one embodies more than Smith, whose uncanny
ability to glean evidence from cartridge casings and shards of bullets has
driven the lab's recent success.

Smith, 46, is so unassuming that "when you first meet him, you almost have to
kick him to see if he will talk," said Joe Zorola, a longtime acquaintance who
is now with Forensic Technology, the contractor that developed the national
database.

On paper anyway, Smith is a low-level technician who sits glued to a chair,
mouse-scrolling through scores of magnified images of cartridge casings on a
computer screen all day. He occupies what, by normal LAPD standards, is a
backwater office job in a department that, by tradition, values street
assignments more than desk work.

It is not the job he wanted. He transferred in by necessity in 1991 after
spending 11 years as a patrol officer in some of the city's toughest police
divisions, a fast-paced, eventful job that he says he loved. A series of
personal setbacks drove him to it ? divorce, a custody battle. He needed fewer
demands, regular hours.

Grunt Work

Smith told himself that it was only temporary.

Located in a drab police station annex in northeast Los Angeles, the lab tests
all kinds of firearms-related evidence. Its experts determine the direction
bullets traveled, catalog seized guns, and compare images of cartridge casings
in search of links between crimes.

Eighty-five percent of Los Angeles murders are committed with guns, many in the
form of drive-by shootings for which cartridge casings or bullets are often the
only physical evidence. Although firearms tests identify guns only ? not
perpetrators ? they are highly useful in backing up eyewitness testimony.

Much of the work is rote and repetitive. Smith was assigned to an especially
boring job ? a file clerk of guns. He soon moved to another boring job ? data
entry for cartridge casings. Then he was promoted to firearms analyst, a
position also noted for tedium, but one that was undergoing a revolution.

The process of linking bullets and cartridge casings to the weapons that fired
them was being transformed by computers. An expanding national database was
allowing investigators to track sometimes elaborate "daisy chains" of weapons
used in multiple crimes.

Suddenly, investigators could compare clues from one crime scene to firearms
evidence from thousands of others, stored in the computer's memory. Matches
could link what had appeared to be unrelated crimes by showing the same gun was
used. Such matches ? cold hits ? became the ultimate goal of the trade.

Workloads in labs swelled as departments scrambled to record images of cartridge
casings collected from crime scenes or test-fired from weapons.

Smith set to work, spending hours at computer screens, clicking through images,
poring over microscopic abrasions on brass.

Sticking with such work requires its own kind of grit, said Doreen Hudson,
supervising criminalist in charge of the lab.

Most people "couldn't sit in that dark room that long," Hudson said. "It takes a
unique temperament. You go through 10 people quickly to find one who can do it."

Smith chafed. But he also was intrigued. As a patrol officer, he had been
witness to one of the most violent periods in the city's history. "I had sat up
on no less than 150 homicide scenes," he said. "I got to see firsthand the
devastation."

Cases that looked routine on paper were to Smith anything but. Their details
evoked painful realities from his patrol days. To this day, "I get a report, and
from that report, I can see in my mind's eye the events?. I can see how it went
down," he said.

He had seen what a 9-millimeter bullet could do, what an assault rifle could do,
"that high powered round," Smith said quietly, "that goes through a car, or a
house, or through a person. And then out the other side."

He began to study the more than 500 types of guns in use, and sought out high-level examiners. It helped him pass the time until he could get back to patrol. "I picked their brains, and sat with them at the scope," Smith said.

More and more, the scores of cartridge-casing images that flashed past him each
week were telling stories.

In their nicks and grooves, he read narratives. He saw the guns' histories ?
their use and manufacture. He discerned the small defects, the tiny variations
that harkened back to the individual machinists who had made them.

After a while, he could tell whether that machinist was trained in Germany. He
knew whether parts of a gun had been sculpted in Spain.

His superiors began to notice. Smith had always had sharp vision. But as he grew
more engaged, it was clear that he had something else.

"That eye," Hudson said. "You could almost liken it to people looking at 3-D
holographs. Some see it, and some don't. It takes an ability to relax the eye
enough to go beyond physical shapes?. He has that."

To understand why Smith's skills mattered, consider the difficulties of finding
microscopic similarities between guns, bullets or cartridge casings across the
three-quarters of a million entries in what is now the nationwide database
managed by the federal Bureau of Alcohol, Tobacco, Firearms & Explosives.

Fast Analysis

The computer analyzes the images at lightning speed. But the electronic eye is
simply not as discerning as the human eye. The computer offers little more than
possibilities ? spewing out lists of 100 and more candidates for matches.

Matches often are not among the computer's top 10 ? or even top 50. Last month,
Smith found a match of breech-face marks on a casing ranked 247th on the
computer's list.

Learning to sift through images in seconds, as Smith does, takes years. Even
then, outcomes vary based on individual skill. Smith's total of 132 confirmed
hits last year is more than double anyone else's in the lab.

"I refer to him as 'the Guru,' " said Zorola, formerly of Santa Ana's firearms
analysis unit. "His level of accuracy, his eye ? I had so much confidence in
that eye."

He said Smith was the best cartridge-casing analyst in California, and probably
in the top five nationally.

Several years into the job, the LAPD was facing budget woes, and Smith's
position was downgraded. He was told that he would have to transfer, probably
back to patrol, or take a pay cut.

But by then, he had found a new calling. He chose the pay cut of about 5% ? from
$55,000 a year to $52,250. "I felt like I was becoming fairly good at what I was
doing and having some success at it," he said.

Last year, the backlog of requested firearms tests had grown to 2,400, meaning
detectives might have to wait months, or years, for results.

To cope with the demand, the lab implemented rigid systems for prioritizing
tests, giving priority to cases about to go to court. This meant that the cases
were months old by the time Smith and his colleagues were asked to weigh in.
Meanwhile, promising, live cases were being pushed to the back of the line.

Change was needed. So Hudson came up with "walk-in Wednesdays." Finally,
detectives called the shots, bringing any case they sensed was hot ? a vandalism,
a murder ? no questions asked. They could sit at the computer with Smith or one
of his colleagues and talk to them directly about what they saw.

The results surprised everyone.

Using the Database

Since the inception of the ATF's database, called the National Integrated
Ballistic Information Network, firearms experts had struggled with the question
of how to maximize hits on the system.

Only 8,300 cartridge casings have been matched on the system out of about half a
million entries.

But the LAPD had stumbled onto an answer: Nearly 40% of the casings detectives
brought in on walk-in Wednesdays have matched other crime investigations or
seized guns ? an astonishingly high hit rate. A monthly federal report earlier
this year showed that the LAPD was posting more hits than any agency in the
country.

The lesson is that the instincts of street-smart detectives trump formal
procedures, Hudson said. Evidence handpicked by them was yielding far more hits
than the old, formal priority system. And the new slew of hits exposed
intriguing patterns: In Los Angeles, at least, crime guns appear to be used in
spurts, change hands rapidly, cross gang lines, then rapidly pass out of use.

Moreover, because Smith and his colleagues were communicating directly with
detectives, sharing their preliminary findings, the tests became timely and
relevant.

In this way, the LAPD has become "a trailblazer" in using the technology
effectively, said ATF Special Agent Lisa Kincaid, western regional coordinator
of the database.

A case this year epitomized the change: On Feb. 14, 23-year-old Jorge Lua was
killed as he was driving to a party. A few hours later, 12-year-old Gregory
Gabriel was killed outside a South L.A. nightclub; he had sneaked out on his
first sleepover with friends.

Smith made the match. The same gun was used in both crimes.

But there was no suspect.

Then, a West Los Angeles LAPD detective came the following walk-in Wednesday
with a seemingly unrelated case: a nonfatal attack on a witness.

Such a case wouldn't have warranted a second look under the old system. It would
have been pushed to the back of the list. Now, though, Smith held the casing
under a microscope, giving it a quick manual scan before imaging.

Then he froze.

He had seen perhaps five dozen casings since the Lua case, and more than a week
had passed. Still, there was something familiar: He had seen that firing pin
before.

It was a cold hit on par with a winning home run in the bottom of the ninth
inning.

With his eyes, and no database other than his brain, Smith had recognized the
subtle microscopic pattern on the West L.A. casing as matching those on casings
from the Lua and Gabriel killings.

Because of the match, detectives who otherwise wouldn't have spoken compared
notes. There was a common name, a description ? enough for detectives to close
in on two suspects, who were arrested. If Smith hadn't linked the cases, it is
likely these detectives never would have put their respective pieces of the
puzzle together.

The National Academies of Science are studying the feasibility of a new system
that might include data on all guns manufactured in the United States ? a
controversial and potentially expensive plan, said John Rolph, professor of
statistics at USC, head of the study panel.

Smith and Hudson say the LAPD's track record suggests that such massive
endeavors may be misplaced. Working closely with street-level detectives, moving
fast and comparing guns and cartridges recovered close together in time may
prove more effective.

"The difference," Smith said, "is that you see the results."

Copyright 2004 Los Angeles Times