Media attention into the shooting death of Gail Knisley on Interstate 270 was intense, both locally and nationally. One of my responsibilities, in addition to acting as case agent and directing the FBI investigation, was to prepare Chief Deputy Steve Martin of the Franklin County sheriff's office for daily press briefings.

Editor's note: In this special three-part series exclusive to ThisWeek Community News, retired FBI special agent Harry Trombitas chronicles his experience as a lead investigator into the fatal shooting that occurred November 2003 on the south corridor of Interstate 270 in Franklin County.

Media attention into the shooting death of Gail Knisley on Interstate 270 was intense, both locally and nationally. One of my responsibilities, in addition to acting as case agent and directing the FBI investigation, was to prepare Chief Deputy Steve Martin of the Franklin County sheriff's office for daily press briefings.

The public needed to be updated on the progress of the investigation on a regular basis to help stem the fear gripping central Ohio. We as investigators needed the public's help and support to solve the case.

In any multijurisdictional investigation, it is paramount that there is one, and only one, voice that speaks for all of the investigative agencies. Martin made the mistake of heading to the bathroom as task force representatives from the various investigating agencies were about to make the decision on who would be the spokesperson. When he returned, he was told that he had been elected without dissent.

Martin certainly was no stranger to the media, having served as spokesman for many of the sheriff's office's investigations over the years, but this was different.

I had extensive media training with the FBI, and I knew the importance of working with the media on major investigations. Our plan was to meet with the media twice a day, as long as we had information to share.

Early in our investigation, keeping the twice a day schedule was not a problem.

Later, during the "dog days" of the investigation when we would go for days with no known shooting incidents to report, we held one press conference a day.

One of the hot topics that continually surfaced with the media involved the profile of the shooter. Everyone wanted to know what our shooter would look like, how he or she would act and what the public could watch for.

One of my areas of interest when I joined the FBI in 1983 was criminal profiling. Although the terminology has changed from one thing to another over the years, I started to receive my FBI training in profiling in 1985 by attending training and participating in case consultations at the Behavioral Science Unit at Quantico, Va., the home of the FBI Academy.

Though I never was assigned full-time with the unit, I became a profiling coordinator in the field. Based on my training and experience, I knew that coming up with a profile of our shooter would be difficult.

This was confirmed during my almost daily briefings with the National Center for the Analysis of Violent Crime's Behavioral Analysis Unit. Generally, we are all over the board when it comes to profiling sniper-type offenders.

I had been the case agent on the Thomas Lee Dillon case, a serial murder case that ran for three years from 1989-1992 in southeastern Ohio in which Dillon used a high-powered rifle to randomly shoot and kill five men who were hunting and fishing in remote areas. Dillon was a white male, age 42, at the time of his capture. In contrast, Muhammad and Malvo were black males, one older, one younger.

During our consultations with the Behavioral Analysis Unit, we determined that although we could not arrive at a profile of our offender, there were several behavioral characteristics we were comfortable releasing to the public to aid in the investigation. Some of the characteristics we released included:

* The offender appeared to be comfortable operating in the south end of Franklin County, meaning that area was his or her "comfort zone." This seemed to indicate that our offender most likely lived or worked in the area where the majority of shooting incidents had taken place.

* The offender appeared to be nonconfrontational, since he or she preferred to shoot at vehicles and buildings from a safe distance.

* The offender could come and go from his or her residence or workplace with minimum supervision, since the shooting incidents occurred at various times and days of the week.

* The offender clearly had access to a weapon.

* The offender most likely drove a non-descript vehicle, since there were no reports from possible eyewitnesses who observed a "high-profile" vehicle fleeing the scene of any shootings.

Releasing these behavioral characteristics seemed to appease the media and helped generate more interest in the case and more phone calls from the public.

Regarding the weapon the offender used, the media repeatedly asked why we wouldn't release the caliber of the round (9 mm) that killed Knisley or talk about the type of weapon used by the shooter.

The answers were very simple. First of all, why would we want to talk about the caliber of the bullet? What would be the benefit of releasing that information? By talking about the specific caliber, it could have caused the shooter to switch to a different weapon. As long as the shooter was using the same weapon, we were able to link cases and track the offender.

Second, as far as what type of weapon was used, we really didn't know for sure. Most people who are familiar with firearms know many different kinds can fire a 9 mm round, including rifles and handguns. Although our firearms expert believed that most likely the round that killed Knisley was fired from a Beretta handgun (later determined to be accurate), why speculate and be wrong? We prided ourselves on releasing only known facts, choosing to stay away from speculation.

During the four months of the I-270 shooting investigation, task force investigators followed 5,750 leads phoned in by the public. More than 1,400 people were identified as possible suspects.

Each of the leads had to be followed and the people mentioned as possible suspects investigated. By March 2004, investigators were tired and disheartened because we had failed to solve the Knisley homicide and bring the shooting incidents to a close. The public was frustrated and still terrified that they or someone they loved would be the next victim.

As an investigative task force, our worst fears were that the shooter would simply stop and disappear for a period of time, only to reappear and kill again.

We needed a break in the case. We were about to get one.

Next week: Tipster leads to the big break in the case.

The opinions expressed in this article solely are those of the author, retired Special Agent Harry Trombitas, and not those of the FBI or any other agency. Trombitas writes the Case in point column for ThisWeek Community News. He currently is system vice president of security operations for OhioHealth and a senior consultant for Armada, a security consulting company, in Powell.