For murderers and those who pursue them, crime and punishment in Baltimore is a crapshoot -- a hit-or-miss proposition in which nearly a third of the people who kill in this city will never be caught and only four in 10 of the killers will ever see the inside of a jail cell.

A case-by-case analysis by The Sun of 1988's murders -- the latest year for which prosecutions are largely complete -- shows that only a quarter of those who committed them received sentences of more than 10 or 15 years, which generally translates to no more than three to seven years before parole.

And, there is this: A murderer in Baltimore stands less than a one in 10 chance of being sentenced to life in prison.

"It doesn't surprise me because I've been living as a judge or prosecutor since 1972," Baltimore Circuit Judge John N. Prevas said. But, he said, he once believed "that murders are easy to solve and the police waited until they had solid evidence before they attempted to take somebody to trial. Now that I know how hard [murders] are to solve and that the police don't wait, I'm much less surprised."

The statistics for the arrest and conviction rates from The Sun's analysis of Baltimore's 1988 homicides differ from those compiled by police and prosecutors. In attempting to more accurately measure a murderer's chances of being caught, convicted and sent to prison, The Sun estimated the number of people involved in homicides -- solved as well as unsolved. (See box on page 8A.)

Many factors contribute to what is happening in the city's criminal justice system.

* The ever-increasing homicide rate -- up 35 percent in the last four years -- is putting pressure on every point in the criminal justice system.

"Let's hope the system isn't going to break," says Timothy J. Doory, who heads the violent crimes unit of the state's attorney's office. "How much can we handle? I don't know."

* The nature of violent crime continues to change. Forty percent of the 1988 murders were the direct result of the city's $1-million-a-day drug trade. Another 15 percent occurred in stranger-to-stranger crimes such as robbery, burglary and rape. Those slayings, which account for an ever-greater share of the total, are typically harder to solve and prosecute than the domestic murders and bar arguments that once made up a larger share of the city's violence.

* The Police Department's investigative sections and the state's attorney's trial division have been hampered by inexperience and a lack of resources in recent years. Members of Baltimore's defense bar routinely recount courtroom victories they believe should have been losses. In 1988, defense attorneys won nearly half the murder cases brought to trial.

* Prosecutors and detectives, in turn, blame public perceptions and declining civic responsibility for jury acquittals and failed prosecutions. They say city juries are simply less willing to convict defendants than juries in the surrounding counties.

Some say this reluctance to convict stems in part from a lingering distrust of the police and the criminal justice system in inner-city Baltimore's poor neighborhoods, where the vast majority of slayings occur.

Still more blame television for oversimplifying the system and thus creating unrealistic expectations.

"We all feel more comfortable when things are neatly wrapped," says Stuart O. Simms, Baltimore state's attorney. "The trail leads up to the door, and the door is still open and the person is standing there with the mud on their shoes, but life is not like that."

Some prosecutions, of course, lack conclusive evidence and the resulting dismissals or acquittals can be regarded as reasonable outcomes. And yet an equally common scenario is that of a strong case, lost in a jury decision that defies explanation.

In one instance, a man was acquitted of shooting a bartender, even though two of his alleged co-conspirators testified against him and police found his fingerprint at the murder scene. Another man was acquitted of robbing and killing a cabdriver despite the testimony of his alleged co-defendant, a corroborating witness and a Baltimore City Jail guard to whom he had confessed.

No outcome is guaranteed. People arrested for the same crime are not necessarily guilty or not to the same degree. Take the murder of Kurt Tackett, for which four people were arrested: One pleaded guilty to first-degree murder and was sentenced to life in prison; two pleaded guilty to lesser charges and were sentenced to three and five years; the fourth was acquitted.

James Baskerville shot his girlfriend, pleaded guilty and was given a life sentence; Mildred Hessie shot her boyfriend, pleaded guilty and got a 10-year suspended sentence. Andre Hawkins was given a life sentence for killing a cabdriver; his partner, Anthony Carter, was given a suspended sentence.

That's the reality, the truth behind such platitudes as "use a gun, go to prison." And it's the reality even though the city devotes more of its resources, manpower and talent to the investigation and prosecution of murder than of any other crime.

In sum, the criminal justice system is struggling to maintain a deterrent even as the violence continues to escalate. Last year, city detectives recorded 305 murders -- a toll unseen since the early 1970s when, through the creation of the shock-trauma system, victims who might otherwise have died were saved by medical technology.

More murders mean more work. In the Baltimore homicide unit, detectives often juggle two or three murder files at once, working each for a few days before a fresh case arrives and effectively ends any hope of prolonged investigation. The chance to pursue open cases or prepare cases for trial is limited.

Four or five years ago, highly publicized homicide cases like the March 6 shotgun murder of the night manager at a downtown restaurant would have been worked for months by a squad of investigators. Now these cases get barely two weeks' attention from one detective.

The Police Department's ballistics laboratory is so swamped that there is not enough time to compare newly recovered guns and bullets with those seized in earlier shootings. The trace evidence lab is months behind in its examinations of hair, blood and fibers. In lab freezers, blood and semen samples become putrid before they can be tested.

Meanwhile, prosecutors are inundated not only with murder cases but with dozens of other felonies. Postponements are common, and plea bargaining to lesser offenses has long been a necessary tool in a system that would strangle on its caseload if even half the murder defendants sought jury trials.

Privately, law enforcement officials blame the city's fiscal poverty and a lack of leadership for the declining deterrent. Many detectives and prosecutors believe that the city's political leaders have reacted slowly to the escalating violence.

In 1982-1983, when Mayor Kurt L. Schmoke was state's attorney, his office received 21,300 criminal and juvenile cases. By 1989-1990, the office was handling an additional 6,300 such cases. But except for salaries and benefits, the operating budget for the trial division and felony units has remained essentially unchanged. In the Police Department's homicide unit, substantive increases in personnel and equipment were not undertaken until this spring, when the homicide rate was fast approaching one a day.

In addition to this lack of resources, The Sun's case-by-case review reveals other systemic problems that routinely place some prosecutions at risk.

For example, homicide cases where the victim lingers in a hospital before dying are less likely to be solved because the case is initially handled by district officers rather than trained detectives. As a result, evidence is less likely to be preserved, and witnesses are less likely to be identified.

Similarly, once someone has been arrested for a city murder, the case is considered closed. The detective moves on to new calls. Little time or money is set aside to help prepare and maintain the case for trial, where the standard of guilt beyond a reasonable doubt must be met before someone can be convicted.

The station house

In every murder, the only given is the corpse.

Whether there will be a suspect, a witness, corroborating evidence or a conviction is partly dependent on the murderer's skill and intelligence, partly on random luck. But with each murder, the criminal justice machine sputters and lurches into action, beginning with the search for evidence.

In Baltimore, that search starts with a homicide detective who may have two or three open murder files on his desk and who will likely handle between 10 and 15 such cases a year. Add to this the unending flood of aggravated assaults, unattended deaths, suicides, overdose cases and police-involved incidents -- all of which must be reviewed, monitored or investigated by the homicide unit.

Despite a more than one-third increase in the murder rate between 1987 and 1990, the number of investigators assigned to the homicide unit remained constant. Only last month were two additional six-man squads assigned to help with the growing caseload. But veteran detectives say it will take months to train ,, these new men in the subtle ways of death investigation.

"About half of these guys are going to make it. They're learning," said one veteran of the unit. "The other half won't be here in a year. You just can't absorb so many new guys at once."

With experienced investigators at a premium, time becomes a constraint in every investigation, and every investigator contends with competing priorities. In early March of 1988, for example, the city recorded five homicides in 24 hours.

Detective Robert Bowman handled two of those murders: one a drug killing in Northeast Baltimore; the other an earlier westside drug shooting in which the victim had later died at the hospital. Faced with two open cases on opposite sides of the city, Detective Bowman made a decision: He pursued the murder in Northeast Baltimore at the expense of the earlier shooting.

"The relatives and witnesses in that case were talking, and it seemed like it had a better chance of being solved," he explained.

The homicide unit practices its own form of triage, allowing the nature of the crime to determine the extent of the investigation.

In the 1988 molestation and murder of La-Tonya Wallace, an 11-year-old Reservoir Hill girl, the primary detective aggressively pursued his investigation for almost a year, interrogating one particular suspect three times, although no one was ever charged.

Other unsolved cases -- the shotgun slaying of a Pimlico youth, for one -- received several weeks of work before the primary detectives became burdened with fresh crimes. But those cases are the exception.

In Baltimore, the average murder generally receives the attention of one or two detectives for several days or a week at best -- long enough to work the basic leads and evidence, but not enough to pursue the case much further than that.

Still, just over 70 percent of the 1988 homicides were solved, equaling the national average. But that statistic, though gratifying to departmental leadership, ignores what happens after an arrest is made.

"Clearance rate shouldn't mean a damn thing," said Judge Prevas. "I'd rather hear that the Police Department arrested people in four out of 10 cases and every one of those four went to jail than to hear the Police Department arrested eight out of 10 and four of those had to be let go."

The department's rate for solving cases also obscures some systemic problems. For one thing, there is no specialization in the homicide unit.

Drug murders, 35 percent of which remain unsolved, are not channeled to detectives familiar with violent drug gangs. Robbery slayings from the city's west side are not tracked to an investigator who monitors robberies in those districts and therefore might know similar holdups or potential suspects.

Adding to the problem is a lack of paid informants. No money is budgeted for homicide investigators to maintain knowledgeable informants. The detectives don't even have a desktop computer system that could allow them to systematically track everything from drug-trade connections to bullet calibers. Instead, such intelligence is buried in case files that will eventually be reduced to microfilm.

Police officials say they are considering a computer system for the homicide unit, but it will take months, perhaps years, to compile a viable data base.

As a result, "the detective starts from scratch each time," says Harry Edgerton, a 10-year veteran of the unit.

Another problem, according to investigators and prosecutors, is that non-lethal assaults that eventually result in death receive less attention from the Police Department, particularly in the critical period immediately following the crime. In October 1988, for example, a 38-year-old nurse named Theresa Roseborough was robbed and critically wounded while standing at a Walbrook Junction bus stop. She spent a month in the hospital before succumbing to her wounds.

Yet when the Southwestern District's reports on the case were sent to the Criminal Investigations Division downtown, the paperwork consisted of nothing more than calls to the hospital by officers checking on the victim's condition. The case remains unsolved.

Lastly, officials in the department's crime laboratory concede that the investigative support units are now all but overwhelmed. In one double slaying from 1988, the blood samples recovered at the crime scene could not be compared with the defendant's because, after months of delay, they had become putrid in the freezer. Despite this, the defendant was convicted.

In the department's firearms unit, bullets and cartridges are no longer routinely compared with shooting incidents involving similar caliber weapons. There isn't time for that. Instead, the two trained examiners limit themselves to those comparisons specifically requested by detectives. And because the increase in homicide cases has consumed almost all the available time and resources, little time is left for the larger number of other shootings handled by district officers.

The courthouse

What wins a murder case?

"Give me a credible, believeable, no-ax-to-grind eyewitness every time," says Ara Crowe, chief of the state's attorney's trial division. "That's tough to beat."

Other corroborating evidence -- particularly physical evidence -- is often hard to come by.

In more than seven of every 10 city homicides, no physical evidence was found -- no fingerprints, no blood, no hair or fibers. And for every 10 cases in which the victim was shot, police recovered the gun less than three times. Yet the myth of criminology as indisputable science -- the stuff of television shows and crime fiction -- still holds sway among jurors, say veteran prosecutors. To deal with those expectations, prosecutors routinely use fingerprint examiners as witnesses in cases where no fingerprints were recovered, merely to explain that decent prints are found in only a small number of criminal cases.

On occasion, science does help prove a defendant's guilt, but more often than not the most powerful and compelling evidence comes from someone who knows what happened. Sometimes, that turns out to be the suspect.

In more than half the 170 homicide cases in which someone was arrested, suspects voluntarily talked to detectives. They made outright confessions, or told long, involved tales that allowed police to discover lies and knock down alibis.

But if eyewitnesses are the most precious commodity in a murder investigation, they are rarely accorded that status in Baltimore. In most cases where state witnesses are in need of pretrial protection, prosecutors and police can do little: The city budgets no money to move and house witnesses.

"[Witnesses] are less willing to say in court what they were

willing to say out of court, and that's the fear factor," said Mr. Doory, of the violent crimes unit. "Just think what we're saying: 'Someone that you know is a killer, and you're going to have to testify against him.' That's frightening."

That's what happened in the drug-related murder of Maurice Ireland. The state had witnesses, but none wanted to come into court. Consequently, the case was never prosecuted. However, the people allegedly involved in Ireland's murder were later convicted for another drug slaying.

"I had solid eyewitness identifications, but what I also had was solid fear by those eyewitnesses," said H. Jerome Briscoe, who handled Ireland's case. "[Witnesses] want these people off the street. They want the killing to stop, but they lack the courage to come forward and put themselves on the line. . . . Drug murders are a class by themselves. Normally they are done as a signal to the community that this particular organization means business and will deal with all complaints in the same manner."

Three witnesses to city murders were slain in 1988. In two of those cases, prosecutors believe the witnesses were killed to keep them from testifying.

"I can't tell you how many times we hear: 'I'd like to help you, but I got to go back there to live,' " says Mr. Doory. "It is frightening that the first thing we can offer to somebody who says, 'We're afraid' is to sleep on a bench in the homicide department. We definitely need more resources in that area."

Even though law enforcement officials know pretrial contact with authorities can reassure witnesses and improve conviction rates, Baltimore's detectives and prosecutors don't have the time for such things. With one case following close on the heels of another, pretrial preparation is limited to whatever case is about to go to court.

The result is predictable: Nearly 15 percent of 1988's homicide cases were dismissed -- either at arraignment or after indictment -- because of insufficient evidence. Seven out of 10 such cases were drug-related murders in which witnesses changed their stories.

Confronted with an experienced cadre of defense attorneys, supervisors in the state's attorney's office try to assign the more complex cases and important prosecutions to the more experienced and effective prosecutors.

For their part, defense attorneys say some of those charged with prosecuting the city's most serious crimes are knowledgeable and effective; others, they say, are poorly trained and frequently lose cases where the evidence is sufficient.

"A state's attorney without a lot of experience will make an incoherent closing argument, throwing in everything and not stressing key points or responding to the defense's argument," said M. Gordon Tayback, a Baltimore defense lawyer. "I have seen cases won or lost on closing arguments regardless of the evidence and the strength of the evidence."

Some courtroom losses are part and parcel of the criminal justice system, where the state has the burden of proving its case "beyond a reasonable doubt."

"It would be far easier . . . to stand up and say you can prove every case, convince every judge and every jury," said Mr. Simms, the city's chief prosecutor. "If we are taking a case forward, we believe that evidence exists to prove this person guilty."

Still, more than 45 percent of all murder trials ended in acquittals.

Basis of study

The Sun's figures estimating the conviction and arrest rates differ from those compiled by police and prosecutors. All slayings from 1988, including some cases which were ruled homicides by medical examiners the following year, were counted.

The arrests counted were those made for homicides occurring in 1988, even if the arrest was made in a subsequent year.

In an attempt to accurately portray the workings of Baltimore's criminal justice system, it was necessary to estimate a total number of people involved in the 238 slayings occurring in 1988. In the 170 homicide cases cleared by authorities, 199 suspects were identified -- a ratio of about 1.2 suspects for every homicide. Using a projection based on the number of people involved in the cleared cases, it is reasonable to estimate that another 79 suspects committed a homicide but were never identified.

Thus, it is reasonable to estimate that there were 278 people responsible for the 238 homicides committed in Baltimore in 1988.