The victim in the five-bullets case opened his apartment door one afternoon. A man with a revolver fired at him and fled. Police quickly arrested the shooter who was after revenge because the victim had testified in a grand jury against the shooter’s brother about a robbery, and that brother was now in jail.

          The shooter told the police that my client, who lived in the same housing complex as the victim, had given him instructions on where to find the victim. My client told me that he had been out in the building’s yard when the shooter approached him and asked if he knew the victim. My client said yes, and the shooter asked where the victim lived. My client pointed to the building and said on the fifth floor. The shooter departed. My client said that he had never met the shooter before; he did not see the shooter with a gun; and he had no idea what the shooter was going to do.

          If this was true, my client, like the unwitting driver of the bank robber whom I had used as an example when writing about the Texas abortion law, was not an accomplice to the crime. I suspected, however, that there was a bit more to my client’s story than he was telling. He may not have met the shooter before, but the shooter was a well-known bad guy who evoked fear in the neighborhood. My client may have known the shooter by sight and been intimidated by him. Furthermore, my client may have heard the stories circulating about the victim’s testimony against the brother and at least suspected that the shooter’s inquiry was not a friendly one. Even so, however, I thought I could mount a good defense, especially since my client had never been in any trouble with the police, was an earnest student, and worked two part-time jobs to help his single mom and three siblings with family expenses.

          Alas, I was not to hear a jury announce a not-guilty verdict in what I thought was a winning case. When the shooter’s case went to trial, the prosecutors wanted my client’s testimony, and they agreed to give him immunity. It only made sense for him to accept it.

I remember this case for other reasons besides my lost opportunity for a victory. The victim after being shot somehow made it to the phone and called 911. Later, as was routine, I obtained a copy of that recording. I heard his voice reporting that he had been shot multiple times. His voice sounded calm. There were no screams or pleas for help. Just an affectless recitation of what had happened and where he was. He hung up. The recording continued with the 911 operator contacting a squad car. She said what she heard, but she added. “I don’t know. It sounded like a prank. I don’t think it happened. But I guess you have to go and check it out.” And he had five bullets in his body.

The case had yet another memorable wrinkle. Shortly after my client was arrested, a preliminary hearing was held. The victim was in a hospital, so the hearing was held there. He still had five bullets lodged in his body, three of which were in his skull awaiting surgical removal. I did not know how he survived the shooting, something I wondered about when it was my turn to question him. Multiple tubes were running in and out of his body, and it was hard to hear his responses with the numerous medical devices making different sounds. He, of course, did not know what had occurred between the shooter and my client, but in response to my query, he said he had always considered my client to be a friend, buoying my hopes for the trial that never was.

This was not as difficult as another cross-examination I once had, this one in a drug case. My client was charged with selling heroin to an undercover officer. In what was then a standard practice, the arrest was not made immediately after the sale, but weeks later, when the undercover pointed out the supposed seller to other members of his team, who then swooped in to grab the client. There was always the chance for a mistaken ID in these situations, and I won some cases on that ground but usually only if I could rather conclusively prove that my client had not made the sale. For example, I established that one client was sorting mail with many other workers at the main post office at the time of the transaction. Most often, however, these drug sales were tough to defend. The public, including jurors, generally thought that trained officers were unusually adept at making these identifications. However, data indicate that they are no better than the rest of us, and study after study have documented that mistaken identifications are one of the chief causes of wrongful convictions. My client, a street person, had no solid proof where he was at the time of the sale, and all I could do was to suggest that the undercover could not conclusively ID the seller, a defense that depended on cross-examination of the cop and almost never worked. The difficulty of that task, however, increased because the undercover between the time of the sale and trial had been shot and paralyzed in an incident unrelated to my case. He was wheeled into the courtroom on a hospital gurney with an IV drip in place. Go ahead: You try to cross-examine under these circumstances and show that that cop’s credibility could not be trusted and that there was a reasonable doubt that your client was guilty. I lost, of course. For the sale of two $5 bags of drugs, which he may or may not have done, my client got fifteen years in prison .