Texas has made many headlines in recent years for the spate of exonerations of wrongfully convicted men. In most of these cases, fortuitous turns of events, along with the hard work of innocence advocates, led to solid proof that eyewitness evidence was mistaken. The same is true in the latest case, in which a Dallas judge released two men based on evidence developed by students at two of the state’s university-based innocence projects.

Claude Simmons Jr. and Christopher Scott were released from custody in Dallas on October 23 based on new evidence of innocence, including the corroborated confession of one of the true perpetrators. According to prosecutors, it was mistaken eyewitness testimony that convicted the men for a 1997 murder. The two had already served over a decade of their life sentences when the innocence project students persuaded Dallas D.A. Craig Watkins to review the cases and pursue the exonerating evidence.

These latest exonerations are noteworthy for the lack of DNA evidence in the case. To date, forty-one innocent Texans have been cleared by DNA. By all accounts, those men were lucky, despite the profound injustices they suffered. In each of their cases, biological evidence existed with the potential to clearly identify the perpetrator. That evidence was collected and, unlike the evidence in thousands of Texas cases, was not lost or destroyed. Finally, innocence advocates agreed to take their cases from the thousands of requests for assistance that they receive every year. The stars had to align to give those men the opportunity to be vindicated.

It is only because the crime was solved these many years later that Simmons and Scott walk free today. It is easy to imagine that the true perpetrator who confessed to the crime might not have ever done so. It is easy to imagine that the innocence advocates might never have accepted their case from the sea of requests because of a lack of forensic evidence. It is easy to imagine that the D.A.’s office would not be inclined to cooperate like Watkins’ office did. In short, it is all too easy to imagine that the injustice inflicted on Scott and Simmons would never have come to light.

But it did. And their good fortune gives us yet another example of how fallible eyewitness evidence can be, and one more reason to insist that we implement the kinds of police procedures that have been proven to reduce the risk of eyewitness error.

The Dallas Police Department deserves great credit for implementing new eyewitness identification procedures earlier this year that include best practices designed to increase accuracy. Several other departments across the state, including the Richardson Police, have also embraced more accurate lineup protocols in response to their experience with a wrongful conviction. The vast majority, however, still handle identification procedures with the traditional protocols developed without the benefit of any of the extensive scientific research in recent decades on eyewitness memory and how it can go wrong.

Accurate evidence is too important to be optional, and Texas needs a law requiring that police follow more reliable eyewitness protocols. Without it, we can be sure to discover other innocent people in prison, with luck, like needles in a haystack. We can also be sure that there will be others we will never know about.

Edwin Colfax is the State Policy Director of The Justice Project, a nonpartisan organization that works to increase fairness and accuracy in the criminal justice system.