How an Unproven Forensic Science Spread Through the Criminal Justice System

Bloodstain-pattern analysis has been accepted as reliable evidence by appellate courts in one state after another with little or no examination of its scientific accuracy.


The modern era of bloodstain-pattern analysis began when a small group of scientists and forensic investigators started testifying in cases, as experts in a new technique. Some of them went on to train hundreds of police officers, investigators and crime-lab technicians — many of whom began to testify as well. When defendants appealed the legitimacy of the experts’ testimony, the cases made their way to state appeals courts. Once one court ruled such testimony admissible, other states’ courts followed suit, often citing their predecessors’ decisions. When discussing the reliability or accuracy of the technique, judges typically relied on their own — or the testifying expert’s own — assessment. Rarely, if ever, have courts required objective proof of bloodstain-pattern analysis’ accuracy.


1954

Sam Sheppard, an Ohio doctor, is convicted of murdering his wife in a case that attracts widespread attention. Paul Leland Kirk, a renowned scientist and criminalist who worked on the Manhattan Project, studies the bloodstains in the Sheppard home and, the following year, offers an interpretation of events that the defense believes exonerates Sheppard.

1957

People v. Carter

The Supreme Court of California affirms that bloodstain-pattern analysis is a proper area for expert testimony and that Kirk is a qualified expert in the field. The case is possibly the earliest instance of an appellate court explicitly accepting bloodstain-pattern analysis as an appropriate field of expertise.

Testifying for the prosecution, Kirk said that if the blood on the defendant’s clothing was spattered at the time of the murder, then whoever was wearing the clothes was within two-and-a-half feet of the victim. Kirk told the jury he had performed experiments to learn about blood dynamics, including beating a contraption made of wood, “sponge rubber” and a thin sheet of plastic to learn which actions create different sizes and shapes of blood spots. California Supreme Court Justice Jesse W. Carter dissented in part from the majority’s opinion, writing: “It most certainly cannot be said that an object made of wood, sponge rubber and a plastic sheet constituted the same thing as a human head.”

Sets precedent in California.

1966

Sam Sheppard is retried, and Kirk’s findings play a central role in Sheppard’s defense. Sheppard is found not guilty.

Pedersen v. State

The Supreme Court of Alaska holds that bloodstain-pattern analysis is an acceptable area for expert testimony. An Alaska State Police officer uses blood spatters on a crab-fishing ship to determine where the victim was standing when shot.

The court notes that trial judges have discretion regarding the admission of experts, and a decision to admit an expert should only be overturned on appeal if it is so far outside the reasonable range of decisions that it constitutes an abuse of discretion. This deferential standard is commonly used by appellate courts on the issue of expert witnesses.

Sets precedent in Alaska.

1971

The United States Department of Justice publishes a report, “Flight Characteristics and Stain Patterns of Human Blood,” by Herbert L. MacDonell, an instructor at a two-year college in New York with a master’s degree. It becomes a foundational text in the field of bloodstain-pattern analysis.

1973

MacDonell teaches his first Bloodstain Institute, a weeklong workshop, in Jackson, Miss., training police officers as bloodstain-pattern analysts. Months later, he teaches a second institute in Elmira, N.Y. Over the next few decades, MacDonell will train more than 1,000 new analysts.

1979

Compton v. Commonwealth

The Supreme Court of Virginia rules that bloodstain-pattern analysis is a proper area for expert testimony. A Danville police officer testifies that bloodless circles on the floor helped determine that the victim was probably sitting at a table when shot.

Sets precedent in Virginia.

1980

People v. Erickson

An Illinois appellate court upholds a man’s conviction for murdering his wife. MacDonell testifies that blood patterns on the defendant’s clothes suggest that his wife’s blood spattered on him while he attacked her. On appeal, the court doesn’t rule on the issue of admissibility of experts in bloodstain-pattern analysis, because the defendant did not specifically appeal MacDonell’s admission as an expert.

The case does not set bloodstain-pattern-analysis precedent in Illinois, but it is later cited by courts in Michigan and Texas to support the admission of experts in the field.

State v. Hall

The Supreme Court of Iowa affirms the admission of bloodstain-pattern evidence and the admission of MacDonell as an expert. The court refers to MacDonell’s field as “relatively uncomplicated” and, as a result, does not require extensive proof of its reliability. The judges write, “The evidence offered to show the reliability of the bloodstain analysis included: (1) Professor MacDonell’s considerable experience and his status as the leading expert in the field; (2) the existence of national training programs; (3) the existence of national and state organizations for experts in the field; (4) the offering of courses on the subject in several major schools; (5) use by police departments throughout the country in their day-to-day operations; (6) the holding of annual seminars; and (7) the existence of specialized publications.” The court does not acknowledge that MacDonell himself is the source of almost all these indicators of reliability.

One skeptical judge, Mark McCormick, wrote that he was unable to agree “that reliability of a novel scientific technique can be established solely on the basis of the success of its leading proponent in peddling his wares to consumers.” But McCormick ultimately agreed to admit MacDonell, citing his technique’s grounding in “principles of physics.”

Sets precedent in Iowa.

1981

State v. Hilton

The Supreme Judicial Court of Maine discusses the testimony of an expert in bloodstain-pattern analysis but does not actually rule on the issue of admissibility of such testimony.

The case does not set precedent in Maine, but it is later cited by courts in Idaho and Texas to support the admission of experts in bloodstain-pattern analysis.

1982

State v. Melson

The Supreme Court of Tennessee affirms a man’s murder conviction and death sentence and finds that bloodstain-pattern-analysis testimony of MacDonell was properly admitted.

The court cites the size and established nature of MacDonell’s field, noting he has trained over 300 people and rejects the defense’s argument that MacDonell’s testimony should have been excluded because, among other reasons, his experiments were not conducted on humans. The court quoted MacDonell’s own testimony from trial. “In physics it makes very little difference as to what the origin of the blood is—whether it is polyurethane soaked with blood or it is a human hand—it really doesn’t matter. You transfer energy to blood hydrostatically and you get an eruption of blood spatters.”

Sets precedent in Tennessee.

1983

MacDonell holds his first Advanced Bloodstain Institute. Twenty-two graduates of the advanced class go on to form the International Association of Bloodstain Pattern Analysts, and Tom Bevel, an Oklahoma police officer and former student of MacDonell’s, becomes its first president.

Farris v. State

The Court of Criminal Appeals of Oklahoma rules that bloodstain-pattern analysis is a proper area for expert testimony. In a murder case under review, Bevel had testified that the bloodstains showed that the defendant struck the victim multiple times after shooting him and that the victim tried to defend himself.

Sets precedent in Oklahoma, citing cases in Alaska and California.

1984

People v. Knox

An appellate court in Illinois accepts bloodstain-pattern analysis as a proper area of expertise and holds that a police officer who takes a three-week course with MacDonell is qualified to testify as an expert.

The court finds that a deep knowledge of physics is not required for bloodstain-pattern analysis, calling the technique “essentially one of pattern recognition and reconstruction.” Justice Allan L. Stouder dissents, writing that he doesn’t believe “3 weeks training” qualified the police officer as an expert and that “the study of blood flight characteristics is based upon the laws of physics.”

Sets precedent in Illinois.

1985

Jordan v. State

The Supreme Court of Mississippi holds that even if a court erred in admitting a bloodstain-pattern analyst, the testimony was not influential enough to merit a retrial. The expert in the case attended a one-week institute with MacDonell in 1973, the first year the course was offered.

1987

Lewis v. State

An appellate court in Texas affirms the reliability of bloodstain-pattern analysis and MacDonell’s qualification to testify as an expert witness.

Sets precedent in Texas, citing cases in California, Illinois, Maine and Tennessee.

Fox v. State

The Supreme Court of Indiana upholds a conviction and finds that the expert in bloodstain-pattern analysis who testified in a murder case is qualified because he met Indiana’s requirements for expert witnesses: His knowledge exceeded that of an “average layperson,” and he had sufficient knowledge and skill to aid the judge and jury at trial. The detective who testified had attended one course on bloodstain-pattern analysis in 1980 and had never testified about such evidence before.

Sets precedent in Indiana.

1990

State v. Moore

The Supreme Court of Minnesota rules that bloodstain-pattern analysis is a proper area for expert testimony and that a serology expert who has never taken a course on bloodstain-pattern analysis is qualified enough to testify as an expert.

Sets precedent in Minnesota, citing cases in Illinois, Iowa and Texas.

1991

State v. Rodgers

The Supreme Court of Idaho decides that bloodstain-pattern analysis is an appropriate area for expert testimony. The two experts in the case had trained with MacDonell and Bevel. The court affirms the men’s admission as experts, saying they each meet Idaho’s legal standard, which requires only that experts be more knowledgeable on a topic than an average juror.

The lone dissenting judge, Stephen Bistline, wrote that “The danger presented by expert testimony interpreting blood spatter evidence is that the prosecution is provided with an expert who appears to be able to reconstruct precisely what happened by looking at the blood left at the scene of a crime. However, a quick review of the ‘science’ relied upon by the expert suggests that we would be better off proving guilt beyond a reasonable doubt without the help of such experts.”

Sets precedent in Idaho, citing cases in Maine, Oklahoma, Tennessee and Texas.

1995

State v. Goode

The Supreme Court of North Carolina rules that bloodstain-pattern analysis is admissible as expert testimony and upholds the death sentence for a defendant convicted of two counts of first-degree murder. The expert, Duane Deaver, worked in forensics for the State Bureau of Investigation and had studied bloodstain-pattern analysis with former students of MacDonell and Kirk. Deaver testified that the defendant, who appeared to have no blood on his clothes, actually had minuscule blood spots on his boots.

Sets precedent in North Carolina, citing cases in Idaho, Indiana, Iowa, Oklahoma, Tennessee and Virginia.

People v. Haywood

The Michigan Court of Appeals affirms the admission of bloodstain-pattern-analysis testimony and holds, among other reasons, that because bloodstain-pattern analysis is not a novel technique, extensive proof of its reliability is not required.

Sets precedent in Michigan, citing cases in California and Texas.

2001

State v. Halake

The Tennessee Court of Criminal Appeals overturns a conviction of first-degree murder, finding that a police officer who had never attended a full course in bloodstain-pattern analysis was improperly admitted as an expert at trial. The court contrasts the officer’s credentials with experts admitted by other courts, including MacDonell.

2004

Holmes v. State

The Texas Court of Appeals holds that a detective who attended one weeklong bloodstain-pattern analysis institute is sufficiently qualified to testify as an expert witness and upholds Texas’ longstanding acceptance of bloodstain-pattern analysis as a reliable technique fit for expert testimony.

In affirming the reliability of bloodstain-pattern analysis, the court writes: “Have any courts held blood spatter analysis to be invalid? The short answer is no.”

Sets precedent in Texas, citing cases from 15 other states.

2009

A groundbreaking National Academy of Sciences study finds serious deficiencies in the field of forensic science in the United States and notes that “the uncertainties associated with bloodstain-pattern analysis are enormous.” The authors note that “in general, the opinions of bloodstain-pattern analysts are more subjective than scientific.”

A federal court finds that Duane Deaver, who testified as a bloodstain analyst in the 1995 North Carolina case State v. Goode, had performed inadequate testing. An audit of the state forensics lab where he worked will find the next year that Deaver provided misleading information on his reports for years. By the time these findings are public, judges in Tennessee and Texas will have already have cited State v. Goode in deciding to admit bloodstain-pattern analysis as a reliable field.

2011

MacDonell runs his 76th, and last, institute.

2018

In January, the Texas Forensic Science Commission holds a one-day hearing, prompted in part by its investigation into the role that a minimally trained bloodstain-pattern analyst played in the convictions of Joe Bryan. In February, the commission decides to create an accreditation requirement for bloodstain-pattern-analysis experts in Texas.


Leora Smith was a senior research fellow at ProPublica. She graduated from Harvard Law School in May 2017. Leora has worked with and provided legal support to the Bronx Defenders, the Service Employees International Union, Amnesty International Canada and other international human rights organizations. Her work can be found on the labor law blog On Labor, The Canadian Broadcasting Corporation’s “The Fifth Estate” and in the Harvard Civil Rights-Civil Liberties Law Review.