STEVEN FORTIN,
Defendant-Appellant.

Argued: December 9, 1998 - Decided: March 1, 1999

Before Judges King, Wallace and Fall.

On appeal from the Superior Court of New Jersey,
Law Division, Criminal Part, Middlesex County.

Anderson D. Harkov argued the cause for appellant
(Mr. Harkov, on the brief; Robert A. Obler, of
counsel and on the brief).

Nicholas Ruggiero, Assistant Prosecutor, argued
the cause for respondent (Glenn Berman, Middlesex
County Prosecutor, attorney; Mr. Ruggiero, of
counsel and on the brief).

The opinion of the court was delivered by

FALL, J.S.C. (temporarily assigned)

In these consolidated, accelerated appeals in this capital murder case, pursuant to R. 2:2-4 we granted defendant's motions for leave to appeal from two pre-trial evidential rulings. The Law Division judge held other-crime evidence can be admitted against defendant during his trial, pursuant to N.J.R.E. 404(b), on the contested issue of identity. We affirm that ruling. The judge also ruled that through use of a "linkage analysis," a former F.B.I. agent, modus operandi and ritualistic crimes expert, can testify at a trial pursuant to N.J.R.E. 702 that the same person who committed a subsequent crime also committed the crimes in this case. We reverse that ruling.

I

Defendant, Steven Fortin, was indicted on September 6, 1995 and charged with first-degree purposely or knowingly murdering Melissa Padilla on August 11, 1994, contrary to N.J.S.A. 2C:11-3a(1) and (2) (count one); first-degree felony murder, while in the course of committing the crime of robbery, contrary to N.J.S.A. 2C:11-3a(3) (count two); first-degree robbery, contrary to N.J.S.A. 2C:15-1 (count three); first-degree felony murder, while in the course of committing the crime of aggravated sexual assault, contrary to N.J.S.A. 2C:11-3a(3) (count four); and first-degree aggravated sexual assault, contrary to N.J.S.A. 2C:14-2a (count five).

The State seeks imposition of the death penalty, upon conviction. On March 18, 1997, pursuant to N.J.S.A. 2C:11-3(c)(2) and R. 3:13-4(a), the State filed and served a list of aggravating factors it intends to prove at the required sentencing hearing if defendant is convicted on count one.

At some time after 11:29 p.m. on August 11, 1994, Melissa Padilla, an age 25 mother of four children, was sexually assaulted, severely beaten, robbed, and strangled to death in Avenel, New Jersey while walking home, after making purchases at a Quick Chek convenience store.

On April 3, 1995 in the State of Maine, Vicki Gardner, a Maine State Trooper, age 34, was sexually assaulted, severely beaten about the face, and strangled by defendant, who pled guilty to this attack and was sentenced to a twenty-year term of imprisonment in Maine.

Based on evidence linking defendant to the scene of the Padilla murder, certain similarities between the New Jersey and Maine incidents, and other circumstantial evidence, the State moved before the trial court to admit evidence on the issue of identity of defendant's attack on Trooper Gardner in Maine during defendant's trial for the New Jersey murder of Melissa Padilla, pursuant to N.J.R.E. 404(b). Defendant opposed that motion and moved to exclude the testimony of Robert R. Hazelwood, offered by the State as an expert on modus operandi and ritualistic behavior, that the person who committed the Maine crimes against Trooper Gardner was the same person who committed the murder of Melissa Padilla.

The judge conducted a pre-trial hearing pursuant to N.J.R.E. 104(a) concerning the admissibility of the other-crime evidence and on the admissibility of Hazelwood's expert testimony. The hearing took place in May 1998. Eight witnesses testified, including investigating police officers from New Jersey and Maine, the medical examiner who performed the autopsy on Marie Padilla, an orthodontist, a medical expert on the diagnosis of sexual assault, a forensic odontologist, and Hazelwood.

The judge issued comprehensive written opinions on each issue, entering separate orders on July 17, 1998 providing, in pertinent part:

On appeal, defendant presents the following arguments for consideration:

II

During the N.J.R.E. 104 hearing various witnesses testified. The trial judge summarized the witnesses' testimony in his written opinion.

Lieutenant Lawrence Nagle, an investigator with the Middlesex County Prosecutor's Office, testified for the State. He was one of the officers called to the Padilla murder scene near Route 1 and Wylie Street in Avenel, Woodbridge Township, New Jersey during the early morning hours of August 12, 1994. Investigation at the scene revealed that Hector Fernandez notified the police he found the body of his girlfriend, Melissa Padilla, inside a large concrete sewer pipe located approximately ten feet from the Route 1 roadway, in a vacant wooded lot. The victim, age 25, was naked from the waist down, wearing only a shirt and no bra. Fernandez, along with Padilla and her four children were residing at the Gem Motel, located along Route 1, approximately five-hundred feet from the location where Padilla's body was discovered.

At approximately 11:00 p.m. on August 11, 1994, Padilla left the Gem Motel and walked in a northerly direction along a foot path parallel to Route 1 to a Quick Chek convenience store, located on Route 1 about two-tenths of a mile from the murder scene, to purchase food and other items, leaving Fernandez to watch the children. A receipt from the store found at the murder scene indicated purchases were made by Padilla at 11:29 p.m. A bag and various food items purchased at the store were found strewn throughout the area of the murder. Padilla's shorts and underwear, still attached to each other, were found near the scene, thrown into a hedge area. The wooded area where Padilla's body was found contained a pathway where people travelled while walking parallel to Route 1.

Nagle described the injuries to Padilla, stating she had trauma to the head, was beat about the face, was manually strangled, her pants were ripped down, there was rectal tearing and there were marks on her left breast area and on the left side of her chin that appeared to be bite marks.

In April 1995, it came to Nagle's attention that defendant had been arrested and charged in Maine with the assault of a female Maine State trooper, involving bite marks, anal penetration, facial beating, and sexual assault. His office was contacted because the Maine officials knew defendant was from Woodbridge Township, New Jersey. Nagle's investigation of defendant's activities in New Jersey established that at approximately 9:00 p.m. on the night of the Padilla murder, defendant and his girlfriend, Dawn Archer, walked through the area of the murder scene.

Archer gave a detailed statement to an investigator on April 20, 1995. She related the following. On August 11, 1994 defendant and Archer were cohabiting at the Douglas Motel, located along Route 1 in Avenel, not far from the Gem Motel. Archer knew Padilla casually, having seen her a few times in the area where they both resided. On the evening of August 11, 1994, defendant and Archer decided to visit their friend, Charlie Bennett, who also resided along Route 1 at the Five Oaks apartments. Sometime around 9:00 p.m., they walked from their residence, south on Route 1, along the north-bound shoulder. They first stopped at the Quick Chek convenience store, made some purchases and then proceeded further south, passing the Gem Motel, then across the parking lot of Bud's Hut to Bennett's apartment. They stayed at Bennett's apartment for about one hour, when defendant wanted to leave. Archer wanted to stay. They were all consuming alcoholic beverages. At defendant's insistence, they left Bennett's apartment, proceeding back across the parking lot of Bud's Hut. At that point, an argument ensued between Archer and defendant. Archer alleges defendant struck her, threw her to the ground and began choking her. He then fled the scene. After the assault, Archer ran into Bud's Hut for help, and the police were called at approximately 10:30 p.m. Archer reported the incident to the police, who arrived in a patrol car at approximately 10:32 p.m. The police report notes Archer was "very intoxicated." Archer had minor injuries and was transported to Perth Amboy Hospital for medical attention but, upon arriving, she left the hospital without receiving treatment. Instead, Archer went to her mother's home in Perth Amboy and stayed there.

A day or two later, defendant came to visit Archer at her mother's home, apologizing for his conduct. Archer observed scratch marks on defendant's face. They left and went back to the Douglas Motel. Later, while showering, Archer observed additional scratch marks on defendant's chest. Archer contends she did not inflict any scratches on defendant as a result of their fight on August 11, 1994. Upon inquiry, defendant explained that after the incident in Bud's Hut parking lot, he fled the area, was afraid the police were looking for him due to the assault, and therefore slept in the wooded area behind their motel, incurring the scratches while running through the woods. Archer accepted that explanation, although noting in her statement the scratches were of the type inflicted by one person scratching another.

Archer described being assaulted by defendant during their four-month relationship on approximately four occasions. She stated defendant would "just snap," without provocation, and would usually throw her to the ground and choke her. She denied any deviant sexual acts or anal sex in their relationship.

Bennett, who was interviewed on May 9, 1995, advised Nagle that after defendant and Archer left his apartment, defendant returned about thirty to forty-five minutes later. No other testimony was elicited regarding Bennett's statement that defendant returned to his residence. Nagle interviewed defendant April 24, 1995 in Maine. He denied assaulting Archer on the night of Padilla's murder, stating Archer may have fallen, and also denied any knowledge of Padilla's murder.

Although it was determined through the investigation that victim Padilla had sold crack cocaine to at least three individuals on the day she was murdered, no crack cocaine was found near the area of the murder. No money was found on Padilla's person and a gold necklace usually worn by her was missing. Nagle testified this particular area along Route 1 is located in close proximity to the Adult Diagnostic and Treatment Center, where convicted sex offenders are evaluated, imprisoned and treated, and where many paroled sex offenders live in the various motels and other housing units in the area. He also testified that the area where Padilla's body was found has a high level of illegal drug activity. Marvin Shuster, M.D., the Medical Examiner for Middlesex County, performed the autopsy on Padilla's body on August 13, 1994, and testified at the admissibility hearing. Padilla was pronounced dead at the scene of her murder on August 12, 1994 at 1:06 a.m. Dr. Schuster identified the cause of death as the result of "asphyxiation; assault and strangulation."

There were several notable findings in the autopsy report. No lacerations, injuries or lesions were found in the vaginal area. A vaginal smear demonstrated "few and scattered spermatozoa, some of which are intact and many of which are represented only by sperm heads." The statement given by Hector Fernandez reveals he and Ms. Padilla had vaginal intercourse on August 11, 1994. He also told the investigator he had anal intercourse with Padilla about a week before her death. However, DNA analysis of the spermatozoa did not match defendant or Fernandez.

An examination of the anus and rectum of Padilla disclosed "linear, vertically-oriented, superficial lacerations of the mucosa. There are several of these, most of which are anteriorly located. A small amount of blood is present within the anus, along with a few fragments of stool material. The rectum contains no blood." Microscopical examination of the anal smears taken was "negative for the presence of spermatozoa," and the sections of anus disclosed "focal, superficial mucosal hemorrhage. Also noted are foci of dense chronic inflammatory cell infiltrates." Dr. Shuster testified the anal tearing injuries were acute and therefore attributable to the attack resulting in Padilla's death. The chronic anal inflammation found has no relevance to the issues posed on this appeal.

On the day the autopsy was performed, Dr. Shuster consulted Jay S. Kartagener, D.M.D., a forensic odontologist, who examined two sets of circular marks found on the lower left chin of Padilla, and one set of circular marks found on her left breast area. Dr. Kartagener concluded the marks on the chin "strongly resemble human bite marks and it is the opinion of this Examiner that they are indeed the result of human dentition." He also observed that "[t]he markings on the breast may be the result of heavy and firm suckling with two teeth marks being observed, but with their shape and size too indistinct to measure."

Numerous samples from evidence collected at the murder scene and from Padilla's body were sent to a laboratory for polymerase chain reaction (PCR) testing, more commonly referred to as "DNA testing." These included vaginal smears, a cigarette butt, fingernail clippings from both hands, a dollar bill with blood stain, along with blood samples taken from the victim, Fernandez, and from defendant. The report from the laboratory concluded defendant was excluded as a source of the DNA obtained from the dollar bill and right hand fingernail clippings. The data obtained was not consistent with defendant being a primary source of the DNA obtained from the left-hand fingernail clippings, but he could not be excluded as a secondary source of the DNA obtained from those fingernail clippings. The laboratory analysis found defendant could not be excluded as the primary source of the DNA obtained from the cigarette butt. The other testing yielded inconclusive results.

Sergeant Gerard P. Madden and Detective Jackie Therriault, from the Maine State Police, testified that on April 3, 1995, Maine State Trooper Vicki Gardner, traveling in a marked police vehicle, noticed a vehicle on the southbound shoulder of Route I-95 facing in a northerly direction. Upon arriving at the scene, Trooper Gardner determined defendant was the driver of the vehicle. Defendant produced a temporary driver's permit, and Gardner determined through a license check that defendant's license in New Jersey was suspended. After smelling alcohol from defendant's breath, and upon administering balance coordination tests and a field alcohol analysis, Gardner determined defendant was under the influence of alcohol, arrested defendant, and began processing him on a charge of operating a vehicle under the influence of alcohol. Gardner also called for another officer to finish processing defendant, as she was off-duty.

At that point, defendant appeared to "just snap" and attacked Gardner. He beat her about the face, sexually assaulted her, and strangled Gardner into unconsciousness. He then placed Gardner in the passenger's seat of the police vehicle, and drove the police vehicle down the highway. Gardner regained consciousness and partially jumped, and was partially pushed by defendant, from the vehicle. Further down the highway, defendant lost control of the vehicle, it turned over, he fled the scene on foot, and was later apprehended at a rest area about a mile from where the police vehicle turned over.

Gardner's face was badly battered, and she suffered a broken nose; bite marks to her chin, to her left breast nipple and on the outer left side of her left breast; and vaginal and anal tearing and lacerations. After defendant's attack, Gardner was naked from the waist down. Her pants were pulled off, with her underwear inside out. When interviewed later, defendant said Gardner pulled her own pants off while attempting to make sexual advances to him.

On April 26, 1995, a search warrant was issued by the Maine District Court for the search of defendant's parents' home in Maine for any belongings owned by defendant and, specifically, for Padilla's gold necklace. The search failed to yield the necklace.

Dr. Lawrence Ricci, who was qualified as an expert in the field of diagnosis of sexual assaults, confirmed Gardner was sexually assaulted. He examined Gardner on April 5, 1995, and found a vaginal tear. He also found extensive bruising outside the anal area and, further inside, large tears of anal tissue. In comparing the autopsy photographs of the injuries to Padilla, Dr. Ricci offered what he referred to as "a broad opinion" that the injuries depicted in the photographs are consistent with the injuries to Gardner in that they were inflicted "with significant force, a blunt-penetrating trauma." However, the injuries to Gardner's anal area were more severe and extensive than those suffered by Padilla.

As a result of the April 3, 1995 attack against Gardner, defendant was charged in a seven-count information (similar to an accusation in New Jersey) with kidnapping, robbery, aggravated assault, assault on an officer, attempted gross sexual assault, unlawful sexual contact, and criminal operation of a motor vehicle under the influence of intoxicants. On November 29, 1995 defendant pled guilty to all charges and was sentenced to an aggregate twenty-year term of imprisonment.

Lowell J. Levine, D.D.S., a forensic odontologist retained by the State, reviewed the autopsy report, photographs of the bite marks and various investigative material concerning the Padilla murder. He also reviewed photographs of the bite marks found on Gardner and various reports of the Maine incident. Dr. Levine was also provided with dental casts and various bite-mark samples of defendant. His findings concerning comparison of the bite marks on Padilla's body with defendant's dental casts and teeth-mark samples are, in relevant part:

In comparing the bite marks found on Padilla with those resulting from defendant's attack on Gardner, Dr. Levine concluded:

Dr. Levine could not conclude within a reasonable degree of scientific certainty that the bite marks to Padilla's chin were caused by defendant.

Robert R. Hazelwood, the State's proposed expert on modus operandi and ritualistic behavior, summarized his extensive schooling and training. He is a former F.B.I. agent, has over thirty-two years experience as a law enforcement professional, a seventeen-year affiliation with the National Center for the Analysis of Violent Crime, and has consulted on more than 7,000 crimes of violence. He has published approximately thirty articles on topics of homicide, rape, serial rape, other types of sexual assault, and various other criminal behavior topics. Hazelwood has authored or co-authored three books on the subject of criminal or deviant sexual behavior, and has taught at various academies and at a few universities.

At oral argument, we requested copies of articles, or book excerpts, authored by Hazelwood concerning "linkage analysis," the method he used in this case to determine whether the same perpetrator committed both crimes. We were provided a copy of Chapter 7, entitled "The Relevance of Fantasy in Serial Sexual Crime Investigations" of a book authored by Hazelwood titled Practical Aspects of Rape Investigation, A Multidisciplinary Approach (1995). The primary topic of that chapter is the role of "fantasy" as the link between the underlying motivations for sexual assaults and the behavior exhibited during the crime. Hazelwood only briefly discusses the use of "linkage analysis" through examination of the modus operandi and "signature," or "ritualistic," aspects of sexual crimes. Hazelwood explains that crime analysts consider modus operandi, but rely on the signature or ritual of the crime to link a series of cases together. Hazelwood testified use of modus operandi and ritualistic behavior was first published in 1992 in an article by John Douglas, who was also an agent with the FBI, and contends it "is accepted within the law enforcement community." Hazelwood has qualified to testify as an expert on modus operandi and ritualistic behavior in murder cases in California and South Dakota.

Hazelwood was retained by the State to review and analyze materials pertaining to the Padilla murder and the attack on Gardner for the purpose of determining whether the two crimes were committed by the same offender, based on a "linkage analysis" of both incidents. In his October 17, 1997 report, reviewed during testimony, Hazelwood explained the methods used in completing this analysis:

Based on Hazelwood's review of all available reports and documents, he determined that the modus operandi of the crimes involving Padilla and Gardner demonstrated some fifteen similarities:

Hazelwood's report discusses ritualistic, or signature, behavior of the violent offender, as follows:

Hazelwood concludes that the crimes committed against both Padilla and Gardner were anger-motivated, and that the offender demonstrated anger through the following identified "ritualistic" or "signature" behavior in both crimes:

In determining that the attacks on Gardner and Padilla were committed by the same person, Hazelwood concluded:

In his testimony, Hazelwood explained that "linkage analysis" is not a science, but rather is based upon training, education, research, and experience in working on thousands of violent crimes over an extended period of time. Hazelwood testified he has been qualified as an expert on the subject of modus operandi and ritualistic behavior in the courts of California and South Dakota, and that two of his former F.B.I. colleagues, John Douglas and Larry Ankron, have been similarly qualified in other states. Hazelwood has also testified approximately fifteen to twenty times as an expert witness on violent crime.

Hazelwood testified that a "linkage analysis" is a criminal investigative analysis where a series of crimes are reviewed and studied for consistency of modus operandi and ritualistic behavior, explaining further:

Hazelwood emphasized the similarities in the modus operandi of both crimes, particularly in the combination present, coupled with the similar signature, ritualistic behavior in both, combined to form the basis for his opinion that both crimes were committed by the same offender. The judge accepted Hazelwood as an expert under N.J.R.E. 702, qualified to offer an opinion that the same offender committed both crimes.

The judge also ruled that subject to the exclusion of defendant's guilty plea in Maine, evidence of the attack by defendant against Gardner is admissible as other-crime evidence in the capital trial of defendant on the issue of identity, pursuant to N.J.R.E. 404(b).

III

Defendant contends that the judge erred in finding admissible evidence of his subsequent attack on Trooper Gardner in Maine at defendant's forthcoming New Jersey trial concerning the Padilla murder. N.J.R.E. 404(b) provides:

The party seeking admission of other-crime evidence must satisfy the following four-part test:

When reviewing a trial judge's determinations on the admissibility of other-crime evidence under N.J.R.E. 404(b), the scope of review is relatively narrow. Harris v. Peridot Chem. (NJ), Inc., 313 N.J. Super. 257, 278 (App. Div. 1998). The trial court "is given great deference in its discretion of determining the admissibility of other crimes evidence." State v. Marrero, 148 N.J. 469, 483 (1997). The appellate court "must review the record in light of the contention advanced, but not initially from the point of view of how [it] would decide the matter if [it] were the court of first instance." Id. at 278-79. The trial judge is in the "best position to engage in th[e] balancing process" required because of his intimate knowledge of the case. Id. at 279 (quoting State v. Ramseur, 106 N.J. 123 (1987), cert. denied, 508 U.S. 947 (1993)). "Only where there is a `clear error of judgment' should the `trial court's conclusion with respect to that balancing test' be disturbed." State v. Marrero, 148 N.J. at 483 (quoting State v. DiFrisco, 137 N.J. 434, 496-97 (1994), cert. denied, 516 U.S. 1129, 116 S.Ct. 949, 133 L.Ed.2d 873 (1996)).

Thus, the trial judge's decision is accorded deference and reviewed under the "abuse of discretion standard." State v. Crumb, 307 N.J. Super. 204 (App. Div. 1997), certif. denied, 153 N.J. 215 (1998). In addition, deference must be accorded a trial judge's determinations "that are substantially influenced by [his] opportunity to hear and see the witnesses and to have the `feel' of the case which [reviewing judges] cannot enjoy." Harris v. Peridot Chem. (NJ), Inc., 313 N.J. Super. at 279.

We find no abuse of discretion by the judge in his analysis on this issue in his comprehensive written opinion dated June 3, 1998. Other-crime evidence is only admissible if relevant to prove some fact that is genuinely in issue. State v. Marrero, 148 N.J. at 484 (1997). Here, the State offers evidence of the Maine incident on the issue of identity. Evidence is relevant if it has a tendency in reason to prove or disprove any fact of consequence to the determination of the action. N.J.R.E. 401. Defendant denies any knowledge or involvement in the Padilla murder, placing "identity" squarely in issue. This satisfies the first prong of the test.

However, when the issue is "identity", other-crime evidence must generally meet a stiffer standard for admissibility. State v. Reldan, 185 N.J. Super. 494, 502 (App. Div.), certif. denied, 91 N.J. 543 (1982). The prior criminal activity with which defendant is identified must be "so nearly identical in method as to earmark the crime as defendant's handiwork." State v. Sempsey, 141 N.J. Super. 317, 323 (App. Div. 1976), certif. denied, 74 N.J. 272 (1977). The conduct in question "must be so unusual and distinctive as to be like a signature." State v. Inman, 140 N.J. Super. 510, 516 (App. Div. 1976). Where there is proof of recent similar crimes by the accused where "a crime has been committed by some novel or extraordinary means or in a peculiar or unusual manner," the evidence is admissible on the issue of identity. Id. at 517.

Here, there was sufficient evidence presented for the judge to reasonably conclude that this "standard of similarity" was met, and that the eight-month lapse of time between the incidents did not effect the probative value of the other-crime evidence, satisfying the second prong of the test.

The proof presented of defendant's involvement in the Maine incident was clear, convincing and unequivocal, satisfying the third prong of the test.

The judge concluded that the probative value of the evidence was not outweighed by its apparent prejudice. In balancing the prejudice against the probative value, we must also consider whether the other-crime evidence is necessary on the issue of identity. State v. Marrero, 148 N.J. at 482; State v. Brown, 138 N.J. 481, 533 (1994). The burden of convincing the court that the other-crime evidence offered should be excluded because its probative value is substantially outweighed by the risk of undue prejudice is on the party urging exclusion. State v. Carter, 91 N.J. 86, 106 (1982); State v. Collier, 316 N.J. Super. 181, 193 (App. Div. 1998).

Other-crime evidence is generally simultaneously highly probative and extremely prejudicial. State v. Stevens, 115 N.J. 289, 300 (1989). This is particularly true where the relevant issue is identity. Therefore, if identity can be proven through other evidence, the balance in weighing the probative value of the other-crime evidence against its resultant prejudice may tip in favor of exclusion. In State v. Hardaway, 269 N.J. Super. 627 (App. Div. 1994), we reversed admission of defendant's participation in a prior armed robbery to prove defendant's presence at the scene of a homicide where the same handgun was used in commission of both crimes. While possession of the handgun at the robbery within three weeks of the homicide shooting was evidence of defendant's presence at the shooting, there was less prejudicial evidence available to establish defendant's prior possession of the handgun. Id. at 630. We explained:

The inquiry of whether the fact in issue can be proved by less prejudicial evidence than introduction of the other-crime evidence actually focuses on the genuineness of that issue, and whether admission of the other-crime evidence is really necessary to establish that element of the crime. See State v. Stevens, 115 N.J. at 301. Pertinent to this case, in view of the available evidence, can the State prove identity in the Padilla murder prosecution without introducing evidence of defendant's conduct in the Maine incident?

Defendant argues the State already has evidence of defendant's identity in the Padilla murder, tipping the scales in favor of exclusion. Defendant points to the testimony of Dr. Levine, who was able to make a comparison of the bite mark found on Padilla's left breast to defendant's teeth marks, without any reference to the Maine incident. Defendant also points to DNA and other circumstantial evidence linking him to the crime scene. The trial judge carefully considered this argument in his written opinion, stating, in relevant part:

Given our standard of review, we are satisfied the trial judge's decision was not "so wide of the mark that a manifest denial of justice resulted," or that his ruling constitutes an abuse of discretion. State v. Marrero, 148 N.J. at 484 (quoting State v. Kelly, 97 N.J. 178, 216 (1984)). The judge carefully applied the four-prong test outlined in Cofield in determining whether the proffered other-crime evidence was admissible.

The potential for prejudice by admission of the other-crime evidence in this case is great. Therefore, while we are in accord with the judge's ruling permitting its admission, at trial the judge must "sanitize" the other-crime evidence by confining its admissibility to those facts reasonably necessary for the probative purpose of "identity." See State v. Collier, 316 N.J. Super. at 195. To an extent, the judge ruling inadmissible defendant's guilty plea in Maine, goes to this effort of minimizing the prejudicial effect. In Collier, we explained:

Here, the State offered the other-crime evidence to establish the identity of defendant as the perpetrator of the Padilla murder. The similarity in the two crimes is the linchpin to the admission of the Maine evidence. Evidence of the similarity of the assault and the resultant injuries to Trooper Gardner are the signature warranting admission. The other details of the Maine incident, such as defendant driving away in the police car, attempting to push Trooper Gardner from the vehicle, the vehicle flipping over, his flight from the scene and ultimate apprehension are highly inflammatory, and do not go directly to the issue of "identity." We cite these facts by way of example and not in limitation of the sanitizing effort that must be undertaken by the trial judge prior to submission of the other-crime evidence to the jury.

Before trial, counsel should confer in an attempt to agree as to what should be presented to the jury concerning the Maine incident to reasonably satisfy the State's need to present evidence on the issue of identity, while still protecting defendant's right to a fair trial. If the parties cannot agree, the judge must conduct another hearing pursuant to N.J.R.E. 104(a) and, after balancing the State's probative needs against the resultant prejudice, "appropriately sanitize the evidence to assure that the rights of both the State and defendant are protected." Id. at 196.

IV

Defendant also contends the judge erred in qualifying Roy Hazelwood as an expert on modus operandi and ritualistic behavior, pursuant to N.J.R.E. 702, to express his opinion that the same person who committed the Maine assault against Trooper Gardner is the same person who committed the Padilla murder. We agree.

The trial judge makes the preliminary determination whether an expert's testimony would enhance the juror's understanding. State v. J.Q., 252 N.J. Super. 11, 26 (App. Div. 1991), aff'd., 130 N.J. 554 (1993). Such evidential discretionary rulings should not be disturbed unless the trial judge abused his discretion. State v. Harvey, 151 N.J. 117, 166 (1997). "In determining the general acceptance of novel scientific evidence in one case, the court generally will establish the acceptance of that evidence in other cases." Id. at 167 (citing Jones v. United States, 548 A.2d 35, 40 (D.C. 1988)). However,

Here, as the judge noted, Hazelwood testified "this analysis is not based on science, but based on his training and experience with violent crimes." While not based on science in the technical sense, his linkage-analysis methodology is certainly founded in the area of behavioral science, in that it analyzes the conduct of the crime perpetrator in two or more crimes to determine whether there is sufficient consistency of behavior to conclude that one person committed both crimes.

We conclude that the same detailed analysis regarding admission of scientific evidence is applicable and necessary in determining whether linkage-analysis expert testimony is admissible. Theories or methods of explaining human conduct and behavior have consistently been subject to significant scrutiny and analysis by our courts when admission is sought. See State v. J.Q., 130 N.J. at 566-71 (examining relevant authorities and literature in admitting expert testimony concerning the Child Sexual Abuse Accommodation Syndrome (CSAAS) to explain to the jury why victims of sexual abuse recant, but not admissible to prove sexual abuse occurred); State v. Kelly, 97 N.J. at 190-97 (analyzing and accepting expert testimony on the Battered Women's Syndrome in a criminal trial as relevant to the honesty and reasonableness of defendant's belief she was in danger of death or serious injury as relevant to the claim of self-defense); State v. Cavallo, 88 N.J. 508, 516-26 (1982) (analyzing and rejecting proffered psychiatric testimony that defendant did not have the psychological traits of a rapist, based on failure of defendant to establish that such evidence is based on reasonably reliable scientific premises). The admission of linkage-analysis testimony has serious consequences, as it is essentially ultimate-issue evidence. Here, defendant has admitted to the Maine assault on Trooper Gardner. The State proposes for Hazelwood to testify that the same person who committed the Maine assault also committed the Padilla murder. The jury could easily interpret that testimony as an expert conclusion that defendant committed the Padilla murder. Whether the jury accepts that testimony is a different issue. Our Supreme Court has viewed such evidence with skepticism:

Here, the judge applied the requirements contained in State v. Berry, 140 N.J. 280, 289-90 (1995), for admission of expert testimony:

In his written opinion, the judge provided a comprehensive discussion of the criteria listed in Berry. We certainly agree with his conclusion that Hazelwood's testimony would assist the trier of fact in understanding how crimes can be linked together through criminal investigative analysis of the modus operandi and ritualistic behavior associated with the crimes. We also agree that Hazelwood has extensive expertise in the area of criminal investigative techniques. Where we differ with the judge is his conclusion that Hazelwood's linkage analysis is sufficiently reliable for admission in this capital murder prosecution.

The State argues that expert linkage-analysis testimony has been admitted in other jurisdictions as sufficiently reliable. In State v. Pennell, 602 A.2d 48 (Del.Supr. 1991), the defendant was indicted and tried on three counts of murder. The murders were committed on three separate dates in the same general area over a ten-month period. There was substantial physical and circumstantial evidence linking defendant to two of the murders, including DNA and fiber matches. There were also marked similarities in the three crimes. All three victims were known female prostitutes and all had similar injuries, including ligature strangulation marks; numerous skull injuries; pattern bruising of the left breast and nipple; and wrist injuries, suggestive of bondage. The cause of death for two of the victims was identical, strangulation and blunt force injury. The cause of death of the third victim could not be determined because the body, found in the water, was submerged for too long. However, the injuries to her body were remarkably similar to those found on the other two victims.

Defendant was repeatedly observed driving a blue van with no side windows and rounded headlights in the area of the murders. A prostitution-decoy operation and surveillance of defendant resulted in sufficient probable cause for obtaining a search warrant, yielding fibers matching those found on one of the victims; duct tape similar to that found in the hair of another victim; a blood stain in the van; pairs of pliers consistent with pinch-type bruises on the abdomen of one of the victims; and a bag of unused flexicuffs.

Additionally, while there was insufficient evidence to link defendant to the murder of a fourth prostitute, there was evidence received at trial that just before her disappearance, a prostitute was seen entering a blue van with no side windows and rounded headlights. That evidence was received under Delaware Rule of Evidence 404(b) (virtually identical to N.J.R.E. 404(b)), to explain why the police conducted the investigation as they did, and what information they had while conducting the investigation.

The trial court permitted the State to offer agent John Douglas, Director of the F.B.I.'s Behavioral Science Unit, as an expert in the area of serial murder, pursuant to Delaware Rule of Evidence 702 (identical to N.J.R.E. 702). Douglas reviewed the three murders and opined that they were all committed by the same person.

Defendant was convicted of two of the murders, but the jury was unable to reach a verdict on the third murder of the victim whose body was found in the water. On appeal before the Delaware Supreme Court, the convictions were affirmed. In ruling on the testimony of agent Douglas, the court stated, in relevant part:

In reflecting on his role in the Pennell trial, Douglas wrote in his book Mindhunter:

The development of the linkage-analysis approach to crime analysis had its origins in the serial crime studies undertaken by the Behavioral Science Unit of the F.B.I., later known as the Investigative Support Unit, conducted under the supervision of Douglas beginning in the late 1970s. Id. at 32. Douglas and others, including Hazelwood, conducted exhaustive prison interviews of convicted serial killers to collect and analyze profile and modus operandi data, developing a behavioral approach to criminal-personality profiling, crime analysis and prosecutorial strategy. Id. at 31. The F.B.I. then successfully field-tested this method beginning in 1978, by offering psychological profiling consulting services to local law enforcement agencies, resulting in the solving of many serial crimes. Id. at 159. Douglas concluded that "the more behavior we have, the more complete the profile and analysis we can give to the local police." Id. at 31.

Douglas explained the development of this investigative technique, as follows:

In Pennell, by definition, Douglas was analyzing the signature behavior of a serial killer, with distinctive behavioral data from three crime scenes. There was overwhelming evidence of commonality, linking all three murders. While there were differences and similarities in the modus operandi of each murder, the signature was the same. Here, in contrast, there are as many differences as there are similarities in both the modus operandi and signature aspects of the Padilla murder and Maine assault, and there is insufficient evidence to conclude defendant is a serial offender. The basis for development of the linkage-analysis method of identification is the study of serial offenders and three or more crime scenes. See Stephen G. Michaud with Roy Hazelwood, The Evil That Men Do, 177-87 (1998).

In State v. Code, 627 So.2d 1373 (La. 1993), cert. denied, 511 U.S. 1100, 114 S.Ct. 1870, 128 L.Ed.2d 491 (1994), the defendant was charged with eight counts of first-degree murder occurring during three separate incidents over a three-year period in the same general neighborhood. This case involved the trial of defendant on four of those murders occurring on a single date. Defendant was convicted and sentenced to death.

There were overwhelming similarities between the eight murders. All occurred in the same neighborhood where defendant lived; there was substantial physical evidence linking defendant to all three murder scenes, including matching latent fingerprints and similar electrical cord and duct tape used at each scene seized from defendant's home during a consensual search; all but one of the eight victims had their hands tied behind their backs through use of a unique handcuff ligature, with use of a distinctive knot; electrical cord from inside each of the three residences was used to tie most of the victims; most of the victims were gagged, some secured by duct tape; at two murder sites the victims were stabbed multiple times and strangled, and at the third site they were shot and strangled; the murderer was right-handed, as was defendant; the method of binding used on the victims demonstrated the use of total control over the victims during the murders; defendant was observed in the crowd gathering to view the murder scene after each murder; the throat of three of the victims was cut in a distinctive manner, nearly severing the head; at each of the murder sites, the clothing found on one victim was turned inside out; and the coroner's reports and testimony identified the various signature elements of the murders.

The trial judge permitted evidence of the murders occurring on the two other occasions under La.Code of Evid. art. 404(B)(1), virtually identical to N.J.R.E. 404(b), as "other crimes" evidence on the issue of modus operandi. On appeal, the Louisiana Supreme Court recited the requirements for admission of such evidence:

At a pretrial hearing concerning admissibility of the "other crimes" evidence, the state introduced substantial proof showing the connection between defendant and the three murders, in addition to the testimony of the coroner that all three crime scenes were signature crimes of one person. The signature aspects found by the coroner were the handcuff-type bindings with electrical cord, the use of a knife to stab or cut and the immobilization of the victims.

In Code, agent Douglas of the F.B.I. testified at the pretrial hearing as an expert in the field of criminal investigative analysis. Douglas testified the murders were the work of one person. In reaching that conclusion, Douglas relied on several identical ritual aspects of the murders, the most important being the distinctive handcuff ligature, one he had never seen in all his years of experience. Another ritual aspect of the murders he relied upon in reaching his conclusion was the killer's need for manipulation, domination and control over his victims. Other ritual aspects identified by Douglas were the overkill present in each crime scene, with the victims either multiply shot or stabbed, almost decapitated, and the predominant use of a knife.

After hearing this pretrial evidence, the trial judge ruled it admissible, finding its probative value was not substantially outweighed by unfair prejudice. In affirming this ruling, the Louisiana Supreme Court stated:

The pretrial testimony of Agent Douglas was introduced to support the State's position that evidence of the other crimes should be admitted to show modus operandi on the issue of identity in the current murder trial. His testimony concerning the similarity in the ritual aspects of all three murder incidents was introduced to satisfy the requirement for admission of the other crimes evidence of proof demonstrating that "the modus operandi employed by the defendant in both the charged and uncharged offenses must be so peculiarly distinctive that one must logically say they are the work of the same person." Id. at 1381, citing, State v. Hatcher, 372 So.2d 1024, 1033 (La. 1979). Apparently, Douglas did not testify before the jury and directly opine that the same person committed all eight murders. The sole use of Douglas's testimony was in relation to admission of the "other crimes" evidence on the issue of "identity."

We have also been provided with the transcript of a preliminary hearing conducted in a criminal case in California on January 18, 1995, where Hazelwood was qualified as a modus operandi expert and permitted to testify to a linkage analysis he completed concerning eight separate sexual assaults, and to express his opinion that six of the eight appear to have been committed by the same individual.

In that California case, the basis of Hazelwood's testimony and opinion was his analysis of the similarities between the modus operandi and similarities in the sexual rituals in the series of eight sexual assaults analyzed. He explained therein that the modus operandi of the crime has three distinct purposes: to ensure success; to protect identity; and to facilitate escape. Examples are method of entry; type of weapon; binding of the victim; wearing a mask, or the like. He also examined the "ritual" used, unique to the individual, also referred to as a "signature" of the crime. Hazelwood found sufficient similarities in six of the sexual assaults analyzed to support his conclusion the same person committed all six. Some of those similarities observed in those six sexual assaults were: all of the victims were attacked within their own residence; in each instance, the offender brought a knife with him and left with it; the offender in all six wore a mask; nothing of intrinsic value was stolen; there was no forced entry; all of the crimes occurred after dark; in each, methods were used to conceal the offender's identity by using a condom or wiping away seminal fluid discharged; there were similar demographics in the areas where the assaults occurred; two assaults occurred in one geographic area and four assaults occurred in another area; all victims were white females between the ages of 19 and 32; the method a "surprise assault" was used in each; the same methods of control were used in each incident; the sexual behavior during the assault was similar; and the verbal behavior in each instance was similar.

We are satisfied that the quantity of signature evidence in that California rape prosecution was overwhelming, and its nature far more distinctively similar than those identified by Hazelwood in his testimony in this case.

The out-of-state cases are not persuasive when applied to the facts and circumstances of this case. Those cases involved serial killers or rapists, with multiple crimes, and multiple evidential identifiers clearly linking those defendants to the crimes charged. Here, there is an attempt to link behavior in two crimes, under circumstances where there are as many differences as there are similarities, starting with the fact defendant was not charged with the attempted murder of Trooper Gardner. There are differences in the age, race, weight and height of the victims. There is a significant difference in the status of each victim. Trooper Gardner is a professional police officer and a potentially dangerous target for someone to perpetrate a crime against, particularly when defendant knew, prior to the assault, that his identity was made known to the state police dispatcher by Trooper Gardner. There are also differences in the type of assault. Trooper Gardner was anally and vaginally assaulted, while Padilla was assaulted anally, but not vaginally.

Hazelwood testified that modus operandi and ritualistic behavior are the subject of articles and books, and that he teaches this criminal investigative technique to law enforcement personnel. Hazelwood stated the use of modus operandi and ritualistic behavior analysis is accepted in the law enforcement community as a method of perpetrator identification. He said the method is included within the Violent Criminal Apprehensive Program, the Criminal Justice Training Program, and the Homicide Evaluation and Assessment (HEAT) Program in New Jersey.

We have no doubt that these methods are valid and have great value in performing the very difficult task of criminal investigation. We are not persuaded, however, that these techniques are sufficiently reliable for an expert in those fields to testify that the same person who committed one crime committed the other under the analysis of the facts and circumstances presented in this case.

Our examination of the authorities and literature authored by Hazelwood convinces us that a linkage analysis as a foundation for the expert behavior identification testimony proffered in this case is wholly inappropriate. In the recent book The Evil That Men Do, Stephen G. Michaud with Roy Hazelwood (1998), there is significant discussion concerning the application of linkage analysis in the identification of serial offenders. Id. at 177-97. Hazelwood defines therein a serial offender as one who has committed three or more offenses. Id. at 122. Additionally, as described in this book, the use of linkage analysis leading to identification of the perpetrator also involves an evaluation of the personal history and background of the suspected perpetrator, to develop a profile. Id. 177-87.

We are simply not convinced the State has satisfied its burden to establish that "the field testified to [is] at a state of art such that an expert's testimony could be sufficiently reliable[.]" State v. Kelly, 97 N.J. at 208.

Further, given the conclusive nature of Hazelwood's testimony, we find there is no acceptable limiting instruction that could be given to the jury to avoid the prohibition against an expert expressing his opinion "in such a way as to emphasize that the expert believes the defendant is guilty of the crime charged under the statute." State v. Odom, 116 N.J. at 80.

V

In summary, we affirm the decision permitting the introduction of other-crime evidence, after it is properly sanitized, and reverse the order permitting Hazelwood to give expert testimony that the same person who committed the Maine crime also committed the New Jersey crime.

Affirmed in part, reversed in part.