Opinion
HUFFMAN, J.
This case presents the elementary question whether evidence obtained from “Sherlock,” an in-house computer system maintained by the sex crimes unit of the San Diego Police Department (SDPD) for investigative purposes, has been properly introduced and used by the prosecution in this criminal action. We conclude such method of data analysis, though a valuable investigative tool, does not meet the test of admissibility in the guilt phase of a criminal trial. We find the error to be prejudicial and that reversal is required under
Watson.
Preface
Kenneth Hernandez was convicted by jury of six violent sex offenses involving two different victims. As to Jane Doe, Hernandez was found guilty of forcible oral copulation accomplished by kidnapping (Pen. Code,
§§ 288a, subd. (c), 667.61, subd. (d)(2)), of kidnapping for sexual purposes (§§ 207/208, subd. (d)),
of assault with the intent to commit a felony upon her (§ 220), and of sexual battery by restraint (§ 243.4, subd. (a)). As to Monika B., Hernandez was found guilty of sexual battery by restraint (§ 243.4, subd. (a)) and of assault with the intent to commit a felony upon her (§ 220).
At a subsequent court trial, the court denied Hernandez’s motion to dismiss all prior conviction allegations and found true two prior serious felony convictions (§§667, subd. (a)(1), 1192.7, subd. (c)(23)) and that those same priors constituted strikes under the three strikes law. (§§ 667,
subds. (b)-(i), 1170.12 et seq.) The court sentenced Hernandez to an indeterminate term of 110 years to life.
Hernandez appeals, contending the trial court committed reversible error by denying his severance motion, by admitting into evidence police computer data via the testimony of crime analyst Karen Goodman by excluding opinion evidence of a sex crimes detective regarding the lack of modus operand! (M.O.) between the crimes involving the two victims, by committing cumulative errors, and by improperly sentencing him under the three strikes law because his two earlier Florida convictions are not “strike priors” under the initiative version of the law and they predated the operative statute which defines “serious felonies” for purposes of the law.
On this record, we conclude the trial court abused its discretion by admitting into evidence Goodman’s testimony concerning her data base search on Sherlock, the computer containing police report data regarding sex crimes, and that such error was prejudicial. We therefore reverse his current felony convictions and the true findings on the priors dependent upon those convictions. In light of our holding, it is unnecessary to address Hernandez’s other appellate contentions.
Factual and Procedural Background
Pertinent in Limine
Motion:
Shortly before Hernandez’s trial on charges he had committed violent sexual acts upon Jane and Monika, the court entertained several defense motions, including a motion to exclude evidence regarding the crime lab analysis of computer data which the prosecutor hoped to use in its case-in-chief. The prosecution offered to use the computer search to show there had been no similar crimes in the geographic areas where the attacks against the two victims in this case had occurred before Hernandez moved to California and after his arrest. Hernandez contended the evidence was more prejudicial than probative, it was without scientific basis or foundation, and no discovery had been given to the defense concerning the matter.
The court held an evidentiary hearing at which crime analyst Goodman testified she had worked for the SDPD for five years, accessing “various computer systems with law enforcement information to identify suspects, victim profiles, trends analysis on crimes.” She had a bachelor of science degree and had received on-the-job training in her unit from people who had developed the various computer systems, from other crime analysts and from miscellaneous investigative units. She explained the computer systems contained law enforcement information “that’s captured on crime cases, arrests,
field interviews, citations” and that the actual computer system on which she ran the search for this case was named Sherlock.
Goodman described the Sherlock system as: “[A]n in-house data base that was developed by crime analysis. It was defined by the sex crimes unit, variables that we capture in there. Information that is put into the sex crimes file is from a sex crimes log, which each case that is assigned to a sex crimes detective has very specific information that’s put down on a sex crimes log. [ID In turn, that log is given to us and we enter it, give it to a clerical support person in the crime analysis unit who then enters each item into the Sherlock system, into the Sherlock sex crimes files.” The entries to Sherlock were generally done within three to four days after a reported sex crimes incident.
Sherlock was used daily in the normal course of business for the SDPD and its sex crimes data base included every reported sex crime that went to a sex crimes detective. Goodman did not know whether there were any sex crimes not assigned to detectives which might not be entered into Sherlock or whether some sex crimes reported to the San Diego Sheriff’s Department which occurred in the SDPD jurisdiction would necessarily be included. Sherlock’s programs were limited to crimes occurring within the SDPD’s jurisdiction. The SDPD frequently relied upon Sherlock, it was regularly used by the sex crimes unit, and it included separate data base files on homicides and one that “captures unique descriptors of suspects.” Goodman assumed a murder case in which a rape had occurred would be cross-referenced in both the homicide and sex crimes files.
Goodman explained that a crime analyst accessed information from the sex crimes file of Sherlock by picking certain variables and defining the criteria needed in a particular search and then selecting those from the system. Examples of the variables contained in Sherlock were physical descriptions of victims and suspects, such as height, race, sex; variables concerning the crimes, such as date, time, location, weapon; and various M.O.’s like “binding or taping or things that would be key searchable items.” For this case, she did a search on the sex crimes file contained in Sherlock; she did not search in any other data base files.
Regarding the search, Goodman prepared a diagram showing the two police beats, beats 125 and 114, where the crimes in this case had occurred. She searched Sherlock for these two areas for sex crimes involving the specific criteria that the crimes have “suspects that were not Black,” a “stranger,” i.e., no relationship between victim and suspect, and were committed between January 1994 and June 1995. Her initial search produced eight separate incidents. After reviewing each, she narrowed her inquiry
down to two cases which matched the M.O. she was given, based on place of attack and suspect description and “the ‘M.O.’ as well.” The two cases were the ones with which Hernandez is charged here. Goodman explained that the other six cases were eliminated because they did not match the specific search criteria and did not involve a victim who had been attacked while jogging. For example, she eliminated those crimes where the attack took place other than in a canyon, i.e, in a vehicle or residence, where there were multiple suspects, where the victim was a male, and where the reported victim did not pursue the case.
Goodman first worked on this case sometime after February 6, 1995, when she was asked to do so by Detective Gregg. At that point Gregg had one crime that had occurred within a canyon and that is how she came up with the other incident in this case for Gregg. Goodman opined the information from Sherlock was relied upon for the conclusion there was another connected case.
On cross-examination, Goodman testified the original information for the data in the sex crimes log comes from initial police officer reports by the investigating officer who responded to the crime scene who would turn that report over to the sex crimes unit for entry into the sex crimes log.
She explained the original information “is transferred from the crime case to the log by the clerical support or the sex crime detective and it’s identified with the specific incident, location, time, and also a key synopsis of the case summary. That would be key words such as suspect actions or victim actions or statements.” Assailant descriptions and any modifications would also be put in the log. After the log is prepared, a copy is given to Goodman’s unit for crime analysis and inputting into Sherlock.
Goodman agreed the information actually put into Sherlock is not “exactly what’s told to the original reporting officer. It’s either . . . what’s in the 911 tape, or . . . what’s told to the reporting officer, or ... it’s an edited version of what is told to the sex crimes detective [.]”
Goodman stated the sex crimes file contained in Sherlock is not connected to any other law enforcement agency computer systems. She was not aware of anyone who periodically checked to ensure that all sex crimes occurring in certain beats within the SDPD jurisdiction were included in Sherlock. She knew of times where a person reported a sex crime that had occurred in the past, for instance over three years ago, that would be input in the sex crimes log only after it had been written up in a police report. Goodman was unsure
of when the information on the December 7 incident against Monika in this case had been entered into the log.
Going to the reliability of the system, Goodman said her supervisor made “sure that the operations within the office are going as procedure permits.” Any editing authority as to the sex crimes log would be with the detective assigned to a particular case. Goodman further stated: “The only reliability that I can attest to here is that of all the items that come from the sex crimes log they’re put in a chronological order . . . and if something falls out of sequence or is not there, our clerical support would contact the sex crimes [unit] and identify a potential problem . . . that would mean that every case that was assigned to the sex crime detective . . . would be reported on there.”
Goodman also acknowledged that a crime occurring outside the boundaries of the two beat areas for which she was asked to limit her search, even one block outside, would not be included in the data retrieved from Sherlock. She further admitted she did not rely solely on the computer data to arrive at her conclusion there was a similarity between the two incidents in this case; she looked into the sex crimes files and handwritten police reports to check on other variables that she had not used in retrieving data from Sherlock.
On redirect, Goodman testified the data entered into Sherlock daily from the sex crimes logs was not entered for the purposes of this litigation, but rather to assist her and other SDPD crime analysts and detectives with their jobs. On recross, Goodman conceded no other police agency used Sherlock.
The court denied the defense motion to preclude Goodman’s testimony, stating: “Well, there are certain arguments [that] could be made with regard to reliability or accuracy and you’ve made those arguments . . . and I’ve considered them all and I’ve considered the cases, but I think [it] goes to the weight rather than admissibility. Those are arguments you certainly may make to the jury and you certainly may cross-examine her about it, but I think the jury has a right to hear this evidence and then you can argue its weight. But I will allow [the prosecutor] to mention it in his opening statement. [<][]... [A]t least for the present I will not allow, and I will sustain your objection to the computer printouts, Exhibit No. 2. That objection will be sustained on the basis of the cases that you’ve supplied and it has a problem with the intelligibility of those documents and I will sustain your objection. But her testimony will be allowed and that will be the order of the court.”
Trial Evidence:
A.
The Prosecution Evidence:
Monika’s testimony revealed that on December 7, 1994, about 5:20 p.m., she was attacked as she jogged on the sidewalk up Gilman Drive toward the University of California at San Diego (UCSD) in La Jolla, in the same direction as traffic, about 100 yards from the Marriott Residence Inn (Inn). Her assailant, whom she identified at trial as Hernandez, came “out of nowhere” from the bushes area between two condominiums, butting his head into her stomach and knocking her off balance.
As she struggled with him on a grassy area to the right of the sidewalk, they fell to the ground and he laughed “a real sick laugh.” He put his hand up her sweatshirt, touching her on her breasts over her bra and put his other hand down her pants, touching her vagina. Although it was getting dark outside, there was enough light to see Hernandez’s features as his face was only a few inches in front of Monika as they struggled for three to five minutes while cars passed by. Monika broke free from Hernandez, who was unarmed, and ran to the Inn, asking the front desk receptionist to call 911. Hernandez, meanwhile, ran down the hill away from the Inn. The 911 tape was played for the jury.
On February 24, 1995, Monika attended a six-person live lineup and picked Hernandez as the person who attacked her. Later that afternoon she identified him at the preliminary hearing and at trial she was “sure” Hernandez was the person who attacked her.
On cross-examination, she testified she had only “tentatively” identified Hernandez at the live lineup as the person who attacked her because his hair was more unkempt than that of her assailant. While she initially told the 911 operator her assailant was five feet one inch tall, she was pretty upset at that time and called back later to say she had made a mistake, he was really about five feet nine inches tall.
Jane’s testimony revealed that on January 2, 1995, shortly after noontime, she was attacked by a man while she took a break from her jog that day to meditate in Rose Canyon. As she started on the canyon trail, she stopped to help a woman and her small daughter calm their dog. While she was petting the dog, a man with long hair in a ponytail in back and wearing a black ski hat with green, yellow and orange stripes and a black ski coat and black shoes walked by and entered the canyon. Thinking something was not right with the man, that he might be a “criminal, robber, not balanced,” because he
was “dressed for the dead of Wisconsin winter," Jane told the woman it was “a good thing to have a dog—[you] just never know what’s out there.” As she spoke, she briefly saw the man’s face as he glanced over his shoulder at her. At that time, Jane was wearing her glasses.
Over defense objection, Jane was allowed to tell her story via narrative and answering leading questions. She described how the man kept walking and she followed about 50 yards behind him, as she looked for a place to sit quietly and relax. Jane lost sight of the man after he crossed a bridge and she stopped to go a few feet off the path. After she took off her glasses and leaned back on her arms, she saw: “coming back down the path—coming eastward this time was Kenneth Hernandez.[
] He . . . was shooting down the path and he looked like ... he was in a world all of his own. He had this really strange look on his face like he was lost in this little world all of his own and I became a little startled, and at first it looked like he may [sz'c] run by me, but no, he didn’t. He ran right up to me and I put up my fists and I said, ‘Don’t hurt me, don’t hurt me.’ ”
Hernandez then delivered a very sharp blow to the left side of her face, causing her to fall sideways. When he came around and started dragging her by her shirt collar across a clearing and into some bushes, Jane screamed and thought, “Oh, my God, my ticket’s up. This is what rape is about. And when it’s over I’m gonna be dead. Isn’t that what rapists do?"
After dragging her about eighty-seven feet, Hernandez threw her hard on the ground, told her to shut up and that he did not want to hurt her, giving her two to three more blows to the left side of her face. Jane started blacking out, was terror ridden, and was in shock, thinking she had to befriend this man and talk him out of what he was doing. She told Hernandez he reminded her of her ex-boyfriend, that they had just broken up and she was missing him so much that she was happy when Hernandez ran up to her. When Hernandez got on top of her, with his butt resting on her stomach and his hands on her shoulder area, Jane screamed she could not breathe because her sweatshirt was over her head and she wanted to kiss him so she could see his face. Hernandez just “chuckled,” and said “Okay, lady, I’m gonna turn you over now."
Hernandez rolled her to her stomach, ripped down her sweatpants to her ankles, pulled off one of her pant legs and pulled down her underpants. Jane said, “He spread the crease of my buttocks and started licking me from top of the crease down to the vagina and I thought I was gonna throw up.”
Hernandez, fully clothed, then lay full body on top of Jane, with his head over her right shoulder and rubbed his lower self up and down on her. He suddenly stopped, said, “Okay, lady, I’m gonna go now. Don’t shout and don’t scream,” and ran south.
Jane got up and started screaming. She told the man who came to her rescue she had been raped. Jane suffered severe headaches, neck trouble, a swollen face, scratches, scrapes and bruises as a result of the attack.
Concerning the issue of identification, Jane testified she made a 75 percent sure photographic identification of Hernandez as her attacker on February 7, 1995. She did not identify him as her assailant at the live lineup because she was not able to see his yellow, crooked teeth. She also could not get his eye color right—thinking his eyes were either hazel, blue, or green.
On cross-examination, defense counsel reviewed with Jane the various descriptions she had given to police at different times during the investigation of the incident. She told police the attacker could not have been much taller than herself at five feet three and one-half inches tall, was thin, with a dark complexion, possibly Hispanic, had a narrow face, had brown hair, which was long in back but was tucked under a hat at the time of the attack, was about twenty to thirty years of age, had green or hazel eyes, had dirty fingernails and “stank.”
After working with Detective Gregg to draw a composite of her attacker and to do a reenactment of the incident at the crime scene, Jane questioned whether she had the attacker’s height and eye color right. After a series of counseling sessions and making handwritten notes about her descriptions of her attacker, she finally remembered his eyes were brown.
When pressed about her identifications, Jane told defense counsel she was very frustrated with his questions and asked him, “Have you ever been raped? Have you ever been assaulted? Have you ever been knocked around? . . . [W]hen you’re in the middle of it you don’t remember a whole lot right away.” While there was no doubt in her mind her attacker was Hernandez, Jane conceded she had never said her attacker had brown eyes nor that he had a ponytail before this trial. She also admitted that at the preliminary hearing she had said she would “never forget [his] sparkling blue eyes.”
Jane further acknowledged that shortly before the preliminary hearing, she had gone through six runs of a live lineup which included Hernandez in the third position out of six men. In the lineups, the men were asked to turn slowly around, quarter by quarter, to stand oblique, to lean forward to show
their eyes, to take their caps off, to speak loudly the words the attacker said to Jane, to open their mouths to show their teeth, to hold out their hands so that their nails could be seen from all angles, and to pull their pants closer to their bodies. During these run-throughs, Jane took off her shoes and got up on the stage to stand next to the glass for closer viewing. She tentatively identified numbers 3 and 4 as looking like her attacker, but told the prosecutor she was pretty sure her attacker was number 4. She could not identify Hernandez after the lineups because the light was not bright enough to see his teeth. She had worn glasses since the age of four and needed them to drive and read.
At the close of her testimony, Jane viewed Hernandez’s teeth close up and exclaimed, “He’s the man. He’s the one.” Later, the jurors filed by the defense table and viewed Hernandez’s teeth.
In addition to the testimony of the two victims, the prosecutor presented an array of witnesses to bolster its case against Hernandez. Several police officers testified about their interviews with Jane, as did the doctor who examined Jane after her attack and the criminal lab expert who analyzed the samples from Jane’s physical examination. An expert opined Hernandez was within 50 percent of the population who could have given the saliva sample found on Jane during her examination. A detective from the SDPD sex crimes unit also testified Hernandez refused to let a photograph of his teeth be taken.
A man and woman who were walking together on the street above Rose Canyon at the time Jane was attacked testified they heard her scream and saw her dragged across the clearing to some bushes. Neither saw the assailant’s face well enough to identify him. The woman described the attacker as a “scruffy” individual, who was thin, not tall, and who wore a heavy dark jacket, dark pants and a multicolored stocking cap. She noted Hernandez’s size was consistent with the height and build of the person who attacked Jane.
A friend who had known Hernandez in Miami, Florida, in 1987-1988, testified Hernandez had moved in with him to his La Jolla apartment, which was located near UCSD and the Rose Canyon jogging trial on December 3 or 4, 1994. He did not know Hernandez’s daily schedule or activities. He said Hernandez had a four- to five-inch prominent “tail” in his hair in the back. On December 12, 1994, he bought Hernandez a pair of cheap black shoes so he would not have to get his good work shoes dirty. Hernandez lived with him until arrested in February 1995.
Two police officers testified to the circumstances of that arrest. Aware of the recent crime against Jane, one officer became suspicious when he drove
on patrol near Rose Canyon around 7 a.m. on February 6, 1995, and noticed a man who had been walking near by break into a run toward the canyon where a female jogger had just run. He radioed for backup and stopped to investigate.
The uniformed officers walked into the canyon area to check whether the man was stalking or following the woman jogger. As they continued along the path they tried to warn another lone female jogger wearing a headset about the suspicious man, but she just thanked them and continued on. When she was about 75 yards ahead of them, the man appeared from some bushes near the trail, watched the woman jogger “intensely,” and started running in her direction. When he was about five to ten yards behind her, the police radios inadvertently came on—loud enough for the man to hear—and he stopped, turned around to face them, pulled back the hood of his sweatshirt and started stretching, as if doing exercises. He walked toward the officers as they walked toward him. The man’s shoes and six to eight inches of the bottom of his pants legs were very damp. His footprints were seen behind the bushes.
Both officers identified Hernandez in court as the suspicious male and said he was wearing a sweatshirt, dark pants, a stocking cap and dark tennis shoes at the time of his arrest. One officer added that Hernandez had a very thick four-inch ponytail, curly hair, a scar in the shape of an “X” on the bridge of his nose, a dragon tattoo on his right forearm, and was wearing a silver eyeball wedding ring. After his arrest, Hernandez was photographed and had blood was drawn.
B.
Karen Goodman’s Testimony:
The parties stipulated the defense would have a continuing objection to Goodman’s entire testimony. After explaining her SDPD crime analyst job in which she assisted the sex crimes unit with the identification of suspects and victims, her educational background and experience, Goodman testified about the process of how information from the sex crimes log was put into the in-house computer Sherlock daily, and on how it was used to search relevant data, stating that “M.O.” referred to method of operation or similarities.
In this case, Goodman helped the sex crimes detective search significant factors to find other similar crimes. Using facts of a particular time range, January 1, 1994, to June of 1995, and two specific police areas within the city jurisdiction, beats 114 and 125, she searched Sherlock for a certain race of suspect and for a type of relationship between suspect and victim. Her
input of the certain factors, i.e, “suspect race not [equal] to Black," and “stranger,” for the relevant time period and geographical location, turned up eight cases, including the two instant crimes Goodman showed on a chart she had made of beats 114 and 125. She eliminated six cases based on additional information by printing out the data for all eight cases and looking at the variables to weed the others out. For example, she ruled out any case where the place of attack was residential or not in an open area, where weapons were used, where there were multiple suspects, and where glass was broken.
Goodman then used the exhibit No. 2 printout to read to the jury the relevant factors of each of the six crimes she excluded because they did not fit the M.O. of the crimes in this case. She pointed out the third and fourth crimes found by Sherlock on the printout were the two instant crimes against Monika and Jane, which she did include based on the criteria given to her by the detective. In other words, she found only these two cases fit the same M.O. regarding a female jogger being attacked from Sherlock’s data base of all reported cases during the year-and-a-half time period for the geographical areas searched.
C.
The Defense Case:
Hernandez’s defense was mistaken identity.
In addition to an identification expert who testified about the various problems in accurately identifying people after observing them in poor lighting, for a short time period, and under stressful circumstances, various San Diego police officers testified about factors affecting Monika’s and Jane’s identifications of Hernandez as their attacker.
One officer testified Jane first described her assailant as being five feet three inches tall and having greasy hair of unknown length. Detective Gregg then testified Jane initially said her assailant was five feet three inches to five feet four inches tall, thin and had green or hazel eyes; on February 7, 1995, Jane said he was five feet four inches to five feet five inches tall, of medium build, and had hazel or blue eyes. When shown the photographic lineup, Jane thought another person looked as familiar as Hernandez as her attacker. She was not sure about Hernandez, saying his eyes were too dark, she needed to see his crooked, overlapping teeth and hear his voice.
Two other officers testified about inconsistencies in Monika’s identification of Hernandez. Monika told one her assailant was in his 20’s, weighed
180 pounds, was wearing a short-sleeved shirt and that she did not know his eye color. She told the other, who interviewed her a few days after her incident, it was dark out and the lighting was poor at the time of the attack. She was uncertain whether her attacker had facial hair as he kept his face concealed.
Stipulations were read that the sun had set at 4:42 p.m. on the day of Monika’s attack, “the definition of sunset is that sunset is considered to occur when the upper edge of the disk of the sun appears to be exactly on the horizon,” and “[t]he California Vehicle Code requires that all drivers of motor vehicles turn their headlights on one half hour after sunset.”
D.
Further Trial Proceedings:
In closing argument, the prosecutor posited both victims had clearly identified Hernandez in court as the attacker, arguing any inconsistencies in such were minor, and that this direct evidence was supported by the circumstantial evidence in the case. He reviewed the direct evidence and outlined the circumstantial evidence for the jurors: Hernandez was not excluded as the attacker because of the saliva; he was wearing black shoes when arrested; he lived in close geographic proximity to where the two attacks occurred; he moved to the area on December 3 or 4, 1994, shortly before the assaults; he was arrested February 6, 1995, a short distance from the trail where the attack on Jane took place; and there were no similar attacks in the area before December 3, 1994, and none after Hernandez’s arrest.
The prosecutor reminded the jury about Goodman’s testimony, arguing there was no evidence to show that other similar crimes occurred in the area of Monika’s and Jane’s attacks either before Hernandez moved into the area or after his arrest and opining the similarities of the two attacks and the incident when Hernandez was arrested showed he was the one who committed each crime. Each involved a lone jogging female near Rose Canyon with the attacker coming out of some bushes or dragging his victim into bushes, the victims were strangers, and the attacker surprised his victim each time, using physical force rather than a weapon.
The prosecutor stressed the only logical inference the jury could make from the evidence provided by Goodman, a crime analyst who “knows what she’s doing,” that there were no similar crimes in the area for the period between January 1, 1994, and December 3, 1994, and from February 7, 1995, until the time of trial, was that since Hernandez had not moved to the area until December 3, 1994, and he had been in custody since February 7, 1995, he must be the one who committed the charged crimes in this case.
The jury returned verdicts finding Hernandez guilty on both counts involving Monika and on four counts involving Jane.
Discussion
On appeal, Hernandez contends the trial court committed reversible error when it admitted into evidence Goodman’s testimony about her search and analysis of data contained in Sherlock from which the prosecutor argued circumstantial evidence of his identity and guilt. He specifically argues the court abused its discretion in allowing such into evidence under the business records exception to the hearsay rule (Evid. Code, § 1271)
as the requirements of that exception were not met, i.e, there was no proper foundation for its admission, the data contained in Sherlock was based on multiple levels of police reports and therefore was hearsay and unreliable. He also asserts the evidence was nothing more than unqualified expert testimony based on statistical probabilities similar to the mathematical data found improper in
People
v.
Collins
(1968) 68 Cal.2d 319 [66 Cal.Rptr. 497, 438 P.2d 33, 36 A.L.R.3d 1176]. He additionally claims such evidence was and is improper scientific evidence, without proper foundation or procedures.
Hernandez further objects to the admission of Goodman’s analysis, as he did below, on grounds such was more prejudicial than probative under Evidence Code section 352. By emphasizing his custody status he claims its admission also violated his constitutional due process rights. Cumulatively, he claims its admission denied him his basic rights to a fair trial with the right to confront and cross-examine adverse witnesses.
The People respond that Goodman’s testimony regarding her computer search of sexual assaults in two police beats in the La Jolla area between January 1994 and June 1995 was relevant circumstantial evidence of identity and the court carefully exercised its discretion in admitting it into evidence.
While a trial court generally has broad discretion to admit proffered evidence, and its discretion will not be disturbed on appeal absent a showing
of an abuse of that discretion (Cal. Const., art. VI, § 13; Evid. Code, §§ 352-354), in this instance we conclude the trial court abused its discretion in admitting the evidence. In doing so, we recognize that there are times when computer data may properly be admitted as a business record (see
People
v.
Lugashi
(1988) 205 Cal.App.3d 632 [252 Cal.Rptr. 434]) and that the best evidence rule will not foreclose the admission of the computer recorded information or programs in such cases (Evid. Code, § 1500.5). This, however, was not one of those cases.
Admittedly, many of Hernandez’s arguments going to Goodman’s qualifications to testify about the Sherlock system and explain the resultant data have been resolved against him in other reported cases. (See
People
v.
Lugashi, supra,
205 Cal.App.3d at pp. 638-642.) The data bases in those cases, however, generally arose out of some statutory or business duty to record the particular data during the regular course of business
(id.
at pp. 642-644; see also
MacLean
v.
City & County of S.F.
(1957) 151 Cal.App.2d 133, 143 [311 P.2d 158]), and the challenge to their admissibility went to alleged foundational shortfalls, not to the defectiveness of the data base itself.
(People
v.
Lugashi, supra,
205 Cal.App.3d at p. 644.)
Unlike many of those cases, the challenge here boils down to the basic question of whether the sources of information for the data base of Sherlock’s system “were such as to indicate its trustworthiness." (Evid. Code, § 1271, subd. (d).) As noted earlier, the prosecutor argued the information in Sherlock was trustworthy because it was relied upon by the sex crimes detectives on a daily basis to do their jobs solving sex crimes. Such explanation, which the trial court apparently accepted, completely ignores the fact the business records exception has been held inapplicable to admit police reports into evidence for the sheer reason such are or might be based upon the observations of victims and witnesses who have no official duty to observe and report the relevant facts
(Carlton
v.
Department of Motor Vehicles
(1988) 203 Cal.App.3d 1428, 1433 [250 Cal.Rptr. 809];
Hoel
v.
City of Los Angeles
(1955) 136 Cal.App.2d 295, 310 [288 P.2d 989]). The data base in Sherlock was taken from the sex crimes log prepared from the purported “relevant facts” from original police reports, whatever those may be.
This situation is similar to the observation the court in
Carlton
made that even though there may be a duty to file certain reports, such requirement does not transform a police officer’s report of an accident into competent, reliable, trustworthy evidence merely because the information contained in the reports is transferred to a computer. We simply cannot find that the police department’s daily internal procedure of having an employee take
“facts" from police reports of sex crimes, put those “facts” into a sex crimes log, and in turn input those “facts” into Sherlock converts those facts contained in the police officer’s sex crimes reports into competent, reliable, trustworthy evidence that is admissible at trial. (See
Carlton
v.
Department of Motor Vehicles, supra,
203 Cal.App.3d at p. 1433.) In other words, the fact that hearsay evidence is put into a log and then again into a computer in the normal course of business does not render such evidence nonhearsay when it is retrieved from the computer even when most of the requirements of Evidence Code section 1271 are met.
Nor does the fact the court excluded the exhibit No. 2 printout from the search of Sherlock as unintelligible convert Goodman’s testimony from hearsay to nonhearsay. Goodman was allowed to explain the data on that document and interpret its results, which she testified also included additional information from the sex crimes detective who asked her to do the search in preparation for this trial. Hence, in addition to the multiple hearsay the data base contained in Sherlock, the added factors given Goodman were also hearsay and included, as Hernandez points out, an unsubstantiated “unique M.O.” which the prosecutor stated was common to these cases and the circumstances of Hernandez’s arrest.
We therefore conclude the court erred in admitting over Hernandez’s hearsay objection Goodman’s testimony.
The question remains whether such error was prejudicial. We think so. In addition to Goodman’s testimony being improperly admitted into evidence under the business records exception to the hearsay rule, the record is fraught with other problems surrounding its admission. Her testimony gave a false aura of computer infallibility in its identification of Hernandez as the perpetrator of the crimes against Monika and Jane. While we do not find her testimony to be the type of expert scientific evidence envisioned by
People
v.
Kelly
(1976) 17 Cal.3d 24, 30-32 [130 Cal.Rptr. 144, 549 P.2d 1240], the prosecutor presented her as if she were an expert without being so qualified, with her telling the jury her educational background and experience, and the prosecutor stressing her expertise as a crime analyst during final closing argument. And though her testimony may not be exactly the same as the
mathematical probability evidence disapproved of in
Collins,
its admission as substantive evidence is equally unacceptable. (See
People
v.
Collins, supra,
68 Cal.2d 319.)
We believe the crime analyst’s testimony is also somewhat analogous to the inherently prejudicial “profile” evidence which was held to be inadmissible for purposes of determining guilt or innocence in
People
v.
Martinez
(1992) 10 Cal.App.4th 1001 [12 Cal.Rptr.2d 838]. In
Martinez,
the trial court had allowed into evidence expert police officer testimony regarding how auto theft rings operate to show the defendant had knowledge he was driving a stolen car. The officers testified the defendant’s actions and the documents in his possession when arrested were similar to those of known auto thieves.
(Id.
at pp. 1004-1006.) The Court of Appeal reversed, finding the thrust of the admitted evidence to show the defendant “fit” a certain “profile” constituted “improper expert evidence similar to the drug courier profile evidence”
(id.
at pp. 1005-1006), which the federal courts have held to be inadmissible to prove guilt because: “ ‘the admission of this evidence is nothing more than the introduction of the investigative techniques of law enforcement officers. Every defendant has a right to be tried based on the evidence against him or her, not on the techniques utilized by law enforcement officers in investigating criminal activity.’ ”
(Id.
at p. 1006.)
The court in
Martinez
also found the evidence improper under California authority by analogy to several cases concerning the admission of other crimes evidence.
(People
v.
Martinez, supra,
10 Cal.App.4th at pp. 1006-1007.) We find these authorities instructive.
First, in
People
v.
Jackson
(1967) 254 Cal.App.2d 655 [62 Cal.Rptr. 208], where evidence of other crimes not proven to be connected to the defendant was found prejudicial, the court noted that: “Evidence of other similar crimes linked to no one at all is clearly inadmissible to prove any element of the crime charged against a defendant, even though the crime occurred within reasonable proximity of time and place. [Citations.] The perversity of the Attorney General’s argument is apparent when carried to its logical conclusion; for example, pursuant to his reasoning a defendant charged with robbing a service station could be convicted by proof that four other service stations were robbed in the same city on the same night by persons unknown, simply because the
modus operandi
as to each of the other crimes was the same as that employed to commit the crime charged.”
(Id.
at p. 658, original italics.)
Then in
People
v.
Long
(1970) 7 Cal.App.3d 586 [86 Cal.Rptr. 590], where the evidence of other forged checks written by another person but
similar to the check passed by the defendant was found improper, the court stated: “ ‘Circumstantial proof of a crime charged cannot be intermingled with circumstantial proof of suspicious prior occurrences in such manner that it reacts as a psychological factor with the result that the proof of the crime charged is used to bolster up the theory or foster suspicion in the mind that the defendant must have committed the prior act, and the conclusion that he must have committed the prior act is then used in turn to strengthen the theory and induce the conclusion that he must also have committed the crime charged. This is but a vicious circle.’ ”
(Id.
at p. 590, quoting from
People
v.
Albertson
(1944) 23 Cal.2d 550, 580-581 [145 P.2d 7].)
As in Martinez,
Albertson
and
Long,
a similar “vicious circle” is at work here where the data base of Sherlock is used to help law enforcement identify persons that fit certain “suspect profiles.” The purpose of the crime analyst’s testimony was to show circumstantially Hernandez’s identity as the person who committed the crimes against Monika and Jane. To do so, the prosecution via the detective on the case gave the analyst certain factors based on those crimes to limit her search of the sex crimes data base in Sherlock. The results of the search based on such factors would necessarily include those two crimes and any others that fit the M.O. or “profile” factors inputted into Sherlock for the search. However, identity was the issue for trial, i.e., it was not known whether Hernandez was the person who committed those crimes. Thus, in essence, the prosecution via the analyst’s testimony was using evidence of other crimes not yet linked to Hernandez to prove his identity and provide that link. Such circuitous “bootstrap” reasoning to bolster the prosecution case is impermissible and allowed the prosecutor to get before the jury what was in effect other crimes evidence without first requiring proof that Hernandez’s “signature” was connected with the other crimes.
Moreover, while the trial court said it considered all arguments and authorities submitted by the parties on the motion to exclude the analyst’s testimony, we do not find on the record the required affirmative showing that the court did in fact weigh the possibility of prejudice against the probative value of her evidence. (Evid. Code, § 352;
People
v.
Meacham
(1984) 152 Cal.App.3d 142, 155 [199 Cal.Rptr. 586]; see also
People
v.
Ewoldt
(1994) 7 Cal.4th 380, 404 [27 Cal.Rptr.2d 646, 867 P.2d 757].)
Under the circumstances of this case, where the outcome of the trial turned on the credibility of the two victims, whose descriptions of their attacker varied before and at trial and contained many inconsistencies and contradictions, we cannot say the trial court’s error in admitting the analyst’s testimony was harmless.
(People
v.
Watson, supra,
46 Cal.2d at p. 836.) The
devastating effect on Hernandez’s right to a fair trial by the admission of such “pseudo-scientific” testimony, which basically elevated multiple layers of hearsay spit out by a computer system named Sherlock to truth, to bolster such credibility cannot be overstated. We can only conclude there is a reasonable probability that a result more favorable to Hernandez would have been reached in the absence of the court’s error in admitting the challenged testimony and reverse his convictions.
Disposition
The judgment is reversed.
Benke, Acting P. J., and McDonald, J., concurred.
A petition for a rehearing was denied June 9, 1997, and respondent’s petition for review by the Supreme Court was denied August 27, 1997.