*123OPINION OF THE COURT

This appeal requires us to clarify the circumstances in which points can be assessed to a sex offender, for purposes of determining his or her risk level pursuant to the Sex Offender Registration Act, under risk factor 7, the category addressing the offender’s relationship with the victim. Specifically, we conclude that, in the circumstances of this case, the People failed to prove that defendant established or promoted his longstanding close relationships with the child victims for the primary purpose of victimization. Because points should not have been assessed under risk factor 7, we reverse.

I.

Defendant was charged, in both Queens and Richmond Counties, with committing numerous sex offenses against four children, who were between the ages of 5 and 12 at the time of the abuse. Three of the victims were siblings, and all four of the victims were children of defendant’s childhood friends. As part of a global resolution of the charges, defendant pleaded guilty to four sex offenses in satisfaction of two Queens County superior court informations, and one sex offense in satisfaction of a Richmond County indictment. In accordance with the plea agreements, the courts imposed all of the sentences to run concurrently, for an aggregate sentence of 15 years in prison, followed by four years of postrelease supervision.

When defendant’s release date was approaching, the Board of Examiners of Sex Offenders (the Board) prepared a case summary and risk assessment instrument (RAI), as required by the Sex Offender Registration Act (see Correction Law art 6-C [SORA]), addressing the charges and convictions in both counties. The Board recommended a score of 125 points, which presumptively falls within risk level three. As relevant here, the Board did not recommend any points under risk factor 7, which the RAI entitles “Relationship with victim.” A letter submitted to the court by the Richmond County District Attorney (DA) recommended the same allocation of points as *124recommended by the Board, similarly not including any points under risk factor 7 (see Correction Law § 168-n [3]). At a scheduling conference held a few days before the SORA hearing, Supreme Court, Richmond County clarified that defendant was contesting the Board’s assessment of points only under risk factor 6 (victim characteristics). The court also invited the parties to address risk factor 7, as well as a possible upward departure because defendant had requested a downward departure. The DA thereafter submitted a second letter, arguing that points should be assessed under risk factor 7 and, in the alternative, seeking an upward departure.

At the SORA hearing, defendant did not contest the 105 points recommended under risk factors 2 through 5. The court disagreed with the Board and the DA on risk factor 6, so it did not assess 20 points under that factor. However, the court did assess 20 points under risk factor 7, finding clear and convincing evidence that defendant had “groomed” his victims, and had changed his relationships with them to enable him to sexually abuse them. Thus, the court assessed defendant a total of 125 points, which rendered him a presumptive risk level three sex offender.1 The court stated that, in the event it was subsequently determined that points should not have been assessed under risk factor 7 (which would then render defendant a presumptive risk level two sex offender), there was a basis for an upward departure to risk level three, due to defendant having victimized a child with a disability, who was an especially vulnerable victim.2

On defendant’s appeal, the Appellate Division affirmed, concluding that Supreme Court did not err in assessing points under risk factor 7 (128 AD3d 927 [2d Dept 2015]). After determining that defendant did not establish his entitlement to a downward departure, the Court had no occasion to consider *125the People’s request for an upward departure. This Court granted defendant leave to appeal (26 NY3d 908 [2015]).

II.

One of the legislature s principal goals in enacting SORA was “to protect the public from the danger of recidivism posed by sex offenders” (People v Stevens, 91 NY2d 270, 275 [1998] [internal quotation marks omitted]). To achieve this goal, SORA contains a detailed system of registration and community notification, with each sex offender’s registration and notification obligations emanating from his or her designated risk level within a three-tiered classification scheme (see Correction Law §§ 168-h, 168-i, 168-j; People v Mingo, 12 NY3d 563, 570-571 [2009]; Stevens, 91 NY2d at 275; see also Correction Law §§ 168-p, 168-q). Under SORA, the Board must recommend a risk level for each sex offender nearing release from incarceration (see Correction Law § 168-1 [6]), based on guidelines the Board created at the legislature’s direction (see Correction Law § 168-1 [5]; Sex Offender Registration Act: Risk Assessment Guidelines and Commentary [2006] [Guidelines]).

When it drafted the Guidelines, the Board created the RAI, which is the tool used to assess points to sex offenders under various risk factors in an effort to determine their risk of re-offending and degree of danger to the community. The SORA court must hold a hearing, in which the People have the burden of proving, by clear and convincing evidence, the facts supporting the sought-after determinations (see Correction Law § 168-n [3]; Mingo, 12 NY3d at 571). The court is not bound by the Board’s recommendations but, rather, must make its own determinations based on the evidence (see People v Gillotti, 23 NY3d 841, 852 [2014]). The court is also required to set forth in writing its determinations, along with the facts and legal conclusions supporting those determinations (see Correction Law § 168-n [3]).

Only risk factor 7 is at issue here. The Guidelines provide that 20 points should be assessed under risk factor 7 if “[t]he offender’s crime (i) was directed at a stranger or a person with whom a relationship had been established or promoted for the primary purpose of victimization or (ii) arose in the context of a professional or avocational relationship between the offender and the victim and was an abuse of that relationship” (Guidelines, factor 7). Here, it is undisputed that the only relevant issue within this risk factor is whether defendant *126promoted his relationships with the victims for the primary-purpose of victimization.

This Court has noted that “ [i] t is plain from the face of factor 7 that it is meant to focus on the relationship, or absence of a relationship, between the offender and his [or her] victim before the crime was committed” (People v Johnson, 11 NY3d 416, 420 [2008]). The Commentary to the Guidelines explains — based on cited research and the Board’s expertise “in the field of the behavior and treatment of sex offenders” (Correction Law § 168-1 [1]) — that the circumstances allowing for the assessment of points under risk factor 7 raise “a heightened concern for public safety and need for community notification” (Guidelines at 12). The Commentary emphasizes that this “is not meant to minimize the seriousness of cases where the relationship is other than that of stranger or professional — e.g., familial. The need for community notification, however, is generally greater when the offender strikes at persons who do not know him well or who have sought out his professional care” (Guidelines at 12 n 8 [emphasis added]). The Guidelines further clarify the language of risk factor 7 at issue here, and provide the following examples:

“The phrase ‘established or promoted for the primary purpose of victimization’ is adopted from the Act itself (Correction Law § 168-a[9]). An uncle who offends against his niece generally would not fall into this category. A scout leader who chooses his profession or vocation to gain access to victims and ‘grooms’ his victims before sexually abusing them would qualify” (Guidelines at 12).

Based upon the foregoing, we must determine in this case where on the spectrum to place defendant’s relationship with his victims. In his relapse prevention plan — which he was required to complete as part of his sex offender counseling and treatment program — defendant wrote that “[a] 11 four victims are the children of my childhood and family friends who I grew up with since we were adolescents]. . . . [W]e all socialized together.'I spent most of my leisure time with these families and their children.” The mother of the three sibling victims confirmed that defendant was a longtime family friend of the victims. She informed the police that, in order to give the parents some respite, defendant had been bringing her children to his home and watching them on weekends for approximately eight years, much of which time elapsed before the *127abuse was alleged to have begun. One victim was defendant’s godson, who he had known since the child’s birth. The record further reflects that, three to six months prior to the instant offenses, defendant was forced to move from his residence and, at the recommendation of the fourth victim’s family, he rented an apartment in the same building as that victim’s grandmother and became a neighbor of that victim and his family. Defendant spent significant amounts of time at that family’s home, engaging in activities such as watching movies, playing games and eating together. Thus, as with the other victims, the record establishes defendant’s substantial nonsexual contact with the fourth victim before he began offending against that victim. Overall, defendant socialized with his four childhood friends — the parents of the victims — and their families on a daily or weekly basis for a lengthy period of time.

It is undisputed that defendant did not “establish [ ]” a relationship with these children in order to victimize them (Guidelines, factor 7). To the contrary, the record reflects that his relationships with them commenced during their infancy, due to his close and longtime friendships with their parents. The question is whether the People proved that defendant “promoted” a relationship with the children “for the primary purpose of victimization” (Guidelines, factor 7).

The word “promote” has several common definitions, including “to advance in station, rank, or honor,” and “to contribute to the growth or prosperity of” (Merriam-Webster Online Dictionary, promote [http://www.merriam-webster.com/dictionary/ promote]). However, we do not look at the word “promoted” in a vacuum; we must discern its meaning in the context of the Guidelines and the related Commentary. The Commentary references the greater need for community notification when the offender chooses victims “who do not know him [or her] well” (Guidelines at 12 n 8). The Commentary also provides contrasting examples: an uncle (who would not be assessed points under risk factor 7), and a scout leader who chooses that position to gain access to and grooms his victims before sexually abusing them (who would be assessed those points) (see Guidelines at 12).

In arguing that points should be assessed to defendant under risk factor 7, the People conflate the concepts of grooming a victim and promoting a relationship for purposes of victimiza*128tion.3 It is clear that points were not intended to be assessed under that risk factor based on grooming, in and of itself; instead, the assessment of those points should be determined based on the nature of the relationship in which the grooming takes place. If risk factor 7 were interpreted to require the assessment of 20 points for every offender who groomed a victim — in addition to offenders who are strangers or professionals — then the vast majority of offenders against child victims would be assessed those points. Such a blanket assessment of points is inconsistent with the purpose of the Guidelines, namely, to require enhanced community notification where abuse occurs in more distant relationships, which indicate an increased risk of reoffending.4

We are also unpersuaded by the People’s argument that the abuse of trust in a relationship is key to the assessment of points under risk factor 7. The Guidelines and Commentary refer to the abuse of trust only in the context of a professional relationship (see Guidelines, factor 7; Guidelines at 12). Neither the Guidelines nor the Commentary refer to the “abuse of such relationship” (Guidelines, factor 7; Guidelines at 12) with regard to the second category of offenders — i.e., those who established or promoted the relationship for the primary purpose of offending. If abuse of trust was a fundamental consideration under risk factor 7, the Guidelines would potentially require the assessment of points in almost all relationships, including familial relationships which are specifically excluded by the Commentary and by the plain language of the Guidelines (see Guidelines, factor 7; Guidelines at 12). However, we must apply the Guidelines logically, as they are *129written, and assess points under risk factor 7 only in situations where the need for notification is enhanced (see Johnson, 11 NY3d at 418; Guidelines at 12 n 8). Indeed, we have held that, to avoid anomalous results, courts should not distort the plain meaning of the risk factors, but must assess points as required under the Guidelines; then, if the resulting risk level does not accurately reflect the offender’s danger to the community, the court may exercise its discretion by granting an upward or downward departure to vary from the presumptive risk level indicated by the risk factors (see Johnson, 11 NY3d at 418).

It is noteworthy that the Board did not assess any points against defendant under risk factor 7, and the DA did not request the assessment of such points until prompted by the court to do so. This certainly indicates that the Board interpreted its own Guidelines to mean that points were not intended to be assessed under the circumstances of this case. Given the Board’s expertise in completing RAIs (see Correction Law § 168-1 [1]), and the fact that it drafted the Guidelines at the direction of the legislature (see Correction Law § 168-1 [5]), courts should give careful consideration to the Board’s interpretations of those Guidelines (see Guidelines at l).5

The People bore the burden of establishing by clear and convincing evidence that defendant promoted his relationship with one or more of the victims for the primary purpose of sexually abusing them (see Correction Law § 168-n [3]; Mingo, 12 NY3d at 571). That burden was not met here. The record reflects that he had long-term, preexisting relationships with the children, continued those relationships in the role of a close family friend who regularly spent substantial amounts of time *130with the children and their families, and did not begin to offend against them until the eldest child was approximately 11 years old (see People v Montes, 134 AD3d 1083, 1083 [2d Dept 2015], lv denied 27 NY3d 904 [2016]; see also Guidelines at 12 n 8). Therefore, the evidence in this record does not support Supreme Court’s determination that defendant “promoted” his relationships with these children for purposes of victimization (see Arthur Karger, Powers of the New York Court of Appeals §§ 13:7, 13:9 at 470-473, 479-485 [3d ed rev 2005]), as opposed to redirecting his long-standing close and involved relationships with them in such a way as to allow for sexual abuse.6

While we do not minimize the harm incurred by the interjection of sexual abuse into already-established close relationships, the eventual introduction of abuse into these relationships is not, alone, sufficient to assess points under risk factor 7 against an offender, such as defendant, who knew his victims well (see Montes, 134 AD3d at 1083; Guidelines at 12 n 8). If the legislature had intended for everyone who sexually offended against children to be designated risk level three, it could have imposed a mandatory override under such circumstances; it did not do so. On this record, we conclude that the lower courts erred in assessing 20 points under risk factor 7.7 Subtracting those points results in the assessment of a total of 105 points, rendering defendant a presumptive risk level two sex offender.

We reject the People’s invitation to review Supreme Court’s alternate finding that an upward departure to risk level three *131would be appropriate. Because the Appellate Division held that defendant was presumptively a level three sex offender, it did not consider that alternate ground (128 AD3d at 928). Accordingly, the Appellate Division order should be reversed and the case remitted to the Appellate Division for consideration of issues raised, but not determined, on the appeal to that Court.

(dissenting). Defendant pleaded guilty to, among other things, first-degree sodomy, second-degree rape, and first-degree sexual abuse, for sexually abusing four children: three boys under the age of 11 and a 12-year-old girl suffering from cerebral palsy. The Board of Examiners of Sex Offenders (the Board) and three lower courts all designated defendant a level three sex offender, posing a “high risk” of danger to the community. The majority now reverses this determination and presumptively classifies defendant as a level two, “moderate risk” offender. I dissent.

The Sex Offender Registration Act (SORA) was created “to protect the public from sex offenders” (People v Mingo, 12 NY3d 563, 574 [2009]). To that end, pursuant to SORA, a sex offender nearing release from incarceration receives a numerical risk level — ranging from level one (low risk) to level three (high risk) — based on a variety of risk factors set forth in the Sex Offender Guidelines (Sex Offender Registration Act: Risk Assessment Guidelines and Commentary [2006] [Guidelines]) created at the direction of the legislature (see Correction Law § 168-1). The sole issue in the instant appeal concerns whether the record contains clear and convincing evidence that risk factor 7 (“Relationship Between Offender and Victim”) applies.

I agree with the majority that “we must apply the Guidelines logically” and “as they are written” (majority op at 128-129). Accordingly, I cannot agree with the majority’s conclusion— contrary to the plain language of factor 7 — that defendant did not promote his relationship with any of his victims simply because those relationships were “preexisting” (majority op at 129).

I.

Risk factor 7 is designed to consider the “Relationship [b]e-tween [o]ffender and [v]ictim” (Guidelines, factor 7; see also People v Johnson, 11 NY3d 416, 420 [2008]). Specifically, factor 7 provides that an offender is assessed 20 points if “[t]he offender’s crime (i) was directed at a stranger or a person with whom a relationship had been established or promoted for the primary purpose of victimization or (ii) arose in the context of a *132professional or avocational relationship between the offender and the victim and was an abuse of that relationship” (Guidelines, factor 7).

By its plain language, risk factor 7 covers a broad range of conduct. It covers, for instance, conduct that is “directed at a stranger,” as well as offenders who “established” a relationship for the “primary purpose of victimization” (id.). But factor 7 is not limited to these types of new or unfamiliar relationships; it also encompasses certain established, or preexisting, relationships. For instance, under subdivision (ii), factor 7 covers conduct involving abuse of “a professional or avocational relationship” between offender and victim (id.).

Even more broadly, under subdivision (i) — the provision at issue here — points are assessed under factor 7 not only where the offender “established” a new relationship with a victim, but also where the offender “promoted” an existing relationship for the “primary purpose of victimization” (id.). As the majority notes, the word “promote” has a well-known, commonly-understood meaning (majority op at 127). Necessarily, the factor’s use of the term “promoted” presumes a preexisting relationship; an offender cannot “promote [ ]” a relationship that does not exist (Guidelines, factor 7). To conclude otherwise would read “promoted” out of risk factor 7 altogether.

Yet the majority concludes that — by virtue of his “preexisting relationships” with the victims — defendant’s conduct is automatically excluded from risk factor 7 (majority op at 129). Not only does this conclusion contravene the plain language of the risk factor, but it undermines the precise goal of the Guidelines: to identify offenders that present a unique threat to public safety, whether based on “(i) the offender’s likelihood of reoffense,” or “(ii) the harm that would be inflicted if he did reoffend” (Guidelines at 2). As the Guidelines recognize, certain categories of sex offenses — by their very nature — present a high level of risk based on the “offender’s danger to the community” (Guidelines at 1). Because “the need for notification is enhanced” (majority op at 129), these high-risk offenses will generally trigger certain risk factors and, by design, will yield a higher total score. For instance, an offender convicted of first-degree rape “[b]y forcible compulsion” (Penal Law § 130.35 [1]) will — by and large — trigger a provision of risk factor 1 (“Use of Violence”), resulting in an assessment of at least 10 points (see Guidelines, factor 1).

*133Child sex offenses similarly present a uniquely grave threat to public safety, warranting special attention under the Guidelines. This category of sex crimes poses a risk of extraordinary harm stemming from reoffense — an element that is given particular emphasis under the Guidelines: “It is important to note that the risk level seeks to capture not only an offender’s risk of reoffense but also the harm posed by a particular offender should he reoffend” (Guidelines at 2). The Guidelines therefore single out “the child molester” as a high-risk offender in light of the substantial “harm that would be inflicted” if a child molester were “to reoffend” (id.). Given their high-risk nature, cases involving child sexual abuse will generally trigger a number of risk factors, including, as relevant here, the “Relationship Between Offender and Victim” factor (Guidelines, factor 7). Accordingly, it is by design that “the vast majority of offenders against child victims” may be “assessed . . . points” under risk factor 7 (majority op at 128).

II.

The record is replete with clear and convincing evidence that defendant “promoted” his relationships with the child victims for the primary purpose of abusing them. Particularly when “look[ing] to the most serious wrongdoing” — as the Guidelines instruct (Guidelines at 6) — defendant’s conduct falls squarely within the plain language of risk factor 7.

As to one victim, the little girl, defendant’s confession specifies the precise moment that he began promoting the relationship for the primary purpose of victimizing her: “I then[,] while driving!,] told [the victim] that I had to tell her something that she should know. I told her that I fell in love with her.” Defendant continues:

“As time went on we slowly became closer and better friends, we would play around and tap kiss each other. She would ask me to get this for her and do that for her and bring this or that to her, and I would jump to do it for her . . . Somewhere down the line kissing became more serious.”

Defendant further describes how he planned a trip with her to Florida shortly after telling the young girl he was “in love” with her.

As to the fourth victim, a young boy who was not related to the other three victims, defendant describes in the “My Instant *134Offense[s]” section of his “Relapse Prevention Plan” how he promoted his relationship with this victim:1 “I selected the ten year old male victim as he became good friends with my [godson]. Access was made easier to this boy because he would enjoy visiting [my godson].” In another section, defendant continues: “My ten year old victim was beginning to be allowed to go places with people other than his parents. I made myself available to him as he was eager to get out, especially to spend time with his best friend, my [godson].”

In lieu of examining defendant’s relationship with each victim, as required by the risk factor, the majority — focusing on defendant’s relationship with the victims’ parents — imposes a blanket ban on “preexisting” relationships (majority op at 129).2 For instance, with regard to the fourth victim, there is no indication that defendant had much contact with the victim whatsoever before he began promoting the relationship for the primary purpose of facilitating the abuse. Nonetheless, the majority somehow concludes that defendant did not “promote [ ]” his relationship with this victim for the primary purpose of victimization, noting that defendant had a close and longstanding relationship with the victim’s parents (majority op at 127, 129-130).

Pursuant to its plain language, risk factor 7 is properly evaluated based on an individualized assessment of an offender’s conduct vis-a-vis his victim(s). On this record, there is clear and convincing evidence that defendant “promoted” his *135relationships with these victims “for the primary purpose of victimizing]” them (Guidelines, factor 7).

III.

Given SORA’s compelling purpose — “to protect the public from sex offenders” — “an accurate determination of the risk a sex offender poses to the public is the paramount concern” {Mingo, 12 NY3d at 574). By needlessly limiting the plain language of risk factor 7, the majority inhibits SORA courts statewide from using the Guidelines in a manner that “fully capture [s] the nuances of every case” so that “the instrument will result in the proper classification” (Guidelines at 4).

Constrained from invoking the risk factors, SORA courts— tasked with reaching an assessment that accurately reflects the threat posed by a sex offender — are compelled, with increasing frequency, to resort to “[t]he departure concept,” an already “common aspect of SORA litigation” (see People v Wyatt, 89 AD3d 112, 119 [2d Dept 2011]). The departure concept not only distorts the scheme established by the Guidelines — which advise that departures should be “the exception” and “not the rule” (Guidelines at 4) — but also constrains appellate review of an offender’s risk level determination (see People v Knox, 12 NY3d 60, 70 [2009] [reviewing departure under “abuse of discretion” standard]).

In this case, the Board classified defendant “at a Level III (High) risk” for reoffending.3 Both the Richmond County and Queens County sentencing courts agreed and adjudicated defendant a level three, sexually violent offender (see People v Cook, 129 NY3d 114 [2017] [decided herewith]). In reaching its determination, the Richmond County SORA court found clear and convincing evidence to support factor 7 points, concluding that “defendant brought that relationship to another level in order to accomplish his goals with each of these children.” The Appellate Division affirmed the level three determination (128 AD3d at 927-928). By reversing this holding and subtracting *136those points, the majority presumptively classifies defendant as a “moderate risk” offender and prevents the lower courts from imposing a level three assessment unless an upward departure is imposed.

IV.

Contrary to the majority’s claim, “the interjection of sexual abuse into already-established close relationships” — by “redirecting” a “close and involved relationship! ]” so as to “allow for sexual abuse” — certainly may fall squarely within the plain meaning of risk factor 7 (majority op at 130). In rejecting a commonsense approach, and ignoring defendant’s conduct towards each victim, the majority reduces defendant to a presumptive level two offender, apparently believing this score to be “an accurate determination of the risk” that defendant — a convicted child sex offender — poses to the public (Mingo, 12 NY3d at 574).

Given the plain language of risk factor 7, child molesters may certainly be more likely to receive factor 7 points {see I., supra). That, of course, is the point.

Order reversed and case remitted to the Appellate Division, Second Department, for consideration of issues raised but not determined on appeal to that court.