472 F.Supp.2d 104 (2007)
UNITED STATES of America
Criminal No. 06-10165-GAO.
Jeremy QUINN, Defendant.
United States District Court, D. Massachusetts.
February 6, 2007.
George W. Vien, United States Attorney's Office Boston, MA, for United States of America.
O'TOOLE, District Judge.
After he pled guilty to possessing more than 500 grams of cocaine with the intent to distribute it, I sentenced the defendant Jeremy Quinn to a term of 60 months imprisonment, the statutory mandatory minimum sentence for the crime of conviction but well below the sentence range recommended by the United States Sentencing Guidelines. I stated my reasoning for imposing the sentence both on the record and in summary form in the judgment. This memorandum supplements those explanations.
Quinn and his codefendant, Cindy Yang, were arrested on November 14, 2005, and charged in a criminal complaint with conspiring to distribute cocaine in violation of 21 U.S.C. § § 841 and 846. Both were released on bond, and both provided some cooperation to government agents in the investigation of drug distribution offenses by others. On June 5, 2006, the government filed a one-count information charging each of the defendants with possession of more than 500 grams of cocaine with the intent to distribute it. On July 19, 2006, both defendants waived their right to indictment and offered pleas of not guilty to the information. Both indicated an intention to plead guilty, but with the government's agreement, the tendering of guilty pleas was postponed to permit the preparation of a pre-plea presentence report (PSR) for each defendant. Yang pled guilty and was sentenced on December 19, 2006. I accepted Quinn's guilty plea and sentenced him on January 18, 2007.
Both defendants were arrested on November 14, 2005, as they emerged from an apartment in Quincy, Massachusetts that was leased by Yang and her brother. Agents with the DEA Drug Task Force had been maintaining surveillance on Quinn and Yang as a result of information supplied by a cooperating witness, and agents had set up and monitored at least one controlled purchase of cocaine from Quinn and Yang.
When the defendants were arrested, agents seized 499 grams of cocaine from a cosmetic bag being carried by Yang. Agents also executed search warrants for Yang's Quincy apartment and for Quinn's residence in Holbrook, Massachusetts. From the Quincy apartment, agents seized an additional 676.5 grams of cocaine and related paraphernalia, including four digital scales, five bottles of inositol, three bottles of acetone, and a box of plastic baggies. They also seized a number of hydrocodone/acetaminophen tablets, 23.7 grams of marijuana, and some unidentified pills thought to be prescription medication. From the Holbrook residence, agents seized 7.7 grams of cocaine and miscellaneous other drugs, including oxycodone and diazepam.
The cooperating witness told agents that he had been buying cocaine, as well as some marijuana and illegal pills, from Quinn for about five years and that Yang worked closely with Quinn. He also said that Yang had herself made about 15 sales of cocaine to him in the year between October 2004 and October 2005. Quinn and Yang also supplied information to agents in subsequent proffer sessions about their own and each other's drug selling activities between approximately 2000 and 2005.
The Guidelines Offense Level
Yang's sentencing hearing was held about a month before Quinn's. As usual, the PSR prepared by the probation office proposed a calculation of suggested sentencing ranges under the United States Sentencing Guidelines. In calculating the offense level for drug crimes, the starting point, and typically the most influential step, is to determine the quantity of illegal drugs that should be attributed to the defendant. See U.S. Sentencing Guidelines Manual ("U.S.S.G.") § 2D1.1(c) (2006). Yang's PSR attributed to her the cocaine she was carrying when arrested November 14, 2005, the cocaine found that day in the searches of the Quincy apartment and the Holbrook residence, and the miscellaneous drugs found at both places. After the appropriate conversions, the total quantity of illegal drugs attributed to Yang was 246.13 kilograms of marijuana. That corresponded to a base offense level of 26 in the table at Guidelines § 2D 1.1(c). The PSR then proposed downward adjustments of 2 points because she qualified for the "safety valve," see U.S.S.G. § 5C1.2, and 3 points for acceptance of responsibility, see U.S.S.G. § 3E1.1(a)and (b). Her total offense level, therefore, was proposed to be 21. As a result of the fact that she had no prior criminal record, her criminal history category was I. The Guidelines range for imprisonment for an offense level of 21 and a criminal history category I is 37-46 months.
In contrast, Quinn's PSR, prepared and reviewed by different probation officers from those who had prepared Yang's PSR, presented a much different proposed offense level computation. In addition to the drugs seized in the arrests and searches on November 14, 2005, Quinn's PSR proposed to attribute to Quinn an estimated quantity that reflected, in essence, several years of cocaine selling. On the basis of information presented to the probation office by the government, which came largely from the cooperating witness, the PSR authors estimated that Quinn should be responsible for having bought and sold a kilogram of cocaine every two weeks for three years. They therefore attributed 78 kilograms of cocaine to him, well above the quantities actually seized by agents at the time of his arrest. When the various drug quantities were then converted into marijuana equivalents, the PSR attributed 15,609 kilograms of marijuana to Quinn (compared with the 246.13 kilograms attributed to Yang). This gave him a base offense level of 36 pursuant to the table at Guidelines § 2D1.1(c). Because he had a minor criminal record, he was not entitled to the "safety valve" adjustment that Yang received. The PSR proposed adjusting his offense level downward by 3 levels for acceptance of responsibility to a total offense level of 33. His record put him in criminal history category II, and accordingly the Guidelines range for imprisonment was 151-188 months.
Determining the Drug Quantity under the Guidelines
Under the Guidelines, the base offense level for the crime of possession of cocaine with intent to distribute is determined by the quantity of cocaine "involved" in the offense. See U.S.S.G. § 2D1. 1 cmt. (background). Quantities of drugs not directly part of the crime of conviction may be used to determine the base offense level if those quantities are part of the "relevant conduct" of the defendant with respect to the offense of conviction. Id., cmt. n. 12 ("Types and quantities of drugs not specified in the count of conviction may be considered in determining the offense level. See § 1B1.3(a)(2) (Relevant Conduct).").
As pertains to the issue at hand, "relevant conduct" is defined by § 1 B 1.3 of the Guidelines as follows:
(a) . . . [T]he base offense level where the Guidelines specifies more than one base offense level . . . shall be determined on the basis of the following:
(1) (A) all acts and omissions committed, aided, abetted, counseled, commanded, induced, procured, or willfully caused by the defendant; and
(B) in the case of a jointly undertaken criminal activity (a criminal plan, scheme, endeavor, or enterprise undertaken by the defendant in concert with others, whether or not charged as a conspiracy), all reasonably foreseeable acts and omissions of others in furtherance of the jointly undertaken criminal activity,
that occurred during the commission of the offense of conviction, in preparation for that offense, or in the course of attempting to avoid detection or responsibility for that offense.
(2) solely with respect to offenses of a character for which § 3D1.2(d) would require grouping of multiple counts, all acts and omissions described in subdivisions (1)(A) and (1)(B) above that were part of the same course of conduct or common scheme or plan as the offense of conviction.
Under subsection (a)(1), the "conduct" that is "relevant" to determining the base offense level for a defendant to be sentenced includes not only his own personal conduct (subsection (1)(A)), but also the "reasonably foreseeable" conduct of any confederate of his in joint criminal activity (subsection (1)(B)). Since Quinn and Yang jointly engaged in distributing cocaine, Quinn would be responsible not only for the cocaine he directly possessed or distributed, but also for any cocaine that Yang possessed or distributed in the course of their joint venture. This is why it is appropriate to attribute to Quinn the 499 grams of cocaine carried by Yang when the two were arrested, and why it is appropriate to attribute to Yang the quantity of drugs seized at Quinn's Holbrook residence.
Subsection (a)(2) "provides for consideration of a broader range of conduct" than under subsection (a)(1) "with respect to . . . drug offenses for which the guidelines depend substantially on quantity." U.S.S.G. § 1B1.3 cmt. (background). In a case where a defendant might be convicted of several drug offenses under separate counts, the Guidelines'"grouping rules" call for the aggregation of drug quantities for each count to determine the offense level for the "group" of similar counts. See id. § 3D1.2(d). The "relevant conduct" Guidelines extends this aggregation principle not just to multiple counts of conviction but to "all acts and omissions . . . that were part of the same course of conduct or common scheme or plan as the offense of conviction," id. § 1B1.3(a)(2), regardless of "whether multiple counts are alleged" or proven. Id., cmt. (background.). Quinn's PSR writers, following these instructions, accepted the cooperating witness' description of Quinn's cocaine sales over several years and arrived at an estimated aggregate quantity of cocaine distribution from his "course of conduct" that dwarfed the amounts involved in just the count of conviction. Correspondingly, the Guidelines sentencing range calculated on the aggregate quantity was substantially higher than it would have been without the aggregation.
Quinn's sentencing range, based on the aggregation of drug quantities, was also substantially higher than the range for Yang, which did not take account of any drug quantities from estimates of her prior distribution activity. As noted above, Yang was sentenced before Quinn. At her sentencing, neither the PSR nor the government suggested that the offense level calculation under § 2D1.1(c) should be based on a drug quantity that included amounts involved in her prior "course of conduct" in addition to the quantities involved in the offense of conviction. In retrospect, this is somewhat surprising because the Guidelines on their face would seem to call for the inclusion of such amounts. Further, in my experience, both PSR writers and prosecutors have typically used the Guidelines regime to take account of quantities from prior "relevant conduct" of drug defendants where they thought there was sufficient reliable information available to do so. Even in a case like Yang's where the sentence ultimately imposed was not within the Guidelines range, the range commonly serves as tentative proposed sentence, a starting point for the consideration of all the other statutory factors that need to be considered. The Guidelines range is thus likely to have some significant influence on the ultimate sentence, whether that sentence is within or outside the proposed range. In Yang's case, it is highly probable that she would not have received the sentence she did if her range had been calculated by including estimates of drug quantities from past distribution activity.
The difference between the methods used to calculate the respective Guidelines sentencing ranges for Yang and Quinn, and the ranges themselves, did not become apparent until Quinn's sentencing. Even then, the difference was not raised by anyone until I addressed it by asking the government to explain it. The government denied intentional manipulation of the Guidelines to produce the difference, and that may well be the case; I have no basis for concluding otherwise. The government's explanation for the difference was that it had more reliable information about Quinn's prior drug dealing than it had about Yang's and conveyed that information to the probation officers writing the Quinn PSR, who then used it as the Guidelines would direct to determine Quinn's offense level.
This explanation is not particularly convincing, especially in light of the statements of offense conduct appearing in both Yang's and Quinn's PSRs. In each PSR, the following identical language appears under the heading, "Other information from the CW" [cooperating witness]:
From approximately October 2000 through at least October 18, 2005, QUINN sold the CW large amounts of cocaine, marijuana, and illegal pills. Specifically, the CW paid QUINN $650 per ounce of cocaine. The CW typically bought that cocaine in two-ounce bags. The CW reported that he also bought "hundreds of pounds of marijuana" from QUINN over that time period. The CW initiated these purchases by calling QUINN. Either QUINN or YANG would then make the delivery and sale. According to the CW, QUINN often utilized YANG for deliveries. YANG "worked closely with QUINN" and "was not ignorant as to the contents of the delivery." The CW noted that he met with YANG for deliveries as often as he met with QUINN. The CW said that YANG made about 15 sales to the CW between approximately October 2004 and October 18, 2005. The CW did not specify how many sales QUINN made to the CW.
Presentence Report for Cindy Yang, ¶ 8 (last revised December 11, 2006); Presentence Report for Jeremy Quinn, ¶ 8 (last revised December 12, 2006) (emphasis supplied). The quoted language rather clearly refutes any suggestion that the government lacked information about Yang's prior significant participation in the distribution of cocaine that could have formed the basis for an estimate of quantities attributable to her as "relevant conduct." The government's proffered explanation also ignores the fact that Quinn himself was an available source of information about Yang's prior relevant conduct.
There are two principal reasons why the Guidelines call for consideration of "relevant conduct" in the sentencing judgment, one philosophical, one practical. The philosophical reason is to assure that the sentence properly accounts for all the "harms" actually caused by the defendant's offense, and only those harms. See, e.g., U.S.S.G. ch. 1, pt. A, ¶ 4(a), p.s. (1987) ("A pure real offense system would sentence on the basis of all identifiable conduct. A pure charge system would overlook some of the harms that did not constitute statutory elements of the offenses of which the defendant was convicted."). The practical reason is to diminish the opportunity for prosecutors to control the sentencing process by selective charging decisions. Id. ("The Commission also recognizes that a charge offense system has drawbacks of its own. One of the most important is its potential to turn over to the prosecutor the power to determine the sentence by increasing or decreasing the number (or content) of the counts in an indictment."). In the broad run of cases, the compromise works tolerably well and the punishment imposed reasonably fits the real-world crime.
Even in the execution of the prescriptive mechanisms of the Guidelines, however, some range of judgment is inevitable when it comes to deciding what conduct is sufficiently connected to the offense of conviction to amount to "relevant conduct." What is the "same course of conduct"? What acts of others were "reasonably foreseeable" to the defendant? What acts were "counseled" or "induced" by the defendant? Because "relevant conduct" need not be proved with the rigor necessary for proof of criminal guilt — the Rules of Evidence do not apply and the burden of persuasion is less demanding — it is more likely in a Guidelines calculation than in a formal trial that different evaluators of particular information will form different conclusions about it. It should not be unexpected, therefore, that occasionally different prosecutors, PSR writers, and judges might, on the same information, reasonably come to different conclusions about whether particular uncharged conduct qualifies as "relevant conduct" that ought to affect the Guidelines offense level.
In the case of a single defendant, there might be alternative proposals about whether to treat certain facts as "relevant conduct" which should influence the sentence, but in the end only one of the alternatives will be adopted. In the case of multiple defendants, there is a very real possibility that inconsistent proposals about "relevant conduct" could be adopted and applied severally to similarly situated defendants. The sequence in which similarly situated codefendants are sentenced might lead to this outcome, as this case illustrates. Another way it could happen would be for defendants who had similar roles in the same "common scheme" to be charged in separate cases and thus sentenced by different judges. In such circumstances, a reasonable judgment — defensible on the merits — that certain prior activity qualified as "relevant conduct" might conflict with another reasonable judgment that the same activity was not "relevant conduct." The possibility of inconsistent resolutions of essentially the same question with respect to two separate but similar defendants is a structural problem within the Guidelines' manner of addressing "relevant conduct."
Moreover, because the "relevant conduct" inquiry is adjunct rather than central to the question of criminal culpability it is possible that it will be pursued by different investigators with different levels of vigor and thoroughness. In other words, the Guidelines are susceptible to the possibility that the effect of "relevant conduct" on the sentencing range can depend on something as impossible to know as how aggressively someone, whether prosecutor or probation officer or perhaps even judge, has probed to learn information about a defendant's past illegal activities. Indeed, the government explains the anomaly in this case in just such terms: its investigation was pressed harder as to Quinn than as to Yang, so that information raising Quinn's offense level was discovered while Yang benefitted from a less diligent search for information about her past activity.
The essential scandal of the anomaly as it works in this case is that it directly subverts one of the fundamental objectives of the Guidelines: to reduce disparity in sentences given to similarly situated defendants.
Dealing with the Anomaly
There are several statutory rules of law that are binding with respect to the sentencing of Jeremy Quinn upon his conviction for the offense of possession of more than 500 grams of cocaine with the intent to distribute it. First, the statute that criminalizes the conduct establishes the range of permissible penalties that may be imposed: "In the case of a violation . . . involving . . . 500 grams or more of a mixture or substance containing . . . cocaine . . . such person shall be sentenced to a term of imprisonment which may not be less than 5 years and not more than 40 years. . . ." 21 U.S.C. § 841(b)(1)(B). Second, "[e]xcept as otherwise specifically provided, a defendant who has been found guilty of an offense described in any Federal statute . . . shall be sentenced in accordance with the provisions of this chapter so as to achieve the purposes set forth in subparagraphs (A) through (D) of section 3553(a)(2) to the extent they are applicable in light of all the circumstances of the case." 18 U.S.C. § 3551(a). Those purposes are "(A) to reflect the seriousness of the offense, to promote respect for the law, and to provide just punishment for the offense; (B) to afford adequate deterrence to criminal conduct; (C) to protect the public from further crimes of the defendant; and (D) to provide the defendant with needed educational or vocational training, medical care, or other correctional treatment in the most effective manner." Id. § 3553(a)(2). Third, "[t]he court shall impose a sentence sufficient, but not greater than necessary, to comply" with those purposes. Id. § 3553(a). Fourth, the court must consider the nature and circumstances of the offense, the history and characteristics of the defendant, the sentencing range determined under the Guidelines, any applicable policy statements issued by the United States Sentencing Commission, and the need to avoid unwarranted sentencing disparities among defendants with similar records who have been found guilty of similar conduct. Id. § 3553(a)(1)-(6).
There was formerly a fifth requirement, that the court impose a sentence "of the kind, and within the range" determined under the Sentencing Guidelines, see id. § 3553(b)(1), but that requirement was overridden by the Supreme Court's decision in United States v. Booker, 543 U.S. 220, 259-60, 125 S.Ct. 738, 160 L.Ed.2d 621 (2005). Courts are now required to consider the Guidelines range in deciding what sentence to impose, id., but are not required to impose a sentence that falls within that range, so long as the sentence actually imposed is, in light of all the pertinent and required considerations, not an unreasonable one. See id. at 261-62, 125 S.Ct. 738.
In this case, my consideration of the proposed Guidelines range led me to recognize that imposing a sentence on Quinn within the range would produce a very substantial sentencing disparity between Quinn and Yang. The range proposed by Quinn's PSR was driven not by the conduct in the offense of conviction but by Quinn's past (uncharged) "relevant conduct." In contrast, the range proposed by Yang's PSR was essentially limited to the charged conduct, with no computational effect given to prior "relevant conduct," even though some of that conduct was described in the report. If I were to accept the advice of the Guidelines as proposed for Quinn, I would be agreeing to employ a method of determining his sentencing range that I did not employ in determining Yang's range. To do so would produce an "unwarranted sentence disparit[y]," see 18 U.S.C. § 3553(a)(6), and thus result in a sentence that could fairly be called "unreasonable."
Post-Booker, the "relevant conduct" provisions of the Guidelines are no less advisory than the rest of the Guidelines. But advice to employ those provisions to determine a sentencing range for Quinn when they had not been employed to determine a sentencing range for his codefendant joint venturer would be, in essence, advice to use a double standard. I have obviously considered the advice, as I am required to do, but I have rejected it, as I also am required to do when its acceptance would be unreasonable.
In determining Quinn's ultimate sentence, in conjunction with all the other required factors, I used the same scope of "relevant conduct" that I had used in sentencing Yang. That is, I considered what the sentencing range would have been if the drug quantities used in consulting the table in Guidelines § 2D1.1(c) were limited to those quantities seized in conjunction with defendants' arrest and the ensuing searches of their residences. Those quantities, after conversion, were between 100 and 400 kilograms of marijuana, indicating a base offense level for Guidelines purposes of 26. See § 2D 1.1(c)(7). At this level, given Quinn's criminal history category of II, the sentencing range is 51-63 months, narrowed in this instance by the statutory minimum mandatory sentence to a range of 60-63 months.
Other Sentencing Factors
I explained my consideration of the other factors specified in § 3553(a) on the record at the time of sentencing, and it is not necessary to repeat them here. All things considered, I determined that an appropriate sentence would be imprisonment for a term of 60 months followed by four years of supervised release with conditions.
 Although he co-signed the lease, Yang's brother lives in New Jersey and, so far as it appears, had no connection with his sister's criminal activity.
 In order to tote up the quantity when different drugs are to be taken account of, the Guidelines require the amount of each separate drug to be converted into a marijuana equivalency.
 Although I adopted the Guidelines calculation presented in the PSR, Yang was sentenced below the guidelines range for reasons not directly relevant here, having to do with the impact of her documented psychological problems on her commission of the offense. Cf. U.S.S.G. § 5K2.13.
 To add to the usual difficulties of accurate estimation of quantities involved in a prior course of conduct, the principal source of the information, the cooperating witness, was deceased by the time of sentencing.
 The Guidelines' premise for including "relevant conduct" as well as "offense conduct" in the computation of the recommended sentencing range is that the "principles and limits of sentencing accountability . . . are not always the same as the principles and limits of criminal liability." U.S.S.G. § 1B 1.3 cmt. n. 1.
 Perhaps I deserve criticism for not including amounts from Yang's prior "relevant conduct" in the offense level calculation even though none of the other participants in the sentencing process argued for doing so. I would make two responses to that possible criticism.
First, though there are undoubtedly occasions when it must be done, a court should be cautious about disregarding what appears to be the unanimous view not only of the adversaries but also of the neutral probation officer. For one thing, the prosecutor, the defense counsel, and the PSR writer are likely to have done a more thorough and detailed investigation of the facts of the case than the sentencing judge, and their consensus is typically a sign that an issue they have not raised is not, in fact, an issue. For another, the adversary process is the preferred engine of truth-finding, even in sentencing. We normally trust the participants in the adversary process to make a case for a judicial ruling to their benefit when the facts and law would provide a basis for such a ruling. In this respect, I am reluctant to out-prosecute the prosecutor, especially in light of the harshness of some Guidelines sentences. Furthermore, a judge who too readily disregards what appears to be the unanimous consensus of all other participants flirts with reintroduction of the idiosyncratic use of the sentencing power, one of the evils the Guidelines regime was supposed to guard against.
Second, these general considerations have particular force in a case, like this one, when the question whether to include particular "conduct" as "relevant" is a reasonably debatable one. It was not plainly unreasonable to limit the relevant drug quantity for the crime of possession of cocaine with intent to distribute to the amount of cocaine Yang was in possession of, actually (on her person) and constructively (in the two residences), on the date of the offense specified in the information. It would not have been unreasonable to do the same in determining the quantity for Quinn. What was unreasonable, but did not become apparent until the Quinn sentencing, was using a broad calculation of "relevant conduct" for Quinn, having used a more limited one for Yang.
 According to the government, investigators focused more on Quinn because he appeared to be the leader in the Quinn-Yang team and because they wanted to see if he could provide information that would lead them "up the chain" to other drug suppliers.
 Each PSR recites that its statement of offense conduct "was compiled from the statement of facts submitted by [the] Assistant United States Attorney . . . and from various Drug Enforcement Administration ('DEA') reports and records." Presentence Report for Cindy Yang, ¶ 1 (last revised December 11, 2006); Presentence Report for Jeremy Quinn, ¶ 1 (last revised December 12, 2006).
 Each PSR also recounts information from a named person (presumably not the unidentified cooperating witness) who said that he had been a regular cocaine customer of Quinn's from 2000 to 2005. That information omits any reference to Yang, and it may be that she had no role in the distribution of that cocaine. It may also be that whether she did or not was either under-investigated or underreported.
 See supra, n. 5.
 An additional consideration, germane in some cases but not here, is the need to provide restitution to victims of the offense. Id. § 3553(a)(7).
 I explain here how I arrived at the sentence actually imposed, outside the range Guidelines would have set. As I stated on the record at the sentencing hearing and as the judgment indicates, using the quantities not disputed by the parties (not the quantities proposed by the PSR) and applying the Guidelines on their own terms without adjusting for the anomaly I have identified would result in attribution as "relevant conduct" of between 5 and 15 kilograms of cocaine (or between 1,000 and 3,000 kilograms of marijuana), resulting in an offense level of 32 before adjustment for acceptance of responsibility.