My good friend Judge Keenan has written a very fine opinion. I assume her effort does not run afoul of the prohibition on advisory opinions by federal courts. See Preiser v. Newkirk, 422 U.S. 395, 401, 95 S.Ct. 2330, 45 L.Ed.2d 272 (1975) (stating “a federal court has neither the power to render advisory opinions nor to decide questions that cannot affect the rights of litigants in the case before them”) (internal quotation marks omitted).1 Accordingly, I *518am pleased to join that part of her opinion which adds to the extant circuit split on the issue of whether, under the residual clauses of federal sentencing enhancement provisions, assault and battery on a law enforcement officer is a crime of violence.
I am compelled to dissent, however, from the majority’s conclusion that the sentencing error in this case is insufficiently “clear” under existing law, Henderson v. United States, — U.S. -, 133 S.Ct. 1121, 185 L.Ed.2d 85 (2013), such that the error cannot plausibly be held “plain” under Rule 52(b). Ante, at 515-17.
Imagine that our panel had on its docket a second case presenting substantially identical issues as this one on substantially identical facts and procedural history. One option for us would be to hold the second case until we issue our opinion in this case so that we can find the error “plain” in the second case. Such an outcome would be required by Henderson. A second option (i.e., the approach taken by the majority in this case), given the imperative that we be “fair” to each appellant in the two cases (and, I suppose, to the two district judges), would be to issue both opinions simultaneously, thereby declining to find the error plain in either one (because the error would not be “clear” until at least one of the opinions had been filed). A third option would be to find the error “clear” and thus “plain” in both cases, regardless of which one was filed first. I believe, given our current understanding of the applicable law, as so well laid out by Judge Keenan, the correct option is to find the error plain in both cases.
In his strongly-worded dissent in Henderson, Justice Scalia scolded the majority for its “mistaken understanding that the only purpose of Rule 52(b) is fairness,” and insisted that the majority had rendered “the plainness requirement ... utterly pointless.” 133 S.Ct. at 1132-1133. In so arguing, Justice Scalia anticipated the very circumstance we face in this case:
Consider two defendants in the same circuit who fail to object to an identical error committed by the trial court under unsettled law. By happenstance, Defendant A’s appeal is considered first. The court of appeals recognizes that there was error, but denies relief because the law was unclear up to the time of the court of appeals’ opinion. Defendant B’s appeal is heard later, and he reaps the benefit of the opinion in Defendant A’s case settling the law in his favor. What possible purpose is served by distinguishing between these two appellants?
*519See id. at 1132 (Scalia, J., dissenting). The six-justice majority was not persuaded by Justice Scalia’s protestations.
It is clear that, not surprisingly, the dissenters in Henderson were most concerned with issues of finality and wasted judicial resources potentially arising from plenary review of forfeited trial errors, the correction of which might upset convictions and make retrials necessary but problematic:
Until today, however, the objective of correcting trial-court error has been qualified by the objective of inducing counsel to bring forward claims of error when they can be remedied without overturning a verdict and setting the convicted criminal defendant free. To overlook counsel’s failure to object, spend judicial resources to conduct plain-error review, and set aside a criminal conviction where retrial may be difficult if not impossible, is exactly the “ ‘extravagant protection’” that this Court has up until now disavowed.
Id. at 1134 (emphasis removed).
In any event, I think the answer to Justice Scalia’s hypothetical, at least regarding errors that result in lengthy illegal sentences, is clear. As only he could, Justice Scalia pooh-poohed the majority’s “disbelief that a lawyer would deliberately forgo objection”:
The Court sees no harm in its evisceration of the contemporaneous-objection rule, disbelieving that a lawyer would ‘deliberately forgo objection now because he perceives some slightly expanded chance to argue for “plain error” later,’ ante, at 1128-1129. It is hard to say whether this conclusion springs from a touching faith in the good sportsmanship of criminal defense counsel or an unkind disparagement of their intelligence. Where a criminal case always has been, or has at trial been shown to be, a sure loser with the jury, it makes entire sense to stand silent while the court makes a mistake that may be the basis for undoing the conviction. The happy-happy thought that counsel will not ‘deliberately forgo objection’ is not a delusion that this Court has hitherto indulged, worrying as it has (in an opinion joined by the author of today’s opinion) about “counsel’s “ ‘sandbagging the court’” by ‘remaining silent about his objection and belatedly raising the error only if the case does not conclude in his favor.’ ”
Id. (citation and emphasis omitted). But the good justice must be forgiven; he’s never conducted a sentencing hearing. There is no sandbagging at sentencing, only errors, sometimes by counsel, sometimes by the court, and sometimes, as in this case, by both the court and counsel. See United States v. Escalante-Reyes, 689 F.3d 415 (5th Cir.2012) (en banc):
[T]he purpose of plain error review in the first place is so that justice may be done. The contemporaneous objection rule is, in part, intended to prevent lawyers from deliberately withholding an objection in an effort to gain another ‘bite at the apple’ on appeal in the event that they are unsatisfied with the court’s ruling. But the plain error rule recognizes that not all failures to object are strategic. Indeed, some (maybe most) of the time, the failure to object is the product of inadvertence, ignorance, or lack of time to reflect.
Id. at 422 (citations, quotation marks, and footnote omitted).
Appellate courts should not hesitate to remediate failures to object at sentencing when those failures result in the imposition of unlawful sentences and the unlawfulness is sufficiently clear at the time the appeal is decided, regardless of the state of the law up until that time. Henderson unequivocally so holds. See 133 S.Ct. at 1130-31 (“[W]e conclude that whether a *520legal question was settled or unsettled at the time of trial, it is enough that an error be ‘plain’ at the time of appellate consideration for [t]he second part of the [four-part] Olano test [to be] satisfied.”) (alterations in original) (quotation marks omitted). We should do so here.
Specifically, I have no hesitation in concluding that the error here is “plain” in light of the wisdom revealed by the combination of United States v. Hampton, 675 F.3d 720 (7th Cir.2012) (holding Illinois offense of assault and battery on a law enforcement officer is not categorically a predicate crime of violence under residual clause), and Rozier v. United States, 701 F.3d 681, 687 (11th Cir.2012) (Hill, J„ dissenting) (same, as to Florida offense of assault and battery on a law enforcement officer), cert. denied, — U.S. -, 133 S.Ct. 1740, 185 L.Ed.2d 798 (2013).2 Manifestly, as the majority opinion makes perfectly clear, Hampton is the most insightful and well-reasoned of the out-of-circuit cases treating the issue of the impact of a law enforcement victim on the analysis of common law-type assault and battery offenses under a “residual clause” determination.3 Judge Keenan’s forceful rejection of the government’s “powder keg” metaphor is as powerful as it is wise and commonsensical.4
*521Additionally, there is authoritative circuit precedent for finding the error here to be plain. In United States v. Boykin, 669 F.3d 467 (4th Cir.2012), we found a district court’s unpreserved procedural sentencing error to be sufficiently clear to merit the “plain error” label. The circumstances surrounding the district court’s misapprehension of a rule of law in that case (circumstances both at the time of the error and at the time of the appeal) were even more opaque than the “crime of violence” determination presented to us in the case at bar.
In Boykin, without objection by the defense to the substantial accuracy of the underlying facts, cf. Fed.R.Crim.P. 32(i)(3)(A) (providing that, at sentencing, the district court “may accept any undisputed portion of the presentence report as a finding of fact”), the district court relied on a presentence report (“PSR”) to determine that the defendant had been convicted of two (of the required three) predicate offenses on “separate occasions” as required by the Armed Career Criminal Act. Namely, he had been convicted of the murder of one victim using one firearm, and (moments later) the assault by shooting of another victim using a second firearm. See 669 F.3d at 469. Under rather obscure circuit precedent, viz. United States v. Thompson, 421 F.3d 278 (4th Cir.2005), which was being interpreted and applied for the first time in a published opinion, and which the district court had actually mentioned at the sentencing hearing (believing it was acting in accordance with it), such reliance was justified only if the information was derived from “Shepard-approved sources.” Boykin, 669 F.3d at 469 (citing, in addition to Thompson, Shepard v. United States, 544 U.S. 13, 125 S.Ct. 1254, 161 L.Ed.2d 205 (2005)).
The panel rejected Boykin’s argument that review was de novo and accepted the government’s contention that plain error review applied. Boykin, 669 F.3d at 469-70. In applying plain error review, the Boykin panel began: “The question is ... whether the facts detailed in the PSR bear [ ] the earmarks of derivation from Shepard-approved sources.” Id. at 471 (brackets in original) (ellipsis added) (quotation marks omitted). The panel did not ask whether the facts were accurate. The panel answered its question by stating, “First, there is no indication in the PSR itself that the information therein came from Shepard-approved sources,” id., contrasting that circumstance with those in a case in which the PSR did happen to disclose the source of information. Id. (citing United States v. Vann, 660 F.3d 771, 817 (4th Cir.2011) (en banc) (Niemeyer, J., dissenting)). Of course, had there *522been such an indication in the PSR, the issue would not have been presented on appeal in Boykin at all.
The Boykin panel then reasoned, “Second, the factual details of the encounter are not typically found in Shepard-approved sources.” Boykin, 669 F.3d at 471. But see Thompson, 421 F.3d at 285.5
Finally, the Boykin panel reasoned that the record on appeal did not contain any documents that
could have conceivably revealed the level of detail of the confrontation as recounted in the PSR and accepted by the district court. As such, we simply cannot determine which facts contained in Boykin’s PSR related to his prior convictions ‘bear[] the earmarks of Shepard-approved documents.’
Thus, while it was not error to use the PSR to determine that two crimes had in fact been committed by Boykin — that information is something that would exist in an indictment or other Shepard-approved source — it was error for the district court to use the PSR’s factual details of the encounter to apply the ACCA enhancement to Boykin’s sentence.
Boykin, 669 F.3d at 471. Of particular relevance to this case, the Boykin panel then concluded that
[t]he error was also plain. There is nothing in the record to show that the PSR’s recounting of the circumstances surrounding the two 1980 convictions exist in Shepard-approved sources. Although some of the information might well appear in such sources, most of it would not, particularly since the sources could not include a plea colloquy or bench findings.
Id. at 471-72. But see United States v. Gillikin, 422 Fed.Appx. 288, 289-90 (4th Cir.2011) (stating, without elaboration, that “[ajlthough the presentence report did not indicate the source the probation officer relied [on] to conclude that the conviction was a violent felony,” the PSR “bears the earmarks of derivation from Shepard-approved sources”). The Boykin panel thus found that a sentencing error by the district court was plain, i.e., clear, even in the face of a rule of criminal procedure that authorized the district court’s finding of facts whose basic accuracy was never challenged by the defendant. And it did so even though prior (unpublished) decisions of this Court had excused the absence of validating source identifiers in the infor*523mation provided by a probation officer in the PSR.
The Boykin panel got plain error review right. See also United States v. Maxwell, 285 F.3d 336 (4th Cir.2002) (in a case of first impression in the Fourth Circuit, finding a sentencing error “plain” where the existence of error hinged on the interpretation of the word “any” in a federal statute). As the Fifth Circuit has explained:
[T]he focus of plain error review should be whether the severity of the error’s harm demands reversal, and not whether the district court’s action ... deserves rebuke. The plain error rule is protective; it recognizes that in a criminal case, where a defendant’s substantial personal rights are at stake, the rule of forfeiture should bend slightly if necessary to prevent a grave injustice.
Escalante-Reyes, 689 F.3d at 423 (citations, quotation marks, and footnote omitted).6
The need for a more enlightened conception of plain error review has recently been well articulated. See, e.g., Dustin D. Berger, Moving Toward Law: Refocusing the Federal Courts’ Plain Error Doctrine in Criminal Cases, 67 U. Miami L.Rev. 521 (2013). Perhaps Henderson signals a step down the road to enlightenment. But enlightenment is not needed in this case; faithful adherence to existing doctrine would do just fine.
For years now, all over the civilized world, judges, legal experts, social scientists, lawyers, and international human rights and social justice communities have been baffled by the “prison-industrial complex” that the United States has come to maintain. If they want answers to the “how” and the “why” we are so devoted to incarcerating so many for so long, they need only examine this case. Here, a 26-year-old drug-addicted confessed drug dealer, abandoned by his family at a very young age and in and out of juvenile court starting at age 12, has more than fourteen years added to the top of his advisory sentencing guidelines range (387 months rather than 211 months, see ante, maj. op. at 508 & n.3), because, as a misguided and foolish teenager, he spit on a police officer. His potential sentence thus “anchored” and “framed”,7 at the high end, between 17 *524and 32 years, Carthorne may or may not feel fortunate to have received “only” 25 years (300 months) in prison. I do not believe he is “fortunate” at all.
I respectfully dissent from the majority’s refusal to find the error in this case “plain.”
. We have not ordinarily followed the practice the majority follows here. That is, when we conduct plain error review, we do not purport to announce a "holding” that the district court indeed committed an error but then, at step two of the plain error analysis, decline to find the error plain. Our normal approach is consistent with the principle that we lack “power to ... decide questions that cannot affect the rights of litigants in the case before [us].” Preiser, 422 U.S. at 401, 95 S.Ct. 2330. Rather, we have taken one or more different paths.
Often, we have simply announced, ambiguously, that there was no "plain error” and left it at that, i.e., without separately deciding whether there was error but that the error was not "clear enough” to be plain. See, e.g., United States v. Strieper, 666 F.3d 288, 295 (4th Cir.2012) (“Because the district court followed the reasoning of the Eighth Circuit regarding an issue on which we have not ruled directly, it did not commit plain error____").
On many other occasions, we have assumed there was error but have relied on Olano step three or step four (see maj. op., ante, at 510) to deny relief. United States v. Jackson, 327 F.3d 273, 304 (4th Cir.2003) (opinion of the court by Motz and King, JJ., on the relevant issue) ("We, along with several of our sister circuits, have frequently disposed of a plain error issue by analyzing either the third or fourth prong of Olano after assuming, without deciding, that there was an error and that it was plain.") (collecting cases).
On at least one or two other occasions, we have reasoned that there was no error that was "plain”, United States v. Wynn, 684 F.3d 473, 480 (4th Cir.2012) ("It is therefore apparent that the issue has not been resolved plainly.”), only to go on to say there was in fact no error at all, id. ("Moreover, on the facts of this case we do not even find error.”), or that Olano step three or four was not satisfied, United States v. Johnson, — Fed.Appx.-, No. 12-4155, 2013 WL 3069776, at *8 (4th Cir. June 20, 2013) (holding that sentencing error was not plain but further holding that defendant failed to satisfy step three of Olano, stating: "Unfortunately for Johnson, even assuming arguendo that the district court’s failure to conclude that USSG § 5G1.3(b) applies to advise the district court that sixteen months of Johnson’s 151-month sentence on Count 1 should run concurrent with his Undischarged State Sentence constitutes error that is plain, thus satisfying the first two prongs of Olano's plain error test, Johnson cannot satisfy the third prong.”).
*518A strikingly odd exception to our practice is the pre-Booker case of United States v. Rouse, 362 F.3d 256 (4th Cir.2004), cert. denied, 543 U.S. 867, 125 S.Ct. 209, 160 L.Ed.2d 112 (2004). In Rouse, plain error review applied to a district court’s failure to impose a concurrent federal sentence; the court imposed instead a departure sentence of ten years to be served consecutively to an active six-year state sentence defendant was then serving. Remarkably, the government conceded that the district court had committed plain error. See Br. of the United States at 8, 2003 WL 25315119 (May 22, 2003) (’’[T]he United States concedes that if the defendant’s [federal] sentence was erroneously imposed [as a consecutive sentence] and that he was [therefore] erroneously sentenced to serve a longer sentence than the law allows, then the district court committed plain error.”). The government contended, however, that in light of the significant "substantial assistance” departure the district court had awarded the defendant, the case should be remanded to allow the district court to reduce the magnitude of the departure in order to achieve the effective sixteen year sentence it intended. Without even mentioning the government’s concession or its alternative argument, the Rouse panel agreed with Rouse that the district court had erred but it simply called the error "not plain.” 362 F.3d at 264.
No previous or subsequent panel of this Court has employed such reasoning so far as I can discern.
. To its credit, the government has not remotely suggested that Olano steps three and four are unsatisfied in this case. There is no doubt that they are satisfied.
. Our own circuit precedent consists entirely of unpublished opinions. See United States v. Baker, 326 Fed.Appx. 213 (4th Cir.2009) (unpublished); United States v. Lowe, No. 94-5792, 1995 WL 440410 (4th Cir. July 26, 1995) (unpublished); United States v. Alston, No. 94-5498, 1995 WL 331095 (4th Cir. June 2, 1995) (unpublished). This is telling; none of our colleagues felt the issue was of sufficient import to deserve a precedential determination. As Henderson makes clear, of course, an error arising out of "unsettled law” can be plain in the light of intervening authority. I do not find anything in Henderson to suggest that new authority that "settles” "unsettled” circuit law must come from the Supreme Court itself or from within the circuit. In other words, applying hindsight, as we must under Henderson, I would conclude that we should find the error here, as of today, sufficiently clear that it rises to the level of plain.
. It bears mention that "assault and battery on a law enforcement officer” is not the proper name or title of Virginia Code § 18.2-57(C), the statute before us. That law, a multi-section statute, has been amended several times since the date of Carthome's conviction, but it presently provides as follows:
[I]f any person commits an assault or an assault and battery against another knowing or having reason to know that such other person is a judge, a magistrate, a law-enforcement officer as defined in subsection F, a correctional officer as defined in § 53.1-1, a person directiy involved in the case, treatment, or supervision of inmates in the custody of the Department of Corrections or an employee of a local or regional correctional facility directly involved in the care, treatment, or supervision of inmates in the custody of the facility, a person directly involved in the care, treatment, or supervision of persons in the custody of or under the supervision of the Department of Juvenile Justice, an employee or other individual who provides control, care, or treatment of sexually violent predators committed to the custody of the Department of Behavioral Health and Developmental Services, a firefighter as defined in § 65.2-102, or a volunteer firefighter or any emergency medical services personnel member who is employed by or is a volunteer of an emergency medical services agency or as a member of a bona fide volunteer fire department or volunteer emergency medical services agency, regardless of whether a resolution has been adopted by the governing body of a political subdivision recognizing such firefighters or emergency medical services personnel as employees, engaged in the performance of his public duties, such person is guilty of a Class 6 felony and, upon conviction, the sentence of such person shall include a mandatory minimum terms of confinement of six months.
Va.Code § 18.2 — 57(C). Notably, as well, the definition of "law enforcement officer” under the statute is exceedingly broad:
*521"Law-enforcement officer” means any full-time or part-time employee of a police department or sheriff’s office that is part of or administered by the Commonwealth or any political subdivision thereof who is responsible for the prevention or detection of crime and the enforcement of the penal, traffic or highway laws of the Commonwealth, any conservation officer of the Department of Conservation and Recreation commissioned pursuant to § 10.1-115, any special agent of the Department of Alcoholic Beverage Control, conservation police officers appointed pursuant to § 29.1-200, and full-time sworn members of the enforcement division of the Department of Motor Vehicles appointed pursuant to § 46.2-217, and such officer also includes jail officers in local and regional correctional facilities, all deputy sheriffs, whether assigned to law-enforcement duties, court services or local jail responsibilities, auxiliary police officers appointed or provided for pursuant to §§ 15.2-1731 and 15.2-1733, auxiliary deputy sheriffs appointed pursuant to § 15.2-1603, police officers of the Metropolitan Washington Airports Authority pursuant to § 5.1-158, and fire marshals appointed pursuant to § 27-30 when such fire marshals have police powers as set out in §§ 27-34.2 and 27-34.2:1.
Id. § 18.2-57(F).
. Although Boykin understood Thompson to have relied on a limited collection of documents “bearing] the earmarks of derivation from Shepard-approved sources,” 669 F.3d at 469 (quoting Thompson), Thompson's actual (and highly ambiguous) recitation of the information on which it held the district court properly relied is far more fulsome:
The trial judge was entitled to rely upon the PSR because it bears the earmarks of derivation from Shepard-approved sources such as the indictments and state-court judgments from his prior convictions, and, moreover, Thompson never raised the slightest objection either to the propriety of its source material or to its accuracy. The PSR details three separate state court judgments, entered on different dates, in which Thompson was sentenced for burglarizing a residence. These three judgments encompass seven different counts of felony breaking and entering, taking place on six different days. And even if they had all occurred on the same day, the PSR further reveals that Thompson’s court proceedings occurred in two separate jurisdictions (Davidson County and Randolph County) and that the residences he burglarized were owned by seven different people living in three different towns.
421 F.3d at 285. This listing of factual details seems to be captured by the Boykin panel's observation that “factual details ... are not typically found in Shepard-approved sources.” Boykin, 669 F.3d at 471.
. In keeping with its office — substantial justice and fairness — the manifest elasticity of plain error review is made clear by the very cases relied on by the majority in its refusal to find the error here plain. See ante, at 516-17 n. 14 ("It is possible for a district court to commit plain error in the absence of controlling authority.” (citing United States v. Neal, 101 F.3d 993, 998 (4th Cir.1996))). And see id. at 517 ("Plain error may arise on occasion when our sister circuits ‘have uniformly taken a position on an issue that has never been squarely presented to this Court.’ ” (quoting United States v. Whab, 355 F.3d 155, 158 (2d Cir.2004))). This is one of those "exceedingly rare” instances in which we should do so. Id.
. Cf. United States v. Jones, 762 F.Supp.2d 270, 284-85 (D.Mass.2010):
If downward departure or variance is appropriate in this case — and I believe it is, how far ought the Court depart? This is the most difficult and offender — specific calculus of all. Is it more appropriate to calculate the departure from the bottom of the 232 month guideline range the so-called "anchoring” to the guidelines principle of which the courts, see e.g., United States v. Docampo, 573 F.3d 1091, 1105 n. 5 (11th Cir.2009); United States v. Bohanon, 290 F.3d 869, 876 (7th Cir.2002), and commentators speak, Sarah M.R. Cravens, Judging Discretion: Contexts for Understanding the Role of Judgment, 64 U. Miami L.Rev. 947, 962 (2010); Jelani Jefferson Exum, The More Things Change: A Psychological Case Against Allowing the Federal Sentencing Guidelines to Stay the Same in Light of Gall, Kimbrough, and New Understandings of Reasonableness Review, 58 Cath. U.L.Rev. 115, 125 (2008); Nancy Gertner, What Yogi Berra Teaches About PostBooker Sentencing, 115 Yale L.J. Pocket Part 127 (2006); Chris Guthrie, Jeffrey J. Rachlinski & Andrew J. Wistrich, Inside the Judicial Mind, 86 Cornell L.Rev. 777, 787-94 (2001); Kate Stith, The *524Arc of the Pendulum: Judges, Prosecutors, and the Exercise of Discretion, 117 Yale L.J. 1420, 1496 (2008)? Or will better justice be accomplished by ratcheting up from top of the appropriate non-career offender guideline out of deference to the congressional mandate which gave rise the career offender concept? 28 U.S.C. § 994(h) (mandating that a "career offender” as defined in the statute receive a sentence at or near the maximum term authorized).