Stream Energy, its marketing arm Ignite, and a number of other defendants (collectively, the “Defendants”) appeal the district court’s order certifying a class of some 150,000 plaintiffs (the “Plaintiffs”) in this civil action brought under the Racketeer Influenced and Corrupt Organizations Act (“RICO”), 18 U.S.C. §§ 1961-68. The Plaintiff investors are Independent Associates in Ignite’s multi-level marketing program, who are claiming to be victims of an illegal pyramid scheme. Specifically, the Plaintiffs claim that the Defendants induced the Plaintiffs to participate in the scheme by misrepresenting that Ignite is a legitimate business opportunity, causing them to suffer monetary losses.
The Defendants argue both that Ignite is not an illegal pyramid scheme and, more significantly relevant here, that class certification is inappropriate because individualized questions of reliance and knowledge predominate over any common issues, defeating class certification under Rule 23(b)(3) of the Federal Rules of Civil Procedure. The district court rejected the Defendants’ argument and certified the case as a class action. This Court granted the Defendants leave to file this interlocutory appeal under Federal Rule of Civil Procedure 23(f). After full briefing and argument, we VACATE the district court’s *147class certification order and REMAND the case for the entry of a proper order not inconsistent with this opinion and for such further proceedings as may be appropriate.
I.
Stream Energy began in 2004 as a venture to provide energy services in deregulated energy markets. Stream does not own energy infrastructure. Instead, it resells gas and electricity that it buys from other utilities. According to Stream, it can provide consumers with cheaper services through this arrangement. Stream began its operations in Texas after it received approval from the Texas Public Utility Commission in 2005. Beginning in 2008, Stream sought to expand beyond Texas, and it has expanded operations to other states, including Georgia, Maryland, New Jersey, New York, and Pennsylvania. According to Stream, it has over one million energy customers, and it has sold billions of dollars in electricity and natural gas. It claims that the vast majority of its revenues come from energy sales, not from the profits it receives from its multi-level marketing system.
This appeal, however, primarily involves Ignite and its multi-level marketing venture designed to promote Stream’s energy services to consumers.1 To participate in Ignite’s marketing program, a willing individual pays a fee, typically $329, and may also pay an additional, but optional, monthly fee for an Ignite-based website, or “homesite,” to promote his or her Ignite marketing efforts. In return, the individual becomes an “Independent Associate,” or “IA,” within the Ignite program and receives marketing materials along with opportunities to attend training sessions hosted by Ignite executives and other successful IAs. The IAs may then recruit potential energy customers for Stream as well as additional IAs to join the Ignite program.
Ignite compensates IAs in three primary ways. First, as the Defendants emphasize, IAs receive a monthly commission based on the number of customers they have recruited to purchase energy from Stream. Ignite calls this income Residual Income or. Monthly Energy Income (“MEI”). Second, IAs receive compensation for recruiting other IAs into Ignite, which Ignite calls Leadership Income. Finally, Ignite also compensates IAs for completing an initial recruitment of energy customers and IAs in a prompt manner. Ignite has developed a “3 & 10” model, through which a new IA recruits three new IAs and ten new customers. By meeting various targets, an IA is entitled to receive various payments of what Ignite calls Quick Start Income.
An IA’s success depends primarily on recruiting a “downline” of other IAs who, in turn, recruit other IAs and customers into the Ignite program. As an IA recruits more IAs into the Ignite program, the IA proceeds up an Ignite ladder of leadership positions. All IAs start out as Directors, the lowest level of the Ignite leadership. By recruiting more IAs, the IA can move up three additional leadership levels, to Managing Director, then to Senior Director, and finally to Executive Director. By building a downline, the IA also receives MEI for the customers whom the downline IAs recruit to join Stream, along with bonuses for recruiting additional IAs. As Ignite touts in its marketing *148materials, “the power of Ignite’s Leadership Income plan is that these bonuses are paid not just to five levels, but on every level to unlimited depth. That’s geometric growth to infinity!”
For its top recruiters, Ignite also developed a “Presidential Director” level. Presidential Directors received luxury cars and other perks from Ignite. Many of these individuals promoted the opportunities of the Ignite program at events across the country.
Ignite has promoted its multi-level marketing program through many forms of media. Ignite developed a magazine called Empower, which featured profiles of the most successful IAs along with other stories encouraging, prospective IAs to join Ignite. Presidential Directors promoted Ignite through presentations fir IAs and prospective IAs. For example, Presley Swagerty, known as the “Coach,” and Randy Hedge, known as the “Cowboy,” were particularly prolific in promoting Ignite through videos, presentations, and conference calls. Ignite also produced a series of videos and presentations explaining the basic structure of the program, and IAs were encouraged to show these presentations to prospective IAs to inform them about the program.
In addition to its own promotional activities, Ignite drew attention from a number of outside media sources. The Plaintiffs allege that, as early as 2005, the Dallas Morning News published a story on Ignite that included a quote from a marketing professor suggesting that Ignite was a pyramid scheme. In years following, the Dallas Morning News, the Atlanta Jour-nalr-Constitwtion, and other media outlets began to feature stories indicating that Ignite may be a pyramid scheme. Indeed, IAs reported to Ignite executives and the Presidential Directors that many prospective IAs asked them to address rumors that Ignite was an illegal pyramid scheme.
Although the parties appear to dispute the numbers, the clear majority of IAs have lost money as a result of participating in Ignite. In contrast, a small number of individuals have made significant sums of money.
This suit was brought by former IAs Juan Ramon Torres and Eugene Robison, who allege that Stream, Ignite, and various individual defendants have violated RICO. They have sought to certify a class consisting of those IAs who have lost money as a result of participating in Ignite’s program. The district court granted the Plaintiffs’ motion and certified a class.2 In its certification order, the district court considered whether the Plaintiffs could establish the proximate cause element of their RICO claim through common evidence of reliance. The district court concluded that the. Plaintiffs could not establish classwide reliance on any particular misrepresentation; but it certified the class because it ruled that the Plaintiffs were entitled to an inference of reliance, which a jury could draw from the fraudulent and illegal nature of a pyramid scheme. Thus, the district court held that, if the Plaintiffs can prove that Ignite is a pyramid scheme, which the parties concede requires only common proof, then the jury is entitled to infer that the Plaintiffs only invested in the pyramid scheme in *149reliance on an implicit representation that Ignite is a legitimate business. This interlocutory appeal followed.
Thus, to summarize, the Plaintiffs seek to certify a class action for victims of an alleged pyramid scheme. The underlying cause of action is brought under RICO. The Plaintiffs allege that they were defrauded because the Defendants misrepresented to them that Ignite was a legitimate company when it was not. Ordinarily, the Plaintiffs must show that the class relied on this misrepresentation in making their investment. The Plaintiffs have not offered evidence that such an actual representation was ever made or that they relied on such a misrepresentation. They argue, however, that such a general representation, and reliance thereon, can be inferred, essentially because such a representation is inherent in all investment opportunities and it is only on such a reliance that a rational investor would invest in Ignite.
To establish a class action, the Plaintiffs must show that the evidence of reliance is common to the class and predominates over individualized issues of reliance under Rule 23(b)(3). In this connection, the Plaintiffs must show (if inferred reliance is indeed a viable theory) that there is no other reasonable scenario that could explain the investors’ decisions to invest, other than the inferred misrepresentation that Ignite offered a legitimate business opportunity. Here, we hold that the Plaintiffs have not met this standard.
This appeal involves several complex and overlapping issues. First, we will discuss the standard for class certification under Rule 23(b)(3), along with the substantive elements of the Plaintiffs’ RICO claim, which bears on the class certification issue. We will also explain the district court’s basis for class certification, which the Plaintiffs adopt on this appeal. Then, we will describe the typical aspects of a pyramid scheme, along with the specific representations, which suggest that Ignite might be a pyramid scheme. Finally, we will consider the relevant legal authorities and explain why the Plaintiffs’ case falls short under these precedents. For these reasons, we will conclude that, the Plaintiffs’ class must be decertified.
II.
A.
The Defendants’ appeal seeks an interlocutory review of the district court’s ruling on class certification. Thus, we begin with a discussion of the standards applicable to our review of class certification orders.
District courts exercise substantial discretion when deciding whether to certify a class, and we will reverse only if the district court abused its discretion or applied an erroneous legal standard. Mullen v. Treasure Chest Casino, LLC, 186 F.3d 620, 624 (5th Cir.1999). At the same time, we are mindful that “[t]he class action is ‘an exception to the usual rule that litigation is conducted by and on behalf of the individual named parties only.’ ” Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 131 S.Ct. 2541, 2550, 180 L.Ed.2d 374 (2011) (quoting Califano v. Yamasaki, 442 U.S. 682, 700-01, 99 S.Ct. 2545, 61 L.Ed.2d 176 (1979)). Consequently, a plaintiff seeking to certify a class “must affirmatively demonstrate his compliance” with Rule 23 of the Federal Rules of Civil Procedure. Id. at 2551. The Plaintiffs have the initial burden of demonstrating that the litigation should proceed on a class-wide basis. See Howard v. City of Greenwood, 783 F.2d 1311, 1313 n. 2 (5th Cir.1986) (concluding that “the plaintiffs failed to sustain their burden of proving” the necessary common*150ality to support class certification under Rule 23(b)(3)).
On appeal, the Plaintiffs have focused their argument to contend that class certification was appropriate specifically under Rule 23(b)(3).3 “A class may be certified under Rule 23(b)(3) only if it meets the four prerequisites found in Rule 23(a) and the two additional requirements found in Rule 23(b)(3).”4 Mullen, 186 F.3d at 623. The parties do not presently dispute that the Plaintiffs meet the requirements of Rule 23(a). Instead, the arguments address whether the Plaintiffs have satisfied Rule 23(b)(3), which permits class certification if “the court finds that the questions of law or fact common to class members predominate over any questions affecting only individual members, and that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.” Fed.R.Civ.P. 23(b)(3). Although Rule 23(b)(3) requires both “predominance” of common questions of law and fact and “superiority” of a class action as a remedy, the Defendants here focus only on the predominance requirement.
“The Rule 23(b)(3) predominance inquiry tests whether proposed classes are sufficiently cohesive to warrant adjudication by representation.” Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 623, 117 S.Ct. 2231, 138 L.Ed.2d 689 (1997). In short, “[w]here the plaintiff seeks to certify a class under Rule 23(b)(3), the Rules demand ‘a close look at the ease before it is accepted as a class action.’ ” Madison v. Chalmette Ref., L.L.C., 637 F.3d 551, 554 (5th Cir.2011) (quoting Amchem, 521 U.S. at 615, 117 S.Ct. 2231).
B.
We must consider the predominance issue under Rule 23(b)(3) in the light of the elements of the Plaintiffs’ cause of action. See Castano v. Am. Tobacco Co., 84 F.3d 734, 744 (5th Cir.1996) (recognizing that “a court must understand the claims, defenses, relevant facts, and applicable substantive law in order to make a meaningful determination of the certification issues”). The Plaintiffs’ claims here are RICO claims; thus, we turn to discuss the elements of a civil RICO claim.
RICO provides, inter alia:
*151It shall be unlawful for any person employed by or associated with any enterprise engaged in, or the activities of which affect, interstate or foreign commerce, to conduct or participate, directly or indirectly, in the conduct of such enterprise’s affairs through a pattern of racketeering activity or collection of unlawful debt.
18 U.S.C. § 1962(c). Additionally, RICO prohibits conspiracies to violate § 1962(c). Id. § 1962(d). A plaintiff may bring a civil action for RICO violations under § 1962 if he or she is “injured in his business or property by reason of a violation of section 1962 of this chapter.” Id. § 1964(c) (emphasis added).
This appeal thus implicates § 1964(c), which we have held requires “a showing that the fraud was the ‘but for’ cause and ‘proximate’ cause of the injury.” Sandwich Chef of Tex., Inc. v. Reliance Nat’l Indent. Ins. Co., 319 F.3d 205, 218 (5th Cir.2003). The Plaintiffs alleged a pattern of racketeering activity consisting of acts of mail and wire fraud. See 18 U.S.C. § 1341 (mail fraud); 18 U.S.C. § 1343 (wire fraud). We have traditionally required a plaintiff presenting a civil RICO claim based on predicate acts of mail and wire fraud to establish proximate cause by showing that he or she relied on a defendant’s fraudulent misrepresentations. See In re Mastercard Int’l Inc., 313 F.3d 257, 263 (5th Cir.2002) (“[A]lthough reliance is not an element of statutory mail or wire fraud, we have required its showing when mail or wire fraud is alleged as a RICO predicate.”). The Supreme Court has since held, however, “that a plaintiff asserting a RICO claim predicated on mail fraud, need not show, either as an element of its claim or as a prerequisite to establishing proximate causation, that it relied on the defendant’s alleged misrepresentations.” Bridge v. Phoenix Bond & Indent. Co., 553 U.S. 639, 661, 128 S.Ct. 2131, 170 L.Ed.2d 1012 (2008). Although Bridge dispenses with first party reliance, “none of this is to say that a RICO plaintiff who alleges injury ‘by reason of a pattern of mail fraud can prevail without showing that someone relied on the defendant’s misrepresentations.” Id. at 658, 128 S.Ct. 2131. The extent to which Bridge alters the reliance requirement in RICO class actions is not at issue on appeal, however, as the Plaintiffs concede that proximate cause in their case depends on reliance. The Plaintiffs argue instead that they have set forth an adequate common theory of reliance.
C.
The Plaintiffs must establish that they can prove reliance through common evidence, as we have said that a class action cannot be certified if proof of reliance will depend on individualized evidence:
[A] district court [considering a motion for class certification] must perform sufficient analysis to determine that class members’ fraud claims are not predicated on proving individual reliance. If the circumstances surrounding each plaintiffs alleged reliance on fraudulent representations differ, then reliance is an issue that will have to be proven by each plaintiff, and the proposed class fails Rule 23(b)(3)’s predominance requirement.
Unger v. Amedisys Inc., 401 F.3d 316, 321 (5th Cir.2005). The Defendants argue that the Plaintiffs’ theory of reliance is necessarily an individualized inquiry.
Relying on the extensive record, the Defendants point out that the Plaintiffs were subject to abounding representations about Ignite, including: (1) positive and negative treatment in the popular press; (2) Ignite’s standard forms and marketing materials, which new IAs received; and (3) *152varying presentations from Presidential Directors who attempted to recruit new IAs to join Ignite in presentations throughout the country. Because the record establishes that each Plaintiff was subject to different representations about Ignite, the Defendants argue that each Plaintiff must establish causation by: identifying a particular misrepresentation that he or she received; and then showing that the misrepresentation caused the Plaintiff to invest in Ignite, thereby causing his or her loss. Similarly, the Defendants argue that even if the Plaintiffs can make this showing, they are also entitled to rebut this evidence with other evidence in the record, which might suggest that the Plaintiffs knew that Ignite was an illegal pyramid scheme. See Sandwich Chef, 319 F.3d at 218-19 (recognizing that knowledge, which is actually a defense to causation, is a relevant consideration when addressing class certification). In sum, the Defendants contend that the nature of the proof in this case on the issue of proximate cause will necessarily be individualized, meaning that common issues of law and fact will not predominate over this significant individualized issue.
The district court recognized that the Plaintiffs could not show through common proof that they received an actual common misrepresentation about Ignite. Instead, the district court acknowledged that the Plaintiffs would have to show the receipt of a misrepresentation through individualized proof and that it was certainly possible that some class members may have known from Ignite’s marketing pitches that it was a pyramid scheme. Nonetheless, the district court certified the class on a second ground, that is, it concluded that a jury could “infer” reliance if the Plaintiffs could establish that Ignite was a pyramid scheme. The district court explained its decision as follows:
Although the litany of reasons that any individual class member signed up to become an IA may vary, common sense compels the conclusion that every IA believed they were joining a lawful venture. That the defendants’ business opportunity is allegedly an unlawful pyramid scheme in which the vast majority of participants are sure to lose money, gives rise to an inference that the only reason the class members paid the $329 sign-up fee (and possibly other fees) is because the true nature of the ‘opportunity’ was disguised as something it was not. As such, establishing proximate cause would not be an individualized inquiry; rather, it could be determined as to all the class members at once. Because it can rationally be assumed (at least without any contravening evidence) that the legality of the Ignite program was a bedrock assumption of every class member, a showing that the program was actually a facially illegal pyramid scheme would provide the necessary proximate cause.
On appeal, the Plaintiffs defend class certification on this basis, arguing that they can establish proximate cause merely by establishing that Ignite was a pyramid scheme.
The Plaintiffs’ theory relies not on a particular misrepresentation, but instead on a “common sense” inference of reliance, which exists from the nature of pyramid schemes. According to the Plaintiffs, a pyramid scheme is a unique species of fraud because pyramid schemes are both illegal and require participants to profit in the scheme by victimizing others, which, in the context of Ignite, were most often friends and family. Thus, the Plaintiffs argue that the fact-finder is entitled to infer that the Plaintiffs relied on a misrepresentation regarding Ignite’s legitimacy if the Plaintiffs can prove that Ignite is a pyramid scheme, which the parties agree can be done through common proof. *153Thus, the common proof that the Plaintiffs offer in this case is evidence that Ignite is actually a pyramid scheme; and this evidence, they claim, is sufficient to establish causation as well.
In response to this argument, the Defendants argue that the individualized representations are still relevant. Even if Ignite was a pyramid scheme, they say, it provided investors at the top of the scheme with an opportunity to profit. Some individuals who lost money might still have invested in the hope that they would be near the top of the pyramid. In this connection, pyramid schemes are little different from other species of fraud — some knowing participants in the fraud will profit, whereas many others will lose money. Thus, the Defendants urge us to decertify the class so that the Defendants can rebut the Plaintiffs’ common theory of reliance through individualized trials.
For the reasons that will follow, we conclude that the Plaintiffs’ claimed common theory of reliance does not hold together.
III.
First, the Plaintiffs’ theory of reliance depends on the premise that a pyramid scheme is a unique type of fraud. We thus begin with a brief discussion of pyramid schemes and turn to the actual representations about the Ignite business, which are part of the record in this case.
A.
The Plaintiffs rely on Webster v. Omntrition International, Inc., 79 F.3d 776 (9th Cir.1996), to define the basic characteristics of an illegal pyramid scheme. We now turn to the description of pyramid schemes in that case.
Initially, we should be clear that a “pyramid scheme” can be distinguished from the many types of businesses organized in a “pyramid-shaped” hierarchical structure. A true pyramid scheme, as that term is used here, refers to a type of illegal and fraudulent activity, structured in a fashion that it “must eventually collapse.” Id. at 781. Pyramid schemes, unlike pyramid-structured organizations, will collapse because such schemes are designed to produce income from the continuous recruitment of new members into a constantly narrowing sales market and not upon sales revenue from a legitimate product to consumers in a normal market. . Id. at 781-82. In short, the scheme collapses when those recruited to sell dwarf the market of those available to buy.
There are typically two elements to such a pyramid scheme: (1) payment to an entity in return for the right to sell its product; and (2) the right, in exchange for the payment, to receive rewards from the entity that are based almost exclusively on the recruitment of new program participants. Id. at 781. Under this standard, some businesses that engage in retail sales may still be a pyramid scheme if “[t]he promise of lucrative rewards for recruiting others tends to induce participants to focus on the recruitment side of the business at the expense of their retail marketing efforts, making it unlikely that meaningful opportunities for retail sales will occur.” Id. Thus, the primary factor in deciding whether a business is a pyramid scheme is whether the business focuses exclusively or almost exclusively on recruiting as opposed to sales.
Pyramid schemes, however, are not losing propositions for all investors. Instead, “pyramid schemes may make money for those at the top of the ... pyramid, but ‘must end up disappointing those at the bottom who can find no recruits.’ ” Id. at 781 (quoting In re Koscot Interplanetary, Inc., 86 F.T.C. 1106, 1181 (1976)). Thus, an individual who participates in a pyramid *154scheme necessarily takes a gamble that she will be reasonably near the top of the pyramid. Although an individual may lose money if it turns out that she invested at the wrong time, this misjudgment does not, a fortiori, mean that the individual is irrational. Such an investor may have rationally assumed both that the business was a pyramid scheme and that the investment was worth the gamble of being near the top of the pyramid. So, with this background, we turn to examine some of the representations regarding Ignite in this case.
B.
The record suggests that Ignite often promoted its multi-level marketing program as just this sort of gamble to prospective IAs. Ignite’s Presidential Directors, who traveled the country promoting Ignite to IAs and prospective IAs, implied that Ignite was a pyramid scheme. In presentations to IAs and prospective IAs, these officers repeatedly underscored that the way to make money was by recruiting other IAs, not recruiting customers. The record shows, for example, that Greg McCord admonished IAs in one presentation that “if you keep concentrating on customers, you won’t make money.” Although these Presidential Directors did not use the term “pyramid scheme” to describe Ignite, a reasonable prospective IA could reasonably construe these representations as the hallmarks of a pyramid scheme: Ignite predominately pushes recruiting over selling, and thus expanding the number of IA participants, over customer acquisition.5 See Webster, 79 F.3d at 782.
Some representations were even more direct. Presidential Director Randy Hedge repeatedly referred to the multilevel marketing business as a “pyramid.” To illustrate, he told his audience on one occasion: “I don’t care if you call [Ignite] an octagon, parallelogram, rectangle— they’re sending me a check.” In another presentation, he shared an anecdote about recruiting an individual into Ignite after calling it a “pyramid deal” because the prospective IA was only really interested in whether the deal was “makin’ any money.” Similarly, various media outlets began to investigate whether Ignite was a pyramid scheme. The Plaintiffs suggested in their complaint that the Dallas Morning News published a story in 2005, which contained an indication that Ignite could be an illegal pyramid scheme. Other media outlets produced similar critical reports about Ignite in 2010 and 2011.
These promotions, although supportive of the Plaintiffs’ contention that Ignite is a pyramid scheme, also buttress the Defendants’ position in opposition to class certification, i.e., these comments suggest that the Plaintiffs will have to prove RICO causation by relying on individualized, and not common, proof of reliance. The Plaintiffs argue that these individualized representations about Ignite drop from the case, however, based on the strength of their proposed inference.' Specifically, the Plaintiffs claim that a jury should infer that the Plaintiffs did not rely on these representations because a rational investor would not participate in a pyramid scheme.
IV.
Turning to the Plaintiffs’ argument that reliance may be inferred, we hold that reliance cannot be inferred merely because *155a business is alleged to be a pyramid scheme, particularly when the record in this ease suggests that investors were told that it was a pyramid scheme. Such an inference is unsupported by our precedents or by the precedents in other circuits.
A.
1.
We begin with a discussion of our relevant precedents. Generally, proximate cause of the alleged injury (here, misrepresentations caused monetary loss) is a distinct element of a RICO claim, which must be established separately from proving an underlying fraud. Thus, even if the Plaintiffs can establish through common evidence that the Defendants engaged in fraudulent or illegal conduct, common issues of law and fact do not predominate over individualized issues unless the Plaintiffs can establish through common evidence that the fraudulent conduct caused, their injury. See Patterson v. Mobil Oil Corp., 241 F.3d 417, 419 (5th Cir.2001) (“While there may be an issue of fact common to all class members — the question of whether or not Mobil was a valid subscriber to the workers’ compensation system — that question does not predominate over the question of whether or not each member of the class suffered a RICO injury.”). In Patterson, we concluded that a plaintiff could establish proximate cause, and thus prove a RICO injury, by showing “that she could have and would have sued Mobil, but did not do so because the asserted false statements led her to believe her suit to be barred by the workers’ compensation regime.” Id. Obviously, such a showing of proximate cause would depend on the individual circumstances and motivations of each plaintiff; and these types of individualized inquiries “defeat the economies ordinarily associated with the class action device.” Id.
This point is illustrated by a case that bears a striking resemblance to this case. Sandwich Chef, 319 F.3d at 224. In Sandwich Chef, the district court certified a class action against a group of insurance companies, on the basis that they had charged excessive premiums by sending inflated invoices to policyholders and misrepresented the correctness of the premium charged. Id. at 211. Evidence in the record also suggested that the charged rates were illegal. Id. at 212. Nonetheless, we decertified the class. Id. at 224. We reasoned that the plaintiffs in Sandwich Chef could not prove proximate cause through common proof, because individualized issues of knowledge and reliance would overwhelm any common proof. Id. at 220-21. Specifically, we pointed out that the plaintiffs and the defendants negotiated the insurance policies in individualized transactions; and evidence in the record suggested that the plaintiffs could have voluntarily assented to the illegal rate structures so that they could receive other benefits in return. See id. at 212-13, 220-21. Because the proof suggested that at least some of the plaintiffs could have knowingly participated in the fraud, we held that the defendants were entitled to undercut the plaintiffs’ evidence of reliance “with evidence that might persuade the trier of fact that policyholders knew the amounts being charged varied from rates filed with regulators and that they had agreed to pay such premiums.” Id. at 220.
In sum, our precedents do not support an inference of reliance from fraudulent conduct, even when the fraudulent conduct at issue is illegal. Instead, we have recognized that, in most cases, reliance will naturally turn on evidence that will differ from case to case. Individual plaintiffs will receive different pitches to join a busi*156ness, and they will have differing expectations in terms of what they expect to receive from the business. Generally, the defendants are entitled to probe these differences at trial by presenting evidence that the plaintiffs knew of the fraud, yet nonetheless participated in it because they believed that it would benefit them.
2.
The Plaintiffs argue, however, that precedents in other circuits allow for an inference of reliance in certain RICO fraud cases; and they further contend that such an inference is warranted on the facts of this case. The Plaintiffs primarily rely, on appeal, on three decisions from other circuits, to which we now turn.
First, they point to Klay v. Humana, in which the Eleventh Circuit approved the certification of a class of physicians who alleged that a group of health maintenance organizations (“HMDs”) defrauded the physicians out of adequate reimbursement for their services rendered by programming their computer systems to pay the physicians less than they were entitled. 382 F.3d 1241, 1260 (11th Cir.2004). The Klay court concluded that the plaintiffs’ claims could be certified as a class action because a jury could infer reliance, stating:
It does not strain credulity to conclude that each plaintiff, in entering into contracts with the defendants, relied upon the defendants’ representations and assumed they would be paid the amounts they were due. A jury could quite reasonably infer that guarantees concerning physician pay — the very consideration upon which these agreements are based — go to the heart of these agreements, and that doctors based their assent upon them.... Consequently, while each plaintiff must prove reliance, he or she may do so through common evidence (that is, through legitimate inferences based on the nature of the alleged misrepresentations at issue).
Id. at 1259. The Second Circuit reached a similar conclusion in a case involving fraudulent overbilling. In re U.S. Foodservice Inc. Pricing Litig., 729 F.3d 108 (2d Cir.2013). Relying on Klay, the Food-service court allowed the case to proceed as a class action, in part because “payment may constitute circumstantial proof of reliance based on the reasonable inference that customers who pay the amount specified in an inflated invoice would riot have done so absent reliance upon the invoice’s implicit representation that the invoiced amount was honestly owed.” Id. at 120.
Finally, the Tenth Circuit in CGC Holding Co. v. Broad & Cassel confronted a certified class of prospective borrowers who paid a non-refundable “loan commitment fee” for a loan that the lender never intended to issue. 773 F.3d 1076, 1082 (10th Cir.2014). There, the court concluded that the class could proceed because a fact-finder could infer that the plaintiffs paid the fee in reliance on a misrepresentation that the transaction was legitimate. See id. at 1091-92. Specifically, the court reasoned that such an inference was appropriate in significant part because the victims of the fraud “were completely deprived of any benefit from their transaction.” Id. at 1093.
In sum, these cases allow for class certification based on an inference of reliance when all individualized issues truly drop out of the case. In each of these cases, a class of plaintiffs paid a sum of money or declined full payment for services rendered without receiving anything of value in return. Additionally, there was no evidence in the cases to suggest any other rational explanation for the plaintiffs’ behavior other than that they were duped by the defendants. Thus, those courts allowed class certification because the only *157reasonable explanation for the plaintiffs’ behavior was that they relied on a misrepresentation.
B.
Turning to the record in this case, we conclude that the Plaintiffs’ evidence does not support a sufficient inference of reliance. Individuals may knowingly choose to invest in a pyramid scheme such as Ignite for any number of reasons, most notably because Ignite provides an opportunity to make money. Thus, the class cannot be certified under our precedents or the precedents cited by the Plaintiffs because individualized issues of reliance and knowledge will be relevant to each Plaintiffs case.
1.
First, the mere fact that this case involves a pyramid scheme does not take this case outside our well-settled precedents regarding predominance in both Patterson and Sandwich Chef. Although the Plaintiffs may be able to establish common proof of a fraud, the common evidence that a fraud existed is not common evidence that the Plaintiffs were injured by the fraud.
,This case is less compelling for class certification than Patterson. In Patterson, there was a common misrepresentation, i.e., the defendant allegedly misrepresented to them that it had workers’ compensation insurance. Here, it is not clear that all Plaintiffs were told that Ignite was a lawful business, given the differing pitches to differing prospective IAs. It appears that some prospective IAs received only versions of the pitch that Ignite provided an opportunity for them to make significant sums of money. Additionally, even if the Plaintiffs here could establish an actual common misrepresentation that Ignite was a legitimate business, they would still have to show that they were injured by the misrepresentation. In Patterson, we recognized that the plaintiffs had to show that they would have sued Mobil had they known that the misrepresentation about its insurance was false. 241 F.3d at 419. Just as there are many reasons why a party would choose to file or not file a lawsuit, there are many reasons why someone would choose to join or not join a pyramid scheme. The evidence here suggests that investing in Ignite was quite similar to gambling — individuals could have become IAs as “a form of escape, a casual endeavor, a hobby, a risk-taking money venture, or scores of other things.” Poulos v. Caesars World, Inc., 379 F.3d 654, 668 (9th Cir.2004).
Nor is a pyramid scheme unique because it is illegal.' In Sandwich Chef, the plaintiffs accused the defendants of lying to state regulators and charging illegal rates. 319 F.3d at 212. Nonetheless, we also pointed out that the evidence suggested that the plaintiffs could .very well want their insurance policies to deviate from filed rates because such deviations could actually benefit the plaintiffs in other respects. Id. at 213. Thus, we decertified the class because the defendants were entitled to show through individualized evidence that the plaintiffs “knew the amounts being charged varied from rates filed with regulators and that they had agreed to pay such premiums.” Id. at 220.
A pyramid scheme is no different from the insurance regime in Sandwich Chef. By joining Ignite, an IA had the opportunity to make money, perhaps even significant sums of money, by building a large pyramid beneath them. Although the Plaintiffs suggest that they would not join a pyramid scheme like Ignite because such a 'scheme would depend in large part on defrauding friends and family members, this supposed distinction is unavailing. *158First, an individual could rationally believe that he could make money for friends and family members if they were all investing at the top of the pyramid. Indeed, the record reflects that the Presidential Directors regularly told prospective IAs that they had enriched their spouses, children, and friends by bringing them into the Ignite program. Second, the same arguments could be made about gambling, i.e., that spending money on gambling harms an individual’s family. But gambling is just the type of activity where no such broad assumptions can be made about the reasons for human behavior. See Poulos, 379 F.3d at 668. And finally, the Plaintiffs have cited no case law that has adopted such an elevated view of human nature.
Thus, the Plaintiffs will have to rely upon individualized proof, and not a generalized inference, to establish proximate cause in each particular RICO case.6
2.
In that connection, the Plaintiffs’ cases also fail to support class certification on the basis of an inference of reliance. Klay, Foodservice, and CGC all involved fraudulent schemes in which the plaintiff victims had no hope of recovering their investments. The courts could not point to any evidence that might provide an alternative explanation for the plaintiffs’ conduct other than that they relied on a misrepresentation that they might profit.
By contrast, an investor could reasonably choose to knowingly invest in a pyramid scheme in the hope that they would make money. As we have already explained, a pyramid scheme provides an opportunity for those at the top of the pyramid to profit from their investments. Webster, 79 F.3d at 781. While many of the Plaintiffs might have decided to invest in the scheme in the belief that it was legal, it is equally possible that many of the Plaintiffs chose to invest in the scheme in the belief that, legal or illegal, it provided them with an opportunity to make money-
Additionally, the representations at issue in this litigation are far more varied than the misrepresentations in Klay, Foodservice, and CGC. In each of those cases, the many individual representations essentially said the same thing — the invoices and bills provided either an amount due or an amount paid, representing that the stated amount was correct. By contrast, the representations here vary in their contents. Some of Ignite’s marketing materials touted its legitimacy, whereas other presentations undermined that legitimacy. To recover on their RICO claims, the Plaintiffs must show that they relied upon the former materials, and not the latter; they may only do so through individualized proof. Thus, the class must be decertified.
V.
In sum, the district court erred in certifying the class because common questions of law and fact will not predominate over individualized inquiries into causation and knowledge. The case is therefore REMANDED for the entry of an order, VACATING the order of certification and for such further proceedings as may be appro*159priate and not inconsistent with this opinion.
VACATED and REMANDED.
. Many of the individual Defendants in this appeal came to Ignite after working at Excel Telecommunications, a failed long-distance company that offered long-distance services in the deregulated telecommunications market through a similar multi-level marketing program.
. The district court defined the class more broadly than the Plaintiffs’ proposed definition, .extending the class to "all IAs who joined Ignite on or after January 1, 2005, through April 2, 2011, excluding the IAs subject to the Eleventh Circuit opinion in Betts [v. SGE Management, LLC, 402 Fed.Appx. 475 (11th Cir.2010) ].” Thus, the district court did not explicitly limit the class to consist only of those IAs who lost money by participating in Ignite.
. Rule 23(a) provides as follows:
One or more members of a class may sue or be sued as representative parties on behalf of all members only if:
(1) the class is so numerous that joinder of all members is impracticable;
(2) there are questions of law or fact common to the class;
(3) the claims or defenses of the representative parties are typical of the claims or defenses of the class; and
(4) the representative parties will fairly and adequately protect the interests of the class.
Fed.R.Civ.P. 23(a).
Then, Rule 23(b)(3) provides that the district court may certify the putative class if: the court finds that the questions of law or fact common to class members predominate over any questions affecting only individual members, and that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy. The matters pertinent to these findings include:
(A) the class members' interests in individually controlling the prosecution or defense of separate actions;
(B) the extent and nature of any litigation concerning the controversy already begun by or against class members;
(C) the desirability or undesirability of concentrating the litigation of the claims in the particular forum; and
(D) the likely difficulties in managing a class action.
Fed.R.Civ.P. 23(b)(3).
. Although many of these pitches targeted IAs, the Presidential Directors apparently often gave these presentations at widely-attended, "revival style” events attended by IAs and prospective IAs alike.
. We note as well that, even if the Plaintiffs could establish reliance through an inference, the Defendants would still be entitled to offer the evidence in the record regarding the misrepresentations about Ignite to probe each Plaintiff's knowledge in individualized trials. See Sandwich Chef, 319 F.3d at 220. Knowledge is a defense to a RICO fraud claim, and the Defendants would be entitled to present this evidence on an individualized basis, as pertains to each Plaintiff. See id. at 220-21.