This private antitrust action challenges the legality of the National Football League (NFL) player selection system, commonly called the “draft.” The plaintiff is James McCoy (Yazoo) Smith, a former professional football player who played one season for the Washington Redskins after being drafted by them in 1968. The defendants are Pro-Football, Inc., which operates the Redskins, and the NFL. Smith contends that the draft as it existed in 1968 was an unreasonable restraint of trade in violation of §§ 1 and 3 of the Sherman Act,1 and that, but for the draft, he would have *1175negotiated a far more lucrative contract when he signed as a player in that year. Smith alleges that he has been injured in his business or property2 in the amount of the difference between the compensation he actually received and the compensation he would have received had there existed a “free market” for his services.
After a/trial to the court, District Judge Bryant held that the NFL draft as it existed in 1968 constituted a “group boycott” and was Ithus a per se violation of the Sherman Act.3 Alternatively, he held that the draft, ^tested under the rule of reason, was an unreasonable restraint because it was “significantly more restrictive than necessary” to accomplish whatever legitimate goals the NFL had.4 Judge Bryant awarded Smith treble damages totaling $276,000. The Redskins and the NFL have appealed the finding of antitrust liability; both sides have appealed the damage award. Relying on the rule of reason, we affirm the finding of antitrust liability and remand for recomputation of damages.
I. BACKGROUND
The NFL draft, which has been in effect since 1935, is a procedure under which negotiating rights to graduating college football players are allocated each year among the NFL clubs in inverse order of the clubs’ standing. Under the draft procedures generally followed, the team with the poorest playing-field record during the preceding season has the first opportunity, as among the NFL teams, to select a college player of its choice; the team with the next poorest record has the next choice, and so on until the team with the best record (the winner of the previous year’s “Super Bowl”) has picked last. At this point, the first “round” of the draft is completed. In 1968 there were 16 succeeding rounds in the yearly draft, the same order of selection being followed in each round. Teams had one choice per round unless they had traded their choice in that round to another team (a fairly common practice). When Smith was selected by the Redskins there were 26 teams choosing in the draft.
The NFL draft, like similar procedures in other professional sports, is designed to promote “competitive balance.” By dispersing newly arriving player talent equally among all NFL teams, with preferences to the weaker clubs, the draft aims to produce *1176teams that are as evenly-matched on the playing field as possible. Evenly-matched teajns make for closer games, tighter pennant races, and better player morale, thus maximizing fan interest, broadcast revenues, and overall health of the sport.
The draft is effectuated through the NFL’s “no-tampering” rule.5 Under this rule as it existed in 1968, no team was permitted to negotiate prior to the draft with any player eligible to be drafted, and no team could negotiate with (or sign) any player selected by another team in the draft. The net result of these restrictions was that the right to negotiate with any given player was exclusively held by one team at any given time. If a college player could not reach a satisfactory agreement with the team holding the rights to his services he could not play in the NFL.6
Plaintiff Smith became subject to the draft when he graduated as an All-American football player from the University of Oregon in 1968. The Redskins, choosing twelfth, picked Smith as their first-round draft choice. After several months of negotiations, in which he was represented by an agent, Smith and the Redskins signed a one-year contract — a version of the Standard Player Contract that the NFL requires all players to sign.7 The contract awarded Smith a $23,000 “bonus” for signing, an additional $5,000 if he made the team, and a salary of $22,000, for a total first-year compensation of $50,000.
Smith made the team and performed at a high level of play as a defensive back until he suffered a serious neck injury in the final game of the 1968 season. His doctors advised him not to continue his football career. After his injury the Redskins paid *1177Smith an additional $19,800, representing the amount he would ordinarily have received had he played out the second (“option”) year of his contract.8
Two years after his injury Smith filed suit in the District Court. After finding that the draft violated the antitrust laws, Judge Bryant awarded Smith damages equal to the difference between his actual compensation and the compensation he could have received in a free market. To compute the latter amount, Judge Bryant assumed that plaintiff in a free market would have been able to negotiate a three-year contract with an “injury protection clause,” i. e., a clause guaranteeing payment for the full term of a player’s contract even if he should be incapacitated. Judge Bryant estimated Smith’s annual “free market salary” by taking the annual salary ($54,000) of another defensive back (Pat Fischer) who signed as a “free agent” 9 with the Redskins in 1968. The resulting calculation yielded $162,000 as the contractual value of Smith’s services in a free market. From this sum Judge Bryant subtracted the $69,800 that Smith in fact received, netting actual damages in the amount of $92,200. This figure was trebled to $276,600 in accordance with the antitrust laws.10
II. ANALYSIS
The legality of the NFL player draft under the antitrust laws11 is essentially a question of first impression.12 This case requires us to consider (1) whether the legality of the draft is governed by a per se rule or by the rule of reason; (2) whether the draft, if tested by the rule of reason, is a reasonable restraint; and (3) whether, if the draft violates the antitrust laws, the measure of damages adopted by the District Judge was proper. We discuss these issues in turn.
A. Per Se Illegality
The traditional framework of analysis under § 1 of the Sherman Act is familiar and does not require extended discussion. Section 1 prohibits “[e]very contract, combination ... or conspiracy, in restraint of trade or commerce.” While this language is broad enough to render illegal nearly all commercial understandings, the *1178Supreme Court in Standard Oil13 established a judicial gloss on the statute which made the “rule of reason” the prevailing mode of analysis. Under this rule, the fact-finder weighs all the circumstances of a case in deciding whether a restrictive practice should be prohibited as imposing an unreasonable restraint on competition. The inquiry mandated by the rule of reason, however, is often laborious, and as the courts gained experience with antitrust problems they identified certain types of agreements which were so consistently unreasonable that they could be deemed illegal per se, without elaborate inquiry into their purported justifications. Among the practices that have been deemed so pernicious as to be unreasonable per se are certain “group boycotts.”14
Plaintiff argues that the NFL draft constitutes a “group boycott” because the NFL clubs concertedly refuse to deal with any player before he has been drafted or after he has been drafted by another team, and that the draft is in consequence a per se violation of § 1. The District Court accepted this argument. We reject it. We hold that the NFL player draft is not properly characterized as a “group boycott” — at least not the type of boycott that traditionally has been held illegal per se — and that the draft, regardless of how it is characterized, should more appropriately be tested under the rule of reason.
The classic “group boycott” is a concerted attempt by a group of competitors at one level to protect themselves from competition from non-group members who seek to compete at that level.15 Typically, the boycotting group combines to deprive would-be competitors of a trade relationship which they need in order to enter (or survive in) the level wherein the group operates. The group may accomplish its exclusionary purpose by inducing suppliers not to sell to potential competitors, by inducing customers not to buy from them, or, in some cases, by refusing to deal with would-be competitors themselves.16 In each instance, however, the hallmark of the “group boycott” is the effort of competitors to “barricade themselves from competition at their own level.”17 It is this purpose to exclude competition that has characterized the Supreme Court’s decisions invoking the group boycott per se rule.18
The NFL player draft differs from the classic group boycott in two significant respects. First, the NFL clubs which have *1179“combined” to implement the draft are not competitors in any economic sense. The clubs operate basically as a joint venture19 in producing an entertainment product— football games and telecasts. No NFL club can produce this product without agreements and joint action with every other team. To this end, the League not only determines franchise locations, playing schedules, and broadcast terms, but also ensures that the clubs receive equal shares of telecast and ticket revenues. These economic joint venturers “compete” on the playing field, to be sure, but here as well cooperation is essential if the entertainment product is to attain a high quality: only if the teams are “competitively balanced” will spectator interest be maintained at a high pitch. No NFL team, in short, is interested in driving another team out of business, whether in the counting-house or on the football field, for if the League fails, no one team can survive.
The draft differs from the classic group boycott, secondly, in that the NFL clubs have not combined to exclude competitors or potential competitors from their level of the market.20 Smith was never seeking to “compete” with the NFL clubs, and their refusal to deal with him has resulted in no decrease in the competition for providing football entertainment to the public. The draft, indeed, is designed not to insulate the NFL from competition, but to improve the entertainment product by enhancing its teams' competitive equality.21
In view of these differences, we conclude that the NFL player draft cannot properly be described as a group boycott — at least not the type of group boycott that traditionally has elicited invocation of a per se rule.22 The “group boycott” designation, *1180we believe, is properly restricted to concerted attempts by competitors to exclude horizontal competitors; it should not be applied, and has never been applied by the Supreme Court, to concerted refusals that are not designed to drive out competitors but to achieve some other goal.23
We are guided in reaching this conclusion by decisions in analogous areas of antitrust law. The courts have consistently refused to invoke the boycott per se rule where, given the peculiar characteristics of an industry, the need for cooperation among participants necessitated some type of concerted refusal to deal,24 or where the concerted activity manifested no purpose to exclude and in fact worked no exclusion of competitors.25 In view of the joint-venture charae*1181teristics of the professional football industry and the purpose of the concerted activity here, these decisions support our conclusion that the NFL player draft is not a group boycott which is illegal per se.
Whether the draft is a group boycott, or not, we think it is clearly not the type of restraint to which a per se rule is meant to apply. A per se rule is a judicial shortcut; it represents the considered judgment of courts, after considerable experience with a particular type of restraint, that the rule of reason — the normal mode of analysis — can be dispensed with. As the Supreme Court explained in Northern Pacific Railway Co. v. United States,26 “there are certain agreements or practices which because of their pernicious effect on competition and lack of any redeeming virtue are conclusively presumed to be unreasonable and therefore illegal without elaborate inquiry as to the precise harm they have caused or the business excuse for their use.” A court will not indulge in this conclusive presumption lightly. Invocation of a per se rule always risks sweeping reasonable, pro-competitive activity within a general condemnation, and a court will run this risk only when it can say, on the strength of unambiguous experience,27 that the challenged action is a “naked restraint[ ] of trade with no purpose except stifling of competition.”28
The Supreme Court emphasized the “demanding standards” of Northern Pacific Railway last Term in Continental T. V., Inc. v. GTE Sylvania Inc.29 Reiterating that “[p]er se rules of illegality are appropriate only when they relate to conduct that is manifestly anti-competitive,”30 the Court overruled Arnold, Schwinn & Co.31 which had held certain vertical restraints illegal per se. The Continental Court noted that the vertical restrictions in question possessed “redeeming virtues” in their stimulation of inter-brand competition; that the restrictions were “widely used in our free market economy”; and that there existed “substantial scholarly and judicial authority supporting their economic utility.”32 For these reasons, the Court held that the restraints at issue were to be analyzed not under a per se rule, but under the rule of reason.
For similar reasons we reach the same conclusion here. The NFL player draft, we think, quite clearly fails to satisfy the “demanding standards” of Northern Pacific Railway. Given that the draft’s restrictive effect is temporally limited, we would hesitate to describe its impact on the market for players’ services as “pernicious.” More importantly, we cannot say that the draft has “no purpose except stifling of competition” or that it is without “any redeeming virtue.” Some form of player selection system may serve to regulate and thereby promote competition in what would otherwise be a chaotic bidding market for the services of college players. The Redskins, moreover, presented considerable evidence at trial that the draft was designed to preserve, and that it made some contribution to preserving, playing-field equality among the NFL-teams with various attendant benefits. The draft, finally, like the vertical restraints challenged in Continental T.V., is *1182“widely used” in our economy33 and has both judicial34 and scholarly35 support for its economic usefulness.
This is not to say, of course, that the draft in any one of its incarnations may not violate the antitrust laws. It is only to say that the courts have had too little experience with this type of restraint, and know too little of the “economic and business stuff” from which it issues,36 confidently to declare it illegal without undertaking the analysis enjoined by the rule of reason.
Our conclusion that the legality of the NFL draft should not be governed by a per se rule parallels the conclusion of most courts 37 and commentators 38 that the legality of player restrictions in professional sports should be governed by the rule of reason. In the case most nearly on point, the Eighth Circuit recently declined, for reasons akin to ours, to apply a per se approach to the NFL’s Rozelle rule.39 While we fully appreciate the administrative convenience of a per se rubric, ease of application alone cannot suffice to recommend it.40 In antitrust law, as elsewhere, we must heed Justice Cardozo’s warning to beware “the tyranny of tags and tickets.”41 When anticompetitive effects are shown to result from a particular player selection system “they can be adequately policed under the rule of reason.” 42
*1183B. Rule of Reason
Under the rule of reason, a restraint must be evaluated to determine whether it is significantly anticompetitive in purpose or effect. In making this evaluation, a court generally will be required to analyze “the facts peculiar to the business, the history of the restraint, and the reasons why it was imposed.”43 If, on analysis, the restraint is found to have legitimate business purposes whose realization serves to promote competition, the “anticompetitive evils” of the challenged practice must be carefully balanced against its “procompetitive virtues” to ascertain whether the former outweigh the latter.44 A restraint is unreasonable if it has the “net effect” of substantially impeding competition.45
After undertaking the analysis mandated by the rule of reason, the District Court concluded that the NFL draft as it existed in 1968 had a severely anticompetitive impact on the market for players’ services, and that it went beyond the level of restraint reasonably necessary to accomplish whatever legitimate business purposes might be asserted for it. We have no basis for disturbing the District Court’s findings of fact;46 and while our legal analysis dif*1184fers slightly from that of the trial judge, having benefited from intervening guidance from the Supreme Court, we agree with the District Court’s conclusion that the NFL *1185draft as it existed in 1968 constituted an unreasonable restraint of trade.
The draft that has been challenged here is undeniably anticompetitive both in its purpose and in its effect. The defendants have conceded that the draft “restricts competition among the NFL clubs for the services of graduating college players” and, indeed, that the draft “is designed to limit competition” and “to be a ‘purposive’ restraint” on the player-service market.47 The fact that the draft assertedly was designed to promote the teams’ playing-field equality rather than to inflate their profit margins may prevent the draft’s purpose from being described, in subjective terms, as nefarious. But this fact does not prevent its purpose from being described, in objective terms, as anticompetitive, for suppressing competition, is the telos, the very essence of the restraint.
The trial judge was likewise correct in finding that the draft was significantly anticompetitive in its effect. The draft inescapably forces each seller of football services to deal with one, and only one buyer, robbing the seller, as in any monopsonistic market, of any real bargaining power.48 The draft, as the District Court found, “leaves no room whatever for competition among the teams for the services of college players, and utterly strips them of any measure of control over the marketing of their talents.”49 The predictable effect *1186of the draft, as the evidence established and as the District Court found, was to lower the salary levels of the best college players. There can be no doubt that the effect of the draft as it existed in 1968 was to “suppress or even destroy competition” 50 in the market for players’ services.
The justification asserted for the draft is that it has the legitimate business purpose of promoting “competitive balance” and playing-field equality among the teams, producing better entertainment for the public, higher salaries for the players, and increased financial security for the clubs. The NFL has endeavored to summarize this justification by saying that the draft ultimately has a “procompetitive” effect, yet this shorthand entails no small risk of confusion. The draft is “procompetitive,” if at all, in a very different sense from that in which it is anticompetitive. The draft is anticompetitive in its effect on the market for players’ services, because it virtually eliminates economic competition among buyers for the services of sellers. The draft is allegedly “procompetitive” in its effect on the playing field; but the NFL teams are not economic competitors on the playing field, and the draft, while it may heighten athletic competition and thus improve the entertainment product offered to the public, does not increase competition in the economic sense of encouraging others to enter the market and to offer the product at lower cost. Because the draft’s “anti-competitive” and “procompetitive” effects are not comparable, it is impossible to “net them out” in the usual rule-of-reason balancing. The draft’s “anticompetitive evils,” in other words, cannot be balanced against its “procompetitive virtues,” and the draft be upheld if the latter outweigh the former. In strict economic terms, the draft’s demonstrated procompetitive effects are nil.
The defendants’ justification for the draft reduces in fine to an assertion that competition in the market for entering players’ services would not serve the best interests of the public, the clubs, or the players themselves. This is precisely the type of argument that the Supreme Court only recently has declared to be unavailing. In National Society of Professional Engineers v. United States,51 the Court held that a professional society’s ban on competitive bidding violated § 1 of the Sherman Act. In so holding the Court rejected a defense that unbridled competitive bidding would lead to deceptively low bids and inferior work “with consequent risk to public safety and health,”52 terming this justification “nothing less than a frontal assault on the basic policy of the Sherman Act.”53 Ending decades of uncertainty as to the proper scope of inquiry under the rule of reason, the Court stated categorically that the rule, contrary to its name, “does not open the field of antitrust inquiry to any argument in favor of a challenged restraint that may fall within the realm of reason,” 54 and that the inquiry instead must be “confined to a consideration of [the restraint’s] impact on competitive conditions.”55 The purpose of antitrust analysis, the Court concluded, “is to form a judgment about the competitive significance of the restraint; it is not to decide whether a policy favoring competition is in the public interest, or in the interest of the members of an industry. Subject to exceptions defined by statute, *1187that policy decision has been made by Congress.” 56
Confining our inquiry, as we must, to the draft’s impact on competitive conditions, we conclude that the draft as it existed in 1968 was an unreasonable restraint of trade. The draft was concededly anticompetitive in purpose. It was severely anticompetitive in effect. It was not shown to have any significant offsetting procompetitive impact in the economic sense. Balancing the draft’s anticompetitive evils against its pro-competitive virtues, the outcome is plain. The NFL’s defenses, premised on the assertion that competition for players’ services would harm both the football industry and society, are unavailing; there is nothing of procompetitive virtue to balance, because “the Rule of Reason does not support a defense based on the assumption that competition itself is unreasonable.” 57
We recognize, on analogy with the Supreme Court’s reasoning in Goldfarb58 and Professional Engineers, that professional football “may differ significantly from other business services, and, accordingly [that], the nature of the competition”59 for player talent may vary from an absolute “free market” norm. Given the joint-venture status of the NFL clubs, we do not foreclose the possibility that some type of player selection system might be defended as serving “to regulate and promote competition” in the market for players’ services.60 But we are faced here, as 'the Supreme Court was faced in Professional Engineers, with what amounts to a “total ban” on competition,61 and we agree with the District Court that this level of restraint cannot be justified. The trial judge concluded, with pardonable exaggeration, that the draft system at issue was “absolutely the most restrictive one imaginable.”62 Even though the draft was justified primarily by the need to disperse the best players, it applied to all graduating seniors, including average players who were, in a sense, fungible commodities. It permitted college players to negotiate with only one team. If a player could not contract with that team, he could not play at all.
Without intimating any view as to the legality of the following procedures, we note that there exist significantly less anti-competitive alternatives to the draft system which has been challenged here. The trial judge found that the evidence supported *1188the viability of a player selection system that would permit “more than one team to draft each player, while restricting the number of players any one team might sign.”63 A less anticompetitive draft might permit a college player to negotiate with the team of his choice if the team that drafted him failed to make him an acceptable offer.64 The NFL could also conduct a second draft each year for players who were unable to reach agreement with the team that selected them the first time.65 Most obviously, perhaps, the District Court found that the evidence supported the feasibility of a draft that would run for fewer rounds, applying only to the most talented players and enabling their “average” brethren to negotiate in a “free market”66 The least restrictive alternative of all, of course, would be for the NFL to eliminate the draft entirely and employ revenue-sharing to equalize the teams’ financial resources — a method of preserving “competitive balance” nicely in harmony with the league’s self-proclaimed “joint-venture” status.67
We are not required in this case to design a draft that would pass muster under the antitrust laws. We would suggest, however, that under the Supreme Court’s decision in Professional Engineers, no draft can be justified merely by showing that it is a relatively less anticompetitive means of attaining sundry benefits for the football industry and society. Rather, a player draft can survive scrutiny under the rule of reason only if it is demonstrated to have positive, economically procompetitive benefits *1189that offset its anticompetitive effects, or, at the least, if it is demonstrated to accomplish legitimate business purposes and to have a net anticompetitive effect that is insubstantial.68 Because the NFL draft as it existed in 1968 had severe anticompetitive effects and no demonstrated proeompetitive virtues, we hold that it unreasonably restrained trade in violation of § 1 of The Sherman Act.
C. Damages
The trial court found that plaintiff would have negotiated a more remunerative contract but for the draft as it existed in 1968, and estimated his damages (before trebling) at $92,200 — the difference between plaintiff’s actual compensation and what his “services would have brought in a free market.”69 Plaintiff urges that this estimate was too low. He contends principally that it was error to include in the subtrahend the $19,800 he was paid after his injury, since the Redskins allegedly are seeking to offset this sum against his pending workmen’s compensation claim.70 Defendants urge that the estimate was too high. They contend that Pat Fischer, an eight-year veteran and twice all-pro, was not comparable to plaintiff, an untried “rookie”; and that the trial court’s hypothesis of a three-year “fully guaranteed” contract was speculative, since no Redskins’ first-round draft choice and no defensive back in NFL history — rookie or veteran— had ever negotiated such advantageous terms.
The computation of damages in antitrust cases invariably has a certain Alice-in-Wonderland quality to it. As the Supreme Court has observed, “damage issues in these cases are rarely susceptible of the kind of concrete, detailed proof of injury which is available in other contexts.”71 Detailed proof of injury is particularly difficult here: because the draft has existed continuously since 1935, there has never existed (as there usually exists, for example, in price-fixing cases) a “free market” for players’ services that might serve as a guide to the prices that would have prevailed absent the antitrust violation.72 In view of such difficulties of ascertainment, the fact-finder is permitted to make “a just and reasonable estimate of the damage based on relevant data,” which may include “probable and inferential as well as direct and positive proof.”73 “The essential thing is that the available data be used in rational ways which warrant confidence that the damage figure reached is, in fact, a reasonable if imprecise estimate, rather than a speculative guess.” 74
Making full allowance for the deference that must be accorded the trier of fact, we have concluded after examining the record that his damage calculation must be over*1190turned. Inclusion of the disputed $19,800 in Smith’s “actual compensation,” we think was clearly proper;75 and the comparison with Pat Fischer, while it produced a hypothetical annual salary ($54,000) well in excess of that paid to any rookie defensive back in NFL history — even during the years of the NFL-AFL “bidding war”— was not clearly erroneous.76 But there was simply no evidence to support the judge’s finding that Smith, absent the draft, would have been able to negotiate a contract containing a guarantee of three years’ full salary, regardless of injury.77 No such *1191guarantee had been negotiated by Pat Fischer, the trial court’s object of comparison, and no such guarantee had ever been negotiated by any Redskins’ first-round draft choice, or by any defensive back at any time in NFL history. The evidence established that multi-year guaranteed contracts expose clubs to great financial risk, and that such guarantees in consequence can generally be negotiated only by veteran players of “Hall of Fame” caliber in key positions. There was, moreover, considerable evidence that Smith, like many rookies, was not even interested in a multi-year guaranteed contract, preferring more “upfront” money and the right to negotiate new terms on the basis of his first-year performance.
Nor do we think that the district judge was free to ignore all this evidence on the theory that it was irrelevant to what would have happened under “free market conditions.” Draft or no draft, Smith’s ability to negotiate such lucrative terms would have depended on the Redskins’ capacity to afford them. The Redskins, like the other NFL clubs, presumably were charging for tickets and TV rights approximately what the market would bear, yet their operating income was only $260,000 in 1968 and $300,-000 in 1969. Sheer economics as well as consistent past practice may thus have prevented the Redskins from assuming, in the case of an untried rookie, the financial risks attending the type of contract Smith proposes.
For these reasons, we conclude that the trial judge’s hypothesis of a fully guaranteed, three-year contract can be characterized only as a “speculative guess.” There was no evidence on which this hypothesis could be founded, no evidence from which it could be inferred. We accordingly remand this case to the District Court for recomputation of damages.
Affirmed in part, reversed in part, and remanded for proceedings consistent with this opinion.
. Sherman Act § 1, 15 U.S.C. § 1 (1976) provides in pertinent part:
Every contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the several States, or with foreign nations, is declared to be illegal .
Sherman Act § 3, 15 U.S.C. § 3 (1976) provides in pertinent part:
Every contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce in any Territory of the United States or of the District of Columbia . or between any such Territory or Territories and any State or States or the District of Columbia, or with foreign nations, or between the District of Columbia and any State or States or foreign nations, is declared illegal.
Smith’s complaint also alleged a violation of Sherman Act § 2, which forbids monopolization and attempts or conspiracies to monopolize.
*1175The monopolization issue was not discussed by the District Court, however, and none of the parties has raised it on appeal.
. Section 4 of the Clayton Act confers the right to sue for treble damages on “[a]ny person who shall be injured in his business or property by reason of anything forbidden in the antitrust laws . . . 15 U.S.C. § 15 (1976). Defendants have conceded that the draft operates to depress the starting salary levels of “top college players," Brief of NFL at 7, a class that evidently includes plaintiff, who was the twelfth player in the nation to be drafted in 1968. Defendants nevertheless suggest that plaintiff lacks standing to sue, arguing that any short-run pecuniary loss he may have suffered is more than offset by the long-run benefits — financial and otherwise — that accrue to players from the draft.. Brief of Pro-Football, Inc., at 29 — 44. Putting to one side the consideration that plaintiff, whose football career was terminated by an injury in his first year, was not around to reap this rich harvest of benefits, defendant’s theory that § 4 incorporates a “net loss” provision is frivolous, and we reject it. Whatever relevance “offsetting benefits” may have to the determination of the reasonableness of a restraint at the merits stage, they are utterly irrelevant to a determination of “injury-in-fact” at the standing stage.
Outside the context of professional sports, courts have regularly denied employees standing to sue for antitrust injuries to their employer, generally on the ground that any injury to the employee is “indirect.” See Berger & Bernstein, An Analytical Framework for Antitrust Standing, 86 Yale L.J. 809, 821-23 (1977). Yet the courts have invariably found that athletes have standing to challenge player restrictions in professional sports, since these restraints operate directly on, and to the detriment of, the employee. See generally Robertson v. National Basketball Ass’n, 389 F.Supp. 867, 883-84 (S.D.N.Y.1975) (citing cases).
. Judge Bryant’s opinion is reported at 420 F.Supp. 738 (D.D.C.1976). Since that time the Supreme Court has decided National Soc'y of Professional Eng’rs v. United States, 435 U.S. 679, 98 S.Ct. 1355, 55 L.Ed.2d 637 (1978), and St. Paul Fire & Marine Ins. Co. v. Barry, - U.S.-, 98 S.Ct. 2923, 57 L.Ed.2d 932 (1978), from whose guidance we have benefitted, a privilege denied to Judge Bryant.
. 420 F.Supp. at 746.
. NFL Constitution & By-Laws, Art. 10.2, 12.-1(F) & (G) (1968), reprinted in Joint Appendix (J.A.) at 212, 214.
. The draft and its attendant restrictions govern the entry of college players (“rookies”) into the NFL. A distinct set of restraints inhibits the mobility of players once they have signed with an NFL club. The Standard Player Contract as it existed in 1968 contained an “option” clause which gave a club, in the event that it and a player could not come to contract terms, the right unilaterally to renew the player’s contract for a second year at 90% of his previous year’s salary. If a player, despite the obvious financial disincentive, elected to “play out his option,” he became a “free agent” at the end of his second year and technically was free to negotiate with any other team. At this point, however, a player’s “freedom” was considerably restricted by the “Rozelle rule,” NFL Constitution & By-Laws, Art. 12.1(H) (1968), reprinted in J.A. at 214-15. This rule essentially required the signing club to compensate the player’s former team; if the two clubs were unable to conclude mutually satisfactory arrangements, the Commissioner could award compensation in the form of current players, future draft choices, or both, as he in his sole discretion deemed “fair and equitable.” The Rozelle rule restrained player mobility because of its in terrorem effect on the prospective signing team: teams were reluctant to risk losing key players and valuable draft choices at the Commissioner’s disposition, and thus were reluctant to sign a “free agent” unless they had succeeded in reaching an agreement with the player’s former team — an agreement which the former team, if it were recalcitrant, could preelude. These restraints, upon which we express no views here, are discussed in Mackey v. NFL, 543 F.2d 606 (8th Cir. 1976), cert. dismissed, 434 U.S. 801, 98 S.Ct. 28, 54 L.Ed.2d 59 (1977); Kapp v. NFL, 390 F.Supp. 73 (N.D.Cal.1974), aff’d in part and appeal dismissed in part as moot, 586 F.2d 644 (9th Cir. 1978).
In February 1977 the NFL clubowners and Players Association agreed to a new 5-year contract that works significant changes in the draft, Rozelle rule, option clause, and Standard Player Contract. See N.Y. Times, 18 Feb. 1977, at 21, col. 2; id., 26 Feb. 1977, at 15, col. 4. The modified draft is far less restrictive than the one described in the text; it continues for fewer rounds (thus applying to fewer players), eliminates the selecting team’s “perpetual” right of negotiation with its players, facilitates players’ becoming “free agents,” and establishes minimum salary levels for “rookies.” Relaxations are also introduced into the Rozelle rule (limiting it to “upper salary level” players), the option clause (making the option year salary 110% of the previous year’s and eliminating option clause after four seasons), and the Standard Player Contract (providing certain injury-protection guarantees). These modifications were noted favorably by the District Court in approving a settlement of the Mackey case. See Alexander v. NFL, No. 4-76-Civil 123 (D.Minn. 29 July 1977), slip op. at 26-28. We express no views on the legality of the modified NFL player draft, confining our attention to the draft as it existed in 1968.
. NFL Constitution & By-Laws, Art. 15.1 (1968), reprinted in J.A. at 216-17.
. See note' 6 supra.
. See note 6 supra.
. Clayton Act § 4, 15 U.S.C. § 15 (1976).
. Professional football, like all professional sports excepting baseball, is subject to the antitrust laws. Compare Radovich v. NFL, 352 U.S. 445, 77 S.Ct. 390, 1 L.Ed.2d 456 (1957) with Flood v. Kuhn, 407 U.S. 258, 92 S.Ct. 2099, 32 L.Ed.2d 728 (1972). Restrictive agreements between the NFL and its Players Association, like collective bargaining contracts in other industries, may in proper circumstances be shielded from antitrust scrutiny under the “labor law exemption.” See, e. g., United Mine Workers v. Pennington, 381 U.S. 657, 85 S.Ct. 1585, 14 L.Ed.2d 626 (1965); Local 189, Amalgamated Meat Cutters v. Jewel Tea Co., 381 U.S. 676, 85 S.Ct. 1596, 14 L.E.2d 640 (1965). The District Court found the labor law exemption inapplicable to the NFL draft on the facts of this case, 420 F.Supp. at 741-44, and defendants have not appealed this ruling. See Brief of NFL at 26.
. The Court of Appeals in Mackey v. NFL, 543 F.2d 606 (8th Cir. 1976), cert. dismissed, 434 U.S. 801, 98 S.Ct. 28, 54 L.Ed.2d 59 (1977), held that the Rozelle rule violated the antitrust laws. The District Court in Kapp v. NFL, 390 F.Supp. 73 (N.D.Cal.1974), aff'd in part and appeal dismissed in part as moot, 586 F.2d 644 (9th Cir. 1978), held that the Rozelle rule, the “no-tampering” rule, and the player draft unreasonably restrain trade. The legality of analogous player restraints in other professional sports has been discussed in several district court opinions; most of these decisions, rendered on motions for preliminary injunctions, are inconclusive. See, e. g., Robertson v. National Basketball Ass’n, 389 F.Supp. 867, 893-96 & 67 F.R.D. 691, 693-94 (S.D.N.Y.1975) (suggesting in dictum that NBA player draft and reserve clause may violate antitrust laws); Denver Rockets v. All-Pro Management, Inc., 325 F.Supp. 1049, 1056-58 (C.D.Cal.1971) (granting preliminary injunction), injunction reinstated sub nom. Haywood v. National Basketball Ass’n, 401 U.S. 1204, 91 S.Ct. 672, 28 L.Ed.2d 206 (Douglas, J., in chambers, 1971) (same); Philadelphia World Hockey Club v. Philadelphia Hockey Club, 351 F.Supp. 462, 486, 503-04 (E.D.Pa. 1972) (denying preliminary injunction on per se theory but suggesting that NHL reserve clause may violate rule of reason); Boston Prof. Hockey Ass’n v. Cheevers, 348 F.Supp. 261 (D.Mass.), remanded, 472 F.2d 127 (1st Cir. 1972).
. Standard Oil of N. J. v. U. S., 221 U.S. 1, 31 S.Ct. 502, 55 L.Ed. 619 (1911).
. E. g., United States v. General Motors Corp., 384 U.S. 127, 145-16, 86 S.Ct. 1321, 16 L.Ed.2d 415 (1966); Klor’s, Inc. v. Broadway-Hale Stores, Inc., 359 U.S. 207, 212-13, 79 S.Ct. 705, 3 L.Ed.2d 741 (1959); Fashion Originators’ Guild v. FTC, 312 U.S. 457, 61 S.Ct. 703, 85 L.Ed. 949 (1941).
. L. A. Sullivan, Antitrust 230, 232, 244 (1977).
. Id. at 230. The third type of boycott is comparatively rare: it can be effective only where it is necessary for firms at the boycotting level to deal with one another in order to carry on their business, as is true, for example, in the case of brokers. See Silver v. New York Stock Exchange, 373 U.S. 341, 83 S.Ct. 1246, 10 L.Ed.2d 389 (1963).
. L. A. Sullivan, supra note 15 at 245.
. E. g., United States v. General Motors Corp., 384 U.S. 127, 86 S.Ct. 1321, 16 L.Ed.2d 415 (1966) (invoking boycott per se rule where retailers induced manufacturers not to sell to competing retailers); Klor’s Inc. v. Broadway-Hale Stores, Inc., 359 U.S. 207, 79 S.Ct. 705, 3 L.Ed.2d 741 (1959) (invoking boycott per se rule where retailer induced manufacturers and distributors not to sell to competing retailer); Fashion Originators' Guild v. FTC, 312 U.S. 457, 61 S.Ct. 703, 85 L.Ed. 949 (1941) (invoking boycott per se rule where manufacturers induced retailers not to buy from competing manufacturers). See Worthen Bank & Trust Co. v. National BankAmericard, Inc., 485 F.2d 119, 124-25 (8th Cir. 1973), cert. denied, 415 U.S. 918, 94 S.Ct. 1417, 39 L.Ed.2d 473 (1974) (“group boycott” label, with consequent finding of per se illegality, properly restricted to (1) horizontal combinations to exclude competitors; (2) vertical combinations to exclude competitors; (3) combinations designed to influence trade practices of boycott victims); E. A. McQuade Tours, Inc. v. Consolidated Air Tour-Manual Comm., 467 F.2d 178, 186-87 (5th Cir. 1972), cert. denied, 409 U.S. 1109, 93 S.Ct. 912, 34 L.Ed.2d 690 (1973) (same); Note, 9 Conn.L. Rev. 336, 340 (1977); Note, 55 Nebr.L.Rev. 335, 344 (1976).
. See, e. g., Mackey v. NFL, 543 F.2d 606, 619 (8th Cir. 1976), cert. dismissed, 434 U.S. 801, 98 S.Ct. 28, 54 L.Ed.2d 59 (1977) (noting “joint venture” characteristics of NFL); L. A. Sullivan, supra note 15, at 251-52 (same); Note, The Super Bowl and the Sherman Act: Professional Team Sports and the Anti-trust Laws, 81 Harv.L.Rev. 418, 419-21 (1967); Cf. Levin v. National Basketball Ass’n, 385 F.Supp. 149, 152 (S.D.N.Y.1974) (noting “joint venture” characteristics of NBA); San Francisco Seals v. National Hockey League, 379 F.Supp. 966, 969 (C.D.Cal.1974) (noting “joint venture” characteristics of NHL).
. See Note, 55 Nebr.L.Rev. 335, 344 (1976).
. Cf. Deesen v. Professional Golfers’ Ass’n, 358 F.2d 165, 170 (9th Cir.), cert. denied, 385 U.S. 846, 87 S.Ct. 72, 17 L.Ed.2d 76 (1966) (refusing to invoke boycott per se rule against tournament entry restrictions where purpose was “not to destroy competition but to foster it by maintaining a high quality of competition”).
. The term “group boycott,” as the Eighth Circuit has noted, can be used as “a very broad label for divergent types of concerted activity,” Worthen Bank & Trust Co. v. National Bank-Americard, Inc., 485 F.2d 119, 125 (8th Cir. 1973), cert. denied, 415 U.S. 918, 94 S.Ct. 1417, 39 L.Ed.2d 473 (1974), and there is in consequence “more confusion about the scope and operation of the per se rule against group boycotts than in reference to any other aspect of the per se doctrine.” L. A. Sullivan, supra note 15 at 229-30. When confronted with concerted refusals to deal that do not fit the classic “group boycott” pattern, the courts almost without exception have held the per se rule inapplicable. In reaching this conclusion, however, they have vacillated between two divergent paths. Some courts have adhered to the traditional canon that all group boycotts are illegal per se, and concluded that the concerted activity at issue was not a group boycott. See, e. g., Joseph E. Seagram & Sons, Inc. v. Hawaiian Oke & Liquors, Ltd., 416 F.2d 71, 76-79 (9th Cir. 1969), cert. denied, 396 U.S. 1062, 90 S.Ct. 752, 24 L.Ed.2d 755 (1970). Other courts have held that the concerted activity at issue was a group boycott, but that there were two types of group boycotts — “per se boycotts” and “rule of reason boycotts” — and that the concerted activity at issue fell into the second category. See, e. g., 11 Von Kalinowski, Antitrust Laws and Trade Regulation § 76.02 at 76-11 (1978) (citing cases). Although these divergent paths presumably will lead in most cases to the same destination, their coexistence has not encouraged clarity or consistency of analysis.
The Supreme Court shed some light on this problem in St. Paul Fire & Marine Ins. Co. v. Barry, 438 U.S. 531, 98 S.Ct. 2923, 57 L.Ed.2d 932 (1978). In that case, the Court held that the refusal by a group of insurance companies to deal on any terms with certain policyholders was a “boycott” within the meaning of § 3(b) of the McCarran-Ferguson Act, 15 U.S.C. § 1013(b) (1976). Noting that its own previous decisions “reflectfed] a marked lack of uniformity in defining the term,” the Court decided that “boycott” should be given the broad meaning familiar from labor law and rejected defendants’ attempt to define it as “embracing *1180only those combinations which target competitors of the boycotters as the ultimate objects of a concerted refusal to deal.” 438 U.S. at 542, 98 S.Ct. at 2930 (emphasis original). Although the Court did not delimit “the precise reach” of the term, id. at 545 n.18, 98 S.Ct. at 2932 n. 18, it held that a “boycott” was at least broad enough to include the practice involved in that case, namely, an agreement by which one company “induced its competitors to refuse to deal on any terms with its customers.” id. at 544, 98 S.Ct. at 2931.
Given the Court’s broad definition of “boycott” in St. Paul Fire & Marine, we assume for purposes of this decision that the NFL draft could be described as a “boycott” — although as we noted above, the draft, unlike the agreement in that case, is not strictly an agreement among competitors. See pp.--- of 193 U.S.App.D.C., pp. 1178-1179 of 593 F.2d supra. To describe the draft as a boycott, however, is not to end our inquiry. The Court noted in St. Paul Fire & Marine that “ ‘boycotts are not a unitary phenomenon,’ ” id. at 543, 98 S.Ct. at 2931, quoting P. Areeda, Antitrust Analysis 381 (1974), and stated that the issue before it was only “whether the conduct in question involves a boycott, not whether it is per se unreasonable.” 438 U.S. at 542, 98 S.Ct. at 2930. In rejecting defendants’ argument that a boycott should be defined as “limited to concerted activity . . . against competitors of members of the boycotting group,” id. at 552, 98 S.Ct. at 2935, the Court expressed no view as to whether “a per se offense in this area” should be so defined, Id. at 542 & n.14, 98 S.Ct. at 2930 & n.14. Although the Court has clarified boycott analysis by suggesting that judicial investigations should pursue the second path described above, it left open the question that confronts us here: whether boycotts not designed to exclude competitors are illegal per se.
. L. A. Sullivan, supra note 15 at 232. See id. at 231-32:
Suppose . . . that a group of professional football teams agreed not to hire any player found to have gambled on games. This is a concerted refusal to deal. But, whatever the purposes and effects may be thought to be, it is a concerted arrangement very different from the typical boycott. It does not constitute concerted action by cornpeting firms at one level to exclude other would-be competitors. It ought, then, not to be analyzed under a generic rule which deals with concerted efforts by traders at a given level to insulate themselves from other competition.
Cf. Molinas v. National Basketball Ass’n, 190 F.Supp. 241 (S.D.N.Y.1961) (boycott perse rule inapplicable to NBA suspension of player for placing bets in league games).
. See Hatley v. American Quarter Horse Ass’n, 552 F.2d 646, 652-53 (5th Cir. 1977) (finding no per se illegal group boycott in association’s refusal to register plaintiffs horse, in accordance with its regulation: “[i]n an industry which necessarily requires some interdependence and cooperation, the per se rule should not be applied indiscriminately”); Worthen Bank & Trust Co. v. National BankAmericard, Inc., 485 F.2d 119, 126, 127, 128 n.7 (8th Cir. 1973), cert. denied, 415 U.S. 918, 94 S.Ct. 1417, 39 L.Ed.2d 473 (1974) (refusing to invoke boycott per se rule where “product” represented by bank credit card “requires cooperative relationships among . . . member banks” and it would be impossible for any bank to issue such card on its own).
. See, e. g., E. A. McQuade Tours, Inc. v. Consolidated Air Tour Manual Comm., 467 F.2d 178, 187-88 (5th Cir. 1972), cert. denied, 409 U.S. 1109, 93 S.Ct. 912, 34 L.Ed.2d 690 (1973) (finding no per se illegal group boycott in airlines’ refusal to list tour operator where tour operator was not competitor of airlines and there was no intent to exclude him from market); Bridge Corp. of America v. American Contract Bridge League, 428 F.2d 1365, 1370-71 (9th Cir. 1970), cert. denied, 401 U.S. 940, 91 S.Ct. 940, 28 L.Ed.2d 220 (1971) (finding no per se illegal group boycott in ACBL’s refusal to sanction local tournament where there was no purpose to exclude competition); Joseph E. Seagram & Sons, Inc. v. Hawaiian Oke & Liquors, Ltd., 416 F.2d 71, 76-79 (9th Cir. 1969), cert. denied, 396 U.S. 1062, 90 S.Ct. 752, 24 L.Ed.2d 755 (1970) (finding no per se illegal group boycott where distillers’ agreement to terminate distributor manifested no anticompetitive intent); Jones v. NCAA, 392 F.Supp. 295, 304 (D.Mass.1975) (finding no per se illegal group boycott in NCAA eligibility rule); *1181College Athletic Placement Serv., Inc. v. NCAA, 1975-1 Trade Cas. ¶ 60,117 (D.N.J.), aff’d without opinion, 506 F.2d 1050 (3d Cir. 1974) (same).
. 356 U.S. 1, 5, 78 S.Ct. 514, 518, 2 L.Ed.2d 545 (1958) (emphasis added).
. See White Motor Co. v. United States, 372 U.S. 253, 263, 83 S.Ct. 696, 9 L.Ed.2d 738 (1963); Evans v. S. S. Kresge Co., 544 F.2d 1184, 1191 (3d Cir. 1976).
. White Motor Co. v. United States, 372 U.S. 253, 263, 83 S.Ct. 696, 702, 9 L.Ed.2d 738 (1963), quoted in United States v. Topco Associates, Inc., 405 U.S. 596, 608, 92 S.Ct. 1126, 31 L.Ed.2d 515 (1972).
. 433 U.S. 36, 50, 97 S.Ct. 2549, 53 L.Ed.2d 568 (1977), quoting 356 U.S. at 5, 78 S.Ct. 514.
. Id. at 49-50, 97 S.Ct. at 2558 (emphasis added).
. United States v. Arnold, Schwinn & Co., 388 U.S. 365, 87 S.Ct. 1856, 18 L.Ed.2d 1249 (1967).
. 433 U.S. at 57-58, 97 S.Ct. at 2562.
. Some form of player selection system, or “draft,” is used in each of the major professional sports (i. e., the sports that have amassed the greatest spectator following through nationwide television broadcasts). The draft has been employed in professional football since 1935. A draft is used in professional basketball (see, e. g., Robertson v. NBA, 389 F.Supp. 867, 874, 891-93 (S.D.N.Y.1975)) and in professional hockey (see, e. g., Philadelphia World Hockey Club v. Philadelphia Hockey Club, 351 F.Supp. 462, 477-78 (E.D.Pa.1972)). After many years of exclusive reliance on its “farm system,” professional baseball adopted a draft in 1965 because the farm system was leading to competitive imbalance. See Note, supra note 19 at 422-23.
. See Robertson v. NBA, 389 F.Supp. 867, 892 (S.D.N.Y.1975), and cases cited in note 37 infra.
. See Note, supra note 19 at 419-22, and authorities cited in note 38 infra.
. White Motor Co. v. United States, 372 U.S. 253, 263, 83 S.Ct. 696, 9 L.Ed.2d 738 (1963).
. See Mackey v. NFL, 543 F.2d 606, 618-20 (8th Cir. 1976), cert. dismissed, 434 U.S. 801, 98 S.Ct. 28, 54 L.Ed.2d 59 (1977) (per se rule inapplicable to NFL Rozelle rule); Deesen v. Professional Golfers’ Ass’n, 358 F.2d 165 (9th Cir.), cert. denied, 385 U.S. 846, 87 S.Ct. 72, 17 L.Ed.2d 76 (1966) (rule of reason applicable to PGA tournament entry restrictions); Kapp v. NFL, 390 F.Supp. 73, 81 (N.D.Cal.1974) (per se rule inapplicable to NFL draft and Rozelle rule), aff’d in part and appeal dismissed in part as moot, 586 F.2d 644 (9th Cir. 1978); Philadelphia World Hockey Club v. Philadelphia Hockey Club, 351 F.Supp. 462, 503 (E.D.Pa.1972) (per se rule inapplicable to NHL reserve clause); Flood v. Kuhn, 316 F.Supp. 271, 275-78 (S.D.N.Y.1970), aff’d on other grounds, 407 U.S. 258, 92 S.Ct. 2099, 32 L.Ed.2d 728 (1972) (per se rule inapplicable to professional baseball reserve system) (dictum). Cf. Hatley v. American Quarter Horse Ass’n, 552 F.2d 646, 652-53 (5th Cir. 1977) (per se rule inapplicable to racing association registration rule); Bridge Corp. of America v. American Contract Bridge League, 428 F.2d 1365, 1369 (9th Cir. 1970), cert. denied, 401 U.S. 940, 91 S.Ct. 940, 28 L.Ed.2d 220 (1971) (per se rule inapplicable to ACBL refusal to sanction local tournament); College Athletic Placement Serv., Inc. v. NCAA, 1975-1 Trade Cas. ¶ 60,117 (D.N.J.), aff'd without opinion, 506 F.2d 1050 (3d Cir. 1974) (per se rule inapplicable to NCAA player eligibility rule).
. See L. A. Sullivan, supra note 15, at 231-32, 255 n.5; Note, 41 Albany L.Rev. 154, 159-60 (1977); Note, 9 Conn.L.Rev. 336, 342 (1977); Note, 55 Nebr.L.Rev. 335, 344-47 (1976); Note, 59 Marq.L.Rev. 632, 638 (1976); Note, The Legality of the Rozelle Rule and Related Practices in the National Football League, 4 Ford.Urb. L.J. 581, 586-88 (1976); Note, Player Control Mechanisms in Professional Sports, 34 U.Pitt.L. Rev. 645, 651-53 (1973).
. Mackey v. NFL, 543 F.2d 606, 618-20 (8th Cir. 1976), cert. dismissed, 434 U.S. 801, 98 S.Ct. 28, 54 L.Ed.2d 59 (1977). See note 6 supra.
. See Continental T. V., Inc. v. GTE Sylvania, Inc., 433 U.S. 36, 50 n.16, 97 S.Ct. 2549, 53 L.Ed.2d 568 (1977); Hatley v. American Quarter Horse Ass’n, 552 F.2d 646, 653 (5th Cir. 1977).
. Cardozo, Mr. Justice Holmes, 44 Harv.L. Rev. 682, 688 (1931).
. Continental T. V., Inc. v. GTE Sylvania Inc., 433 U.S. 36, 59, 97 S.Ct. 2549, 2563, 53 L.Ed.2d 568 (1977).
. National Society of Professional Engineers v. United States, 435 U.S. 679, 692, 98 S.Ct. 1355, 1365, 55 L.Ed.2d 637 (1978).
. Handler, Changing Trends in Antitrust Doctrines: An Unprecedented Supreme Court Term — 1977, 77 Colum.L.Rev. 979, 983 (1977). See Continental T. V., Inc. v. GTE Sylvania, Inc., 433 U.S. 36, 49, 97 S.Ct. 2549, 2557, 53 L.Ed.2d 568 (1977), citing Chicago Bd. of Trade v. United States, 246 U.S. 231, 238, 38 S.Ct. 242, 62 L.Ed. 683 (1918) (under the rule of reason, “the factfinder weighs all of the circumstances of a case in deciding whether a restrictive practice should be prohibited as imposing an unreasonable restraint on competition”). Cf. id. at 50 n.16, 97 S.Ct. at 2558 (in determining whether a particular commercial practice should be prohibited per se, “[t]he probability that anticompetitive consequences will result from [the] practice and the severity of those consequences must be balanced against its pro-competitive consequences”).
. L. A. Sullivan, supra note 15 at 187, 188. See National Society of Professional Engineers v. United States, 435 U.S. at 691, 98 S.Ct. at 1365, quoting Chicago Bd. of Trade v. United States, 246 U.S. at 238, 38 S.Ct. at 244, (“The true test of legality is whether the restraint imposed is such as merely regulates and perhaps thereby promotes competition or whether it is such as may suppress or even destroy competition.”)
. The bulk of our colleague’s dissent is devoted to a recitation and analysis of detailed facts relating to the history of professional football, some in the record before the district court and some supplemental thereto. See note 49 infra. Our colleague then makes his own findings of fact as to the benefits derived in 1968 and currently from the professional football draft, and its effect on competition, these findings of fact differing from those of the trial court. For example, “[s]ome argue that this [competitive balance] has been caused by other factors, but the preponderance of the factual testimony and record evidence supports a conclusion that the college player draft was the key factor which produced the competitive balance of the teams . . There was no showing to the contrary in the 2,000 page record.” Dis. op. at -, - of 193 U.S.App.D.C., at--, 1198-1199 of 593 F.2d.
First, we believe that the NFL’s theory of competitive balance, even were it empirically valid, is legally wide of the mark in a rule of reason inquiry, see pp.---of 193 U.S. App.D.C. 1185-1189 of 593 F.2d, infra. Moreover, there was a showing in the record which contradicts the findings of the dissent as our colleague himself realizes: “The court concluded that the evidence on whether the draft was essential to competitive balance was ‘at best equivocal,’ and that no correlation was demonstrated between the draft and survival of the League.” Dis. op. at- of 193 U.S.App.D.C., at 1211 of 593 F.2d.
We suggest this finding of fact by the trial court, as recited by our colleague, must be affirmed if there is evidence in the record to support it, and that an appellate court cannot make a different finding of fact, particularly when the record evidence obviously lends itself to the conclusion reached by the trial court. See Daniels v. Hadley Memorial Hospital, 185 U.S. App.D.C. 84, 91, 566 F.2d 749, 756 (1977).
First, to the extent the defendant’s theory relies on the draft as a device to allocate player talent evenly, without which wealthy owners would be free to corner the market, Judge Bryant found “abundant convincing testimony” that factors other than money are often decisive in players’ choice of teams. 420 F.Supp. at 746. See J.A. at 945 (unrelated business opportunities); id. at 1049-52 (same); id. 1073-74 (racial discrimination and general community atmosphere); id. at 1080-84 (preference for NFL over AFL); id. at 1086, 1091-92 (dispute with owner); id. at 1112 (climate); id. at 1212 (educational opportunities); id. at *11841214-15 (disagreement with coaching staff and management). Further, the evidence with regard to the movement of free agents and former World Football League players who were ungoverned by the draft did not reveal any trend for the best players to become concentrated with teams “offering money, glamour and success.” 420 F.Supp. at 746, citing Free Agents’ First Wave On the Move, Washington Post, 7 July 1976, at D1.
Second, it is doubtful whether the draft was effective in maintaining whatever competitive balance did exist in the League. Although from year to year the worst teams usually got somewhat better and the best teams somewhat worse, it is uncertain how much, if any, of this is explained by the draft. See J.A. at 939-40, 990, 1975 (testimony of Dr. George Burman); id. at 1968-69 (testimony of Bruce Caputo).
Without by any means going into the minute detail furnished with such relish by our colleague, we think that two other factors contribute at least as much as the player draft to producing and maintaining a competitive balance in the league — television revenues and coaching changes.
The player draft originated in 1935 and was in full effect from then on, but none of the wonderful development of the professional football game pointed to by the appellant, and encored by our colleague, occurred until the mid-1950’s. The advent of professional football on television created a synergistic effect between television and the sport. Television exposure brought more fans, produced more revenues, which in turn produced better paid players, more profitable franchises, and generally a better entertainment feature for television. This sort of ratchet effect started working in the mid-1950’s with the advent of television and has continued over the years. See Proceedings, Conference on the Economics of Professional Sports, Washington, D.C., 7 May 1974, at 30, J.A. at 290.
It is possible to say that in the 1935-1955 period the player draft had some importance in preserving a kind of competitive balance, preventing rich owners from running away with the league (if there were any independently rich owners of the then only nine professional football teams). But by 1968, as now, the impact of the player draft is clearly different from what arguably it had been in 1935-1955.
The great equalizing feature now is not the player draft, but the league’s arrangement to place the teams financially on nearly equal footing — equal sharing of television revenues. In 1968 each team in the NFL received nearly one third of its revenues from television and radio, averaging $1.13 million. See J.A. at 618. (In 1974, the most recent year for which figures appear in the record, such revenues averaged $2.28 million and were 34.5 percent of total revenues. Id. at 655). We were told at oral argument by counsel for the NFL that such revenues now represent over half of average team income. If one realizes that at the present time all teams in the league will receive more than one half of their revenues in exactly equal shares from television, and nearly every team in the league is able to sell out its stadium for all the league games, then it appears that the financial sinews for a competitive player balance have been equalized. See J.A. at 1978 (testimony of Dr. George Burman).
The second factor producing competitive balance (or lack of it) is the impact of the coach on the success or failure of the team. Despite the claims to “competitive balance” from whatever factors, it is clear that a small number of the teams have dominated the play-offs year after year. The record shows nine teams as occupying twenty-two out of the available twenty-four places in the play-offs in the 1973-1975 seasons. See 420 F.Supp. at 746, citing N.Y. Times, 6 June 1976, at 56. These nine teams include teams with long records of success year after year with the same coach: (1) Dallas Cowboys with Landry; (2) Washington Redskins with Lombardi and Allen; (3) Minnesota Vikings with Grant; (4) Oakland Raiders with Madden; (5) Miami Dolphins with Shula; (6) Pittsburgh Steelers with Noll. These teams have consistently been in the play-offs and won titles, despite the fact that under the draft they have been getting the worst picks year after year. It would appear that the effects of fine coaching swamp whatever effect the draft may have on team performance.
Analysis of the record of coaching changes, or lack of changes, illustrates at least two things. First, that the player draft does not have an equalizing effect to the extent of knocking out the top teams, if the top teams have good coaches. Second, it is certainly arguable that what happens to change the position of teams in the league is not a draft choice advantage, but that the losing teams fire the losing coaches.
Our colleague argues “[i]n only one of twelve seasons from 1964-1975 did those teams in the bottom half of the NFL’s standings fail to win a greater number of games in the next following priority selections in the draft. Statistical analysis shows that this could not have been the product of chance occurring without regard to the draft.” (Dis. op. note 45). Our dissenting colleague did not tell us how many of the teams in the bottom half of the standings not only received the usual priority in draft selection, but also fired their losing coaches. Whatever statistical analysis our colleague is talking about, it is obviously worthless unless it takes account of the changes in the head coaches and the coaching staffs as well.
*1185The changes in position in the league effected by changes in head coaches is easily seen from the record by what happened after the arrival of Lombardi in Green Bay, the arrival of Lombardi and thereafter Allen at Washington, the move of Shula to the Miami Dolphins, and the move of Noll to the Pittsburgh Steelers. See J.A. at 658-76; id. at 950-52, 1978-79 (testimony of Dr. George Burman); id. at 904 — 05 (testimony of Edward Garvey); id. at 1116-18 (testimony of Bart Starr). In each of these four instances teams with a long series of losing seasons, during which they had had the benefit of priority in the player draft, were very shortly upgraded into title contenders after the arrival of these coaches of recognized high ability.
We think the trial court was eminently correct when he “concluded that the evidence on whether the draft was essential to competitive balance was ‘at best equivocal,’ and that no correlation was demonstrated between the draft and survival in the league.”
. Brief of NFL at 9.
. Defendants argue that college football players possess a significant amount of bargaining power, since they can refuse to sign with the team that drafted them, electing instead to sit out several seasons or to play for the Canadian Football League. The first alternative, however, necessitates both a considerable financial sacrifice and a potentially harmful interruption of a player’s career. The second alternative is little better: the evidence established that the opportunities for American players in Canada are both limited (owing to a hiring preference for native players) and significantly less rewarding (lower salaries, fewer promotional opportunities, less “glamor”). The evidence plainly supported the District Court’s finding that the availability of these alternatives failed to furnish a player with much, if any, bargaining power in salary discussions with the drafting club.
. 420 F.Supp. at 746. Our colleague has reinforced his own encyclopedic knowledge of football with repeated references to The NFL’s Official Encyclopedia History of Professional Football (1977), beginning with note 3. Whatever may be the propriety of a trial court taking judicial notice of an “encyclopedia” to establish facts (and an encyclopedia whose occasional inaccuracy our colleague himself corrects, note 23), see McCormick’s Handbook of the Law of Evidence §§ 329-31 (1972) (2d ed.), we think it improper and a bit unfair to ground an opinion urging reversal of the trial court’s principal factual conclusions (that the evidence was equivocal whether the draft was essential to preserve competitive balance, and that the draft had a severely anticompetitive impact) on “facts” culled from a book appearing a year after the trial court’s decision. We suggest the factual bases, and therefore legal validity, of the dissent is similarly flawed by reliance on personal observation (note 23) and common knowledge in the community (notes 51 and 69).
Indeed, there are frequent references to events occurring after the oral argument in this case, e. g., O. J. Simpson’s 1978 record (p. 29), Terry Metcalfs 1978 Canadian contract (note 69), the New York Yankees’ recent success (p. 53). One man’s reminiscences of a popular sport should not be confused with a judicial decision.
If Judge Bryant’s principal factual conclusion is to be set aside, as our colleague urges, it must be on the ground that it was clearly erroneous, see Daniels v. Hadley Memorial Hospital, 185 U.S.App.D.C. 84, 91, 566 F.2d 749, 756 (1977), and, surely, clearly erroneous on the record before him, see Garcia v. American Marine Corp., 432 F.2d 6, 8 (5th Cir. 1970) (per curiam); U. S. v. Summit Fidelity & Surety Co., 408 F.2d 46, 48 (6th Cir. 1969); C. Wright and A. Miller, Federal Practice and Procedure *1186§ 2577 and n.94 (1971), without the wholesale importation of “facts” judicially noticed for the first time on appeal. See Davis, A system of Judicial Notice Based on Fairness and Convenience, in Perspectives of Law 94 (Bound ed. 1964); cf. Landy v. Federal Deposit Insurance Corp., 486 F.2d 139, 150-51 (3rd Cir. 1973).
. Chicago Bd. of Trade v. United States, 246 U.S. at 238, 38 S.Ct. 242.
. 435 U.S. 679, 98 S.Ct. 1355, 55 L.Ed.2d 637 (U.S. 25 April 1978), aff'g 181 U.S.App.D.C. 41, 555 F.2d 978 (1977).
. 435 U.S. at 693, 98 S.Ct. 1355, at 1366 (footnote omitted).
. Id. at 695, 98 S.Ct. at 1367.
. Id. at 688, 98 S.Ct. at 1363.
. Id. at 690, 98 S.Ct. at 1364 (emphasis added; footnote omitted).
. Id. (footnote omitted).
. Id at 696, 98 S.Ct. at 1368.
. Goldfarb v. Virginia State Bar, 421 U.S. 773, 788-89 n.17, 95 S.Ct. 2004, 2013, 44 L.Ed.2d 572 (1975):
The fact that a restraint operates upon a profession as distinguished from a business is, of course, relevant in determining whether that particular restraint violates the Sherman Act. It would be unrealistic to view the practice of professions as interchangeable with other business activities, and automatically to apply to the professions antitrust concepts which originated in other areas. The public service aspect, and other features of the profession, may require that a particular practice, which could properly be viewed as a violation of the Sherman Act in another context, be treated differently.
. 435 U.S. at 696, 98 S.Ct. at 1367.
. Id. See Chicago Bd. of Trade v. United States, 246 U.S. at 238, 38 S.Ct. 242, quoted in note 45 supra.
. 435 U.S. at 695, 98 S.Ct. 1355. The Redskins contend that the draft did not eliminate competition, since it permitted bargaining between the club and the player following the player’s selection in the draft. See note 47 supra. Yet here again the parallel with the argument and decision in Professional Engineers is telling. The Society argued “that competition, in the form of bargaining between the engineer and customer, is allowed under its canon of ethics once an engineer has been initially selected. ... It then contends that its prohibition of competitive bidding regulates only the timing of competition, thus making this case analagous to Chicago Board of Trade . . .” The Court found this reliance on Chicago Board of Trade misplaced, pointing out that “petitioner’s claim mistakenly treats negotiation between a single seller and a single buyer as the equivalent of competition between two or more potential sellers”. Id. at 693 n.19, 98 S.Ct. at 1366. The football player draft produced a much more noncompetitive imbalance in the respective negotiating positions of players and single team than the positions of the parties in Professional Engineers.
. 420 F.Supp. at 746.
. Id. at 747. See Note, supra note 19 at 425 (in order for player draft to be a reasonable restraint, “at least a minimal opportunity to deal with more than one team should be assured”).
. See Kapp v. NFL, 390 F.Supp. 73, 82 (N.D.Cal.1974) (suggesting that NFL draft is unreasonable because it permits boycott of player “even when the drafting club refuses or fails within a reasonable time to reach a contract” with him), aff’d in part and appeal dismissed in part as moot, 586 F.2d 644 (9th Cir. 1978). The NBA has adopted a selection system under which a player is free to negotiate with any team if the drafting team fails within a stated time to tender a contract containing specified minimum terms. See Robertson v. National Basketball Ass’n, 556 F.2d 682, 686 & n.5 (2d Cir. 1977), aff’g 72 F.R.D. 64, 69-70 & n.1 (S.D.N.Y.1976) (approving settlement). The NFL and its Players’ Association have likewise adopted a less restrictive draft under which “the selecting club does not obtain a perpetual right of negotiation with the players” and “players who are unable to come to acceptable contract terms with the club selecting them can become free agents after limited periods of time.” Alexander v. NFL, No. 4-76-Civil 123 (D.Minn. 29 July 1977), slip op. at 27 (approving settlement of Mackey case). See note 6 supra.
. See Note, 55 Nebr.L.Rev. 335, 354 (1976) (arguing that NFL draft would be “reasonable” if conducted in two sessions, with players who fail to sign after second session becoming “free agents” thereafter); Note, supra note 19 at 423, 425 (arguing that professional baseball draft, which allows player to be “redrafted” by another club if the first club fails to sign him within given time, “has proven itself to be a workable alternative to the more restrictive drafts used in other sports and thus should be required in all sports”).
. 420 F.Supp. at 746-47. See Note, 55 Nebr. L.Rev. 335, 353 (1976) (arguing that NFL draft would be “reasonable” if it lasted for five rounds instead of 17). Cf. Mackey v. NFL, 543 F.2d 606, 622 (8th Cir. 1976), cert. dismissed, 434 U.S. 801, 98 S.Ct. 28, 54 L.Ed.2d 59 (1977) (holding Rozelle rule “unreasonable,” inter alia, because it applied not just to better players, but “to every NFL player regardless of his status or ability”); Note, The Legality of the Rozelle Rule and Related Practices in the National Football League, 4 Ford.Urb.L.J. 581, 589-90 (1976) (arguing that Rozelle rule would be “reasonable” if it did not apply to “average” players). The NFL and its Players’ Association recently have adopted a less restrictive draft which “extends to fewer players by continuing for fewer rounds.” Alexander v. NFL, No. 4-76-Civil 123 (D.Minn. 29 July 1977), slip op. at 27 (approving settlement of Mackey case). See note 6 supra.
The NFL suggests that a draft of fewer rounds would not have helped plaintiff, who was a first-round draft choice in 1968. See Brief of NFL at 50. Yet if the teams were forced to compete for the services of average players it is probable that those players’ salaries would rise, thus lifting the “floor” at which compensation for the most talented players, subject to the abbreviated draft, would have to begin. A “free market” for the services of most players, in short, would have pushed the salary levels of all players up.
. See Note, 9 Conn.L.Rev. 336, 344-45 & n.55 (1977).
. The Supreme Court in Professional Engineers, however, suggested that ethical norms designed to regulate competition, and marketing restraints relating to product safety, can be justified only if they “have no anticompetitive effect.” 435 U.S. at 696 n.22, 98 S.Ct. 1355 (emphasis added). See id. at 699, 98 S.Ct. 1355 (Blackmun, J., concurring).
. 420 F.Supp. at 749.
. Paragraph 12 of the Standard Player Contract signed by Smith and the Redskins provides:
Any payments made hereunder to the Player, for a period during which he is entitled to workmen’s compensation benefits by reason of temporary total, permanent total, temporary partial, or permanent partial disability shall be deemed an advance payment of workmen’s compensation benefits due the Player, and the Club shall be entitled to be reimbursed the amounts thereof out of any award of compensation.
Plaintiffs other arguments for a higher damage award, which need not be detailed here, are frivolous, and were properly rejected by the District Court.
. Zenith Radio Corp. v. Hazeltine Research, Inc., 395 U.S. 100, 123, 89 S.Ct. 1562, 1576, 23 L.Ed.2d 129 (1969).
. See Alexander v. NFL, No. 4-76-Civil 123 (D.Minn. 29 July 1977), slip op. at 18 (approving settlement of Mackey case).
. Bigelow v. RKO Pictures, Inc., 327 U.S. 251, 264, 66 S.Ct. 574, 90 L.Ed. 652 (1946), quoting Story Parchment Co. v. Paterson Parchment Paper Co., 282 U.S. 555, 561-64, 51 S.Ct. 248, 75 L.Ed. 544 (1931).
. L. A. Sullivan, supra note 15 at 787 (footnote omitted).
. Plaintiffs “actual damages” obviously were reduced by the $19,800 that he actually received, and his actual receipt of this sum cannot be ignored on the speculative chance that the Redskins may claim and be allowed it as an offset to an unrelated and as yet undetermined liability. The fact that the Redskins have derived a benefit from their payment of this sum in this case, of course, may prevent them, on some sort of estoppel theory, from deriving a second benefit from it in the workmen’s compensation suit; but that question is of no concern to us here.
. Pat Fischer, like plaintiff, was a defensive back who signed with the Redskins in 1968. Although Fischer was a seasoned player, it cannot be said that Smith, an All-American first-round draft choice, was demonstrably out of Fischer’s league. A very rough index of the two players’ comparability was furnished by the NFL Commissioner’s refusal, in a 1968 Rozelle Rule transaction, to award Smith as “compensation” to Fischer’s former team. See 420 F.Supp. at 749 & note 6 supra. Putting to one side the murky question of how “bonuses” are to be allocated, see 420 F.Supp. at 749 n.9, Smith’s actual one-year compensation ($28,000 “bonus” plus $22,000 salary) and Fischer’s annualized one-year compensation ($24,000 “bonus” plus $30,000 salary) are not too far apart anyway. The trial judge was obliged to find some guideline to Smith’s hypothetical compensation under conditions more nearly approximating a free market. He came up with a player who played a defensive position similar to plaintiffs and signed as a “free agent” with the same team in the same year. Bearing in mind “the familiar rule that a wrongdoer is responsible for uncertainty in calculating the damages proximately caused by his own wrongdoing,” Lehrman v. Gulf Oil Corp., 464 F.2d 26, 45 (5th Cir.), cert. denied, 409 U.S. 1077, 93 S.Ct. 687, 34 L.Ed.2d 665 (1972), we cannot say that the trial judge’s comparison was unreasonable.
. The District Court’s findings on this score are contained in 420 F.Supp. at 748:
The evidence showed unequivocally that top players at that time were being offered long-term contracts by some teams, including at times the Redskins. The Court concludes that a player of plaintiffs talent would certainly have been able to negotiate a three-year contract on very favorable terms in the absence of the restraints imposed by the owners. The evidence also showed that such players were able to negotiate contracts which guaranteed payment for the full term of the contract even if the player were injured (“fully vested contract” or “injury protection clause”). On the basis of the evidence of the contracts being negotiated at times elsewhere in the league and by the Redskins during the relevant period, and considering that such protection was shown to be of great concern to players of plaintiffs caliber, the Court also concludes that plaintiff would have been able to include such injury protection in a long-term contract had it not been for the illegal restraints imposed upon him.
Numerous, references appear in the record to so-called “no-cut” contracts, which were frequently negotiated in the league. It is apparently undisputed, however, that such contracts did not guarantee salary payments, in the event of debilitating injury, beyond the year in which the injury occurred. See Brief of Pro-Football, Inc. at 22. Additionally, the record contains references to several multiyear player contracts from which the player’s warranty of “excellent physical condition” was excised. Again it appears that the provision does not assure salary payments, in the event of injury, over the full course of the contract period. Characterizing two such contracts, plaintiff stated that the salary would be paid “as long as [the player] was physically fit.” See Brief of Appellee Smith at 7a (contract of Bill Yearby); id. at 9a (contract of Carl McAdams). Whether this is a sensible construction of such language is, needless to say, not before us. So interpreted, however, the contracts plainly do not support the judge’s finding that Smith, absent the draft, would have been able to negotiate a contract guaranteeing three years’ salary, regardless of injury.
Although we set aside the trial court’s damage award insofar as it depends on a conjectural guarantee of three years’ salary, regardless of injury, we do not intend otherwise to circumscribe on remand the trial court’s discretion in computing plaintiffs damages. We do in passing note with approval the approach taken by Judge Bryant at trial, see 420 F.Supp. at 749 n.9, in ignoring arguable distinctions between salary and *1191bonus for purposes of measuring overall player compensation. In this regard, setting aside a particular item of damages which we think to be without support, we intimate no view on the overall reasonableness of the damage award apart from the particular theory on which it was made.