The decision, dated November 18, 2022, is Aristocrat Technologies Australia Pty Limited v Konami Australia Pty Limited (No 3)  FCA 1373 (Nicholas J). It follows an earlier decision finding the patent in suit (Australian Patent No 754689) valid and infringed, and it’s fairly lengthy. I’ll try to explain the gist of it here, but may return to some of the finer or more fact-specific points in future posts.
The case involves electronic gaming machines (EGMs) having three principal components: a cabinet, a platform, and games (para. 57). Representative claim 1 of the ‘689 Patent reads as follows:
A random prize awarding feature to selectively provide a feature outcome on a gaming console, the console being arranged to offer a feature outcome when a game has achieved a trigger condition, the console including trigger means arranged to test for the trigger condition and to initiate the feature outcome when the trigger condition occurs, the trigger condition being determined by an event having a probability related to desired average turnover between successive occurrences of the trigger conditions on the console.
The court interprets the “substance of the invention” to comprise “a random prize awarding feature with the proportional trigger” (para. 173). The defendant Konami infringed by (1) selling 6,931 EGMs on which infringing games were installed; (2) selling “4,647 conversions for the installation of an infringing game on an existing EGM”; and (3) making 7,380 “no charge conversions” (NCCs). The plaintiff sought an award of the defendant’s profits on the infringing sales, and damages in the form of a royalty for the NCCs.
Most of the decision relates to the award of profits. Long story short, the plaintiff argued that it was entitled to Konami’s entire profit on these sales, on the authority of Australian and English decisions holding that “the infringer cannot avoid accounting for profits by relying on the existence of possible non-infringing alternatives that might have been used” (para. 108, citing Celanese International v BP Chemicals  RPC 203 (Laddie J); the leading Australian decision, also discussed at length, is Dart Industries Inc v The Decor Corporation Pty Ltd (1993) 179 CLR 101). Readers of the blog and my other work may recall that I have criticized this rule on many occasions, e.g., at pp. 200-06 of my book, discussing Dart and Celanese. What’s interesting here, however, is that the court, while adhering to these precedents, nevertheless does take noninfringing alternatives into account in what it refers to as a second stage of the analysis, relating to apportionment.
On this latter issue, the court reads the relevant precedents as also standing for the proposition that that “the plaintiff could not recover profits which were not properly attributable to the use of the plaintiff’s invention” (para. 109), though “apportionment would be inappropriate where, without the infringement, the infringing articles would not have existed, or where the invention was an essential ingredient in the creation of the infringer’s whole product” (para. 116). In particular, the court considers noninfringing alternatives to be relevant where apportionment is necessary to avoid unjustly penalizing the defendant. Here are the key passages:
119 In 2004 the Canadian Supreme Court in Monsanto Canada Inc. v. Schmeiser  1 SCR 902 held that a patentee is only entitled to that portion of an infringer’s profit which is attributable to the appropriation of the patented invention. The Court held that a comparison must be made between the infringer’s profit attributable to its use of the invention and its profit had it used the best non-infringing option available. This approach, which measures profits by reference to non-infringing alternatives, does not reflect the law in either Australia or the United Kingdom. However, it has been suggested that this approach may be justified not on the basis that the defendant’s profits correspond to the difference between those made supplying the infringing product and a product that was non-infringing, but on the basis that a consideration of non-infringing options will provide a useful measure with which to apportion profit between that which is attributable to the use of the invention and that which is not. On that analysis, any consideration of non-infringing alternatives, is only relevant when the court is required to assess what part of the infringer’s profits are attributable to the infringer’s appropriation of the invention. In those circumstances, the use of the term “non-infringing alternative” is referring to a non-infringing course of action that can assist in identifying the value of the relevant appropriation. . . .
181 It is useful to consider that discussion of the principles in the context of the facts in Dart and a hypothetical example. On the findings of fact made in Dart none of the claimed profits would have been made but for the respondents’ use of the patented invention. That is because, on the primary judge’s findings, the existence of the patented feature was essential to the making of any sale. The retailers to whom the product was supplied would never have purchased them had they not included the press button seal. But there may be cases in which, even if it is established that the profit would not have been made but for the use of the patented invention, some part of the profit made is shown to have no reasonable connection with the act of infringement. It is at this stage of the analysis that the onus moves to the infringer. . . .
183 Properly understood, Aristocrat’s argument is not that the relevant EGMs and conversion kits would have been manufactured or supplied by Konami without any feature game, but they would not have been manufactured and supplied with a non-infringing feature game. The argument therefore calls for a counterfactual analysis which involves determining whether there was any market for the relevant EGMs and conversion kits had they incorporated a non-infringing feature game that did not use the 689 trigger. Put another way, and adopting the language used by the Full Court in Dart at 408, was it the feature game incorporating the 689 trigger that created the market for the relevant EGMs and conversion kits manufactured and supplied by Konami?
184 With regard to the counterfactual analysis, Aristocrat submitted that apportionment was not available because the existence or otherwise of alternative products (i.e. the counterfactual) is not relevant to apportionment. I do not accept that submission. As I have stated above, the availability of non-infringing alternative products is not relevant when determining whether apportionment is available. However, the existence of non-infringing alternative products may assist in ascertaining the extent of the profits made by the respondent as a result of its use of the patented invention when apportioning the profits.
185 The profits recoverable on an account of profits are not measured by reference to the difference between the profit actually made by the infringer as a result of its infringement and the profit it would have made had it adopted a non-infringing alternative. That proposition is well settled and is not disputed by Konami. It does not suggest that the profits to be awarded to Aristocrat should be calculated in that manner. However, if apportionment is appropriate then the availability of non-infringing alternatives, and the cost of adopting them, may be relevant in determining the value that should be attributed to the invention that was wrongfully appropriated by the infringer.
186 In the hypothetical example of the braking system in the motor vehicle considered in Dart, it may be appropriate in determining what value to put on the braking system relative to the motor vehicle as a whole to have regard to other non-infringing alternatives that might have been used in place of the infringing braking system. In that situation, if the invention was shown to offer little benefit over other systems available at similar cost, it may be open to the Court to conclude that only a small proportion of the profits made on the sale of the vehicle should be attributed to the wrongful appropriation of the invention. Conversely, if the infringing system was shown to possess significant advantages over such alternatives, or where the system otherwise added significantly to the attractiveness of the vehicle to potential purchasers, then this is a matter that would be taken into account when determining how to apportion the profits made on the sale of the motor vehicle.
Applying these principles to the facts, the court concludes, for reasons I will not recap here, that 35% of Konami’s profit on the infringing sales were attributable to the infringing “prize awarding feature with the proportional trigger” (see paras. 188-290). The court then goes on to deduct, inter alia, the allocable fixed costs that Konami would have incurred from selling non-infringing products:
373 I do not regard this as a case in which the alternative products would have generated a much smaller amount of revenue than the infringing products: cf. Kettle at 469. On the contrary, I am satisfied that had Konami used a different trigger in the 689 Games it would more likely than not have achieved sales not vastly different from those made using the proportional trigger. However, in circumstances where the proportional trigger had advantages over other triggers which, all other things being equal, may have led at least some venues not to purchase the alternative product, I think it is appropriate to make an adjustment in Aristocrat’s favour with the aim of ensuring that Konami accounts for the full profit it derived by selling the infringing products. Although the very large adjustment suggested by Aristocrat is not warranted, I do think that some meaningful adjustment is necessary which allows for the possibility that some not insignificant proportion of the sales actually made would not have been made if a different trigger had been used. In all the circumstances I am satisfied that it would be appropriate to reduce the overhead attributable to the infringing products by 20%.
Finally—and skipping over some details that are not essential to the purposes of the present post—the court, having considered the purportedly comparable licenses submitted by the parties, awards a royalty of $1,000 per unit for the NCCs (para. 562). It declines to award enhanced damages, finding, inter alia, that the infringement was not “flagrant” (para. 571). The total award, as determined at a follow-up order dated December 15, 2022, comes to $16,248,116 in profits, plus an additional $11,642,730 in interest; and $4,995,000 in damages plus another $3,014,008 in interest.
Anyway, though I will continue to mull over the implications of this decision in the coming weeks, I think the court ultimately arrives at the correct conclusion—that noninfringing alternatives are relevant to apportionment—while still adhering (as I assume it must) to the rule that “[t]he profits recoverable on an account of profits are not measured by reference to the difference between the profit actually made by the infringer as a result of its infringement and the profit it would have made had it adopted a non-infringing alternative.” The only problem, as I see it, is that these two principles are contradictory. As I wrote in 2013 (see my book, pp. 199-200):
. . . suppose that Firm B earns a profit of $10,000 from sales of a gizmo that comprises ten components, one of which (Component #5) infringes Firm A’s patent. The issue for the court is to determine how much of the $10,000 profit is attributable to Component #5. The correct economic answer to this question is the difference between the $10,000 profit Firm B actually earned and the profit Firm B would have earned had it not infringed. . . . [F]rom an economic perspective the correct way to go about estimating the profit Firm B would have earned had it not infringed would be to estimate Firm B’s profit had it resorted to the next-best alternative to Component #5—call it “Component #5A.” . . .
Seen in this light, it is not clear how a court could both apportion the contribution of the infringing feature to a firm’s profits and ignore the existence of noninfringing alternatives. To ignore the existence of noninfringing alternatives in effect means that the court assumes, in the example above, that Component #5 is irreplaceable. If Component #5 were both irreplaceable and necessary to the functioning of gizmos, however, then absent the infringement Firm B would have sold no gizmos and rightly should disgorge the entire $10,000 profit; there should be no apportionment. . . . If, on the other hand, a gizmo that simply lacks Component #5 is both functional and marketable, one could try to estimate the profit Firm B would have earned from selling such gizmos and subtract that figure from $10,000. In principle, though, this simply means that the court is assuming that a gizmo without any counterpart to Component #5—rather than a gizmo containing Component #5A—is the next-best alternative to an end product including Component #5. Properly understood, then, apportionment and the relevance of noninfringing alternatives are inseparable concepts. Any attempt to apportion profits by, say, some estimate of the feature’s relative importance to the marketability of the end product, will necessarily imply something about the next-best noninfringing alternative, whether the analyst is aware of this implication or not.
One caveat I would make to the above today is that in some cases multiple components may all be essential to sales of the end product. In cases involving standard-essential patents, for example, the end product may be unmarketable unless it incorporates all of the relevant SEPs, but you wouldn’t say that all of the profit the defendant earned from the sale of the end product is attributable to any one single SEP; and of course the royalty due for the infringement of any one SEP should be only a fraction of that profit, though determining exactly what that fraction is can be a difficult undertaking. See, e.g., Norman V. Siebrasse & Thomas F. Cotter, The Value of the Standard, 101 Minn. L. Rev. 1159, 1198-1228 (2017). In any event, if I’m understanding the facts of the present case correctly, the court actually did assume that, but for the infringement, the defendant would have sold a (slightly smaller) quantity of products incorporating some (not specifically identified) noninfringing Component #5A; and the outcome overall, assuming the correctness of the court’s factual analysis, is consistent with my view that you can’t do apportionment without, explicitly or implicitly, taking noninfringing alternatives into account. With that, I wholeheartedly agree, though it would be nice if we didn’t have to go through so many unnecessary doctrinal hurdles to reach that result.