Innovation Patent Claims Found Invalid in Dynamite Decision

Dynamite Games Pty Limited v Aruze Gaming Australia Pty Limited [2013] FCA 163 (4 March 2013)

Australia’s innovation patent is widely considered to have an unreasonably low threshold of validity – so much so, in fact, that last year IP Australia put forward a proposal to eliminate the ‘innovative step’ standard entirely, and require that innovation patent claims meet the same test for ‘inventive step’ as a standard patent.  You could therefore be forgiven for thinking that pigs might sprout wings and take to the air before an innovation patent would be found invalid on grounds of lack of innovative step.

Yet – until and unless there is a successful appeal – you may need to keep your eyes peeled for porcine aviators, because a judge in the Federal Court of Australia has done just that, striking down all asserted claims of an innovation patent held by Dynamite Games Pty Ltd (‘Dynamite’) for failing to meet the innovative step threshold.  Dynamite had sued Aruze Gaming Australia Pty Ltd (‘Aruze’) for infringement of various claims of a standard patent relating to gaming machines, along with a number of claims of a related innovation patent.

Aruze, in turn, sought revocation of the patents on numerous grounds available under the Patents Act 1990 – that none of the claims is for a manner of manufacture; that the claimed invention is not novel; that the claimed invention does not involve an inventive step or an innovative step; that the claims are not clear and succinct; and that the claims are not fairly based on the matter described in the body of the respective specifications.

The decision is interesting for a number of reasons, starting with the fact that it is a judgment of Justice Emmett, the same Federal Court judge who recently found the ‘business method’ claims of a patent application filed by Research Affiliates, Inc, to be invalid on the basis that they were not patent-eligible under the ‘manner of manufacture’ test.  Justice Emmett considered all of the asserted grounds of invalidity of Dynamite’s patent claims, but declined to formally rule on any of the others after first finding the standard patent claims to be obvious, and the innovation patent claims to lack an innovative step.  However, he did indicate that, had he needed to do so, he would have been inclined also to rule against Dynamite on the grounds of lack of novelty and lack of ‘manner of manufacture’.  In doing so, he reprised some of the themes of his Research Affiliates decision.

Justice Emmett also declined to formally rule on the allegations of infringement, on the basis that he had found all of the asserted claims to be invalid.  Again, though, he indicated that he would have been inclined to rule in favour of Aruze, had he been required to do so.

I cannot help but feel that if I were a patent holder or applicant, with a matter relating to my claims coming before the Federal Court of Australia, I would not want to see Justice Emmett on the bench.  He does not appear to be what one would call a ‘pro-patent’ judge, and he does not seem afraid to adopt an ‘idiosyncratic’ approach to precedent in order to arrive at what he no doubt believes to be the ‘right’ result.

Dynamite’s Patents

Dynamite is the owner of Australian standard patent no. 2004281073 (‘the standard patent’ and Australian innovation patent no. 2009101338 (‘the innovation patent’).  The innovation patent is a divisional of the standard patent, and both relate to 'gaming apparatus and systems'.  More particularly, the patents disclose networked electronic gaming machines, of a type familiar in casinos, clubs and bars.

The aim of the arrangement described and claimed in the patent specifications is to increase interest in a game by guaranteeing that a particular game event – such as a win, or some other feature event – will occur within a predetermined period of game play.  In essence, the system monitors the game, and keeps track of the elapsed period (which may be based on number of games played, number of bets placed, or some other measure), and if the game event does not occur by chance within the predetermined period, the system ‘makes’ it happen.

According to embodiments described in the specification, the amount of further game play required to trigger the guaranteed event reduces, the longer the player continues with the game.

Importantly, the system includes an ‘indicator component’.  This is some form of visual indication, such as a counter, a gauge, a pointer or similar, which allows the player to see how long it will be until the guaranteed event must occur (of course, it could occur earlier, if the player is lucky).

The system is not said to be limited to any one particular game.  Rather, it may be employed in any suitable game, i.e. any game in which events occur which could be monitored and triggered if they have not occurred ‘naturally’ after a predetermined amount of game play.

The Court’s Approach to the Patents

The court appears to have been especially concerned with the inherent breadth of Dynamite’s patent claims.  In describing the content of the specification, the decision highlights certain facts in bold: the applicability of the claimed invention to ‘any suitable game’ executing on ‘any suitable type of gaming machine or platform; and that the play parameters can be chosen ‘in any desired manner’.

As in his recent decision in the Research Affiliates case, Justice Emmett seems greatly concerned about lack of detail in the description of 'how' to do certain things.  In Dynamite’s case, he assumes (quite reasonably) that where the specification does not disclose the necessary implementation details, this must be because they form part of the common general knowledge (CGK) possessed by the ordinary skilled worker in the gaming industry (see, e.g.,[115,116]).

This issue is taken up again, at [169-172], in relation to manner of manufacture.  The court’s objection here appears to be that, even though invention may reside in an idea, so long as it its reduced to a practical form, an alleged invention may be no more than a ‘concept’ if all of the skills required to implement it form part of the CGK.

Significance of Patents in the Gaming Industry

The gaming industry makes extensive use of the patent system.  New gaming ideas trend to look simple, especially with the benefit of hindsight, however they must be carefully designed and mathematically modelled in order to ensure that they are profitable over time.  The evidence of one of the expert witnesses called by Aruze is summarised as follows (at [44]).

…the first step in designing a game is to have a creative idea, such as a mathematical gaming feature or game functionality, or perhaps an idea for a theme. Once the game designer has the initial idea, the game designer works on the mathematical model of the game. That involves development of the specific rules of the game, that is, deciding how the game is going to play and function. Once the rules of the game have been decided, the probability calculations of each winning combination applicable in the game, that is, the win frequency, need to be worked out. Where gaming machines are operated in a regulated environment, the probability calculations are restricted as each game must provide a specific minimum theoretical return to the player. The play table is then developed. That is dependent on the game probability calculations and is also directly influenced by any required theoretical return to the player in a regulated environment.

Similarly, the court summarised the evidence of one of Dynamite's witnesses in the following manner (at [50]).

…what may seem like small variations in game design can be critical factors to the success of a game design. Creating a new feature or introducing a variation to an existing feature type, which distinguishes one game design from others in the market, is difficult, and it is the aspiration of all game designers to achieve that and for their game design to create the desired following in the market.

The aim of developing new games and features is thus to make gaming machines more interesting to players, so that they firstly choose the machine with the new features over other, competing, machines (see, e.g., the decision at [106]), and secondly that they them go on to play for longer than they otherwise would.  Successful new game features are therefore potentially very valuable, no matter how simple and easily copied they may be with the benefit poof hindsight.  This probably explains the extent of use of patents by major players inn the gaming industry.

Attacks on Validity by Aruze

Aruze relied upon four patents, and four prior gaming machines, in an attack on the novelty of Dynamite's claims.  The same prior art formed the basis for attacks on inventive/innovative step, as well as having a bearing on the ‘manner of manufecture’ issue. 

It is usual for a court to consider novelty first, before moving on to look at inventive step, or innovative step.  This is natural, because the assessment of inventive or innovative step starts by identifying the differences between the claimed invention and the prior art, in order to evaluate whether this difference in inventive/innovative.  Clearly, if there is no difference (i.e. if the claims are not novel), then there is no need to proceed further.  On the other hand, if there is a difference, this will usually have been clarified by the novelty analysis.

It is therefore somewhat unusual that Justice Emmett elected to dive straight in with inventive and innovative step, leaving the ‘easier’ analysis of novelty until later.  His approach to the individual grounds could also be described as idiosyncratic – possibly appealably so!

Lack of Innovative Step

The court’s rulings on lack of inventive step in the standard patent, and lack of innovative step in the innovation patent, are difficult to separate.  The approach set out in each case – subsections 7(2) and (3) for inventive step, and subsections 7(4) to (6) for innovative step – is quite different.  Yet, having initially reviewed these provisions (at [92] and [95]-[96]), Justice Emmett appears to evaluate both based on a singe approach to the CGK in the gaming industry.

The patents cited by Aruze were not considered to be available for assessment of inventive step because the court was not persuaded that game designers routinely studied patents (at [113]).

However, the prior gaming machines – which in some cases embodied inventions disclosed in the prior patents – were considered to form part of the CGK.  On this basis, the court considered each claim of the standard patent (at [114]-[144]), and found all to be obvious.

The court then turned to the question of innovative step.  As noted above, the required approach is quite different from assessing inventive step.  As noted above, the process is mandated by subsections 7(4) to (6) of the Patents Act 1990, and requires that the presence of an innovative step be determined with regard to each individual available item of prior art information considered separately.  If there is a difference, then the question to be asked is whether this difference in an immaterial one, which makes ‘no substantial contribution to the working of the invention’.

Justice Emmett simply does not do this in respect of Dynamite’s claims.  Here is his analysis of claim 1 (at [145], with similar findings for claims 2-5 in the following paragraphs):

Claim 1 is for an electronic gaming machine in which a bet is placed, game events occur and wins are returned, and gameplay and game events are monitored, and includes a guarantee feature, wherein if a monitored game event does not occur within a predetermined amount of gameplay, a guaranteed game event is triggered. Claim 1 does not appear to vary from the prior art. If there is any variation, it does not appear to make any substantial contribution to the working of the invention. There is no innovative step.

If claim 1 ‘does not appear to vary from the prior art’, which specific item of prior art information is the court referring to here?  At this point, we have just finished reading a lengthy analysis of inventive step, in which features of a number of different gaming machines, all commonly known to game designers, are considered together.  It is not at all clear what an ‘appearance’ of ‘variation’ might be, and this certainly does not form part of the analysis mandated by subsections 7(4) to (6)!

Furthermore, it is not clear what it would mean for this (nonexistent) variation not to ‘appear to make any substantial contribution to the working of the invention’.  Without identifying a specific difference between the express terms of the claim, and a single source of prior art information (e.g. one patent, or one gaming machine), and then assessing exactly what function, if any, this difference performs, it is simply not possible to apply the innovative step test.

The fact that similar claims were already found to be obvious is not to the point.  As the court said in Delnorth Pty Ltd v Dura-Post (Aust) Pty Ltd [2008] FCA 1225 (upheld on appeal to the Full Court, Dura-Post (Aust) Pty Limited v Delnorth Pty Limited [2009] FCAFC 81):

The idea behind [innovative step] seems to be that a claim which avoids a finding of no novelty because of an integer which makes no substantial contribution to the working of the claimed invention should not receive protection but that, where the point of differentiation does contribute to the working of the invention, then it is entitled to protection, whether or not (even if), it is obvious. Indeed, the proper consideration of s 7(4) is liable to be impeded by traditional thinking about obviousness.

Lack of Novelty?

Had the court conducted a formal novelty inquiry first, breaking down each claim into its separate features, and determining whether or not each of those features in present in each individual source of prior art information, then the innovative step assessment should have been straightforward.

Both parties provided extensive tables comparing the features of the claims with the prior art.  However, in the event, Justice Emmett did not use the analysis in the tables, stating (at [185]) that:

… I would have been disposed to conclude that Aruze has established that the claims relied on by Dynamite are anticipated by the prior art relied on by Aruze. However, having regard to the conclusion that I have reached concerning the lack of an inventive step and innovative step in relation to the Patents, it is not necessary to address the very detailed analysis contained in the tables.

Manner of Manufacture

Justice Emmett provided an extensive discussion of ‘manner of manufacture’ (i.e. patent-eligibility), mostly reprised from his recent decision in Research Affiliates (at [151]-[166]).

The court summarises Aruze’s argument (at [169]) that neither specification:

… teaches the relevant skilled addressee how to implement the claimed invention. In modern electronic gaming machines, features of the claimed type are typically implemented in the software that runs the games. The skilled addressee must be assumed, as a matter of common general knowledge, to be able to, or to have access to programmers who can, write source code to give effect to such features. The claimed invention is essentially no more than a concept. That is significant in the context of the manner of manufacture, as well as lack of innovative step as grounds of invalidity.

There are clear echoes here of the Research Affiliates decision. where an inadequate description of how the claimed invention was implemented contributed to a finding of lack of manner of manufacture.

In my view, this is incorrect.  Adequacy of the description is required under section 40 of the Patents Act 1990, however a claim can clearly be for a manner of manufacture, even if the description is inadequate, or is absent altogether.  A claim directed to a ball-point pen, for example, is plainly for a manner of manufacture, even if it is not novel, and if the specification describes a pencil!

The line of reasoning adopted in this case is concerning in relation to all computer-implemented inventions.  There is a suggestion here that even if the specification teaches the skilled person how to implement the invention (because the necessary programming skills are part of their stock-in-trade), the claims might be found ineligible for patenting because the field is such that a relatively ‘conceptual’ description provides the necessary teaching.

As the court put it – invoking the ‘Microcell principle’ – at [171]:

Rather, on one view, the Patents give no indication as to how the idea is to be implemented. An absence of any teaching as to how to implement the claimed invention might suggest that the claims of the Patents merely describe the use of known products, being gaming machines and their components, gaming systems and gaming software, for which their known properties make them suitable, namely, implementing the rules of the game.

It seems unlikely that ‘programmability’ of a computer was the kind of ‘known property’ of a product that the High Court had in mind when the Microcell case was decided (under the 1903 Patents Act) in 1959.  Furthermore, this reasoning suggests that if the specification had provided further implementation details – even if these were completely within the capabilities of the skilled person to supply for themselves – either:

1. exactly the same claimed invention would magically have become patent-eligible; or
2. the invention would remain ineligible, because the inventor had provided no more than a ‘concept’ along with a description of how it may be put into practice that supplies nothing more than information already known to the skilled person.

With respect, both of these outcomes would be absurd.  It is useful to view this in terms of the US approach to invention, in which the process is divided into ‘conception’ (a word which, notably, shares linguistic roots with ‘concept’) and ‘reduction to practice’.  The inventive step may reside wholly in the conception stage – indeed, with diligent reduction to practice this is considered to be the true date of invention.  The situation is not so different in Australia.  As Justice Emmett points out himself, at paragraph [166]:

Distinctions between the idea or concept or principle informing an invention, on the one hand, and the means of carrying it out or embodying it in a manner of new manufacture, on the other, have long been recognised. Thus, invention may lie in the idea, it may lie in the way it is carried out, or it may lie in the combination of the two. An idea of itself cannot be patented. An idea that is part of an inventive step must end in a new method of manufacture. When an idea is incorporated into means for carrying out the idea, the idea itself can be taken into account when considering validity. Inventiveness may repose largely in the idea. It is almost inevitable that the subject matter of an invention that involves an improvement to a known combination will be described as an idea or as a concept.

So where is the problem here?

Infringement

Again, having declared the asserted claims invalid for lack of inventive/innovative step, it was not strictly necessary for the court to decide the question of infringement.

Justice Emmett nonetheless considered the issue, indicating that he was inclined to a finding of no infringement, due to effectively construing 'indication' to mean an 'accurate indication' (e.g. at [202], [219]).  To my mind, this is a questionable conclusion.  An ‘indication’, in its ordinary meaning, need not be a precise measure.

Conclusion

There is a lot in this decision to question, and to be concerned about, not least the apparent inclination of the court to find against the patentee at almost every turn of the case.

There are points on which almost every aspect of the decision might be challenged on appeal.  The court’s application of ‘obvious to try’, in relation to inventive step (which I have not discussed in detail here) is questionable.  It is unclear that the correct legal approach has been applied in relation to innovative step.  The novelty analysis seems incomplete, despite the provision by the parties of extensive claim charts.  Indeed, if the claims are not novel, then it would have been unnecessary to consider inventive/innovative step.  The comments in relation to ‘manner of manufacture’ are very concerning, and not only for computer-implemented inventions.  And there is certainly potential for a challenge to the construction of the term ‘indication’ in relation to infringement.

There is, as I understand it, a lot of money to be made in the gaming industry.  This seems like a decision that could well be appealed.  If so, this could be a busy year in patent law for the Full Bench of the Federal Court!

Tags: Australia, Infringement, Innovation patent, Litigation, Novelty, Obviousness, Patentable subject matter, Revocation