Tetricide – Struggling to Define Copyright that is not Trademark
Joshua A. Cottle, Texas Review of Entertainment and Sports Law
Introduction: What is Tetris?
Any videogame player worth his salt is familiar with the classic game Tetris and has likely played some version of the basic game, whether on a cell phone or iOS or the Wii downloadable content or wherever a download of the classic game can be had. If you were asked in a bar to describe the basics of the game, you might have a hard time articulating them, simple as they seem to be, but your description might go a little like so: “an empty grid of so many squares fills gradually with a randomized series of four-block clusters, which descend from the top of the grid towards the bottom, where they begin to stack on top of each other to the top of the grid, are rotated by the player in order to form lines across each side of the grid, which formation eliminates that row and buys the player more time to stack more blocks until one edge of a block finally touches the top of the grid, at which point the game ends.”
This is a simple game. We have as elements two-dimensional blocks (simple geometric figures), a grid of so many squares, a falling motion, a terminus, and two mechanisms of control—movement and rotation. The question a would-be cloner would ask himself, as Tetris and its variants have proven popular and timeless since its introduction 20 years ago, is “How can I copy the rules of the game without copying the expression of the game?” That is, how could an individual copy Tetris without violating Tetris’ copyright?
An answer depends on how one defines the meaning of Tetris—what is its nature, character, or identity, legally and otherwise? This requires both a metaphysical inquiry of a sort usually beyond lawyers and judges, and an inquiry into the type of design the law should seek to protect in the copyrightable audiovisual expressions of videogames. However, these questions of meaning formed the basis for a battle waged between an alleged infringer and Tetris Holding, which was resolved in 2012. The court’s answer in this case attempted to balance justice against rip-off clones while preserving the development of spiritual descendants rooted in the notional ancestor. The answer does not particularly impress me, despite the court’s extension of the metaphysical inquiry into identity of a game.
The case is Tetris Holding, LLC v. Xio Interactive, Inc., 863 F. Supp. 2d 394 (D.N.J. 2012). Relevant is this fact of ownership: “Tetris Holding, LLC owns the copyrights to the visual expression of the numerous Tetris iterations and licenses those rights to Tetris Company, LLC, which then sublicenses its rights.” Id. at 396. That is, Tetris as a copyright is a visual copyright only. What does this mean in videogame parlance? Practically everything: the entirety of a videogame is conveyed audiovisually—musically, narratively, textually, and actively. It means that Tetris Holding could not bring a code copyright litigation, but any outright ripoff of the code would depict the game in the same way, so without actual copying of the code, Tetris Holding could only rely on a substantial similarity copyright litigation to defend its property.
Xio Interactive claimed to have used Tetris in order to develop an Apple iOS game similar to the spirit of Tetris in order to make brisk sales. The court was unconvinced from the record that mere inspiration was at work with Xio’s adaptation of Tetris: “Xio was more than inspired by Tetris as Xio readily admits that its game was copied from Tetris and was intended to be its version of Tetris.” Id. at 397. Xio’s game was titled Mino (after domino), and the Tetris pieces are called tetrominos, a historical factoid mentioned in the case. Xio’s basic defense was that it did not copy any copyrighted or copyrightable element of Tetris. Tetris Holding sent Apple a take-down notice against Mino, and Apple initially complied with the request, but following receipt of a counter-notification filed by Xio, told Tetris Holding it would reinstate the game in ten days pending Tetris’ litigation. This case illustrates a weakness in the Digital Millennium Copyright Act: copyright owners send takedown notices to a platform distributor like Apple, which responds by shooing the product off the market; unless the copyright owner is willing to litigate its claim practically immediately, all the individual on notice must do is file a counter-notification to the distributor in order to be reinstated in the market unless the owner files suit. So DMCA offers some bark, but the bite comes from the pocket of the litigant. Large IP owners like Tetris Holding, which sink a lot of costs into aggressively policing its IP for lookalikes, are game for such litigation, and were so in this case, but note the limitations that might exist for smaller IP owners like the typical indie developer.
Both parties moved for summary judgment, and neither contested any material facts, so the determination of the winner would depend on which had legal claim. Xio’s defense, as stated, was that it did not copy a copyrightable element despite its blatant copying of the design of Tetris. Xio conducted legal research before pushing through its development of the product, and rooted its argument in a wonderfully thorny argument of the fundamental nature of games. Xio “argue[d] the elements it copied were not original expression, because they were part of the game itself—the rules, function, and expression essential to the game play—which is not protected.” Id. at 399.
It is a commonplace to copyright law that under 17 U.S.C. § 102(a), computer programs are protectable both in terms of graphics (expression) and code. Apple Computer, Inc. v. Franklin Computer Corp., 714 F.2d 1240, 1249 (3d Cir. 1983). It is also a commonplace that copyright law does not protect ideas—§§ 102(a) and (b) form what is called the idea-expression dichotomy. Alas, this commonplace is difficult to put into practice, especially in the program context. See Whelan Assoc., Inc. v. Jaslow Dental Laboratory, Inc., 797 F.2d 1222 (3d Cir. 1986). Different circuits design different ways of dealing with the dichotomy problem in the programming context. Some courts are loathe to undertake the Jacobean struggle on metaphysical terms. See Computer Assocs. Int’l v. Altai, 982 F.2d 693, 705 (2d Cir. 1992). The Third Circuit’s test in Whelan essentially requires a distinction be drawn between the “purpose or function of a utilitarian work . . . and everything that is not necessary to that purpose or function.” Whelan, 797 F.2d at 1236 (the first being the work’s idea, the second being the work’s expression). The Second Circuit designed an abstraction-filtration-comparison test meant to separate itself from the Whelan test by which the non-literal elements of computer programs were compared for substantial similarity infringement. Altai is not much different in theory from Whelan, and so thought the Tetris judge. Under Altai, the program is abstracted into its basic components, after which a conceptual filter is applied to sift out the unprotected material, leaving a nugget of protected material to the copied work, left intact and unabstracted—substantial similarity between this nugget and the protected work would lead to a finding of infringement.
Alas, Altai is not out of the metaphysical thicket. The first step requires a great parsing of the “principles or ideas driving the program and the essential processes and functions by which it achieves those purposes.” Tetris Holding, 863 F. Supp. 2d at 402. At some point in this abstraction, protection is lost, or else the author would be given protection for ideas, illegal (and therefore, legally-metaphysically impossible) under the Copyright Act. Note that the Whelan opinion judiciously refused to suggest that only one idea could be derived from the underlying work, or that the idea underlying the work will be accomplished through program design such that “structure and organization will therefore always be part of the expression of such works.” Whelan, 797 F.2d at 1238, n.34. The New Jersey court expresses allegiance to Whelan but no significant theoretical split from Altai, and proceeded to try and define the protectable expressions of Tetris, if any existed.
The court explored merger, of course, the doctrine by which an expression and an unprotectable idea are so united as to be inseparable and thus entirely unprotectable. As an anti-monopoly device, merger is appropriate to apply when there is no way or only a very few ways to express an idea. The court also considered whether the doctrine of scènes à faire, by which a particular literary work must by convention follow certain tropes of expression or not properly be that kind of literary work, also applied. Of course, this doctrine must be carefully tailored, or else videogames would never exist.
The court was unpersuaded by Xio’s primary argument, which essentially argued that it had not infringed on copyrightable expressions because the “ideas of Tetris (or the rules of the game) [are] not protectible, [and] neither are the ‘functional aspects’ of the game or expressive elements related to the game’s function or play [sic].” Tetris Holding¸863 F. Supp. 2d at 404. The court concluded that Xio thus conflated merger and scènes à faire in arguing Tetris Holding could not copyright “expression inseparable from either game rules or game function.” Id. The court held that although copyright and patent grant monopolies to different areas of intellectual property, it was not convinced that “any and all expression related to a game rule or game function is unprotectible [sic].” Id. at 404–05.
Xio’s reliance on different copyright laws, including cases involving videogame characters, was determined to be unavailing to their argument. They cited Atari, Inc. v. N. Am. Philips Consumer Elects. Corp., 672 F.2d 607 (7th Cir. 1982) to argue that “functional commonalities” between Mino and Tetris should not be held against them. The Seventh Circuit there held that while some elements of Pac-Man were not copyrightable, it was because of scènes à faire, not commonality, and found protection for the functionality of the main characters in Pac-Man without finding protection for elements like the maze design, the scoring table, or the dotted lines. No comment was offered in Atari about whether these unprotectable elements were beyond the reach of copyright law because of rule-integrality, as Xio sought to argue. The protection offered for the characters is not surprising, and seems to be why the New Jersey court found protection for Tetris: where copying is admittedly (or otherwise) wholesale of some elements and the function of the ripoff remains substantially similar to the original, and the elements copied are expressed in a non-mandatory way, infringement will be found. See Incredible Techs., Inc. v. Virtual Techs., Inc., 400 F.3d 1007 (7th Cir. 2005) (golf games cannot be presented in totally fanciful ways because otherwise the game depicted would not be golf). The New Jersey court refused to find a scènes à faire defense to the settings elements, as the Seventh Circuit did for Pac-Man, because Tetris was wholly fanciful. How Pac-Man is not wholly fanciful as a maze game and Tetris is wholly fanciful as a blocks game is not clear, and can be expected to be further explored by the courts, but it pays to recall the sine qua non of substantial similarity disputes: “How much of my work is similar to a predecessor’s work?” Note well: courts do not care about plagiarists’ dissimilarities in determining infringement, though they do in the division of profits for mixed-copyright works. See Sheldon v. Metro-Goldwyn Pictures Corp., 81 F.2d 49 (2d Cir. 1936); Callaghan v. Myers, 128 U.S. 617 (1888).
In reviewing the weakness of another case used by Xio in support of its position, Midway Mfg. Co. v. Bandai-America, Inc., 546 F. Supp. 125, 152 (D.N.J. 1982), the court indicated its allegiance to the legal opinion therein, by which the court articulated an extension of “look and feel” trade dress theory to substantial similarity copyright infringement. Two games, Galaxian and Space Invaders, were compared for idea and expression, and the high ideas of both games—players defending a virtual base or ship from oncoming aliens—were unprotectable, but “under the umbrella of expression [were] the look and feel of the characters and how they move and act.” Tetris Holding, 863 F. Supp. 2d at 407.
The Tetris court’s opinion gets a little strange where it begins to attack Xio’s legal conclusions, which posits that if a videogame feature is “dictated by functional considerations, regardless of whether there may be a number of different ways to implement that feature’s functionality, copyright does not protect that feature.” Id. at 408. This certainly does sound like a dangerous extension of merger, which is rooted in the inexpressibility of a feature outside of a single way. Xio seems to be arguing that merger applies as well where the feature is inexpressible outside of a cluster of ways—for instance, where does Tetris’ four-block series become substantially different from a five-block series, or if the motion of the pieces is inverted, or the grid dimensions change from 10 x 20 to 10 x 25? This is where the query into the identity underlying Tetris gets interesting. The court found that since Xio did “not dispute that Tetris is a purely fanciful game, meaning it has no grounding in the real world, unlike a video game simulating a karate match or a golf game,” and notwithstanding the fact that Tetris had been released to the public domain, Xio could not therefore avail itself of a scènes à faire argument, because of a lack of “expressive elements ‘standard, stock, or common’ to a unique puzzle game that is divorced from any real-world representation.” Id. Phrased differently elsewhere in the opinion, scènes à faire seems to require for this court “stock or common imagery that must be included” in the piece in order to mount a successful defense, but this is not the most general expression of the notion of scènes à faire, which are “those expressions that are standard, stock, or common to a particular topic, or that necessarily follow from a common theme or setting.” Id. at 412; see also, e.g., Gates Rubber Co. v. Bando Che. Indus., Ltd., 387 F.3d 522, 535 (6th Cir. 2004). So, the New Jersey court appears to have unnecessarily limited scènes à faire as a defense to Xio. Even if the court did not limit the defendant to a world of games where “stock or common imagery” must be included to recreate the look and feel of that particular type of game, the court has temporally removed companies like Xio from a successful defense based on scènes à faire, assuming that not yet has the mass of world history availed to the enterprising legalistic mind certain guiding literary elements by which that mind could raise an effective scènes à faire defense to a game like Tetris—in other words, Tetris was born too early for Xio. The fact that Tetris has been cloned many times in different forms, and judgments have been upheld in favor of such clones because of play differences, such as in Dr. Mario, did not seem to matter to the court for Xio’s scènes à faire defense. One hopes Xio puts this unnecessary limitation on audiovisual scènes à faire into its appeal, though it may be a weak defense, as a proper scènes à faire defense is usually required to antedate the copyrighted work itself. See Krista S. Schwartz et al, Software and scènes à faire, Copyright World 190 (May 2009), at 11, available at http://tinyurl.com/b5pp5x4.
Takeaways: Define the applicable filter of abstraction by changing the comparability of elements
The defensibility of a copycat’s recreation of a notional ancestor like Tetris depends on how abstract a level of definition is given to the game by the determinant of a game’s nature, which is up to lawyers and ultimately a judge, as are all things litigated. The New Jersey court said that “[t]he idea of Tetris does not necessitate the particular characteristics of the audio-visual display.” If Tetris were more specifically articulated than the high-abstraction version the court provides, then it might indeed require these particular characteristics, and so merger or scènes à faire would be available as defenses to defendants. If a game’s rules are phrased in a certain way, then the ways by which these rules are expressed find different limitations. A game like Tetris may therefore turn on its particular expressions, so games copying those expressions would not infringe. This seems dangerous for many businesses developing videogames, which may not be able to afford having their games’ definitions locked in at too specific a level. Crash Bandicoot and God of War, for instance, are both platformers, but no one would confuse the two games. In order to distinguish the games, other elements of expression must be accounted for adjudication (use of a map, for instance; the protagonist; how points are accrued; the storyline, etc.). But if one wanted to create something like Crash Bandicoot that was not quite Crash Bandicoot, some indispensable elements would exist: gems, breakable boxes, a supernatural sidekick, forward-linear gameplay, etc. Designers must take into account how they must express the relationship between the constitutive elements of any particular genre and any particular example in the genre in attempting to define a position in the videogame world with strong roots in a notional ancestor. Dark Sector may have played similarly to Gears of War or Uncharted or even Resident Evil 4, but the games occupied different worlds with different design elements besides the over-the-shoulder camera. In worldless puzzle games like Tetris and Mino, which follow similar rules of play, visual expression of the rules of the game become the key differentiation on which designers should focus. Ultimately, it should be recalled that soccer and football are both sports involving fields with goals at each end and a ball in play, but how the ball moves and how points are tallied and the rules of motion across the field are completely different, thus leading to different games.
Defining a game is difficult, and a judge’s definition is apt to be one defined primarily by commercial proprietarian thought and designer expectation. If Tetris Holding is any indication, courts will be unfriendly to copycat attempts to scale down the definition of abstraction to a more specific level, whereby the rules become the elements of the game. Any judicial definition of a game is likely to differ from a game critic’s or gameplayer’s own, so it is important to survey case law to see how specific of abstract the judiciary’s definitions of games tend to be in order to expect the kind of design distinctions a copycatter must make in its clones to avoid a finding of infringement. The best key may be to avoid having gross similarities, which might then permit the judge to scale the “idea” of the game to a more specific level, before then determining whether the parts in play fit under the idea or exist outside the idea. (Judges will say they determine infringement the other way: by first defining the rule and then finding whether the elements fit the rule, but sensibility should dictate design with an eye to legality: determining whether a dispute even exists in the first part outside of the facts, which are the comparable elements, cannot happen. Even motions to dismiss require more than legal pleadings.)
The New Jersey court chose to provide a high-level definition of Tetris similar in height to the one provided for Space Invaders and Galaxian in the Midway case, and then compared features of both games “‘as they would appear to a layman’”, concentrating “‘upon the gross features rather than an examination of minutiae.’” Universal Athletic Sales Co. v. Salkeld, 511 F.2d 904, 908 (3d Cir. 1975). Since Tetris is a theoretically simple game, the lay gamer might be the exact person for whom a substantial similarity test is intended. Considering how popular the game is, with Tetris Holding hitting over 100M cell phone sales recently, this seems valid. The Third Circuit, which oversees the District of New Jersey, follows the Second Circuit’s substantial similarity test closely. See Dam Things from Denmark v. Russ Berrie & Co., 290 F.3d 548, 562 (3d Cir. 2002). The Second Circuit’s approach in a case like this, where plaintiffs’ work is wholly original, tests substantial similarity against an ordinary consumer’s scrutiny. See Boisson v. Banian, Ltd., 27 F.3d 262, 271–72 (2d Cir. 2001). This scrutiny is essentially pretended to be a casual once-over of the audiovisual elements of both games depicted. The court here determined substantial similarity through a side-by-side comparison of the products, in keeping with precedent. See Value Grp., Inc. v. Mendham Lake Estates, L.P., 800 F. Supp. 1228, 1233 (D.N.J. 1992). This comparison was verbalized by the court in terms of a “squint” test, whereby if a common user has to “squint to find distinctions only at a granular level, then the works are likely to be substantially similar.” In other words, where the “look and feel” of a game is substantially similar to a copyrighted predecessor, infringement is likely to be found. Note that this “look and feel” test is a unique extension of Whelan and Altai, which originated the “look and feel” tests and dealt with software programming, rather than the audiovisual expression of the underlying code, as in Tetris Holding. The cases will be more complicated the less obvious the clone is in its use of precedent materials, and more heated the more specific the abstraction of a given game is. If Xio had changed up some features of Mino so it did not look so much like Tetris, it could have escaped infringement findings. But the judge seemed to have predetermined Xio’s liability by setting the definition of Tetris at so high a level of abstraction—if the abstractions are comparative or equal in their terms, then the literal elements underlying the idea’s expression will also be more strictly enforceable. Copycat disputes over notional ancestry in the Third Circuit, and likely in every other Circuit, will turn on challenges to the level of abstraction chosen under Whelan or Altai.