Substantial similarity, an analysis of the similarity between two works, is the fulcrum of copyright infringement. Recent cases involving Led Zeppelin’s signature song “Stairway to Heaven,” the award-winning movie “The Shape of Water,” and Google and Oracle’s dispute over computer code all require courts to grapple with this fundamental question. This Article reveals that copyright plaintiffs today have only a one in ten chance of winning – the worst in a century, as well as the cause - defendants’ devastatingly effective use of pretrial motions and the rise of lawsuits against non-rival defendants. Scholarly debates on substantial similarity typically revolve around the works of authorship at issue, circuit splits, and leading cases. These discussions entirely miss the unprecedented and unreported demise of substantial similarity. Rehabilitating the doctrine will require a fresh look at how we judge substantial similarity. Judging when defendants appropriated too much can be an inherently opaque and subjective enterprise. To many courts and scholars, substantial similarity is “bizarre,” “ad hoc,” and “a virtual black hole in copyright jurisprudence.” Every creative work borrows some inspiration from other works, whether copyrighted or not. If left theoretically unaddressed, the law will continue adrift without compass or rudder, inhibiting the growth of creative industries. Helping plaintiffs win fairly and viably will be critical in fostering diverse, innovative ecosystems like architecture, literature, movies, and software. For this to happen, three things must change. First, courts must judge works holistically from the vantage of an ordinary observer, and once plaintiffs have proven facial similarity, defendants should then explain why they are not infringing. Avoiding elemental dissections of the copyrighted work makes substantial similarity easier for factfinders to administer and less likely to penalize plaintiffs whose works invariably incorporate some unprotectable elements. Second, jurors must be the default arbiters of substantial similarity. The layperson, and not experts, are copyright’s principal audience. Third, requiring copyright registrants to identify the expressive elements of the works in their registrations will help clarify what they seek to protect and provide better notice to defendants. Beyond copyright, the observations and insights in this Article map to broader scholarly debates on rules versus standards, the centrality of jury trials, and structural biases within the American justice system