Nintendo’s summoning patent: Windows Central, PC Gamer got right what The Verge and a law prof got wrong — and let’s look at a 50-year-old game

There has been a lot of discussion in games-specialized media and the gamer community about it. PC Gamer quoted a videogame IP lawyer from a U.S. firm that has litigated hundreds of patent cases (including some rather significant ones) calling it “an embarrassing failure of the US patent system.” But a tech news website named The Verge published a fundamentally misguided article (“No, Nintendo and Pokémon did not patent ‘summoning charaters and making them battle’”), and one of their reporters, Tom Warren, shared the article on X with the claim that he had “seen plenty of poor reporting on this over the past week” — when The Verge‘s own article is a textbook example of poor reporting.

In the first part, this article addresses The Verge‘s errors. In the second part, we’ll show that Nintendo’s patent description essentially uses (without explicitly saying so) a 50-year-old arcade game for the starting point, which is outrageous in and of itself.

The Verge‘s three biggest mistakes

1. “Complicated” and “more specific”

They say the patent is “complicated” and “more specific” than summoning and battle mechanics. But as we’ll discuss further below, there’s nothing complicated if one focuses on the practical effect of claim 1, the broadest claim of that patent. One can always view the threat level differently, but in principle, the concerns that others voiced over hundreds of games being at risk were valid, regardless of any smokescreens The Verge tries to put up.

Headlines always have to oversimplify. One doesn’t need The Verge‘s permission for that.

2. The Verge is wrong because StarCraft did not have all of the claimed elements, but Windows Central Gaming was right about Diablo and World of Warcraft

The Verge quotes Charles Duan, an assistant law professor from American University Washington College of Law, saying this: “If an old game like Starcraft, which obviously predates the 2022 earliest filing date of this patent, had all the features that this patent claims, then that invalidates the patent under the patent law rules of ‘anticipation’ or ‘obviousness.” That is 100% incorrect. StarCraft did not have “all the features that this patent claims” even if one focuses only on the feature-poorest claim, which is claim 1. The patent is now on Google Patents and you can see that even the simplest claim, claim 1, distinguishes between a “player character” and a “sub character”. That distinction did not exist in StarCraft. In StarCraft, there are player units and enemy units. But there is no particular player character, like a general of the army, that owns “sub characters” as Nintendo calls them. That is different from games like Pokémon and Palworld.

The use of the word “If” does not render the prior art claim by that assistant professor a mere hypothetical. It is clear based on the use of “invalidates” (not “would invalidate”) in the main clause and in the context of the article.

That mistake alone discredits The Verge. It’s all the more embarrassing for a website that makes bold claims about others engaging in “poor reporting.” The question here is not whether the distinction between a “player character” and its “sub characters” justifies the grant of that patent to Nintendo (it doesn’t). But from a patent law point of view, the player character/sub character dichotomy in Nintendo’s patent claim cannot be read out.

You either need better examples to say the patent is rendered non-novel by something that truly has all of the elements or you need to argue that Starcraft combined with a game that has the player character/sub character distinction renders the patent obvious. But saying that (as the assistant prof quoted by The Verge did) StarCraft had “all the feature that this patent claims” is demonstrably wrong. It shows that (assistant and non-assistant) law profs must be careful before they comment publicly on patent matters because it can be inconducive to their reputation, but it also shows that The Verge itself doesn’t really know what it’s talking about in this particular regard. It just wanted to criticize other media instead of thoroughly analyzing the issue.

By contrast, Windows Central Gaming got it right:

In her first article on this patent, Jennifer Young pointed to a couple of other Blizzard games: Diablo and World of Warcraft. Those are games where there undoubtedly is a player character, and certain types of player characters can summon what Nintendo would call “sub characters.”

3. The Verge makes a 100% false claim about when Pocketpair first showed gameplay footage from Palworld

Maybe some people are too busy thinking about how they can bash their colleagues and competitors than doing their job.

Here’s another falsehood that is not just on the verge of being wrong, but simply 100% incorrect:

US Patent No. 12,403,397, by comparison, was filed in the US in March 2023, before Pocketpair had shown any gameplay footage from Palworld. (Palworld was released in early access in January 2024.)

Maybe The Verge should read other media before criticizing them. That would educate them. A few months ago, Windows Central Gaming reported (as did various other media, all of which were pointed to it by games fray) that Pocketpair first showed gameplay footage of Palworld in June 2021.

If the people at The Verge don’t want to read Windows Central Gaming and other websites, they should consider using AI bots for fact-checking. That might help. Take Perplexity, for example:

If Nintendo filed for a patent in May 2023, was that before or after Pocketpair showed any gameplay footage from Palworld?

Pocketpair first showed gameplay footage and revealed Palworld as early as June 5, 2021, with further presentations including trailers and gameplay content released before the game’s early access launch on January 19, 2024. Nintendo filed for their patent in May 2023, which is after Pocketpair had already shown gameplay footage of Palworld. Therefore, Nintendo’s patent filing in May 2023 was after Pocketpair showed any gameplay footage from Palworld, which began in 2021 and continued through 2022 and early 2023 with public trailers and presentations.

That is not to say that one could rely on what such a chatbot says, but when it provides sources, one can verify. Here, the bot got it right, but The Verge did not.

Atari’s 1978 Outlaw, Midway’s 1975 Gun Fight and the starting point for Nintendo’s patent

To elaborate on the first point in the previous section, but also because it’s generally relevant:

“The name of the game is the claim” as patent professionals sometimes say, but it’s also important to figure out from what prior art and how a patent seeks to differentiate itself before making bold claims.

In the digital industry, patents are always incremental. There is no patent that doesn’t stand on the shoulders of giants. The question is whether it makes a significant contribution to the state of the art. That’s why digital patents (and patents in many other fields) start by describing the state of the art and then go on to explain what’s new and better about the patented technique. Typically, the state of the art that patent applications in this field reference is rather new, given the pace of innovation. Not so here.

The textual description starts on page 25 of the PDF version of the patent. The first paragraph outlines what Nintendo considers the state of the art over which its patent constitutes significant innovation:

Conventionally, as for a game in which a plurality of characters battle with each other, a technique of progressing a battle by causing a character to perform a motion corresponding to a command designated by a player, has been known.

Let that first sentence sink in. They’re talking about a game where two or more (that’s what “plurality” means) characters are engaged in battle, and they move based on commands from the player.

In the early 1980s, I played a game on my Atari 2600 video game console that was already a few years old by then: Outlaw. The game was published by Atari in 1978, and its designer and developer (David Crane) founded Activision not long thereafter. As you can see on YouTube, it had precisely what the above sentence described:

But even Outlaw wasn’t the first one to do it. Three years earlier, in 1975, arcade game maker Midway published Gun Fight, of which Outlaw was essentially an adaptation:

Now let’s look at the second sentence, which is where Nintendo said what it believes its patent covers beyond what Gun Fight and Outlaw already did 47-50 years ago:

[I]f, in addition to a battle by a method in which commands are used, a battle by another method is enabled, variations of battle can be increased.

So they say: instead of having one battle mode, you can have two or more. What they seriously say is that they’ve invented a means of managing two or more battle modes.

Their patent claims don’t cover every game that has two battle modes. But what they told the patent examiner was that prior to their patent application, games had only one battle mode. They don’t say that the battle mode they chose for the starting point already existed half a century ago.

One of the things The Verge didn’t consider is that Nintendo’s patent does not reveal any actual technology that makes automatic battles between characters work. What Nintendo wants a monopoly on is practically a kind of switch between battle modes, but the battle modes themselves are not new. Here’s a quote from The Verge:

“The claims describe a system that sounds a lot like auto-battling from Pokémon Scarlet and Violet, in which the player can send a pokémon into the overworld to automatically fight wild pokémon rather than initiating a typical turn-based battle that the player controls.”

Actually, claims just claim. Descriptions describe.

Claim 1 of the patent (the braodest and therefore most threatening one) covers a high-level switch. It doesn’t even specify how the automatic battle works. And the description? All it says (but only in connection with narrower claims than claim 1) is that there is a “determination range” within which the player character or the summoned character is close to an enemy character and will engage it (otherwise it will ignore it). That’s all. There are drawings and there’s a lot of text, but the most part oft hat is meaningless in the end because they explicitly call all of that just “example[d]” and “exemplary”: that means they show how one might implement the patented idea, but they also want to have a monopoly on all alternative options.

For example, the patent description mentions throwing a ball to summon a character. But that is not required in order for claim 1 to be (allegedly) infringed.

It doesn’t matter that there are more specific claims in the patent (even those are still fairly generic) and that their underlying ideas are described somewhere. If Nintendo sues over claim 1, all that matters is what claim 1 potentially covers. Of course, anyone who would get sued would try to give it a narrow meaning, but Nintendo would do the opposite. The “operation inputs” the patent claims refer to can be anything. If you set a character into “automatic battle” mode, that is an operation input. If you sent it to a given location, that is an operation input. If you tell it to use a certain weapon, that’s an operation input.

Let’s look at it pragmatically: there are countless games (and new games continue to be released all the time) that do all of what claim 1 purports to cover. It was always an oversimplified version, but never unjustified to describe it as a patent on summoning a character and letting it fight. All that the patent claim says is that there must be two different wys of letting it fight, such as activating a “fight automatically” mode on the one hand and pointing the summoned character to a particular enemy. In any game that has a player character capable of summoning “sub characters” you will find more than one battle mode, and there will always be some user interaction involve.

Invalidation is easier said than done

Our original article on this patent already explained that even when there are reasons for which a patent should be revoked, it’s not as easy to achieve that result as it may sound.

We mentioned that the United States Patent and Trademark Office (USPTO) has recently made it more difficult to challenge granted patents. It’s also worth noting that if a validity determination must be made in a U.S. court, it will be put before a jury most of the time, but U.S. juries are extremely hesitant to conclude that a government agency made a mistake. Jurors are routinely shown a video that portrays the USPTO as omniscient and incredibly thorough. The video shows images of patent examiners in lab coats, jointly looking at something, but how it actually works is that patent examiners sit at a desk and have a rather limited amount of time available per patent application.

This Nintendo patent was apparently granted without a patent examiner looking at a single game. Nintendo provided the patent office with a link to a web page about Pokémon. A patent examiner ran a search, but only in a database of patent documents, without even using Google.

Here you can see the references the examiner cited:

Here’s Nintendo’s information disclosure statement, which points to a webpage:

And here are the searches the examiner performed in a patent document database: