Cognition Dissemination: A Straighter Portopia Serial Murder Case Remake Should Still Happen

When Square Enix started registering trademarks for The Portopia Serial Murder Case in countries around the world, it appeared the company was gearing up to release a modern remake of the mystery adventure game to commemorate its 40th anniversary. The title, the very first one spearheaded by a pre-Dragon Quest Yuji Horii, was being redone for the modern audience and released for the first time outside Japan. This both is and is not what they released.

The trademarks were resultingly for the Square Enix AI Tech Preview: The Portopia Serial Murder Case. The name didn’t mean the entire game was mostly made by an AI program and polished by humans — technology isn’t quite ready for that innovation yet.

This is an updated demo version with redone graphics, importantly utilizing a modern tech program known as Natural Language Processing implemented for educational purposes. Unlike the original game, where the player had to use somewhat precise wording to trigger certain scenes to progress, the new AI program applied can “glean meaning from natural language” to more easily understand player intent after they’ve typed in something resembling a solution. PC programs couldn’t figure out much on their own in the early 1980s, but modern programs can process this through AI programs such as this one. Since this was being used for experimental purposes, the game was released on Steam for free.

The AI processing player meaning was the intention, anyway. The idea sounded solid ingenious on paper, an adventure game in which the player wouldn’t need to figure out how to progress through anywhere near as much trial and error (or, let’s be honest here, a strategy guide) in key cases. That’s not how it works. The game’s NLP AI process has trouble deciphering the meaning behind even basic player-typed sentences, more issues than, if you can believe it, the old 1983 title. Multiple Steam reviews can attest to this, the reason why it has a bleeding red “Very Negative” aggregate. The result doesn’t nearly live up to the promise, like everything AI-related thus far. It’s a good thing this was only a free experiment.

There’s something else Square Enix could do from here: It will never be too late to take the assets created for this release to remake of the original game more faithfully, to provide the experience those who saw the trademarks thought was coming in the first place. It would be a shame if the new character models, backdrops, and sound went to waste in a failed experiment, and it’s never too late for the company to capitalize on a small-but-faithful market perpetually hungry for a new adventure game.

The Portopia Serial Murder Case deserves a good remake because of its importance in video gaming history. The original title was highly influential among the early Japanese game development world, one responsible for kickstarting the murder mystery adventure genre in the country. Though Horii has long left this style of game behind with how overwhelmingly successful the Dragon Quest series of RPGs have been and continue to be, the title that gave him his start is an important title to preserve. But a remake should come with a necessary number of enhancements and polishing to make it palatable for the current audience. The effort required would be enough that Square Enix could charge for it, for a price on par with the recent (and reportedly unappreciated) Paranormasight: The Seven Mysteries of Honjo.

I don’t think that scenario will play out. Square Enix is more likely to prioritize improving the AI tech in the current version, feeling that it will one day be up to par with their aspirations. There’s no telling how long that could take, but there are benefits for the company and perhaps the players of their games to reap if they come close to perfecting it. The ideas that gave birth to this experiment were sound, and the company didn’t have to worry much about fan vitriol in this particular case because it’s a free release. The audience that wants to play a more faithfully remade version of the original game instead is vocal, but ultimately infinitesimal.

The other concern I have pertains to the team updating the game to a frightening degree. This adventure game could get so good at discerning player meaning from words that the technology could then be applied to other Square Enix games with no issue, and subsequently adventure games from other publishers if the company decides to sell the technology. Developers responsible for programming mystery solving in other games in larger companies could find themselves reassigned to other purposes if they’re not out of jobs. If you think this is a farfetched and fantastical scenario, keep in mind the rapidly fast strides AI has made and continues to make.

My preferred compromise would be for Square Enix to continue updating this while preparing and releasing a more faithful remake they could charge for. I’m mildly confident in this happening, so I’ll keep my fingers crossed.