Intermittent Mechanism

A belated third entry in my video series on detective games. The pace of these has been slow, but I’m going to have to step it up, as these are intimately related to course prep for a course I’m teaching in the Spring term of 2020.

Script below the jump.

Hello everyone. Ian here. Sorry, I was been busy with a heavy teaching load, but I’m back, after a break, with the third video in a series on detective stories in games. And if all goes as planned, this will be the final video in the series to look at more than one game and make an overarching critical point—from here on in, each video will be a deep dive into one game.

In 2017, Eidos Montreal released the DLC “A Criminal Past” for Deus Ex: Mankind Divided. I consider it to be the high point of the Eidos Montreal Deus Ex prequel games. For one, it’s set in a prison, which is always a great choice of setting for a stealth game. But beyond the basics of its setting and its map, it’s also probably the most genuinely rich and interesting piece of cyberpunk fiction that’s ever been told in a Deus Ex game. It’s certainly miles beyond what the base game offered, with its horrendously miscalculated “Augs Live Matters” attempt at political allegory.

In A Criminal Past, Adam Jensen is tasked with extracting Hector Guerrero, an Interpol agent who is deep undercover, embedded in a prison gang. The gang’s main revenue stream is the illegal sale of prosthetic body parts, which it procures with the help of the prison’s warden. The prison’s guards are systematically killing the most heavily-augmented prisoners, using tiny behavioral infractions as paper-thin justifications, then stripping their bodies of their most valuable prosthetic organs for sale on the black market. Exactly how deeply Guerrero is involved in this scheme is left up to the player to discover. The game has several endings, branching depending on how much Jensen implicates Guerrero in the organ harvesting scheme. There’s a bunch of physical evidence to be found implicating Guerrero, and you can also get him to confess to executing a hit on a fellow prisoner, by asking certain questions in the dialogue tree. These questions only show up, though, if Jensen has previously found two key bits of evidence against Guerrero. It’s quite possible to miss this evidence, and therefore never hear Guerrero’s confession, and also miss a bunch of the other evidence littered around the prison that gives a full picture of the organ harvesting scheme. How many details Jensen discovers about this criminal network are really left up to the player’s own commitment to exploration.

If you successfully get Guerrero to confess to the murder of Ian Wilburg, you get the “Objection!” Achievement—an obvious homage to Capcom’s Phoenix Wright: Ace Attorney series. And while I appreciate Eidos Montreal’s show of cross-publisher franchise love here, I do have to say: “Hold It!”

Because Eidos’ invocation of Phoenix Write here is not a good fit for the sequence I just described. One of the things I especially admire about the Criminal Past DLC is the fact that, considered as a mystery, it is littered with missable clues. Players can be easily mis-led by their own lack of diligence. We’re given the freedom to miss the big picture, and even to come to the wrong conclusion. Phoenix Wright games, by contrast, aren’t like this at all. Mechanically, they’re notable for just how much they trap players in a space, refusing to let them leave until they’ve found every last relevant clue.

I still have nightmares about wandering around aimlessly in the Berry Big Circus location of Justice for All, unable to figure out how to progress. Because of how the game is constructed, I could leave the area at any time, and go back to Phoenix’s office. But I couldn’t actually end my investigation and move on—the game denies players the power to do so voluntarily. The investigation only ends when Phoenix tells Maya it does, after the game has decided you’ve found the requisite clues.

And in Justice for All, the discovery of these clues is often contingent not on sweeping the area thoroughly in the point-and-click sections, but instead by showing character’s profiles to other characters. So, for instance: the ringmaster’s office doesn’t show up on your movement options screen until you show Regina her father’s photograph. And then once you get there there’s this obvious clue—a piece of paper sticking out of the pocket of his tuxedo jacket—that the game refuses to let you pick up.

(What do you mean we just can’t go through people’s pockets?!? This entire game is about searching for clues!!! Just seconds ago I rifled through the documents of someone’s salary negotiations!!!)

Similarly, Moe the Clown’s room doesn’t open up until you show Maximillian’s profile to Detective Gumshoe. That bit of gating feels arbitrary, but at least your dialogue with Gumshoe establishes Moe as a definite lead. Except if you head directly to his room, as the game seems to suggest you do, it’s empty. Moe doesn’t show up in his own room until you first talk to Ben, and Ben doesn’t show up until you’ve presented every character profile to every other character—not only Max’s profile to Gumshoe, and Berry’s profile to Regina, but also Max’s profile to Regina.

So the only way to navigate the space and escape is to successfully wind your way through all of these unintuitive dialogue triggers. My first time through the game I wasn’t thinking that way—I was thinking I had missed an actual, physical clue somewhere. So I just kept pointing and clicking on everything, and meanwhile the game was just continuously mocking me with this damn scrap of paper that Phoenix refused to pick up.

I wouldn’t mind the prominence of these sorts of triggers if they were at least better signposted. For instance, when you return to the circus later on, Moe gives you a scrap of a threatening note and tells you that either Max or Ben should be able to say more about it. But Ben has nothing to say about it. And Max tells you to ask Regina about it. But Regina’s nowhere to be found. In order to get her to show up again, you need to talk to Acro, and then grab that scrap of paper from Berry’s jacket, which Phoenix finally now agrees to pick up. Then Regina finally has something to say about it. But Ben never does, not even when the entire thing’s been re-assembled!

I’ve spent a few minute here complaining about a game I actually really like. I find the Ace Attorney series, as a whole, to be incredibly charming. It has wonderful characters, getting amazing mileage out of simple but wonderful character animations. The absurd escalations of its convoluted mysteries fit its overall soap-opera storytelling style well. But the fact of the matter is that games don’t provide the player with much room to arrive at erroneous conclusions.

You can certainly fail in them—if you exhibit the wrong evidence during court sessions, your hit points will be dinged, and eventually you’ll lose the case. But failure happens quickly, whenever you veer even momentarily off of the game’s script. The games don’t branch and provide contingencies for when a player follows a false lead, as we see in Criminal Past. Its script it linear to a fault. It doesn’t let you leave an area before finding all the clues, because if it did, the game could fall into an unwindable state. Each and every clue has its own unique use within the court scenes to follow.

And this is has always been one of these games’ biggest weaknesses, mechanically speaking. Sometimes the script is so inflexible you end up having to mind-read what the developers were thinking. So, for instance: Right here, this witness is saying that my client wore his TV show costume when he stabbed the victim. But that’s a contradiction: this poster clearly shows that the costume includes gloves, and it’s already been established that my client’s fingerprints were found on the knife. But what do I present—poster, or knife? Poster? Knife? Poster? Knife? Let’s go with poster. Nope, not what the game wanted. So … knife? Yeah, that one’s right. Feels sort of arbitrary, though. These are the sorts of “failures” that Ace Attorney games allow for—minor departures from its script, resulting in immediate penalties. These games don’t allow you to miss clues, or actually come to egregiously wrong conclusions. They’re written as linear courtroom dramas, and just aren’t built for those sorts of contingencies.

I’ve begun this video with this contrast between A Criminal Past and the Ace Attorney games because, as a diptych, they so well illustrate the different ways in which detective games can account for the failure of players to solve the central mystery. This was never really a consideration of detective stories in literature, or in film. In these media, the point is to mislead the reader, or the viewer, to fill the text with false leads and multiple erroneous conclusions, to have them perhaps fail to guess the answer—but in the end the answer is always given by the detective figure, serving as an author stand-in, revealing all. Unlike books or movies, games have to ask: if they player is themselves the detective figure, and they mess up on the way to their conclusion, what then? Do you just let them miss things? Do you just let them be wrong? A Criminal Past’s answer to this conundrum is yes: if players aren’t inquisitive and persistent, they’ll never discover who killed Ian Wilburg. The Ace Attourney games follow a different design philosophy: one in which the player cannot miss any crucial details, because the player’s role in solving the central mystery is too rigged to fail.

In the rest of this video I’ll be looking at this theme of player failure, and the lengths that game systems will go to avoid it. And I’ll acknowledge, upfront: allowing the players to fail is a difficult thing to pull off, in terms of both gameplay and writing. A Criminal Past can get away with it, because it offers a whole mess of immersive sim pleasures outside of the side-mystery of who killed Ian Wilburg. Something like Heavy Rain can do it, because it features multiple protagonists undertaking their own largely independent investigations, and the game can adapt to having one character fail to pick up of piece together all of the clues—just as it can adapt to having one or more characters outright dying before the player reaches the end of the game. But elsewhere—I think as the Ace Attorney games show—there are a plethora of pragmatic reasons to limit the options for player failure. So let’s look at a few more case studies.

First up, we’ll take a look at the most prominent, AAA-budget detective game from the past decade: L.A. Noire.

L.A. Noire uses several tricks to make sure players don’t miss vital clues. The most blunt is simply denying players the means to progress until they have found specific clues—although the game is comparatively subtle in how it achieves this.

When you arrive at the first main crime scene for a case, a distinctive “investigation” music plays. You can step out of the boundaries at any time—the game is not so obnoxious as to throw up any “you are now leaving the crime scene” invisible walls—but, if you do, that music fades out. This helps hone player attention, drawing up a boundary of spacial relevance. As you walk around within this boundary, there are little chimes that play when you’re close to clues, prompting you to hit the “interact” button to examine them. If you’re playing the PC version of the game with mouse and keyboard, by default there’s just that chime. Although it’s also possible to turn on a small visual indicator that lights up on the bottom left of the screen when you’re near a clue. If you’re playing the console version, you also have the benefit of haptic feedback buzzing off as you approach and interact with clues.

But that’s it—and, all and all, it’s relatively subtle. There’s no specialized vision mode or object highlighting to guide you to each detail. The game doesn’t give you a checklist of off all the clues you’ve found (at least, not until the end), nor does that investigation theme music ever turn off, signaling that you’ve found everything. Basically, you inspect what you can, until something happens, and you’re given a new address to head to.

So, for instance: In the case of “The Golden Butterfly” homicide, you have to inspect clues until a patrolman interrupts and gives you the address for the victim, Deidre Moller. The first step in getting him to do this is to find her name tag. Once you find that, he’ll rush over to a police phone and call the station. But then, he just … stands there. He’s scripted to not deliver you the information until you find additional clues—this is the game’s way of gating you’re progress until you’ve found enough key clues to continue. He’ll remain frozen at his spot until you examine the rope marks on the victim’s neck. Then he’ll chime in, and a new address will be added to your list of locations, allowing you to move on.

In addition to this gating of locations, another trick L.A. Noire uses to make sure you don’t miss clues is redundancy. Choosing all four correct dialogue responses in this interview with June Ballard meant that I heard the name “Mark Bishop” four times. And it’s not just dialogue that provides redundancies—physical evidence does, as well. You can jump immediately into your interview with Michelle Moller without examining the Moller residence at all, and it doesn’t really matter. The only thing here is signs of missing jewelry (redundant with the missing jewelry on Deidre Moller’s body) and Hugo Moller’s size-8 boot (redundant with his other boot you’ll find later in the investigation).

Finally, there are clues that are unmissable because the game takes the wheel from us entirely. In the case “The Gas Man,” you discover that the Steffans won a prize from Gulliver’s Travel Agency not by poking around in smoldering rubble, but just by having the winning ticket be placed directly in your hand during a cutscene. In “The Golden Butterfly,” the game makes sure we recognize the titular broach by ensuring that Michelle Moller spills the beans about it even if we fail every single question in her interview.

I was interested in whether the game would ever allow me to fail my way into the wrong conclusion. “The Golden Butterfly” and “The Gas Man” seemed like the best cases to try this on, because, along with “The Studio Secretary Murder,” they’re among the few cases in the game where you end up with two viable suspects, each of which has a reasonable amount of evidence stacked against them.

This being the case, I went out of my way to find as little evidence as the game would allow me to. I walked right past the footprints by Deidre Moller’s body, and was pleased to find that the game let me ignore them, despite how central they are to pinning the crime on Hugo Moller. I noticed that the game gave me the addresses for the three gas service men before I rifled through their lockers, so I skipped the lockers and rushed straight to pick them up.

These were small victories. Most of the time, the game plopped evidence on my lap when I was deliberately trying to avoid it. Either that, or it bottlenecked and made me backtrack. “The Gas Man” gives you a lot of locations to hit up early on, and for a long time I thought I could get away with deliberately ignoring Matthew Ryan’s name on the list of service providers at the Steffens residence. Eventually, however, I ran out of locations to hit up, and the game gated my progress—I wasn’t allowed to continue on to Fire Station No. 32 until I examined more physical evidence at both the Steffens and the Sawyer arson sites.

My hope was that, though deliberate incompetence, I could sabotage the case against one suspect by way of sheer lack of evidence. This didn’t end up working. I missed some bits of evidence, but the game provided other bits to make up for it. I deliberately failed chase sequences, and this only resulted in an immediate game over and reloading of a checkpoint. In the end, both suspects always ended up at the station, with a baseline level of evidence against both of them. It was up to me to sink the case through terrible suspect questioning—and here, I found that the game finally did give me the freedom to fail. I had to wait until the climax of the case, but I finally had the chance to absolutely whiff it.

But even then, I couldn’t fail too much. One dirty trick that L.A. Noire plays with “The Golden Butterfly,” “The Studio Secretary Murder,” and “The Gas Man” is that ultimately neither of the suspects committed the crime—the crime was committed by a third party, a serial offender, who evades your department’s suspicion until the conclusion of the Homicide desk cases and Arson desk cases, respectively. It’s a cop-out that retroactively removes the weight from your earlier decisions on who to charge in these cases. There are things I do like about how these cases play out: for instance, the fact that your bosses always berate you for not pinning the crime on the social undesirables—the communist, the anarchist, the militant homeless man—even when the evidence against them and the other suspect is just about even. It’s a sly touch that highlights the real underlying role of the police force. But in terms of the freedom these cases offer to fail on their own terms, I’m going to have to give the game a rank of … [unbecoming].

Although I’ve played through L.A. Noire several times at this point, I have to admit that I don’t really like it. The game has a weird neither-fish-nor-foul thing going on, which I imagine is an artifact of the beleaguered development path it took on its way to eventually being published by Rockstar. Although the game is set in a massive open city, it doesn’t really feel like an “open world” game, largely because players aren’t given any downtime to free-roam between missions. At the conclusion of one case, you’re always launched directly into the next one, saddled with new responsibilities and new pushy dialogue from your partner. The game feels like sort of like a prototype for another, better game, one that abandons the strict linearity and actually gives you room to breath and explore side-content. I think I’d love a version of L.A. Noire where you could suspend a case if you’d hit a brick wall and run out of leads. Where you could head back to the station, write up a report, and get assigned a new case. Then return to your cold case later. Maybe even put to use some new leads you gathered in another case. Maybe you could play as different characters in different departments—which you actually finally do, in the game’s final section—but you could switch between them at-will, to catch up on their individual cases and push the larger story forward, much like you do in GTA V.

This might sound like a pipe dream, utterly disconnected from the realities of game development. But in fact there’s a game that gets rather close to what I’m imagining in terms of open-world detective story design. And that—perhaps unexpectedly—is The Witcher 3.

The Witcher games are fantasy action-RPGs, so they might not be immediately what springs to mind when thinking about detective games. But in terms of both mechanics and story, if you look at what Geralt actually does in these games, it’s a lot of investigating disappearances, following up on leads, and tracking killers. Albeit, these killers are monsters, because Geralt is a monster-slayer, and not a traditional detective. Still, though, The Witcher 3 is, in addition to being many other things, a master class in open-world mystery design, with a much more robust approach to player curiosity, thoroughness, and failure than something like L.A. Noire.

CDProjekt Red has shown an interest in using The Witcher franchise to tell detective stories from the very beginning. In the first Witcher game, they went so far as to the genre-bending character Raymond Maarloeve—clearly a play on the pulp fiction author Raymond Chandler and his most famous character, Philip Marlowe—who somehow works as a 20^th century American hard-boiled private eye in a setting that is obviously modeled after circa-14^th century Poland. Maarloeve guides us through the second chapter of the game, an astoundingly ambitious detective yarn in which Geralt attempts to identify and expose the leader of a conspiracy.

There are things I really like about this section of the game: chief among them, the fact that Geralt can come to a completely wrong conclusion if the player doesn’t diligently pursue the correct line of inquiry during a crucial autopsy sequence. Even this early on, CDProjekt Red were making generous allowances for player failure in solving their mysteries.

But there are also numerous things I hate about The Witcher’s second chapter. Over the course of the five or so hours it took me to play through it, its central quest line split into nearly a dozen tiny sub-quests that were difficult to keep track of. I also dislike that the process of eliminating suspects is accomplished through an opaque system of dialogue triggers, meaning that suspects will be automatically scratched off your list during dialogue with other characters around town. This creates a feeling of not being fully in control—I would have much preferred a menu where we could access the full suite of evidence and manually strike suspects out of consideration. (But that would require building something entirely new on top of the game’s already-extant menu system, so I understand why they didn’t do it.) As things stood, there were stretches of the chapter where I couldn’t figure out how to definitively rule out a certain suspect, and conclude a specific sub-quest. For awhile I was genuinely afraid that I had broken the main quest. I hadn’t, but the only way to get un-stuck was to talk to everyone, waiting for something to trip—a tedious process which gave me flashbacks to Berry Big Circus in Phoenix Wright.

But that was The Witcher 1. By the time of The Witcher 3, CD Projekt Red had gotten much better at integrating bits of detective work into their game world. There’s a lot that can be said about The Witcher 3, and certainly a lot that already has been said in the 5 years since it came out. I’m going to limit my praise for the game to two basic claims, which I’ve chosen because they intersect the best with the overall theme of detective work in games:

• First, The Witcher 3 excels at rewarding player curiosity.
• Secondly, The Witcher 3 also rewards player perceptiveness and diligence.
  • However (and these are sub-claims):
    • It doesn’t reward these as consistently as it does curiosity, and
    • it avoids negative feedback letting you know that you’ve failed in your perceptiveness.

These last two bits might sound like bad things, but I assure you that they are very, very good, for reasons I hope will become clear.

So, let’s dive into the first bit, about The Witcher 3 rewarding player curiosity. The Witcher 3 has some of the best environmental storytelling I’ve ever seen in an open-world game. If you see something in the world, it often pays to slow down, inspect it, and ask what went down. Your reward is often not some small random encounter, but a full-fledged side quest. For instance, although it’s possible to pick up the “Phantom of the Trade Route” contract through dialogue with a character who is explicitly marked on the mini-map, it’s also possible to stumble on it, by noticing an abandoned cart in the middle of the road, and poking around for clues. The environmental detail scattered in the open world doesn’t just tell a story, it actively leads into a quest.

To fully make your way into this quest, you have to be not just curious, but also persistent. It’s not enough to just stop and take a look at the cart. You also need to specifically smell the wine, and then follow the scent trail of the wine to the location where the rest of the quest will take place.

But the game doesn’t consistently reward this sort of persistence. As an example: the Rioux-Cannes Outpost is an area besieged by harpies. After you kill the harpies and destroy their nest, you can take in the environmental storytelling here. There are mirrors scattered everywhere about, which a note in the area explains were part of an attempt to re-direct sunlight into a nearby low-lying valley. If you put two and two together, you can deduce that the harpies’ love of shiny things caused them to descend upon the area, granting an unhappy ending to this experiment in engineering. But this is just a bit of environmental storytelling—it never expands into a larger quest. It satisfies surface curiosity, and nothing more. Unlike the Phantom of the Trade Route, there’s no need to follow any trail, dig any deeper. You just draw your conclusions about what happened here, and move on.

Structurally, it’s important that areas like the Rioux-Cannes Outpost coexist with areas like the Phantom of the Trade Route crash site. Sometimes the game rewards persistent digging, other times it offers only surface pleasures to whet our curiosity. As players, we’re never quite sure which situation we’re encountering at any given time—and the game rarely outright tells us when we’re mistaking one for the other.

A great example of this is the side quest “Precious Cargo.” It is strategically positioned as one of the earliest possible side quests the player can encounter in the game, and it serves as sort of a statement of principles. A merchant tells Geralt that monsters attacked his horse, so he had to jump off his cart, leaving it to crash in the nearby swamp. He wants a box retrieved. Following the cart’s tracks into the swamp, you have the opportunity to find a lot of clues in the area that contradict the merchant’s story: the presence of arrows indicates that the cart was attacked by humans, not monsters, and there’s a dead man that the merchant left out of his story altogether. If you note all of these clues, you can call the merchant out on his lies, and he’ll admit that the story he told you was fabricated, and he’s a partisan disrupting enemy supply lines. But the game also lets you ignore them all! You can walk straight up to the box, pick it up, and then hand it over, no questions asked. And the game does not punish you for your lack of thoroughness. You don’t get a forced game over. The quest doesn’t even end up in a “failed” state—it goes in the “completed” pile, just like any other successfully completed quest. Sure, if you actually read the text description, the game makes fun of you a bit for taking the whole thing at face value. But aside from that, it lets you be unobservant, and sends you on your blissfully ignorant way.

The greatest strength of the game’s open-world environmental storytelling is that, even when the actual solution to a mystery can’t be found until you’ve proceeded through the proper steps, elements of that mystery will still be out in the open, free for the player to stumble upon out-of-sequence. One especially cherished moment I have of my experience of the game comes from my first play through of the Hearts of Stone expansion. I had been doing a bunch of business in preparation for attending a wedding, and over the course of my travels found myself entering the town of Erde from the north. The village was eerily deserted, to the point where wildlife was wandering around in it, so I jumped off my horse to investigate it. This wasn’t based on any line by Geralt that his medallion was vibrating, or anything. My behavior was motivated by nothing but my own curiosity. The village had two inhabitants, an old man and an old woman. And maybe it’s just that I’ve watched the Star Trek TNG episode “The Survivors” too many times, but this creeped this out: how and why were they just living here, in an otherwise uninhabited village, with no economy to support them? Now I was full of paranoid questions, so I continued exploring, and found pools of blood in a nearby hut. Finally, I had something to confront them on! But they offered a reasonable explanation for it. And I didn’t have anything else to press them on. So I just hopped on my horse, and moved on. My curiosity was still raging, but I had nothing else to go on. I assumed, at the time, that Erde was like the Rioux-Cannes Outpost—a strange and evocative little place, with a tiny bit of story, but not connected to any larger, solvable mystery.

But I was wrong! Very wrong. And I didn’t find out until much later, when I was looking something up on the Witcher wiki. It turns out that things could have gone a very different way, if I had not wandered randomly into Erde, but instead been specifically directed there by the halfling herbalist who grants the quest “Without a Trace.” If I had done that, I would have been told of the disappearance of his apprentice Folkert, found Folkert’s wagon, and picked up his blood trail. This would have opened up additional dialogue for the couple, and also allowed me to find additional forensic evidence in the village, eventually allowing me to discover, after much diligent investigation, that the couple are cannibals who ate Folkert. My intuition about their creepiness was not wrong! I just didn’t see the big picture. And I respect the game enormously for allowing me to get that little, incomplete glimpse. A lesser game would have limited my interactions with the couple unless I had started the side quest from the beginning. It would have made this mystery sidequest into an all-or-nothing push through a linear series of investigative events. I very much appreciated the opportunity to have only glimpsed a small part of it, and failed to see the rest of it through. If anything, it deepened the sense of mystery for me, making my initial brief and inconclusive encounter much more memorable than any of the cases in L.A. Noire.

And that’s the beauty of giving the player room to fail, when it comes to finding all the clues. I realize that this is much easier to do with small bits of side content than it is to do with the central campaign of a game. But nevertheless, moments like the mysterious town of Erde, or the “Precious Cargo” side quest, or getting Hector Guerrero to confess to the murder of Ian Wilburg in A Criminal Past, represent the gold standard of how mystery stories can be told in games.

And I like to have high standards! The next couple of videos I have planned are deep dives into some truly excellent mystery games. Thanks for watching, and thanks for your patience—and I’ll try my best to up the pace of these.