Crimes and Punishments: The Danger of the Infallible Sherlock Holmes

In some ways, it’s what you’d expect from a detective videogame: you gather clues through a series of puzzles and interrogations. The puzzles can be deduction or dexterity-based. Some, like an arm-wrestling match with a drunk sailor, combine the two. That particular one has you watching the man’s face for clues as to his arm-wrestling strategy and then requiring you to hit different buttons accordingly. It’s lower-tech version of LA Noire’s facial expression capture system, and it’s just as effective. Holmes’s observation skills give him an advantage even in what would appear to be a test of brute strength.

The “deduction space” menu system sees the different clues floating around each other. You pick two of them and if they provide an interpretation, it’s added to another screen of connections. They’re drawn as neurons, meaning the opening credit sequence of flying-through-brain-cells is more than just a Fight Club reference (or a Metroid Prime reference, or a reference to all the other cell-based title sequences that oh so many videogames love).

After gathering enough clues, the game will tell you that you can make a deduction. Some pairs of clues can be interpreted in one of two ways, some only one. You chose an interpretation for each and, so long as none of them contradict, you’ll be able to accuse the guilty party.

Well, the maybe-guilty party. Crimes and Punishments lets you proceed with interpretations that don’t match what actually happened. In the game’s first case, there are three possible suspects. Depending on how you interpret the clues, you can accuse one of the three or a pair of them working together. After making that call, you are given a “moral choice”. Keeping with the previous example: if you choose a lone murderer, you are given the choice to “condemn” or “absolve” them. If you choose the pair, then you pick which one is the mastermind and which is the accomplice.

After a resulting cutscene plays out, the game gives you the option to check your conclusion, showing how many clues you found and turning the text green if you’re right, red if you’re wrong (not a very good choice for accessibility reasons, as there’s no indicator to help colorblind individuals). Whether or not you decide to check the conclusion, you opt to stay with your decision or choose a different one.

After confirming that, yes, again, you’re sure you don’t want to change, you’re given a “Personality Ranking” (I got “sympathizing”) and sent on to the next case.

Do you understand how wonderful this is? Holmes can be wrong. The deduction and accusation system is built to allow interpretation without inhibiting progress. It’s a straightforward argument: the outcome of guilt or innocence is not based on a preexisting condition but is the result of a process attempting to determine it. And that point is important enough for the meaning of the game that the developers built different paths that are WRONG. They’re not just different (like, the narrative equivalent of three perfectly balanced classes in multiplayer). You can accuse innocent people and send them to jail by misinterpreting the facts. And it becomes even bleaker when you consider that in the Sherlock Holmes world, he’s not supposed to be an interpreter, but the closest thing to Objectivity that can be achieved—the Western European, male, highly educated, etc, etc, Enlightenment Victorian ideal of masculinity, etc. The stories stop before the trial because in their world the trial is irrelevant—if Holmes has declared you guilty, well, you’ll get one scene where the whole thing is explained either to you or by you.

This Holmes doesn’t have the advantage of the Holmes of Doyle’s stories (or the Holmes of Moffat’s Sherlock): his authors aren’t invested completely in his always being right. Because Frogware wants the player to have some input, suddenly everything is not stacked in Holmes’s favor.

The infallible Holmes is a dangerous idea, like the Genius Who Is Tormented Because He Is A Genius, or the Genius Whose Social Nastiness Is Excused Because He Is a Genius. He provides a point of identification for someone who might see themselves as supremely rational, or someone whose social nastiness or torment is the result of their being a genius.

Ends can justify the means in fiction a lot less messily than they could in real life. The self-contained and scripted nature of fiction means that the cop who violated that person’s civil rights was justified because they were guilty (they are always guilty)—if you’ve got nothing to hide then surveillance shouldn’t bother you (because surveillance is only ever deployed against the already-guilty, you’ve seen the cop shows). Because torture works to save the day, and the only downside is maybe the loss of the “humanity” of the torturer (but they can recognize that loss, and so can be redeemed, which makes for a great story, so, really, there is no downside).

Introducing doubt into the Holmes character, smashing that hundred-year-plus old archetype and trying to remove some of its power, its influence, its ability to justify, is important. Crimes and Punishments decouples the what-happened of the crime from the detective’s reconstruction-of-what-happened and then again from the moral-judgment-of-what-happened. Its concession to its nature as a videogame, having to provide feedback on whether you were right or wrong, is its biggest weakness. But by adding the moral judgment to the end of the system, it undermines its finality. Morality doesn’t flow directly from some group of objectively proven facts, free of any interpretation, of deciding what information is relevant and what isn’t.

And if interacting with a system that models a simplified version of a worldview with nuance and people and processes rather than checkboxable characteristics gives a player some better tools to understand the world around them, well, that’s even more important.

Brian Taylor has a detective problem.