Mystery House Review: When the "Primitive" Label Really Fits

I don't like to throw around the word "primitive" in my reviews because people use it interchangeably to describe the sophistication of a game's native hardware and the quality of its design, but in this case I think it can be used equally well to describe both things.  As much as I enjoyed weaving my own story out of the holes in the plot of Mystery House (1980), a lot of what makes it quirky and confusing is a direct result of the fact that nobody had attempted a graphic adventure before this, and its designers, Roberta and Ken Williams, were doing quite a bit of flailing.  They might have helped themselves by spending some time with comic books or illustrated children's books, but even those examples would only have gotten them so far.  Unlike in a fixed narrative, interactive fiction requires you to account for a wide range of player choices; that is, you can't count on the player taking a single, linear path through the game world.  Previous interactive fiction games had been text-only, so there were conventions for adding cues to the text, but not for the visual interface.  To make things even more difficult, this was the Williams' first interactive fiction game of any kind, so even the textual component had its share of problems.

Normally, I would focus on the big picture in a game review, but here the technical issues are frequent enough that they seriously detract from the gameplay and really need to be addressed in detail.

Let's start with its most obvious shortcoming: the graphics, or, more specifically, inconsistency in the graphic design.  It's not necessarily a problem for graphics to be simplistic or even child-like, but they should work as part of a self-consistent aesthetic.  On a quick glance at the in-game items, shown below, you'll see some things that are neatly drawn (the gun, the small key), and others that are ambiguous or proportioned oddly (the candle, the towel, the sledgehammer).  This inconsistency is not just an aesthetic problem, either; it can also make for a confusing in-game experience.  The player needs to type in commands to interact with what they're seeing, and that can only happen if they know what they're looking at.

And even when it comes to objects that aren't necessary to complete the game, it's still important that players can at least get some idea of what they are.  Otherwise, they might be inclined to spend hours puzzling over how to interact with the lampmas tree of David (I still don't know what that thing is).

Another frequent headache with Mystery House is the failure to provide visual cues that indicate spatial continuity.  In the left picture below, I'm standing in a hallway looking through a door.  The fact that it's a door is pretty obvious, but the problem is that there's no visual indication of where this doorway stands in relation to anything else.  By placing the doorway to the room in the center of the frame with no other cues, it suggests to the player that they should move forward into the room or back where they came (initially from the west).  As it turns out, however, I can move east (right), at which point I'll be faced with an identical doorway.  Not only is there no indication that the hallway continues east, but when I finally did move east, I didn't even realize I was faced with a different room because the image and text the game presented me with was identical.

The right-hand picture illustrates a slightly different problem, where the visual field misrepresents where the player is located in space.  Here, it looks like you're standing at ground level in front of a tall tree, but to leave this screen, you have to either go up to climb the tree, or down to return to the forest.  The former option is reasonable enough, but the latter suggests that you have already climbed the tree partway, or are standing on some elevated ground that's not indicated on the screen.  Either way, it's bound to confuse the player, breaking the illusion of the game world.

And while we're talking about breaking the player from the game world, let's look at the cases where the graphics provided too much information.  The most egregious cases are the bodies.

The X's over the eyes of the corpses are...unnecessary, as are the action lines over the bumps in the head.  If the graphics were incapable of showing what the designers wanted to communicate to the player, they should have put it in the text.  And there must have been a better way to indicate to the player what the telescope was looking at (see below) than drawing a dashed line across the screen. 

All that these extraneous cues serve to do is to remind the player that they're playing a game and not exploring the grounds of a mysterious house filled with dead bodies.

I also had some issues with the text component of the game.  I actually think a two-word parser is a good choice for Mystery House because it simplifies the user interaction and allows the player to focus on the plot, but the problem is that the parser isn't permissive enough.  This becomes clear on the very first screen, when the user is forced to type "Go Stairs" just to walk up to the house.  If the user knows what they need to do, but can't easily communicate it to the game, that's not a puzzle, it's a game design problem.

Still, despite all of its problems, I enjoyed Mystery House.  It had an innocent charm to it that was part child-like and part surrealistic.  I'm not sure that a murder mystery was the best fit for Roberta's imaginative and oft-illogical approach to adventure games, but there was still plenty of evidence for the talent that would later make her famous in the game industry.  If you're curious and want to try it, the game was released into the public domain in 1987 by Sierra and can be played for free in your browser at the internet archive