No Time To Play

The work-to-fun ratio

by on Mar.27, 2011, under Gamedev

You know how adventure games went from being the most popular kind to becoming a minor niche? Much has been written about the why and how of it, and whether it was deserved or not. But nothing compares to discovering for yourself the trade-offs involved in making adventure games versus any other kind.

See, working on a text adventure so soon after an arcade game made me notice a simple fact that should have been obvious in retrospect: the ratio of effort spent to entertainment provided is terrible for the former. So bad, in fact, that I can understand any game developer (whether professional or hobbyst) who decides it’s simply not worth it.

Look at it this way: Buzz Grid supports exactly five commands — north, south, east, west and “skip level”. It also features exactly four types of interactions, one of which you can’t even control. Despite that, it can provide countless hours of fun. By contrast, Catch That Cat has 17 explicitly programmed interactions just in the first room, and I haven’t counted the exits or the many implicit and predefined interactions, every single one of which had to be programmed in or otherwise set up. And you can exhaust all of them in just a few minutes, after which the game loses all its appeal. Not to mention that as the developer I already know them all, so I can’t play my own game.

You could argue that it takes one year to write a novel which you can read in a few days. Worse, it takes two years to make a movie that you can watch in a couple of hours. The ratio is much, much better for an adventure game.

But you see, any other type of game trumps adventures in this regard. And that’s what matters when a developer decides what kind of game to do next.

Can this situation be improved? Likely, yes. There is talk of applying procedural generation to interactive fiction, a technique used successfully in everything from roguelikes to high-profile titles such as EVE Online and Fuel. And even without PCG, murder mystery-themed adventure games such as Blade Runner can offer surprising amounts of replayability. But that’s a rather narrow sub-subgenre. Hopefully with PCG we’ll be able to make any adventure game as replayable as, say, a Japanese RPG (where the only things you can vary are your party build and tactics).

Are we there yet? No way. In fact, I’d say we’re barely starting in that direction, and not for lack of trying. But whoever figures it out first might just trigger a revival of the genre, and make a fortune out of it (thanks, Iain). After all, the success of the recent Monkey Island remakes, or the new Sam and Max series, are evidence that the public still loves adventure games; it’s us developers who have grown disillusioned. And I really hope that new technology will change the equation for us, to everyone’s benefit.

Creative Commons License
The work-to-fun ratio by Felix Pleșoianu is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.

:, ,

2 Comments for this entry

  • Victor Gijsbers

    I’m not entirely convinced by your arguments. I can see how an IF game might take far longer to program than some arcade games. But IF is certainly not near the bottom of a work-to-playing-time ratio. As in IF author, I can write: “Suddenly, a horrible demon appears in the study in a cloud of purple smoke, snatching the sorcerer’s soul before his body has time to hit the ground.” Takes me maybe twenty seconds. As the maker of a 3D game, I have to build a model of a demon, of purple smoke, and of a soul — probably at least a week of work, and that is without worrying about how I can make the body fall naturally, the demon appear in a suitably impressive way, and so on.

    It seems to me that there are two factors that determine the work-to-playing-time ratio. First, how much work is it to make material in the chosen medium? With text, almost no work; with 3D graphics, a lot of work (and even more if you want sound effects, music, voice acting). Second, how much of your game’s content has to be generated by hand? On this dimension, text adventures are a lot of work: we generally want and expect many hand-crafted interactions and events. On the opposite end are infinite or randomly generated games using a (relatively) small number of interactions, such as some arcade games and roguelikes.

    Now games on that latter end of the how-much-generated-by-hand scale are games that use procedural generation. But that simply makes them different games. There is no way you are going to see procedurally generated games that play like the standard IF of today, because having lots of hand-written prose and non-standard interactions it what gives those games their identity. On the other hand, there is nothing stopping you from writing an interactive fiction game with procedural generation today — I am doing that right now, building an IF roguelike — as long as you don’t mind that it will be a game very different from Make It Good, The King of Shreds and Patches and Aotearoa.

  • Felix
    Felix Pleșoianu

    Haha, yes, you make some good points. I was thinking that the 3D demon you mention would probably introduce a battle in some RPG… which battle you can replay many times with different tactics and party structures. And that’s before considering the rest of the game. Besides, what about the part about making the demon act interestingly in the game? That is going to be more complex in text because there are many more possible verbs.

    And no, you can’t truly replace hand-written prose. But if you could devise an advanced method to recombine pre-crafted prose fragments algorithmically…

Posts by date

March 2011
« Feb   Apr »

Posts by month