What Makes IF Enjoyable? — Conclusions Drawn

The two primary factors that make IF enjoyable are:

1.  Playability.  The game should not normally interfere with the player’s attempt to play it, although players will tolerate a high degree of frustration if the game seems to be meant to be frustrating (some of us kept playing _Beta Tester_), and especially if the game is frustrating in a variety of ways.

2.  Emotional Salience.  The thematic material of the game should deliver an emotionally interesting and, ideally, powerful experience.

–the values of player agency, multiple endings, clever puzzle design, meta-cleverness, and all that — these are bogus.  They’re pseudo-values.  Players don’t respond to them.

Not primarily, anyway.  People seem to respond a little to multiple endings, and I think that’s largely because it allows you to pace your involvement in the game.  If you like a game a lot, you can return to it; if you only like it a little, you can set it down and be satisfied.

Now, it seems peculiar that the things we thought we valued in interactive fiction — most notably player agency and puzzles — are irrelevant to player enjoyment.  But there it is:  these values shake out quickly if you look at what people are actually responding to.  _The Grand Quest_ was puzzle-heavy, and most reviewers disliked it.  _The Ascot_ was puzzle-light — not in fact, but in most people’s experience of it — but most reviewers easily forgave it that, although they rated it lower because of the form.  In _Snow Quest_ and _Rover’s Day Out_ you had really no agency whatsoever, but people didn’t complain much about that.

One thing people were right to value is playability, and they were righter than I gave them credit for.  Look at _Eruption_:  For all that I think Richard Bos is a weasle, his game *was* very playable.  In fact, that’s all it was.  But for many, that was sufficient for them to enjoy it at least a little.  Now, if an author can submit a game that he says is half-considered, and players give it tepid-but-positive reviews because it’s highly playable, then that tells you the power of playability.

Puzzles are a bad design choice:  they don’t contribute enough value for the risk they impose of breaking gameplay for your players.  I know a lot of people won’t like to hear that, but that’s what the data tells us.  The two games of this Comp with really amazing puzzles, _Byzantine Perspective_ and _The Ascot_, were often praised for those puzzles, but both were downrated on the same grounds — too short; not enough going on.  That’s the best response puzzles seem normally to get.  At worst, puzzles stymie the player and trap him in the game logic.

For a puzzle to add value for a player, the player needs to enjoy puzzles generally, *and* solve that puzzle successfully; if those two things are not true, you haven’t increased the value of the game for your player.  And, for many players you have probably decreased your game’s value, particularly if the player can’t escape the puzzle and it breaks game-play for him.

If you must include puzzles, the reviews from this year’s IF Comp strongly suggest you should make them easy or optional.

The big value people talk about sometimes that seems to have the most weight is immersion.  But immersion is not a value separate from playablility and emotional salience:  rather, it is the product of the two.  Gameplay that is too difficult, past a certain point, breaks immersion.  And a game that has no emotional salience never creates immersion.

So, there you have it.  To make a good text game, do two things:  First, make a game that is *playable*.  More generally, that means make your game obey the restrictions put on the game-experience by the player’s intelligence.  This is is therefore a “negative value” — you need *not* to screw up.  Second, make a game that is *emotionally salient*.  That means primarily that it speaks of something emotionally interesting in a thematic language that the player understands intuitively.  This is a “positive value” — you need to actively provide this.

Probably, the third, optional, value is agency over outcome — multiple endings which give the player some control over the final emotion of the story.  It’s not really necessary, but it provides a stop-gap, such that a player who is dissatisfied with the first outcome can get a new one.  Therefore, if you bother to write multiple endings, you should spend most of the game hinting what they’d be like, and make them as different from one another as you can while keeping the unity of the narrative, so as to provide the broadest array of options possible.

Advertisements
Published in: on November 9, 2009 at 11:08 pm  Comments (14)  
Tags: ,

The URI to TrackBack this entry is: https://onewetsneaker.wordpress.com/2009/11/09/what-makes-if-enjoyable-conclusions-drawn/trackback/

RSS feed for comments on this post.

14 CommentsLeave a comment

  1. Conrad, you approach is interesting, and I do not want to cast doubt upon the value of the project. But it nevertheless seems to me that you generalise somewhat too quickly from IF Comp results to IF in general.

    The big difference between judges who play game in the IF Comp and general readers/players is that the former do not choose which games to play, while the latter do. Imagine what would happen if you took a random group of book readers and had all of them read through twenty books and give a mark to them. Do you think books like “Ulysses” and “Gravity’s Rainbow” would have a snowball’s chance in Hell to get a good mark? I don’t; but I do consider these two of the best books ever written. They are great, but they are not accessible. Instead of two great literary classics, I could also have chosen two works of, say, modern fantasy: a general reader will not like “hard-core” modern fantasy, because it is not accessible to her.

    So what you will find in an survey of the IF Comp is that accessibility is a great value; and it seems to me that “playability” has a lot to do with “accessibility”, as does not having difficult puzzles. This doesn’t prove that other aspects of game design are less important; merely that they are less capable of making an undifferentiated audience responsive to your work.

    I’m not sure what to make of the emotional salience, though. I like emotional salience, but I do not consider it essential for either accessible entertainment or good art. Hm…

  2. Victor,

    The results are a product of the selection process — sure. But that process — watching how people respond to IF Comp games — is integral to the IF playing and authoring communities. Certainly selected populations, or “niche markets,” could respond differently.

    That might apply to the IF equivalent of literary markets, supposing there was one; or it might apply to the IF equivalent of the erotica/porno market. And there *is* one of those — AIF. Whether this analysis holds true of hypothetical literary IF, or of AIF, I can’t say without data.

    But I suspect games which those populations find have poor playability and experience as being emotionally flat would do poorly. If there are different conventions about what makes a game playable and emotionally salient, that’s probably a cultural issue.

    I’m ambivalent about the term “accessibility,” because its meaning isn’t sufficiently nailed-down. It could mean, “I got what it was about.” It could mean, “I didn’t have to think about it too hard.” It could mean, “I found it comprehensible,” or, “I thought it was entertaining.”

    In any case, it’s a subjective term that people aren’t talking about, and opinions of which, because of its vagueness, I can’t reliably impute to people based on the reviews they’re writing.

    “This doesn’t prove that other aspects of game design are less important; merely that they are less capable of making an undifferentiated audience responsive to your work.”

    If an aspect is less capable of making a general audience respond positively, that does mean it’s not important.

    Specificially, if factors B and C strictly determine whether a game is positively reviewed and highly rated — if games with “lots” of B and C will do well, and those with “little” will do poorly — then factors A and D, while they give people something to talk about, are stylistic devices.

  3. I think these observations are significant, Conrad. Thanks for taking the time to work through the reviews and extract commonalities.

    I would add one positive value for puzzles: Solving a puzzle helps the reader/player feel that he or she is _involved_ in the story, taking an active part in moving the story forward. I’m not sure one would get quite the same sense of involvement in a puzzleless game.

  4. Solving a puzzle helps the reader/player feel that he or she is _involved_ in the story, taking an active part in moving the story forward.

    Only if they solve it.

    Glad you dig the analysis.

    C.

  5. This whole series of “reviews of reviews” is pretty awesome.

    A few questions. First, I wonder if “playability” is not only a negative value, but a value secondary to what you call “emotional salience”. In other words, we only care for playability because it is necessary if we are to get the full thematic experience of the game. A sign is that, as you mentioned, we can tolerate a great deal of unplayability if the emotion or theme demands it.

    Second, I’m a little vague on what “emotional salience” means (though I suspect you have a precise meaning for it). There are few games I’ve enjoyed more than Bad Machine, for example, but I’m not sure precisely what emotion I’d say was evoked. It was more an enjoyment at working out a puzzle, together with the peculiar atmosphere of the factory. Perhaps “emotion” or “theme” means “internal response to something perceived”?

    Third (just to try to reconcile you and Victor) I wonder if we might say: “Playability” is a universal precept necessary for good IF, though sufficiently abstract that it applies to different works in different ways. But “Puzzles should be easy” is a corollary applying the universal precept to a particular “market” as it were, namely the judges of the IF Comp. Some, even many, of the judges are more of a literary than technical background, and so will be more easily stumped by a difficult puzzle than, say, the computer geeks that Infocom games were targeted at (who would enjoy the puzzle). Maybe?

    Anyway, really interesting stuff.

  6. Remember that IFComp is for short IF. You’re only allowed 2 hours of playtime to judge from. Me, I like long, sprawling, epic puzzlefests. But you don’t get a chance to vote on any of those in the comp. And the last time I really cared about an emotional IF experience was when Floyd died (the first time).

  7. Newbot – I’m not feeling sufficiently enterprising today to define “emotion” and “theme.” Nor would I care to define what exactly a puzzle is, for that matter.

    If you follow the “Foster-Harris” tag on my blog, you’ll find my summary of a 50’s-era writing instructor’s ideas of writing emotionally meaningful work. Or if you click on Emotional Logic (Aristotle) under “Resources,” you’ll see how he treats them. For themes, I’d refer you to Wikipedia.

    Playability seems to be its own thing. Yes, it’s true that we have trouble getting to the meaning of a game if the game is not easily playable: but _Eruption_ had no emotional salience at all — didn’t make the attempt — and yet many liked it well enough, strictly because it was playable. Meanwhile, _Beta Tester_ was nearly unplayable, and yet had an emotional signal, often called “snarkiness,” that caused many to like it.

    In terms of reconciling with puzzle-players — I’m just talking about what the general population seems to respond to. If people want to make puzzley games, or play them, I think they should. What I’m saying is, based on what I’m seeing, puzzles don’t add value for most players.

    Nathan said: “Me, I like long, sprawling, epic puzzlefests. But you don’t get a chance to vote on any of those in the comp.”

    But if there were enough people who shared your taste voting and reviewing, then presumably short, focused puzzlefests would out-compete narrative IF, which doesn’t seem to be the case.

  8. Specificially, if factors B and C strictly determine whether a game is positively reviewed and highly rated — if games with “lots” of B and C will do well, and those with “little” will do poorly — then factors A and D, while they give people something to talk about, are stylistic devices.

    Are you saying that there are only two kinds of property, namely (a) things that will make a piece popular among a general, undifferentiated audience, and (b) stylistic devices?

    If you give 100 random readers the Iliad and a Harry Potter book, I’m sure the Harry Potter book will get a higher average rating. I’m also sure that Homer’s greatness cannot be captured by the term “stylistic devices”.

  9. Concerning the first two points: I almost suspect I’m trying to twist your words to my own preconceptions. I’m inclined to agree that “playability” (a negative value) and “emotional salience” (or some such thing, a positive value) are involved in enjoyable IF, but I suspect we don’t mean quite the same thing by those.

    E.g. Eruption. As you say, it is good almost entirely in a negative way: no spelling errors etc. This is what most of the reviews address (setting aside discussion of “the rant” in the About text), sometimes they even call it a technical demarcation rather than a game (e.g. Super Doomed Planet, “As a demonstration this is nice, but [etc.]”). In this sense, they are almost always comparing it to more bug-ridden games. In other words, Eruption is quite playable.

    At the same time, a few reviewers will mention an absolutely minimal positive value to it. Merk, for instance, writes “I think even the author would agree that Eruption is only a snack-sized diversion en route to more substantial works.” Jenni at Pissy Little Sausages writes “I dunno, man. The game might not have been saying much, but it wasn’t entirely a void: the Jack character was almost two-dimensional, and [she then gives other examples]” These reviewers tend, in addition, to impute a minimal interest to it (Merk rates it “Limited Appeal”, and Jenni seems to have a mixed opinion; she calls it “not great”). To take a counter-example, you did not seem to find even a minimal appeal in Eruption, and you say “To my mind, these simply are negative values: We want IFs to be spell-checked and bug-free so that we can play them and enjoy the stories and solving the puzzles.” I.e., you saw nothing positive (not even puzzles) in the game, only negative values, and therefore it did not appeal to you.

    But it would certainly be absurd to call this minimal appeal “emotional salience”, as you point out. It may be relief by comparison to annual clump of buggy Comp games, or it may be mild amusement at learning the geography or whatever, but it certainly has no theme, or emotional investment, or anything of that sort. That is why I wanted to interpret “emotional salience” in aesthetic rather than specifically emotional terms (I’m being rather vague, sorry). It would allow even Eruption that minimal positive appeal some reviewers seemed to give it.

    Which is also why I suspect I’m twisting your words. You are not yet trying to come up with universal principles of good IF, which any reviewer _must_ perceive in a work if he is to like it, even if he may see it where others do not (similar to what you mention in Emotional Logic): rather these are specific conclusions gleaned from this specific Comp. And so by “emotional salience” you do indeed mean “emotional salience”: actual emotional investment or something of that sort. And hence (just to be daring: correct me if I’m misinterpreting you) although a puzzle game may be neither playable nor have emotional salience, that does not mean puzzle-players will not enjoy it: it just means the generality of players will not enjoy it.

    In which case I’m inclined to agree; or at any rate, you certainly have a strong argument. I’d even call it a discovery, if I were not so new to IF: I, at least, would never have suspected that, e.g., player agency has only limited appeal for the generality of IF Comp judges; but that is what the reviews seem to suggest.

    Of course then one wants to go on and ask: How do these conclusions about the generality of IF players apply to IF as such? Is the generality ‘right’ about what’s important and what’s not? Might not the very fact that it is the ‘generality’ result in a bias (e.g. towards easy playing or whatever)? Which seems to be Victor’s tack…

  10. Argh, I’m embarrassed at how long my post is! Sorry Conrad! I’ll stop writing so much, I promise!

  11. Victor: “Are you saying that there are only two kinds of property, namely (a) things that will make a piece popular among a general, undifferentiated audience, and (b) stylistic devices?”

    I’m saying that if a property does not have the power to make a game more or less well-received, then it does not have the power to make it more or less well-received.

    “If you give 100 random readers the Iliad and a Harry Potter book, I’m sure the Harry Potter book will get a higher average rating. I’m also sure that Homer’s greatness cannot be captured by the term ‘stylistic devices’.”

    Yes, and therefore the advice I am giving to people would not be suitable for an author who wants to write the IF equivalent of _The Illiad_.

    But then again, if we resurrected Homer and taught him to code in Inform 7, I don’t imagine the findings here on OneWetSneaker would slow him down.

  12. Newbot,

    I’m fine with long replies.

    “That is why I wanted to interpret ’emotional salience’ in aesthetic rather than specifically emotional terms (I’m being rather vague, sorry).”

    Well, that could be. It’s an interesting question what would constitute aesthetic value without involving emotional salience: I suppose you could could look at prettiness of the language, and so forth. There’s a well-developed body of work on that topic, which you could probably port into IF without much trouble.

    “…rather these are specific conclusions gleaned from this specific Comp.”

    These are general conclusions gleaned from this specific Comp. I’m presuming that any particular IF Comp will run about like this one. The cosmological principle.

    “And hence (just to be daring: correct me if I’m misinterpreting you) although a puzzle game may be neither playable nor have emotional salience, that does not mean puzzle-players will not enjoy it: it just means the generality of players will not enjoy it.”

    Some portion of the playing population may well enjoy a given puzzle, and for these it will add value. For the others, I’m not saying they will not enjoy it, in the sense of resenting it; I’m saying it won’t *add value to the game.*

    In other words, it seems that for every one person who finds a puzzle adds significant value, there are more people for whom the puzzle detracts a little value.

    And, even people who really like the puzzle will tend to down-rate the game if the other factors of playability and emotional salience aren’t there. Then you have to keep in mind that puzzles are costly of development time: you have to work out the puzzle logic, figure out the coding, debug and beta test the thing.

    So my conclusion is that it’s a better design strategy for authors to put their development time and elbow grease into getting the emotional salience and playability right.

  13. Conrad,

    Fair enough. I think I can agree with all that. And while I reserve judgment as to whether what adds the most value for most people is equivalent to what is ‘better’, I do understand (as you’ve been saying all along) that that is not your point. Your point is to discuss what factors will generally cause players to up-rate or down-rate the game, on the basis of how they reacted in this Comp; and to provide concrete design strategies on the basis of those factors. Which makes sense.

    One last question and I’ll be done. Do you think you’ll conduct a ‘reviews of reviews’ series next Comp? E.g. make predictions on the basis of your conclusions and see if they bear out? (I suspect I will.)

  14. By the way, yes I agree these blog posts are interesting! (Hence why I’m responding to a lot of them.)

    I really agree about playability, but I don’t think that means games need to have immediately ‘easy’ puzzles or no puzzles at all (although both are perfectly valid approaches). In terms of puzzles, it seems like some authors really didn’t get that a key job of the author is to help the player solve puzzles, not (just) to make puzzles hard to solve. This is not a new concept.

    For example, I forget what it was called, but there was an SF game where you play a kid in a space mining ship, with a mirror puzzle. In order to solve the mirror puzzle there is one basic concept you had to figure out… um… I can’t say this without spoilers, so stop reading now if you want to avoid them… it’s literally a thinking-outside-the-box solution. Now, obviously I may just be sore because I had to hit the walkthrough, but I don’t think this puzzle was well designed – even though actually, I think the core ‘oh duh’ solution was a pretty good one. The problem was that negotiating the puzzle was so unnecessarily complicated (large and unmarked co-ordinate numbers, angles multiplied by ten for no reason, apparently different angle systems, no display of mirror position/angle, etc) that players – or at least me – spent so much time fiddling with this and trying to set mirror angles to the nth degree of accuracy and trying different numbers blind to see the result, we didn’t notice.

    In fact, even after getting the key point from the walkthrough (so I knew exactly where I wanted to put the mirrors) I had to look at the walkthrough *again* just to figure out the values to set the other ones to (in order to achieve the placement I had scribbled in two seconds on an envelope). That’s just weak. If the puzzle had been clearly displayed, unambiguous, and easy to operate – in other words if it hadn’t given me so much stupid space to play around with unnecessary parameters in the ‘obvious’ part which wasn’t getting me anywhere – maybe I would have thought of the non-obvious puzzle answer. And maybe, having got the correct solution, I’d have actually been able to implement it…

    Another thing is hint systems. Too many games didn’t have a hint system (otoh, kudos to the boy-trapped-in-robot game for its mainly-not-needed-but-welcome hints and especially its funny non-hints). Walkthroughs are walkthroughs, but a good hint system means you still get some satisfaction from solving the puzzles, even where you needed a nudge. If you’re stuck, no problem. I don’t believe many people would rate a game down too heavily because the puzzles were hard, if they could get quality hints that still let them solve some of the puzzle on their own.

    Maybe that’s just how I feel and there are plenty of players who just throw up their hands at first sight of a puzzle, without even bothering to try – or alternatively, who consider using even the subtlest hint system to be as base a depravity as consulting the walkthrough, and so wouldn’t benefit from it. But I think on the basis of this year’s comp games, there isn’t really a way to gain strong evidence of that; most of the really hard puzzles were hard because they were badly done (game I mentioned) or because they weren’t fully hinted (the stage school one).


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s