Notgames Forum

I am looking into the usage of scene based narrative a bit and think that it can be quite effective. What I mean by scene based narrative is that the player is confined into a smaller area and after some kind of criteria is fulfilled the game moves on to the next scene (there can of course be branching scenes and so on). Example games of this would be Heavy Rain and Photopia (http://en.wikipedia.org/wiki/Photopia).

The reason I think it is a useful (meta?) design is that it confides the space of possible interactions, so it is easier for the player to navigate through the game. It also acts as a sort of engage and reward feature. Players feel "progress" when coming to a new scene and feels pushed towards continuing the game since they want to know what the next scene will look like. It also mimics the way many books, films and plays are built up where the scene-design is a very effective mechansim.

First of all I would like to hear your thoughts on this. What do you think of this kind of design? Do you see any drawbacks by using it?

Next, I would like to hear your thoughts on what kind of progression criteria one should use. More "gamey" games could have a puzzle needed to be solved or some other kind of obstacle, but that would not be in the spirit of not games ;) Instead I am interested in mechanisms that does not filter players (in the way a puzzle or skill based challenge do), but instead arer straight forward focus on improving the experience of the scenes. The most simple would be to just have "Next Scene" button and allow players to progress when ever they feel like it. Another way is to require some simple interactions to be made and one could highlight these so they are impossible to miss. To me the previous two ideas does not seem very immersive when just implemented in a straightforward fashion though. Perhaps one could hide them a bit, but how obscured can these things be until they become puzzles? Any other ideas?

Finally, I am very interested in hearing examples of games that use this approach and your thoughts on how the scene-based design worked in these.

Is this related to the String of Pearls (http://staff.jccc.net/mfitzpat/storystructures/htdocs/pearls.htm) structure?

When we were designing "8 (http://Tale-of-Tales.com/8)", we had a big open world that players could explore. But because we were worried that they would get lost (and give up) we added obstacles and puzzles to force them to explore the space bit by bit. Gradually the world would become more accessible and in the end everything would be open. Most of our obstacles were closed doors or blocked passage ways that you had to open by solving a puzzle.

I actually like the idea of having a clear interface on the screen for "Next Scene". I think it might not be a bad idea at all. Like turning the page of a book.

The best way to solve any design problem, in my opinion, is to create the circumstances so that the player voluntarily chooses to do exactly what you consider to be the optimal play style. So perhaps, instead of thinking about means of preventing the player's progress, you can try to think of ways to encourage the player's desire to stay in the same space.

Maybe a story is told in that room and if the player wants to hear it, they need to stay in the room. Maybe an item that the player would like to have appears in this room once in a while. Maybe the player can change the appearance of the room (clean up, for instance).

In Amnesia, you make the player look for hidden items. This does not necessarily prevent them from going to the next room, but it does keep them in the current room a bit longer.

Yeah string of pearls seems pretty much like it. Although he examples are not games I would consider good examples :) The main thing about each scene (to me at least) would be that everything should be very close spatially and would easily give the player a good overview.

Regarding the "Next Scene" button, I guess it could be disguised as a sort some sort of event that the player knows is final. For example: "Click here to enter escape pod" and then just letting the player experience the scene until they feel satisfied. I like your suggestion that instead of blocking the path forward, one could let the activities surrounding seem interesting. A good design might then be to make the "Next scene" button not so interesting at first but more and more attractive as the scene is explored. To go on with the "escape pod"-example, the player could notice how the ship falls apart when it is examined more closely, giving the message that entering an escape pod is a good idea.

I am really interested of doing this in a way that:
1) Makes sure that the player is never stuck (in many games, even without puzzles, one can get stuck at "guess the verb" or pixelhunting).
2) Does not keeps the player at scene through meaningless or additive awards (like achievements)
3) Still makes it possible to tell a story.

I see a little dilemma though. If one has a room where there is a certain number of possible interactions to be made and the player can choose to exit at any moment, what shall one do to limit the player's feeling of missing out on stuff. Either one can make sure that all that can be done is super obvious, but then I think a lot of the immersion is lost (just a matter of working through a list). Or if it is always unknown if all is done (and what can be done), and the player might a) get stuck looking for more interaction b) leave the scene with a feeling of missing important things.
It seems to me that these two things are opposites and fixing one makes the other worse. Or is there some way of solving this?

As for Amneisa: Hidden items work really well to force some exploration, but it also seems a bit cheating to me, so I hope to improve for upcoming games. Nobody seems to mind it, but I guess people are just indoctrinated into gaming standards ;)

I think it's ok to cheat. :)
Because what you're suggesting is a big cheat anyway: imposing a linear structure is cheating!

Ideally, for me, interactive entertainment is open. Architecture is my model, much more than literature or film. Architecture does guide people's movement but rarely enforces it (and when it does, it has narrative implications!).

But it will take us (and the audience) a while to get comfortable with such a model (if we ever will).

You should play The Path (http://Tale-of-Tales.com/ThePath). :)
Because it contains lots of interactions and no puzzles (apart perhaps the minor issue of provoking the Wolf, if you can call that a puzzle). Of course, The Path is about getting lost and feeling helpless. So it doesn't need to entertain the player all the time. (which is a cool side effect of the horror genre ;) )
The thing that happens in The Path is that the players make up their own story as they play. And they are aware of this. So playing becomes exploration of your own interpretations more than of the game world. And that way, everything you discover is something gained.
Maybe you'll find it inspiring.

In response to your 3 points, I'd say that if you don't have obstacles then the player can't get stuck. So that solves point 1 and 2. ;) And as for 3, I'd suggest to let go of plot and backstory and simply fill the world interesting things to see and do. The one problem with this approach is that some players will rush through, thinking that they're playing a game and that they should try to reach the end as soon as possible. I guess you could solve that by not having an end. The player will often still blame the designer, though, and not his own silly behaviour. So this solution generates new problems.

As for your dilemma, do you think the player would still feel like missing out if none of the interactions give them a reward? If the interactions would simply be amusing or interesting in and of themselves. Then it's more about the player choosing their own experience, rather than missing out on what the designer intended the experience to be.

Of course, lots of players are not capable of taking such responsibility. Because they still think in linear terms and are actually mentally passive when playing a game, expecting the designer to serve them everything on a golden platter.
Oddly, they don't expect this from other software, like Photoshop or their web browser. I firmly believe that videogames should be more like computer programs than like movies.

Anyway, for all of this to work, we need to let go of fixed stories. Because if we have a concrete story to tell, we will always feel the need to force the player to hear it (or worse: they'll need to hear certain bits in a certain order). That's a negative approach to design. A positive approach would be to offer opportunities for the player to grow a story in their minds. And then the task of the author is to make sure that this story is interesting, by giving the player the right ingredients.

Need to give The Path a stab one of these days. I have been meaning to for a year or so :)

While I do think that back story and forced narrative are not bad things (and I guess you do neither), I agree that one should not be to hung up on them (as in first create a story and then find a way to force the player through it). Although I have written about this, I seem to kind of stuck in in myself. Once the fixed story is removed a lot of problems go away, such as "what if the player fails to notice this!".

Perhaps each scene could be seen as some kind of experience chamber that shapes itself after the player's interaction. As a sort of play mate (like you have been talking about earlier). That way it should not be a matter of having a set interaction space to explore, but rather to make it inviting for interaction and then provide interesting feedback for the player. Once the experience chamber has nothing more to offer, the player should feel compelled to move on.

Now this sounds all nice, but the problem is of course (as always) to implement it properly! Having a game shape itself after the player is far from a trivial problem, so I guess some sort of cheating is must. And as you say, I am cheating already by imposing a linear scene structure :P The big problem, as I see, is getting away from having an set number of interesting things that the player simply goes through and instead as this personal experience to it. That is something we have covered elsewhere so skipping that for now.

We have to beware of "technological utopianism"! It's almost a reflex for us to think that all problems can be solved by technology. That all that is needed is more and more sophisticated technology. This is not necessarily true. Especially because of the nature of the problem.

In this specific case, the nature of the problem is artistic. It should be solved through artistic means. Through cheating, indeed. ;) We do not need to really create a living machine that responds to the player. All we need to do is activate the player's imagination so that they feel like they're experiencing a uniquely meaningful story. I really believe a lot of this can be accomplished through suggestion (and you do a lot of suggesting in Amnesia, for instance, especially through the sound!).

I think we need to think of the player as an active part of the experience. Many videogames are designed as perfect machines with one cog missing. And then the player is expected to play the part of this cog. Instead, I would advocate trying to build the machine itself in collaboration with the player. Since our technology can never be as sophisticated as the player, we need to somehow encourage the player to use their imagination, and to play along.

Children do this spontaneously when they play with dolls or cars. They know very well that a wooden block is just a wooden block. But they also know that they will have a lot more fun if they pretend that the wooden block is a car and they slide it over the walls and ceilings. If we can get the player of our videogames to take that step, to willingly and actively play with our software, then we can accomplish amazing experiences with relatively little effort.

I'm not sure what the videogame equivalent of a wooden block is. But it sure gives me something to think about. :)

As for a scene based structure, I really think it's best if it makes sense in the story. Otherwise you end up forcing it too much. There should be a plausible reason why the player cannot go the the next room, a reason that makes sense in the story. The simplest device that is used in videogames is the hidden key. But most of the times, there is no plausible reason why the key would be lost, though. So the player is yanked out of the story. A lost key should be meaningful (or inspire imagination), I think. Otherwise, just leave the door open.

If the story does not support a scene-based structure, then I'm inclined to think that it's better to not pretend that the structure has a narrative reason and instead to add a simple totally artificial layer to the game. A completely honest interface that tells the player that they cannot progress yet. I'm even thinking that this could simply be time based. You just tell the player up front that they have to spend a minimum of 5 minutes in each room. Depending on the atmosphere you want to create, you might even add a maximum time as well, after which the game automatically transitions, for instance.
This way the player is completely free in terms of interaction and there's no artificial puzzle to solve. So they can focus completely on the fiction.

Yeah, not being frank about things is really bad... I think a lot of people who might initially be interested in video games are REALLY turned off by that.

DESIGNER: Here's this fantastic world, do what you want in it.
PLAYER: Yay! I want to sit at the chair.
DESIGNER: ... except that particular thing! And that... And that.

To some extent it's about communication (telling the player what to expect), but also not trying to "pull a fast one".

Can you think of any games that are this honest?

As far as I can see, the only thing that seems to send this message is the graphic style: if it's stylized and cartoony, you don't expect much interaction, while if it's realistic, you do. Though that might be conditioning in part.

Sometimes it's fun to try things, to figure out what works. But often it's just frustrating and takes you out of the story. Maybe we should make it fun again. :)

Interesting point about honesty and being very explicit about the artificial constraints we are using. It's probably a good idea at this experimental stage.


This is a great quote, by the way. We should do something with it. ;)

Shall we start a forum thread for "The Collected Sayings of Michaël Samyn"?

Eh, maybe I'm misunderstanding but my little manuscript actually tried to show how most games are being dishonest. They show you stuff that looks like it's working (like a chair for instance) but then you can't use it.

I think there are a lot of ways to be clear about these things (instructional text, graphical clues, etc). But in the end the best thing is to just leave them out and build the game with components that just don't need explanation, I think. Or are we missing out on thing then?

About trying things (as a player) -- tinkering and seeing what works is really fun in certain contexts (and hell in others). Partly it must be about familiarity: If we are in a kitchen we want the stove to work and the fridge door to open (= things should "just work"). If we are in a room with a strange panel with buttons that activate different sounds we have no expectations and are more open about what should happen. There's a good reason so many games are set in a fantasy world.

 :D I'd prefer a forum thread for "The Collected Doings of Everybody Else" ;)

(meaning that I hope my ideas inspire people, because they something stifle me -they may sound good but I often don't know what to do with them...)

Link please?


Hm. But what if the working of the stove is irrelevant to the story?

I really like this thought of scenes. It is certainly something I might look into for my final project. I remember a concept I had for a walkabout game in which you would walk through the house of the media magistrate in The Fountainhead, beautifully designed, and then you would hear radio flashbacks and see newspaper clippings about the past, leading up to the 'inevitable' end of the character in the book. A bit like Salome but more wordy - did you call that a vignette, Michael? I remember you used a nice word for it.

I think an important issue is that the purpose of a place or scene is dictated by expectations. So if the player thinks he starts a game he will look for puzzles. But you can tell him there will be no puzzles, naturally. Downside is that all other media have some limit; film, books, poetry; all have a narrative. Even many walks through the woods follow a route that has a beginning and end. So just dropping the player in a room (or field of flowers if I get my way) will not really communicate a goalless system. But again, you can just tell the player. "Enter the field, leave whenever you will. You can walk around but only Lucy knows the names of the flowers." Sort-of how I start my game currently; "You play the locomotion of Lucy..."

If we are going for realistic narrative I doubt there is any reason for keeping a player in a room - one more lost key and I will personally scream. But in a game such as The Path or Salome the narrative takes a step back, like in a poem. So we could. But;
It indeed is more about the fantasy of the player. Why not let him go if he wants to? In the Netherlands there is an amusement part sort-of like Disneyland but themed around little humanoid creatures (silly Hobbits) called 'Laven' who have their village. There are no attractions I can think of in that part of the park, but as a child I used to love it. In fact, I remember that part of my youth with a huge charm because that little village was somehow 'real' and the attractions were more just excitement. In that area you could stumble upon a bakery where the Laven were making (delicious?) Laaf bread and it would feel like you discovered it yourself; then you would run to your parents (no more than few feet behind) and tell them what you saw and invent a wholelot more. There is also a fairytale forest with just fairytale characters - again with no attractions one can enter. If people enter the park and walk through this area because nothing prevents them from stopping then that would be their loss.

If Disneyland is an excellent example for leveldesigners (as it to my annoyance so often is used) then perhaps The Country of the Laves is our example? A wonderful place of exploration?
Lets all just go there!

I had an idea that started a while back; I am a great enthusiast of the nude side Domai, which is nude but without being sexual. It propagates shameless enjoyable beauty in nature, a bit like naturalism; from the viewpoint that nudity != sexual. Over the past months I had some discussions with Brits about this subject, starting with 'haha, silly prudish Brits' going to ':'(', so I thought of how to use games to challenge this notion. Along came a game set in the sort-of scenic water/sun-lit rock environment Domai often has, in which you walk a (clothed) woman to a bathing spot, where to the player's (feigned) surprise she undresses. So far things to many would be 'sexy'. Then the game would just be about... sunbathing, rock-climbing, swimming. To the point where the player has played in a single area as a nude woman for so long; having no goals; no options that are vaguely sexual; so he would sort-of get simply the enjoyable feeling of the experience of beauty. It bothered me for ages how I could make anything without goals. If I plant a flag at the top of a rock and tell the player to go get it then I have just made a normal game but with the perverted angle of having a nude avatar. If I make the game simply with a nude woman without goals then the experience itself becomes the main part.
When I just read this thread I realize this game is just a scene in which you can spend as much time as you want, looking for all the interactions. Perhaps she can pick flowers and put them in her hair? Carry around lilies? Dry in the sunlight?.. Doing a deferred interaction like in The Path would be very good. Specifically, I just realize this game is a mature Land of Laaf in which you get a scene that is rare to obtain in real life (especially since you control a character that is not you) which you can just explore.

Having a simple scene would make what the player does the focus. Just having an enjoyable environment and an enjoyable character to do it with. I do not need a timer or goals. If you create a pleasant environment I will look around it. I certainly want a mature Country of the Laves.

What if the interaction that gives you the means to go to the next scene is simply very interesting in and of itself? More interesting than getting the reward.

This you need to expand upon :) Are you talking about a sort of like an engaging conclusion to the scene? Or do you just mean that the most interesting interaction (which players will be drawn to) is also the interacts that locks up access to the next scene?

Sorry, it was a quick note that I made via my iPod before going to bed.

I was thinking that rewarding the player is not a bad thing in and of itself. The only problem with rewards in many games is that you do a lot of interactions only because you are given a reward. And then you're interacting with the system, and not with the story. And we don't like that.

But the reward system is a very good one for pacing the experience of the player, as in the scene-based narrative structure that you're suggesting.

So it seems to me that we can have our cake and eat it by simply making the interaction itself more interesting than the reward you get (i.e. the fact that you can progress). So that the player genuinely enjoys playing and the reward feels like a kind of bonus (that may even surprise them).

A great example that comes to mind is Ceremony of Innocence. It is structured as a linear series of postcards. You first see the front of the postcard. Some elements are interactive. It's fun to interact with them. After some specific interactions, something happens in the scene on the card. Then the card flips over so you can continue reading the story (and get access to the next card). But the interactions with the elements on the picture side of the postcard is where the gameplay is. They're often so pleasant that getting the reward is sometimes disappointing (which is subsequently made up for by the clever story and the excellent voice acting).

I think the kind of "free-form" interactions that were designed for CD-Roms like Ceremony of Innocence is something we can learn a lot from. Especially people like you and we who like working with 3D simulations. We have a tendency towards realism that is really not required by the medium. But realistic interactions can be very banal. Maybe letting go of realism is one way to make interactions more interesting in and of themselves.

Michaël Samyn:
Good point! I really like the idea of a player thinking:
"nooo, is the scene already over? Do I have to progress now? Can't I just play a little longer? Pleeeeeassse" :)

To bring in another "game being a play mate"-metaphore

That's a great emotion to aim for in design! :)

I think Crayon Physics Deluxe works like that a lot actually. Many of the levels are really easy to beat by building a really boring solution and just pushing the ball forward. Any one who just runs through the levels in that fashion will surely get tired very fast though. The second you start trying to come up with creative solutions the whole game changes and become so much more enjoyable. Basically you could stay at the first level forever, playing around with the system. And since progress is so easy it could be classified as voluntary and (kinda) immediate.

I remember from Petris talks that he first tried to fight this phenomenon (which is a natural instinct from a game design perspective) but later decided to instead embrace it -- the game is so much better off because of that!

Aren't a lot of games already a scene-based narrative?

A lot of FPSes essentially have you going through very fancy hallways, which block you and force you to clear them in order to continue.  Large portions of Half-Life games are a good example because they're more narrative-focused, but it's very common for almost all FPS games.  Non-scrolling platformers are even better examples - each screen is a scene based on physical limitations.  Kyntt is a perfect non-scrolling platformer example of a scene-based narrative.

Maybe I'm misunderstanding the terminology, though.

My thought was more that there would be very little (if any!) going back for forth between scenes. Also, and idea was the exit of each scene would be fairly obvious and the player could leave when they felt like it. This means that the player could remain and explore a certain scene until they thought it was enough and then move on.

I agree that games like God of War, Devil may Cry, Painkiller etc have this kind of design where they lock you into a very short area and then throw enemies at you, when enemies are killed you can move on. These are a very simplistic and extremely constrained examples though, but the idea is essentially the same. What I am interested in exploring is using this for more not-gameish works and I think that using a limited environment and a clear (or at least easy to find) exit could allow for a good base to build a large variety of different experiences. The idea is also that the scene is very non-linear with many different things that can be done. For example, one simple scene could be a kitchen where the player could try out different equipment and then exit by just going through a door.

Hope that makes it a bit clearer! :)

Ohhhhh, ok.  Gotcha!

So you do each scene once, but you can choose whatever order.  Yeah that makes sense since you have been talking about the idea of experiencing something a second time kind of ruining horror.

I've thought about this too for what I want to make.  But in my example you could go through the scene multiple times.  It'd be nice to figure out a way to present the scene slightly differently each time, though...  Maybe there'd be different details shown?  Anyway, maybe getting off topic.

Thanks for clearing that up. :)

I like how it is done i Forbidden Siren. You revisit the same level several times, but using different characters and things you did on past visits can influence what you see. Also, since the time of visits is not chronological things you end up doing stuff that you previously encountered too (if I recall correctly) :)

I also like the idea of revisiting places but adding slight differences.

What exactly is the advantage of structuring a game in different scenes (as opposed to, e.g., having one big scene)? Is it a way to impose some linearity that can then be used to tell (linear/plot-driven/arched) story? Is it merely for technical reasons (reducing memory requirement and loading times e.g.)? Is it a way to easy the player into the game by limiting the options available at any moment? Or are there other advantages?

For my part, the main reason it is to make user focus and not feel overwhelmed. For example, if game requires the player to explore a house, then just letting the whole house be explorable from the start might feel like "too much" for the player and faced with all possibilities they decide to do nothing instead. The scene based narrative solves this by letting the player explore one place at the time and thus confiding the player to a smaller area and letting her explore without getting lost. This can off course be solved in other ways, such as the metroid style where new areas are blocked until some condition is met, but this easily makes the game more complex and harder to handle.

Another advantage is that it makes it easier to handle change. For example, lets say a game takes place at one location and changing scene is simple going forward in time. Then for each "scene", it can be change according to things made in the previous scene. An example would be a dinner party where the first scene is the appetizer, next the main course, and so on.

It is also easier from a design perspective and lets you easier plan what things that can interact with one another. If the player can explore a huge world then there is no telling what strange things they might do that could break the experience.

Of course, sometimes having a large world might be preferable, but I think scene-based-narrative solves some problems that might arise.

Your reasons are very sound.
The big problem to me with a scene-based structure is that it imposes linearity. Is that something you're trying to solve? Are you trying to find a way to have a non-linear scene-based structure? Or do you like the linearity?

It depends on the game. I can think of game ideas where linear would not be good. But for the ideas that are whirling around in my head now, imposing some sort of linear structure seems appropriate. What I want to be able to do (given my current ideas) is to give the player as much freedom as possible and still be able to focus on conveying a certain message, theme, etc. I believe that this limiting of action-space can be good for gameplay (as I explained earlier, it does not overwhelm the player), but I also believe it is good for design. The way I have worked on games for over 10 years is to be able to have write down a synopsis and use that in order to plan as much as possible. While I like the idea of the game flowing from some simple start, this is not the way I am used to working and when I have tried it, problems have arisen (it was not until we made proper planning for Amnesia that the project really started shaping up).

I guess it depends a bit on the whole authorship thing too. This is really another discussion, but I can kinda sum it up here: When working on a game there is usually a lot of people involved, even for a small team like us, we have a core team of 5 persons and then the same amount (if not more) of freelancers that handles misc stuff. This means, that once content is made, it is hard to have artistic control and I take on the role as a coordinator more than artist. Sure, there is a lot of tweaking that can be done and tons have been made in projects, but it is mostly on lower level as consistency needs to be kept and there is always fear of the dreaded "domino effect". This means that I would like to be able to "see" the game before it is implemented and to limit the scope of possible interaction it lets me do that. If I would have just let it all flow from a simple base mechanics, it would all be a mess and very hard to keep track of. I guess that sort of development works if you are very few working on the project, but they way we make games now it does not.

I think this is a large reason why many (all???) major game productions feel so "soul-less". Even in games like Bioshock, which has an interesting concept, things are extremely shallow and nobody really wants to take "responsibility" (just check interviews with Kevin Levine where he insists that the game is just about having fun and any message is accidental)

As a final note: I have recently began writing a short story (in Swedish! :P) whenever I have time to spare and motivation. There is a vast difference on personality to this work, compared to when I work on game. Even though I pour a lot of myself into a game, I can kind of distance my self from it and not take it so personally. But with a written story, it is very personal and it feels more like a work that I really have to stand for and defend. I would like to have this in a game and to feel that I have really put part of myself into it. Amnesia is a step close, but far from the extent I would like. This does not mean that I need to design everything myself, it just means that I need to be able to "project" myself into at the design process and then make sure it sticks when production starts.

Now this became a really long rant. Sorry bout that :)

Don't be sorry. It's a nice read.

I understand how a scene-based structure makes projects of the size that you're making more manageable. I do think that you should be careful that the structure is not telling too much of the story (at the expense of the author). So I can imagine that you need to do a lot of work on the narrative to make it work with this structure. And I guess that some stories would not work at all.

I'm not arguing in favour of an open world, or letting everything emerge from a simple mechanic. I'm arguing for adapting the form to the content and not vice versa.

I hear what you're saying about the distance you feel as an author creating games (vs writing a story). We actually created this distance on purpose when we were first starting to make games (after years of making deeply personal and almost exhibitionist art pieces). But now we're slowly moving back towards a more personal approach, while still maintaining a desire to communicate with a wide audience.

Interesting turn about 'designing everything'. My current game is practically all by me, and I have reached a comfort zone where I realize that to make larger games I need to get more people to work on it - whilst fearing letting others in who may not understand what I am driving at.

The whole issue of authorship in games is poignant, large productions all seem quite soulless. The small quirky indie-games with their small quirky flaws are far more 'warm'. I think large games should definitely move towards authorship and try and say something. Perhaps modern culture is against this to begin with? I hear so many gamers who say they do not want clever stuff. They'd rather have a re-hash of robots from the 70's.

Interesting irony in interactive media is that if you let it be goalless the player will always assume he may have missed something - and if you give him a signal when the goal has been fulfilled he cannot function without having that final goal. I suppose the better path for a scene-based game would be to make clear the means and nature of interaction (the explorable elements) and let the gameplay be in the manipulation of these elements rather than the exploration. Otherwise there will always be the fear of not having pixel-hunted properly - I hated that most in old adventuregames, I like puzzles but they would always rely on a factor I simply did not notice - and the environments were nice, but I would rather spend time in them because they are nice (as I am prone to do in Riven) rather than because I have to find a button about 10 pixels large (which happened, also, in Riven).

This is the major problem for me. Because when you clearly mark all the interaction options (as in a branching dialog system), it is easy for the player to go compulsive and feel a need to go over evey option. At least this happens to me in adventure game where I fear to miss vital clues to puzzles or storyline and goes through all options even though I would not really want to. But as you say, if the interaction possibilities are hidden, then it is easy to miss out on things or (as you say) give the feeling that they have missed something.

Is way out of this to explicitly show the importance of all interaction, and thus telling the player "you really need to check these things out, but these other things are just minor and can be skipped"? Another problem arise though with having everything visible and "in your face" interaction areas; the player does not feel that they are exploring themselves, but just browsing through a designer's script. Perhaps I have no need to worry though and that a player simply ignores the visibility of the action points?

I have a suspicion that any comment passed on what the player has done may become an evaluation: the player will search for a 100% or at least a 'pass' at 40%... In The Path there were 144 flowers and I never found them all, even if I did spend 5 hours on attempting that. The question is whether or not that could have been accomplished without setting me up for an ultimately impossible goal (I do not dispute 144 is possible, but it is highly unlikely to stumble upon that final flower).
Yes, the developer is sort-of imposing his 'script' again. Which can be OK at times, I suppose, but for the player it must be tempting to ignore his own satisfaction and validate himself by the opinion of the developer.

Some games work well without explicit validation; the original Zen Bondage measured how completely you covered an object with rope up to 100%, but never acknowledging that was the goal. So an implicit measurement of completion may be good. But perhaps this is just window-dressing the same thing to make it less obvious?

I wonder where the point is when we go from 'toy train', which has almost no complex afforandces but provides a long-winded enjoyment to a child to the point where make a choice between saying A and B and feel unnerved we cannot do it both.

Perhaps it actually just is neurotic on the side of the player as well, because lately I have noticed that if a game is not explicitly about finding all the different endings (but still provides them) I just play it once and suppress the impulse of playing it to 'find everything'. Perhaps the enjoyment must be in the actions themselves, not in the result of the choices? I say it may be neurotic on the side of the player as so often with playtesting games the player looks at you as asks if he is doing things well - to which the only real valid response is: "I don't know - are you entertained*?"

* Entertainment in the old sense of the word

Duchamp once said "Il n'y pas de solution parce qu'il n'y pas de problème." I believe you can turn this around as well: if you stumble into a problem, remove its cause and you don't need a solution. In this case: if you remove puzzles and storyline, the problem dissolves.

I also think that everything in the game should be worth doing or it shouldn't be in the game. So you should be happy if players try to find everything. It's their choice. Just make sure that you never require it. So they know it's their choice. They are smart enough to know that if they play like that, they are giving up some of the visceral immersive enjoyment of the game in favour of another type of enjoyment.