Two-Click Interface - which way round?!

[AGA]

Adventure games shouldn't try to do anything new!  They reached their apex in the mid 90s, and any more modern thinking will only lead to bad things.

[Babar]

[ citation needed ]

Well, I meant in terms of minimalism in interface. Every time this topic comes up on the forums, the majority seem to be favouring it. It seems in this thread you're even arguing for a single-button interface?!
..or is it just a vocal minority?! I remember Vince did some sort of GTD thing on interface where he lambasted the Sierra interface (deservedly, I guess ) and favoured as minimal an interface as possible. What does that guy know, though, right ? He took 5 years to make 1 game!

I find that most less-common verbs can be replaced by an inventory item. For instance, rather than putting a dig command on the GUI, give the player a shovel. In the case of pickpocketing, a suitable item could be a sharp coin; this is a coin with its edge sharpened, so that it can be used to cut purses (hence the term "cutpurse").

What would be your suggestion for pushing someone off a ledge? Or pulling someone up from one after you've had a stern talking to with them about how they won't be evil anymore?

But I have also been known to code a special GUI with 40+ verbs on it, just to see if it could work

I think I actually nabbed hat code from that game .

If your game features interesting descriptions of everything (and I'm of the mind that LOOK descriptions are a vital part of adventure games!), and then you have an object you can interact with as well, isn't that automatically 2 interactions?

I'd suggest hovering or some such similar mechanism for use as "look", but I get the feeling tablet users will get annoyed with that .

[Stupot]

So, which is more important?  Look first and then act?  Act and possibly never look?  Does the player want to answer the question, "What's that thing?" before rushing to use it?  Is the game more about exploring before taking action, or taking action with optional exploration? I don't think this is a closed-case for every possible game, even if many people are used to one kind of control.  I can easily imagine a player being frustrated with the default being "walk over there and pick that up" rather than "tell me what it is first."

For me, it's "look first and then act", but I don't do 'look then interact, look then interact, look then interact' on every hotspot in the room.  Rather, when I walk into a new room, I like to do a sweep of the whole screen before I do anything else.  I click 'look' on anything and everything, just to get an idea of what is there, what is clickable and what needs doing.  Only then do I start interacting, start getting things done.  And Left Mouse Button is definitely the 'getting things done' button.

Making descriptions for "look" is more expensive in terms of time, effort, and memory (space).  We certainly don't want to put most of our expense into descriptions if they are unnecessary or skipped, and we certainly don't want to beg the user to please remember to look at items so that they can be entertained.  That would be silly, too.  

Understandable. If a game doesn't have descriptions of every object, I as a player won't necessarily miss them, and I can see your reasons for possibly wanting to avoid spending time on them.  But I personally do read them (regardless which button they're on), and I'm sure I'm not the only person who does.

As for begging.  I have written unique responses for 'look' in my game.  There are a few generic ones in there (a door is a door is a door), but there are also (I hope) some quite funny ones, and I do want people to look at them and potentially be entertained. Without them, the game is less funny, so I'm going to probably have a note in the readme gently urging the player to try looking at everything.

[Trapezoid]

There are big problems with a single action interface.

A. The player can't have a conversation with Nurse Edna before deciding to shove her down the stairs.

B. The player is robbed of that lightbulb-over-the-head moment. Instead of being rewarded for wondering if they can push Edna, they're merely checking to see what would happen if clicked on her, like they do with every hotspot, and the specifics of the resulting action are unexpected. The feeling of control is diluted.

C. Or, as a developer, you've decided from the start that clicking on a character always results in a conversation. Thinking on those terms, it never even occurs to you to make a more physical and entertaining solution. Edna is now talked into leaving her office. Boring.

I'm not advocating that every game needs the full 9-verb lineup, or that you can't combine Use and Pick Up, etc. But you need to encourage the player to have "Oh, what if I tried..." moments, and a selection of verbs gives them options to try other than "clicking on everything once."

[Eric]

If you think it merits its own thread, just start a new one and mention that the topic came up here.

I thought of that, but didn't want to overstep! I guess things are continuing here in new and unexpected directions, so I'll withdraw my request! Thanks, Snarky!

[LimpingFish]

Anyone wanting to make a modern adventure game should avoid using "classic" interfaces. They have aged...poorly, and only continue to make design sense to people who love old adventure games.

As to the original question, left look, right interact.

[Adeel]

I support those who are in the favour of LMB = Interact and RMB = Look. This is the most natural interface I can possibly think of.

I once played a game (can't remember its name properly) which used LMB = Look and RMB = Interact and it was very annoying for me!

As for the question whether look at command should be used or not. Look at command should be a must for adventure games, in my opinion. Only using a single mouse button throughout the entire game is too boring for me!

I personally look at everything first and then proceed to interact with it.

[StillInThe90s]

Being brought up with scumm and similar systems, LMB as "main" or interaction button feels pretty intuitive. Beneath a steel sky always felt a little backwards to me.

Slightly off topic but not quite: I'm currently trying out a setup using single LMB = walk/select+combine inventory, doubble LMB or single MMB = interact, RMB = look/clear inv item. No verb gui or coin. Just an inventory and mousover text. Is doubble click a good way to go?

[Babar]

I personally don't think, at least in your situation, that it is a good idea. While I can understand issues of wanting a separate USE and TALK, as far as walking goes, it doesn't seem necessary.

If you click on something specific, it generally implies you want to interact with it. If you're just clicking on the path, or on nothing, then it implies you want to walk there. You'll never really want to WALK ON the computer or something, and if you want to show the player walking to an item before interacting with it, you can combine those two into interact: the player first would then walk to the item, then interact with it.

I can't think of a situation where walking on to an object and interacting with the object should produce 2 separate necessary results- Pressure plate, elevator platform, door, etc. would all give the same result whether you WALKED on to them, or USED them.

[Radiant]

Slightly off topic but not quite: I'm currently trying out a setup using single LMB = walk/select+combine inventory, doubble LMB or single MMB = interact, RMB = look/clear inv item. No verb gui or coin. Just an inventory and mousover text. Is doubble click a good way to go?

I would not recommend that, it's not very intuitive. And not everybody has three mouse buttons.

[StillInThe90s]

@Babar: The idea of having separate look, walk and interact clicks was to be able to walk around without accidentally looking at or interacting with things. An annoying thing about later, slimmed games is that I often find myself clicking away text that I already have read or sometimes even solving puzzles by accident, which makes the gaming experience feel a bit unintelligent (i.e. Click everything at least once and you're done!). 

I can't think of a situation where walking on to an object and interacting with the object should produce 2 separate necessary results...

I can.  A situation where the player did not want to interact with an object but, for example wanted to walk past it. The game I am working on has use-able snow on the ground, which means that the character would be unable to walk around the scene without a walk function. But I heartily agree with you about not having a separate walk-to-object function 1980:s sierra style: "YOU ARE TOO FAR AWAY TO PICK THAT UP." -Not desirable.


@Radiant: The MMB would do the same thing as a LMB double click. It would just provide an extra alternative.

[Knox]

B. The player is robbed of that lightbulb-over-the-head moment. Instead of being rewarded for wondering if they can push Edna, they're merely checking to see what would happen if clicked on her, like they do with every hotspot, and the specifics of the resulting action are unexpected. The feeling of control is diluted.

I totally agree. It seems to me that most people here prefer the 2-click interface, and that's OK, but I personally really hate it . I always feel that all you need to do is just click on anything once (with the same mouse icon) without any real thought other than "lets see what happens": (hmm, lets try clicking on the waiter; will that initiate a conversation, cause the main character to order some more wine, or cause the main character to grab his tie and choke him to death?...click! Lets find out.).

Bah, I guess I just prefer having a little more "control" when it comes to choosing the possible outcome of interacting with characters + environment...plus I like the different little mouse icons for each action, hehe! 

[Thaumaturge]

Regarding the question of an interface that might allow multiple potential actions, as in having the option of either talking to or punching a character, I'm inclined to suggest a context-sensitive verb list/coin, similar to the interface that I recall being used in Gabriel Knight 3.

Essentially, interaction would be governed entirely by the left mouse button: When only one action is available, the icon for that action is shown (perhaps attached to the cursor in a third-person game) and a single click performs that action. When more than one action is available, a different icon is used (I think that I used ellipsis ("...") in a prototype that I made once upon a time), and clicking opens a small verb list or verb coin populated with the icons for the available actions; clicking on one of these icons performs the action. The verb list/coin can be closed without action by either clicking some icon -- perhaps in the centre if a verb coin is used, perhaps at one end if a list is used -- or perhaps by right-clicking, or clicking outside of the coin/list.

"Looking" could be incorporated either via right-clicking, or as an available action in the list/coin (although the latter might make the interface a little tedious -- it might be a nuisance to have to open the list/coin if there's only one action besides "look").

[Andail]

I've always been a bit unimpressed by the verbcoin, mainly because they've felt clunky and like an unnecessary extra step between me and my verb, but lately I've been swayed a bit.
I think a simple, well-implemented verb-coin is actually pretty nice as long as it's dynamic and updates according to what I'm clicking at. The ability to customize the verbs is pretty cool - I can both look at a document but also read it, or pick it up. I can talk to a person, or push him, or pick his pocket, or look at him, etc etc.

And when it comes to porting, the benefit of a one-click GUI is pretty obvious.

[Stupot]

Knox seems to be in a minority, saying he wants more verbs because he wants to feel that you can do more than just one interaction (or two if you include 'look') on each object/hotspot.  Other people are saying What's the point in adding 9 verbs or 4 verbs, when let's face it, each object/hotspot does generally have one interaction, so two-click is better.

What's the problem here?
Perhaps game designers are being too predictable  Why aren't we using the extra verbs as part of the puzzle?  Boxes and doors don't have to be opened. Levers and ropes don't have to be pulled, NPCs don't have to be 'talked' to.

The older games used to take advantage of this.  Not everything would take the most obvious verb; some lateral thinking was involved. In more recent times, games have shyed away from tricking the gamer like this.  I think for most people, if 'open' doesn't work on the box they'll immediately consult a walkthrough.  Perhaps this is why the two-click interface seems to be the preferred choice: people no longer enjoy playing this guess-the-verb game.  And this falls into the designers' favour because they no longer have to think of clever ways to use all those verbs.

[Snarky]

I don't think there were many games from the "verb inventory" era that used unexpected verbs with any frequency. Maybe a half-dozen times in one game, at most. In the end, it was such an exceptional thing that it almost felt a bit unfair: like the game was breaking the unspoken rules of how to play it.

Also, verb coins are the devil's own UI. Outside of adventure gamers, no one understands them.

I find that most less-common verbs can be replaced by an inventory item. For instance, rather than putting a dig command on the GUI, give the player a shovel. In the case of pickpocketing, a suitable item could be a sharp coin; this is a coin with its edge sharpened, so that it can be used to cut purses (hence the term "cutpurse").

What would be your suggestion for pushing someone off a ledge? Or pulling someone up from one after you've had a stern talking to with them about how they won't be evil anymore?

Probably dialog options. In the first case, I have a hard time imagining a scenario where you know a reason to push a character of a ledge but the puzzle is figuring out that you should do it, so the challenge would probably be more about either establishing motivation (navigating dialog tree? Using object on character to elicit the remark that causes you to push them over the edge?), or eliminating whatever obstacles there might be (arranging a diversion or simply timing it so witnesses are looking away, figuring out how to get up close, etc).

For the second case, just make the first option in the dialog something like a: "I think you've learned your lesson." b: "While I've got you here..."

[CaptainD]

I did quite like that they did with Fate of Atlantis (and perhaps others, but that's the game I really noticed) - in that when you were in the dark for instance, you had different verbs reflecting the fact that you couldn't see things very well


Stupot - I think you have a point, but I think from a developer's point of view it's also down to not having thousands of generic responses / having to write multitudinous responses for every mad possible combination of actions and objects that the player might try.   

[Babar]

Probably dialog options. In the first case, I have a hard time imagining a scenario where you know a reason to push a character of a ledge but the puzzle is figuring out that you should do it, so the challenge would probably be more about either establishing motivation (navigating dialog tree? Using object on character to elicit the remark that causes you to push them over the edge?), or eliminating whatever obstacles there might be (arranging a diversion or simply timing it so witnesses are looking away, figuring out how to get up close, etc).

It was more: Talk with the guy, he says something that shows he needs to die, say something to get him distracted (or say the same thing again if you missed the window to do it the first time), then when his attention is diverted, push him. So your solution would be "PUSH HIM" as a dialogue option?

But yeah, I hated the verb-coin as well .

I actually don't disagree with Knox, I wouldn't dislike multiple possible actions either. When I'm suddenly placed in a situation where I have to start reconsidering this cool puzzle I thought up (or to be honest, the ONLY puzzle I could think up to move the story along ) because the super-simplified interface won't allow for it, or I have to start thinking of unintuitive alternatives to implement it (I think Radiant suggested using a sharpened coin inventory item as a way to interact with a character who I could otherwise only talk to, so as to be able to pickpocket him), I realise something is wrong, and the interface is not working.
Personally I blame laptops and touchscreens for not having properly implementable middle mouse/scroll buttons. That would solve all my problems (until I come up with ANOTHER interaction mode I want that doesn't have space ). WE NEED A NEW INTERACTION DEVICE! Or some inventiveness to come up with a new interface. Gestures or something on touchscreens to allow for multiple interactions?

I'm not sure how unintuitive the Sierra Interface is for someone who's never used it, I suppose some ways to make it more palatable would be to minimise the number of clicks required to do something. So maybe have a sort of hybrid Lucas-Sierra interface. Have the 5 interactions somewhere permanently on screen, have keyboard shortcuts, etc.

This trend towards minimalisation is kinda interesting. You originally had the text parser with its theoretically limitless possible actions, then someone said "Hey, technically there's only a fixed number of things you can do, why not have them all in a drop-down list as well?", then from there 12 verb-list, then 9, then 5, then 3, then 2, now 1, and probably at every step of the way, the authors/creators complaining. I remember reading that the Two Guys from Andromeda REALLY didn't want to switch over from the text parser for SQ4, and fought that decision as long as they could, before finally accepting it, but sticking in the almost (game-progression-wise) useless SMELL and TASTE icons there as a response to "You don't need so many verbs, get rid of them".

[Slasher]

Hi

I don't think there is anything wrong with using the default walk, look, talk and interact options.

There may be scenes where an extra option could be added such as push, pull etc etc giving the player not only a clue but also other options that can be performed.

Having grown up on games of the 90's people nowadays are bound to strive for better options and stuff but if you can remember way back then it seemed simple but great fun and I think more and more people are out to try and impress us with there new ideas so much that they forget the reason why we all play games in the first place.

Having said that,  there are a number of members here that have it sussed

[Andail]

Alright... I think this whole "players today don't understand this and that" is getting a bit, I don't know, presumptuous. The trend now is that everything should be as quick as simple as possible, because apparently people can't learn new rules anymore.

If people can learn how to play games like Dwarf Fortress, or memorize all the gazillion skills in any given mmorpg, I think they can handle an extra click or two on a set of icons. If a game is good, people will learn how to control it.

I think Vince's manifesto has a lot of merit, but a game world isn't an operating system - sometimes it makes sense that the same type of object/character/hotspot should be manipulated/interacted with differently, and in a number of ways. Sure, we can get by with an interact/look GUI, but that's because we've adapted our games so that everything can be interacted with just like that, but that also makes our games increasingly similar. We mainly look at things, pick them up and then combine them with other things.

I have the standard RM/LM for The Ssmaritan Paradox, but there have definitely been moments when I would've prefered a wider assortment of actions - situations I've solved by having an extra GUI pop up (turn the handle clockwise or counterclockwise? things like that).

Mind that I'm not advocating a static set of verbs, like the MI-bar, but a dynamic set, just to give an extra depth to the player-world interaction.