At what point in time do you start thinking about how players will control the virtual characters in your game design? I''ve recently decided to throw some action elements into my magic cauldron of game design (it smells good, but looks horrible... maybe a pinch of ''learn how to program first'' would help), and I''m having lots of trouble controlling the brew. I find that it really helps to think about interface very early on in the design, as it''s helped me not only think of new possibilities, but more importantly it''s helped me to avoid making things unnecessarily complicated. "This is what I have to work with." Personally, I think that the controls of a game are the determining factor in my judgement of a game. Games can look great, and be a lot of fun to play, but if I can''t easily control it my opinion of the game will plummet. So, when do you start to think about how you want players to control your game? How do you make sure that it''s always easy to control it for casual gamers, but still allows for a lot of flexibility and depth for the true die-hards?

That''s usually the first thing I think about, but the second to last thing I implement.

I always base most of my code around how the data will flow from the game to the console.

~Dwarf

This is probably the third thing I look at after raw concept and general design. It''s also what''s been thwarting me for quite some time on my current design. I''ve ended up with, "yeah, this would be cool to be able to do, but how the heck does the interface communicate all of this in a fun and manageable fashion?"

Especially because interface impacts art and coding in addition to gameplay, I''m starting to see that it needs to be very high priority.

quote: Original post by Silvermyst
How do you make sure that it''s always easy to control it for casual gamers, but still allows for a lot of flexibility and depth for the true die-hards?

I wonder if coming up with "control tiers" would help? The idea would be that the game could be played easily, if not optimally, on the first tier, so that players got an idea of what was expected of them. If they wanted to truly excel, they''d need to investigate further options, which would lead them to the deeper tiers. You wouldn''t hide the lower tiers from the player, but you''d maybe provide for "simple" and "complete" interface options.

Combined with difficulty level, multiple tiers might let both the casual and hardcore have fun playing. "Simple" mode combined with easy difficulty would be beatable, but "simple" with hard would not. (btw, Civilization and Alpha Centari, two very deep games, have examples of this in terms of management of citizens and the concept of "advisors" that coach newbies on what to build)

--------------------
Just waiting for the mothership...

Usually, controls and interfaces (and I''d also put viewpoint into the same area) are around the third or fourth thing I start working on. And they''re a very important step, because they can change a game''s feeling pretty tremendously, and will create a lot of the genere defenitions that will be slapped to your game.

Wav''s idea about control tiers is a good one. Me, I''m a break it down into little pieces sort of guy. When I''ve got a lot of commands/actions that need to be done, I ask myself which of those need to be done at around the same time (moving and shooting in an action game), and which can be done in seperate pieces (buying guns in counterstrike vs moving). From there, I guess it''s always been a combination of seeing what works in other games, thinking about how I want my game to play, brainstorming, experimentation, and constant re-evaluation.

If you can get early prototypes done (even limited mockups in easier to use programs) you can test how it would feel to play through parts of your game, and see just what you need and want at any given time.

I''m sure there are people who have done studies on this sort of thing. Though one of th ebest pieces of advice I ever got was "Don''t make something change unless you want it to get the user''s attention."

Hope that helped.

quote: Original post by Silvermyst
How do you make sure that it''s always easy to control it for casual gamers, but still allows for a lot of flexibility and depth for the true die-hards?

I once said on this board "design for PC with console sensibilities". At the time, I meant "limit the number of ''extra'' buttons you throw in; integrate features with the existing control configuration and only expand that if absolutely necessary." It was (surprise) an action/reflex game, so designing around a game pad was excellent.

I actually employ the idea of control tiers in my still-ether control schemes. The idea is that there are a wide variety of triggers, modifiers, toggles, etc potentially available to the player, but only the most relevant (player''s choice) can be assigned to buttons/keys.

Example #1 - Basketball: I could have controls for specifying which hand to hold the ball in, how far out to lean, whether to raise or lower the ball, different types of shots, different types of passes, etc. The player can configure a number of preset schemes between which she can switch on the fly, so she might have set up a scheme which emphasizes passing for a point guard who doesn''t shoot all that well (*cough* Jason Kidd *cough*) but a shooting-/challenge-oriented scheme for an offensive player like, say, Kenyon Martin or Shaq. These schemes can then be associated with players (switching focus selects the last/default scheme for that player) or the user''s control device (maintain scheme across focus switches) - again, player''s choice.

Example #2 - Warsim: Completely dispensing with micromanagement and instituting both chain of command and unit initiative, orders are given by right-clicking on the unit(s) and navigating the pop-up menu to the appropriate command. By default, all the common and generic commands are available in such a fashion as not to clutter the pop-up (perhaps no nested pop-ups). Advanced users could configure their pop-up menu as well as create keyboard shortcuts, record macros, script events, etc, etc. To do this, every possible basic command has to be accessible somewhere and these commands have to be "combine-able" to create more complex instructions.


There''s also the concept of user-configurable "action maps" (in a totally different sense than DirectInput uses it) which invoke context-sensitive behaviors as defined by the user/developer. I outline the concept and provide a (bad) code example here.

quote: Orignal post by Silvermyst
At what point in time do you start thinking about how players will control the virtual characters in your game design?

Once I''ve gotten past basic character design, or during gameplay design. You can determine that the hero will conquer the Demon Knights by jumping up and then doing a reverse split kick, but if your users can''t enter the necessary combination, well... ''Nuff said.

OLUSEYI wrote:

quote: It was (surprise) an action/reflex game, so designing around a game pad was excellent.

I''m actually using the PS2 controller as the base for my current control-design. I''ve really come to like the way this controller works. I''m still debating if the controller is more functional than the mouse-keyboard combo, or if it''s just more functional for certain types of games. (the main reason why it''s more functional is that it only takes one finger to control a joystick, instead of three fingers to control the A-S-D-W keys, although I sometimes think that the PS2 controller joysticks could use some revamping to increase precision)

quote: There''s also the concept of user-configurable "action maps" (in a totally different sense than DirectInput uses it) which invoke context-sensitive behaviors as defined by the user/developer. I outline the concept and provide a (bad) code example here.

Yeah, I remember that thread and some others of yours (I think you posted one or two more about the subject of controlling virtual basketball players). I guess I must''ve listened to you, because subconsciously, I''ve included the context-sensitive behavior in a slightly different form. I''ve created a design where each button can have a slightly different result depending on the situation. I''ve devised some different ''modes'' (for example, ''combat mode'') where buttons take on specialized roles. But no matter the mode, each button retains its overall function. Button 1 does not change from ''control legs'' to ''control arms'' if the situation changes from ''non-combat mode'' to ''combat mode''. That sounds like a superfluous statement, but it is quite a challenge to
a) give each button its own personality
b) give the player maximum flexibility (meaning, giving the player the ability to do everything he wants or can think)

Creating a good controller-design almost feels like a puzzle.

quote: Original post by Silvermyst
I''m actually using the PS2 controller as the base for my current control-design. I''ve really come to like the way this controller works. I''m still debating if the controller is more functional than the mouse-keyboard combo, or if it''s just more functional for certain types of games. (the main reason why it''s more functional is that it only takes one finger to control a joystick, instead of three fingers to control the A-S-D-W keys, although I sometimes think that the PS2 controller joysticks could use some revamping to increase precision)

I''m with you on that. I actually dislike using my keyboard to play games as it gets enough of a pounding from my incessant typing (plus I''ve switched to a Natural Keyboard, which makes game-like control harder). One particularly good thing about the PSX/PS2/Gravis Gamepad Pro controller is that there aren''t too many buttons and yokes to control at once, and they''re fairly accessible. Some combinations are precluded for physical reasons (you can''t ask a player to use the d-pad and left analog stick on a PS2 simultaneously - and you shouldn''t anyway), but with 10 buttons you can do quite a lot.

My current sports design uses the top 4 buttons (L/R 1/2) as modifiers on the face buttons (which gives each of the face buttons five functions straigh away). I then allow for the combination of modifiers to yield secondary modifiers, but never more than two modifier buttons or it gets messy. That is, labelling the face buttons a, b, c and d for simplicty:

   ( ) (L1) (R1) (L2) (R2) (L1+L2) (L1+R1) (L1+R2) (L2+R1) (L2+R2)a   1    2    3    4    5     6       7       8       9       10b  11   12   13   14   15    16      17      18      19       20c  21   22   23   24   25    26      27      28      29       30d  31   32   33   34   35    36      37      38      39       40 

That''s a lot of triggers! I then allow for some of the actions to vary based on whether the button is tapped or held, though the control logic for that is proving challenging to make robust. What I currently do is maintain a counter within the input device object and flag a button as tapped or held based on how long it was down (scaled by clock frequency and user-selected sensitivity).

quote:
Yeah, I remember that thread and some others of yours (I think you posted one or two more about the subject of controlling virtual basketball players). I guess I must''ve listened to you, because subconsciously, I''ve included the context-sensitive behavior in a slightly different form.

Ha! Feel the power of my influence!

Uh... yeah.

quote:
I''ve created a design where each button can have a slightly different result depending on the situation. I''ve devised some different ''modes'' (for example, ''combat mode'') where buttons take on specialized roles. But no matter the mode, each button retains its overall function. Button 1 does not change from ''control legs'' to ''control arms'' if the situation changes from ''non-combat mode'' to ''combat mode''. That sounds like a superfluous statement, but it is quite a challenge to
a) give each button its own personality
b) give the player maximum flexibility (meaning, giving the player the ability to do everything he wants or can think)

Excellent. Each button should control a general aspect of the game so that the state-based changes are logical and make it easier for the player to figure out what to do and how to do it.

quote:
Creating a good controller-design almost feels like a puzzle.

Believe me, it is. I can''t count the number of input schemes I''ve thrown out because they were inconsistent, incomplete, insufficiently expressive... Even with all that I''ve posted here, actually selecting which "impulse triggers" to map to which logical control triggers is a daunting task.

OLUSEYI:

Hey, you completely forgot the L3 and R3 buttons (press down joystick left for L3 and joystick right for R3)! I actually think those are some of the niftiest ones I''ve seen, since you can push them WHILE still using the joysticks to control movement.

The system I currently have uses mainly R1/L1/R2/L2/R3/L3.

R1
-tap
-quick tap
-hold
R2
-tap
-quick tap
-hold
R3
-tap
-hold

That''s 8. Multiply by 2 (for L1/L2/L3) = 16.
Add to that:

R1 + L1
-hold
R2 + L2
-hold

That''s 18 total, without even thinking about the 4 directional buttons (which would need a function that can be used WITHOUT using joystick 1) and the 4 ''arcade'' buttons (which need a function to be used WITHOUT using joystick 2).

I don''t think I want to really create more combinations of buttons, but that''s still a possibility, because that could become part of the ''advanced control''. For rookies the controls have to remain simple and logical.

The R3 and L3 buttons will need similar functions for the ''tap'' and ''hold'' feature, because I think those particular buttons are a little harder to press than all other buttons.

Of course, the ''tap'', ''quick tap'' and ''hold functions of say the R1 button need to all be somewhat similar as well.

I still have the X, O, etc buttons and the directional buttons available, but I''m leaving that for the next piece of the puzzle. For now, I''m focusing on one element, and want to make sure that I can perfect it with just those 6 buttons (R1/L1/R2/L2/R3/L3).

quote: Original post by Silvermyst
Hey, you completely forgot the L3 and R3 buttons (press down joystick left for L3 and joystick right for R3)! I actually think those are some of the niftiest ones I''ve seen, since you can push them WHILE still using the joysticks to control movement.

I''m actually on a PC (Gravis Gamepad Pro) so I don''t have those two yokes and their buttons. How do you distinguish between tap and quick tap though? Do the Sony libraries support them (you are working on a PS2 game, aren''t you)?

quote: How do you distinguish between tap and quick tap though?

Well, it''s not so much that it needs to distinguish a quick tap, as much as that it needs to be able to accurately count the number of taps. In other words, when I press R1 quickly 6 times in a row, the R1-tap function needs to be carried out exactly 6 times.
And yes, for now, the game will need to be for PS2, unless there is (will be?) a controller identical to the PS2 controller for the PC. I simply can no longer do with less than the number of control combinations that the PS2 controller offers.