next »

[Steve Swink]

G'day gents!  So, I'm writing a book about 'feel' in games for Morgan Kaufmann. Here's the sample chapter I've been working on:

http://www.steveswink.com/Misc/Swink_SampleChapter.doc

There are a couple threads around here where people deconstruct various mechanics in  surprising detail and argue over the most amazing minutia, so I thought this would be a great place to post this.  I think everyone ignores indies' opinions on things like mechanic design and feel, but some of the most insightful discussions I've seen on those subjects are here.  And, come to that, some of the best feeling games come out of the indie scene.  So, word, would love to hear your thoughts. 

I’m at the ‘getting people who know what they’re talking about to review and give me feedback’ stage, so I’m interested in questions like the following:

* Is this a valuable thing to be writing about?
* Would this be useful for students and/or other designers?
* Is my approach sound? Is there anything I’m overlooking?  What’s working well and what could be improved?
* What else would you like to see addressed, if anything?
* How well does this stuff comport with your experience designing and implementing game mechanics?


Thanks!

-Steve

[DrDerekDoctors]

It was all going so well until you used the word "proprioceptive".

I shall have a proper read at some juncture, but I think you're right that "feel" isn't something that's been particularly explored. I'll be interested if it's possible to do that in literary form without it being wanky.

[Anthony Flack]

It was all going so well until you used the word "proprioceptive".

I'd like to reiterate this, but without the winky face. I will instead use a stern serious face. Seriously, it was all going so well until you used the word "proprioceptive".

I think it's a potentially interesting subject - in fact, it is possibly the most interesting subject in videogame design. I'll be interested if it's possible to do that in literary form without it being wanky. 

[Bezzy]

I'm glad someone is tackling this. "Feel" and "kinaesthesia", to me, are far more core to games than storytelling or flashy art or sound. Not that those aren't just as lovely/useful, but feel always seems to get short shrift. It's not percieved as important.

I was working on a 3D platform game a while back (which shall remain nameless) and it always bothered me that moving the analogue stick around didn't quite match up to the angle of the movement. When you rolled the analogue around the outer perimiter at a fixed speed, you could see that the motion of the angle of your character wasn't matching up properly. It was almost as if the cardinal directions were "sticky".

I complained about it, on and off for months. Most people said "it's not a big deal" or "I can't tell what you're talking about". The people who did agree that the phenomenon was present didn't really seem all that bothered by it. I kind of accepted that nothing would be done about it, and that maybe I was just "feeling things". It wasn't game breaking, as far as everyone else was concerned, but I knew that it was *going* to be, because I was going to use a sort of aiming mechanic later on, and this would affect that negatively.

Finally a programmer agreed that there was something wrong with it (I think it may have been inconsistent with his AI's interface or something), and took a stab at fixing it.

The problem was to do with deadzoning and/or normalizing. Or atleast, I think it was. I didn't end up solving it myself, but I did discuss it with the guy who did, a bit, and put forward a couple of hypotheses. My memory is a bit hazy as to which was correct (if any were correct). Here's both reasons I remember which might have been valid, though:

1) If you do a deadzone check for EACH AXIS before normalizing the joystick vector, then create this phenomenon of "sticky cardinals". It's because you're cutting a square out of the center of the detail of your joystick's possible positions. Imagine a circle with a square cut out of the center, and then collapsing the shape in on itself so that the hole is filled - it makes a kind of bevelled cross, meaning that all the detail is in the cardinals. Sorry. That's a sort of wishy washy way of describing it. What you need instead is to cut a CIRCULAR deadzone out of the joystick's possible area. So, normalize the vector *first*, and then take the deadzone value away from the length of the vector.

2) If you normalize a 2D vector, one of the two axes will always be 1.0 or -1.0. That means that you end up pushing more tracking detail into the diagonal directions (I... think). The fix is to normalize the joystick vector using a 3D vector, where x=joy.x, y=joy.y and z = MAX_JOY_DEFLECTION as a kind of control. That way, after normalizing, z is always 1.0, and x and y are relative to z. As a result you're not warping the joystick vector into the corners.

If you try Zuma on XBLA, you can feel this problem. Just try rolling the aim slowly, and you'll see that it sticks to those directions, and is less easy to aim diagonally.

Incendentally, I noticed the same problem in Cortex Command.

This can actually be pretty useful for some games - if you want to somewhat separate the axes (for a grid based game with a continuous selector) then this will "bend" your movement toward the cardinals a bit more, so hitting "up" is more likely than hitting "up/right".

WOW! Complete tangent. I'm going to read your chapter now.

[Edit] Oh yeah... the game I worked on there... it um... wasn't exactly canned, but everyone left. It did feel a lot nicer after the fix, and some people commented that it felt more "responsive" somehow.

And that's what I mean when I say it gets such short shrift. If your primary interface into the game is the controls you give people, then you *have* to work at making them feel right. It's their bridge into experiencing another world of arbitrary rules. If the bridge is rickety, it destroys that immersion - the kinaesthesia. It's still seen as less important than getting other things right, though - like people say "It's good enough", but don't recognize the finesse they can bring to the controls.

Whine, whine, sorry.

[Anthony Flack]

I wonder if this is related to the reason why Capcom Classics on the PS2 has such horribly cocked-up controls on the analogue stick. It's almost impossible to pull off diagonals, so without an arcade stick two thirds of the games are ruined, just like that. Silly boys. Fancy expecting people to play Commando when the diagonals don't work properly?

It would never even occur to me that a real programmer would attempt to create a dead zone by simply hacking at the x and y axes to make a square hole. Urgh.

[Bezzy]

I wonder if this is related to the reason why Capcom Classics on the PS2 has such horribly cocked-up controls on the analogue stick. It's almost impossible to pull off diagonals, so without an arcade stick two thirds of the games are ruined, just like that. Silly boys. Fancy expecting people to play Commando when the diagonals don't work properly?

It would never even occur to me that a real programmer would attempt to create a dead zone by simply hacking at the x and y axes to make a square hole. Urgh.

Two words: Unreal Engine.

[Data]

Hey bezzy, can you give the latest version of cortex command another try and tell me if you still feel that issue present?

Because I'm trying to see it myself and just can't really feel it. The aiming angle seems to correspond pretty well and without cardinal bias with the analog stick direction on my 360 controller. My deadzone is calculated on the magnitude (0.1<) of the vector created by the two analog axes, so it should be a hole.

Thanks!

[Steve Swink]

It was all going so well until you used the word "proprioceptive".

I shall have a proper read at some juncture, but I think you're right that "feel" isn't something that's been particularly explored. I'll be interested if it's possible to do that in literary form without it being wanky.

Ok, so, as I was telling Alec, I didn't realize that proprioceptive was so far from common use when talking about sensation - it gets used a lot in behavioral psych/cogsci literature.  It's kinda interchangeable with kinesthesia, but is more appropriate for the context (I think.)  At any rate, if you forgive me that one wank, the rest is very simply written.  Also, I can include a definition of proprioceptive in the sidebar of the book - would that help?  I think it's a useful word for this discussion.  And/or I can alter the language, since apparently it's turning some people off before they read the entire thing . 

@Bezzy: What you're talking about is exactly the kind of thing I want to address.  Small changes in code = large changes in the sensation of control. They're especially hard to come to terms with because the sensation is so subconscious, so difficult to isolate. 

[konjak]

With all these books and articles and tips on how to make perfect gameplay and game feel, it would start to feel pretty limited what you can do once you make a game thinking about all these things. Because, to me, this is very individual and having many people's perception except yours in your mind, they can conflict and you're then making someone else's game...

But I dunno, I just like to rant. Either you make a game and it turns out good or you don't. 

I might be off-topic on this, I didn't read everything too deeply. 

[Steve Swink]

I don't think anyone who writes a book or article would claim that it leads to perfect gameplay; there's no such thing as a formula for perfect gameplay, a perfect painting, a perfect song.  If you going to be an artist, however, you need to understand perspective, figure drawing, how to render light and shadow, and so on.  These are specific, quantifiable skill sets which one needs to learn, and in which one can be instructed.  The 'craft' part as opposed to the 'art' part. So, when I start writing about game feel, I'm simply trying to translate my experience designing game mechanics into some abstraction that could be useful for someone else learning to do the same thing.  Memorizing all the indicator muscles on the human form != the formula for painting a perfect picture, and no one would ever claim otherwise.  The only difference is that games are so new that no one's really wrapped their head around what all the relevant skill sets are, and how to teach them.  Hence the need to write about it.

That said, I agree that once you understand and are proficient in the various skills - however you got that knowledge - you shouldn't listen too closely to what everyone else says.  It's like any creative endeavor, a struggle between canon, knowledge, and individual expression.  I dig what Jon Mak has to say, and agree with it - if you want to make a game, go play bunch of games that you like and then make your own game.  As long as you're making your own game, your own vision, you're doing right.

Guess I like to rant as well .

[Bezzy]

Hey Steve. I've so far only skimmed the document - it has been a busy day. I do intend to give you some more detailed feedback once I've had a proper look.

[Steve Swink]

Hey Steve. I've so far only skimmed the document - it has been a busy day. I do intend to give you some more detailed feedback once I've had a proper look.

Thanks! I look forward to it .  I realize there's quite a bit there to parse through and I really appreciate you guys taking the time to read it.

[DrDerekDoctors]

Ok, so, as I was telling Alec, I didn't realize that proprioceptive was so far from common use when talking about sensation - it gets used a lot in behavioral psych/cogsci literature.  It's kinda interchangeable with kinesthesia, but is more appropriate for the context (I think.)  At any rate, if you forgive me that one wank, the rest is very simply written.  Also, I can include a definition of proprioceptive in the sidebar of the book - would that help?  I think it's a useful word for this discussion.  And/or I can alter the language, since apparently it's turning some people off before they read the entire thing . 

I think it's just the context of it, I can be a fairly sesquipedalien and circumlocutive sod at the best of times but I tend to do it in conversational terms. For me, it's when it's in the context of a book or an essay that it comes across as offensively wanky. I think in your essay, that rather than shortcutting it with a word you should really try and get across what it means in terms of gaming so that a gamer can relate to it.

[Bezzy]

Hey bezzy, can you give the latest version of cortex command another try and tell me if you still feel that issue present?

Because I'm trying to see it myself and just can't really feel it. The aiming angle seems to correspond pretty well and without cardinal bias with the analog stick direction on my 360 controller. My deadzone is calculated on the magnitude (0.1<) of the vector created by the two analog axes, so it should be a hole.

Thanks!

Will do. May take a while because I have a horrible hamster infested internet tube.

[Edit] That was quicker than I expected!

I'm afraid I can still feel it there. It's easiest to tell on the aiming. Just roll the aim around very slowly, and within about five degrees of each cardinal, it seems to "snag". I don't know if that's by design - to help you aim horizontally and vertically. If it is, I found it a bit annoying for fine tuning my aim around the horizon. Certainly, it's not as pronounce a problem in CC as it is in Zuma on XBLA.

I'm also using a 360 wired controller, with drivers.

I should double check with Tom (the guy he fixed it) about how he fixed it. Do try the third control axis idea when normalizing, though. Christ. If this is actually an insurmountable problem, based in hardware, I'm gonna feel like a right cock.

Also, the movement seems to have improved a bit more since I last played. You don't get caught in the terrain quite as much (still a bit, though). Keep it up guys

[Data]

Hm I guess since the analog data is derived form those two normalized scalars (X and Y), it means the total area is square, and thus CC's deadzone hole is also square. I'll look into this later since you're the first person to complain about it (vs a gazillion complaints about other specific stuff). It's def on my to-do list though, so thanks for bringing it to my attention!

[ravuya]

Because of Bezzy's post, I went back and made my deadzone round.

Maybe it's the placebo effect, but it feels much more organic now. Killer tip.

[Steve Swink]

I think it's just the context of it, I can be a fairly sesquipedalien and circumlocutive sod at the best of times but I tend to do it in conversational terms. For me, it's when it's in the context of a book or an essay that it comes across as offensively wanky. I think in your essay, that rather than shortcutting it with a word you should really try and get across what it means in terms of gaming so that a gamer can relate to it.

Whoa, ok. Thanks for the feedback. It would never occur to me that proprioceptive could equal offensive. Let me know if you have any other moments like that .

[Anthony Flack]

I have read it through now. I have to say you kinda lost me with all the Mario talk.

It's not that I didn't understand it... it was more that I already understand these mechanisms quite well - and I suspect that most people with any game programming experience do too - and so I don't really need such a detailed breakdown of the mechanics of Mario in order to get the point. And I don't know that a non-programmer would be interested enough to want to have it described in such detail (but I could be wrong?).

Also, I think people tend to overstate Miyamoto's revolutionary influence. He's a good designer, but he's not god. It may be different from my perspective, since in my country there was no videogame crash, no NES revolution (in fact, the NES had almost no presence at all), most games were played on home computers rather than consoles until the Playstation came along, and the arcades didn't start to die out until some 15 years later. So there was really no earth-shaking arrival of Shiggsy on the scene. And besides, I also find Donkey Kong's stiffness quite endearing...

But I am happy with my new signature quote; thanks!


 

[ravuya]

I find the Miyamoto love tends to be higher in North America, where the console was king. In Europe, they tend to love 8-bit authors like Braben.

It's hard to overstate the Spectrum's influence on the UK.

[Steve Swink]

I have read it through now. I have to say you kinda lost me with all the Mario talk.

It's not that I didn't understand it... it was more that I already understand these mechanisms quite well - and I suspect that most people with any game programming experience do too - and so I don't really need such a detailed breakdown of the mechanics of Mario in order to get the point. And I don't know that a non-programmer would be interested enough to want to have it described in such detail (but I could be wrong?).

Also, I think people tend to overstate Miyamoto's revolutionary influence. He's a good designer, but he's not god. It may be different from my perspective, since in my country there was no videogame crash, no NES revolution (in fact, the NES had almost no presence at all), most games were played on home computers rather than consoles until the Playstation came along, and the arcades didn't start to die out until some 15 years later. So there was really no earth-shaking arrival of Shiggsy on the scene. And besides, I also find Donkey Kong's stiffness quite endearing...

But I am happy with my new signature quote; thanks!
 

That's excellent feedback!  I'm happy you felt the descriptions of the basic mechanisms of Mario were accurate.  In this case, you already understand the mechanisms quite well so this particular chapter isn't useful.  Consider this as a rough template for other chapters; my plan is to break down a whole slew of different mechanics this way. 

Where this could be useful, even for professional designers (and, perhaps to a lesser degree, programmers) is with respect to issues like the one Bezzy/ravuya/Data are discussing above.  I'm not intimately familiar with the problem they're working through because I haven't implemented it myself.  If I were to create a similar feel to what they're describing, understanding what they're talking about would be a great time saver.  Better yet, understanding why a circular dead zone feels better than a square one could help me wrap my brain around these mechanisms in a more abstract way (not limited to aiming with an analog stick or whatever.)  So, yeah, that’s all I’m trying to get at here.  Even the fact that ‘dead zone’ is a common term used to describe what they’re feeling/programming is fascinating.  What other common terms are out there?  Floaty, twitchy, sluggish, and so on…is there a way to pin down what those actually mean to game feel in a generalized way? 

Also, one thing to note is that the intended audience here is more designers and students aspiring to be designers than it is programmers.  I realize that there's a lot of crossover there and that in order to be an effective designer, one needs to learn how to program. Though I do a bit of programming, I consider myself a designer.  I see the programming as a means to an end, which is enabling me to design experiences - feel in particular, because that's my favorite part of design.  So what interests me primarily is not the programming underlying a certain feel, but the thinking of it, the design process.  That's what I'm trying to get into for each of these mechanics.

So let me take a step back and ask these questions:

•   What games do guys you think have the best feel and why?
•   What were the most important leaps forward for game feel?
•   Who created these games and do you know anything about their process?
•   Is feel purely subjective? Are there any inviolable rules for game feel?
•   What are other common implementation issues related to feel (like bezzy’s circular dead zones from above)? 


On the Shiggy stuff, I realize that this rubs some folks the wrong way.  I don't think Miyamoto is God; I can (and probably will) rhapsodize about any number of games and game creators.  I’m pretty excitable in general and will happily prattle on about games and game creators I love for hours on end, and write in more or less the same way I speak.  I don't want to turn anybody off from potentially useful stuff in the process, though, so I appreciate you letting me know when that’s happening.  Also, I'd really appreciate some help filling in the gaps in my knowledge of game history, particularly in the UK .  Stuff I can source would be cool.  Also, what ought I to have played that I haven’t feel-wise?

Thanks!

- Steve