Game grammar in action: AOC’s DPS bug
At this point, everyone is talking about the Age of Conan issue with DPS. In short, they tied doing damage to a trigger in the animation sequence, which meant that slower animations would do damage at a different rate than faster ones. And all the female combat anims are slower, so all the female characters do less damage per second.
From a game grammar point of view, this is a clear example of getting the wrong end of the stick. Recall the distinctions between the “salad” and the “dressing” of a game — the “salad” is the actual game, and it can be represented in many different ways.
In particular, you can certainly take the typical MMO combat game, even the realtime varieties, and model the game proper with cards, dice, or numeric outputs. In fact, hardcore players often do the latter in order to analyze how they are doing, because it’s easier to run statistical analysis on text logs than on 3d graphics.
As I have said before, if you get the “salad” right, then the dressing — the storyline, art, music, characters, setting, theme — can only serve to enhance. Start with the experience design, and if your core is rotten or an afterthought, you’ll be putting lipstick on a pig.
Folks working on teams tend to push for the primacy of their own discipline, and these days, with so many games being primarily about experience design and not about game design, it’s easy to put experience design at a higher priority than the mechanics. An animator is not going to want to be told that his carefully crafted ten second animation may be sped up and played in one second. He will rightly point out that a human being making a motion in one second versus ten stands very differently, and distributes weight in a different way, and that therefore the animation.
But the only reason to do the animation in the first place is to convey that the action happened. It is a piece of systemic feedback, comparable to turning a Magic card 90 degrees, or tipping over the king in a game of chess. You might as well light up a green light over the stick figure’s head. For that matter, the question of whether or not the character is female or not is also purely an aesthetic choice. They could be red or black checkers and play the same.
Just yesterday, I was commenting that there are two rare and vital skills a game designer needs to acquire: the ability to see the game in their head with no dressing at all; and then the ability to see the game in their head with no mechanics at all, as a player sees it.
42 Responses to “Game grammar in action: AOC’s DPS bug”
Sorry, the comment form is closed at this time.
I guess I don’t see this as a salad issue at all. The combat system is great salad as is the dressing, but someone just mixed up teaspoon and tablespoon while reading the dressing recipe. It really says more to me about the difficulties of managing design and development on a large scale project with large teams.
If the system were solid and also independent of any graphics, it wouldn’t have happened, is my point.
This is a common failing of late, and becoming more and more common as games become more visually sophisticated and it seems less and less sophisticated games.
Witness GTA:IV which was loudly and roundly given 10s and accolades due to the stunning visual presentation, and dressing. Voice acting and animation is done supremely well. When you are all said and done, you have a game which is at best half, or even a third of the predecessor GTA:San Andreas. Noone in the review industry appears to be looking for depth. It’s all about production value.
I think this is the net result of a sort of immaturity. Given too much ability to present things beautifully at the expense of the game, the game suffers. This is a ‘function follows form’ error. I believe that as the available resources mature, so games will back down from this grasping at visuals at the expense of the game.
Well, we can hope anyhow.
– Nox
I agree that the system in MMOs such as they are should be independent of any graphics, and that this was a silly thing to let slip without catching it in testing. It got me to thinking though, and now I proceed to derail this totally off-topic.
I think that in the future it will be possible to conceive of such MMO systems where graphics are not mere dressing. The statement that says “The only reason to do the animation is to show that the action happened” is correct but assumes a binary-only representation for actions – “Done or not done”, “Ready or fired”. I don’t know if Age of Conan was reaching in a nascent way for what I’m about to describe, but there might be actions for which there are many, many intermediate states between Action Began and Action Ended, and that each state is significant. The best example would be in one-on-one fighting games like Street Fighter, Tekken, or Soul Calibur. In such games, a single attack has many frames and it is vitally important what frame your opponent is in when you compose your response. The truly elite players break down these fine mini-states and analyze them carefully frame-by-frame. In this sense, the visual presentation is not mere dressing on an action – it is conveying actual mechanic-related information.
Now granted, you could spit out a stream of numbers to a scrolling-text pane to tell you what intermediate frame you were in, so in that sense the fancy 3D is still ‘dressing’, but that would be a lot harder from a usability point of view. My point is that in such games, each frame of animation is not just eye candy, but actually corresponds 1-to-1 with a unique internal state for that mobile which has consequences. I would be excited beyond belief to have Soul-Calibur level combat in an MMO with thousands of different possible moves, but that, of course, would require us all to have ping times measured in nanoseconds to actually work.
Very nicely put Raph.
You’re completely wrong Tim. This is like trying to tow your boat by putting a motor in the trailer and then attempting to drive both vechicles at the same speed without connecting them.
Fundamentally, the game designer needs to be able to say: this attack does X damage and takes N seconds. Doing it the Funcom way, the designer gets to say that exactly once: when the art is specified. Then they never get to tune the timing again without redoing the art. That simply doesn’t work in practice because designers are not omniscient, tuning is what makes the game fun. Designing a system that prevents tuning is made of fail.
It seems like this may simply be a result of immature software development practices, not poor design choices. Granted, the designer probably should have not tried to do it this way in the first place… but for them I don’t think this is a make or break kind of mistake.
However, any Software Engineer/Computer Scientist who’s worth hiring should have looked at the idea of tying combat damage to animation timing and said “no!”. Separating your presentation layer from your logic layer (and that from your data layer) is pretty much the first software design you (should!) learn… so it makes me think that either the devs were sleeping, or there’s a fundamental culture problem where people with little technical knowledge are apparently allowed to make technical decisions.
It absolutely baffles my mind how they could do this. And how they could be unable to solve it without editing the animations.
In the whole salad/dressing simile you’re making the salad out of the dressing.
Don’t they have some method to constrain the animation to a time? Sure it might looks a bit silly for the character to be swinging hyperactively, but at least it would fix the balance issue.
From what I can tell, animations lock out the client character control while they’re playing. That means 2 bad things: First, animations act as a timer. Second, clients give authorative swing times to the server.
Contrast this to WoW. WoW sends a toggle to the server (autoattack on or off). The WoW client sends an ability event to the server when an ability is used. The client constrains the ability to send the request to when an ability is “off cooldown” but few other constraints.
Take AoC then, if you use a program that speeds up process (your typical speed hack), you now attack faster, run faster, cast faster, everything is controlled by the client. If you manage to decompress their files and patch in your own animations or remove theirs somehow, you can now attack at a normal or faster speed. If your framerate is low, your attacks come slower. It’s a poor idea. If you intercepted the data stream and started sending your own packets you could attack as quickly as you wanted.
In WoW, your standard attacks come at intervals determined entirely serverside. Cooldowns are checked serverside. If your client sends a request to attack before it’s available you get an error response. Any speed hacks would only speed up movement speed (because server-authorative movement causes rubberbanding and is terribly annoying) and even then there’s checks that will disconnect you if you do anything illogical. (Or flag your account)
Using animations as a timer function mean a) changes to animations mean changes to timer response, as is the issue here. b) Animations are an actual combat mechanic (which while this could be interesting, it doesn’t work in this combat system) c) Any combat-speed rebalancing requires new assets from the artists (RETARDED)
Likewise, it’s terribly frustrating to lose control of your character due to animation. In a group of monsters, and you figure you can kill one, but then you have to get the heck out of Dodge. You happen to get a fatality. Oh, you have to stand there for 5 seconds lopping off limbs while his friends are beating you instead of running away like you planned.
I imagine this never would have been a problem on a little 3 person team of good developers. Everyone would be aware of how things work and it just wouldn’t have ever gotten to the point of being a problem. Or else the problem would quickly be recognized and corrected.
As Laura points out it’s not the best design in the world. A small team, on recognizing the failure would probably change it very fast once the animator realized they had to redo their work to fix the bug.
Basically little (smart) teams have a lot more freedom to make mistakes, because they can (theoretically) recognize them faster and respond faster. They get to evolve and migrate towards good design because of their greater insight and recognition of flaws plus agility.
Traditional big teams on the other hand are slow to change. They may not realize they’ve gone down a path with a flawed design until … well until after release. And correcting the issue becomes a big deal then, whether the issue is just implementation or more deeply design.
Spaz, if you look at this presentation:
https://www.raphkoster.com/gaming/designingforeverywhere.shtml
You will see that in my diagram I actually show two different kinds of things going into the visual representation.
– direct feedback for actions
– current state representation
https://www.raphkoster.com/gaming/agdc07/Slide52.JPG
That said, both of these can be displayed in many ways, and the simulation should not be coupled to the output.
The Andean Bird game demonstrates this too, where you can compare the earliest version which exposes the stats on the sim, to the last version which does not.
https://www.raphkoster.com/2006/07/11/a-vague-game-idea/
https://www.raphkoster.com/2007/03/09/gdc-07-jon-blows-nuances-of-design-session/
Solution: slow down the male animations, add more powerful attacks. Hasten the female animations, maintain the power of the attacks.
When you try to design with the gender-equality movement in mind, you end up designing by committee. Stereotypes are perfectly acceptable to use in games just as they are perfectly acceptable to use in comedy. In this case, men should be slower, not faster; and women should be more agile. I’d take the extra step and have body structure and weight effect combat performance.
I don’t know how they “tied doing damage to a trigger in the animation sequence”, but I remember some game modification tools have a setting for tweaking the animation rate of models, making doing so as simple as flipping a switch. If the animations are baked, yeah, then you sort of have a “recall” problem.
Morgan, the animation rate is quite tweakable… the problem is that they’ve now got to tweak all of the animations, and there are between 800 and a 1000 of them that need to be modified by hand so that the dps rates end up being the same. A combo is a string of animations, and the system doesn’t treat them as a single whole, so you need to actually go and rebalance every part of the combo string.
The “solution” of altering the animation speeds is basically what they’re being forced to do now, but it’s a terrible and inefficient solution that would’ve been easily avoided if they had dealt with the mechanical design properly.
@Raph,
That’s *really* hard. It’s very easy to do one or the other (though most people are better at the latter), but being able to do both is kind of like being able to look left at the same time that you’re looking right. I’ve become very adept at doing the former, and it’s made it much harder for me to do the latter. Once you pull back the curtain it’s very hard to ignore the man behind it. I can kind of extrapolate, but I rarely trust my instincts there because I’m effectively imagining what other people would think rather than actually having a sense of how *I* would come at it without the deeper knowledge of mechanics. I can guess at it basically, but not much more than that. Anyone that can genuinely compartmentalize their knowledge of mechanics to the point that they don’t see the seams anymore is worthy of a great deal of respect.
You mean just like every other MMOG? Yeah, let’s beat up on the guys for trying something different… š This isn’t the end of the world. Changes are relatively easy and simple to make to virtual things. Blah blah blah.
Morgan has inadvertantly also brough up the other good discussion that this problem has given rise to: sexism. There are far too many people who either feel that this isn’t a bug or that some variant of it should be a feature. In principal I would agree that adding some sort of appearance-related-skill-balancing into a game would be an interesting feature… but it shouldn’t be an accidental consequence of something you believe to be a simple aesthetic decision. When it’s not an advertised, or intended, feature it starts to look like a whole lot like systemic sexism in the design team (although of course in this case we’ve already discovered that it was just poor game construction).
They do actually, and the problem is not with the individual attack animations (those have already been fixed); it’s with the combos, which are a string of attack animations. You can’t constrain the entire combo sequence, because they’re individual animations. That means you need to go and shave a second or two here and there from the individual animations making up the combo until you’ve got the right time. There are close to a thousand individual combat animations due to combo chains, and they all need to be adjusted, and the checked to make sure the dps didn’t stray from where they wanted it. This is a pain.
Secondly, where are you getting the idea that the client is determining the combat steps? It’s possible that this is in fact happening, but it’s not necessarily so. The server itself could very well have access to the animation lengths, and reject any message coming in that tells it to make an attack faster than it believes you should be able to. Speed hacks have historically been defeated by systems like that, where it checks to see if you’ve exceeded the possible bounds of movement in which case it rubberbands you. I don’t know why you’d assume that combat works any differently here without further evidence.
While this is technically true (sorta anyway, the artists aren’t the only people who can adjust the timings, so new assets are only needed if the animations look too strange going faster or slower), it’s also mostly irrelevant. DPS is derived from damage over time, which means that if you need to rebalance the combat system you ideally do it by tweaking the damage numbers, not the speed numbers. The problem here is that there’s no way to change the damage numbers for males or females without changing it for BOTH. Damage numbers are by combo and class not by gender. It doesn’t make it any less silly to do it this way, because you end up with exactly the sort of problem that did happen, but in a perfect world, they never would’ve had to have touched the animation timings at all, ever.
The salad analogy’s nice, but could have been reduced to a link.
I think part of the problem in the latest generation of games is that the 3d hardware is doing more of the job than just presentation. Specifically, it’s doing collision-detection, which is business logic but depends on the actions of the polys, which should be the view layer. If the Conan system is relying on collision detection for attacks, that might explain the dependence on animation timing.
On the other hand, in a client/server system like an MMO, trust of the client should be at absolute 0, and relying on collision detection is a 100x worse sin than it would be in a non-networked environment.
I guess we are thinking it’s not this because then it would be even easier to catch ealier when the coded the different combo times for female avatars.
The damage thing doesn’t bother me so much as the fatality animations that root your character in place and remove your control.
Lord of the rings online has several of these as well, like the loremaster res where he smokes a pipe and blows smoke on the victim. When I want to move I want to move. I don’t really care if feet slide or popping occurs. Learn how to blend.
Or how about a console game I played recently, full of jumping puzzles, where the jump actually occurs after the next run cycle to prevent a pop. This gives you a random interval between each jump that often sends you falling off a ledge. It’s infurating. And the sad thing is I doubt one single person would have noticed the pop.
Rik, why would that be easier to catch? It’s a question of whether or not the server has access to the animation data. It doesn’t actually mean that there’s more or less oversight that the numbers match. They’re not adding the numbers twice, but they may be merging the animation timings with the server logic control.
Kharaza those fatality animations wouldn’t just pop or slide, they’d look ridiculous. You can’t blend movement into an animation that has you grabbing the opponent by the head and slitting his throat. You’d end up pantomiming ridiculously complex kill animations 10 feet away from any enemies. If you’re going to do that you may as well not have the fatalities at all.
What baffled me about this issue is how it didn’t come up during game play balancing. I find it hard to believe that whoever did the game balancing didn’t notice this. I can’t imagine doing any kind of game balance tweaking without a handy spreadsheet of DPS values, Dmg per Cost and other derivative values. To produce such a spreadsheet you’d need to grab the data first. At some point you should notice that there’s something really wrong.
I disagree about the animations needing to be done by hand though. That can be automatic, like how WC3 scales your animation speed to your unit’s attack speed. If dozens of 3d engines can and have done on-the-fly animation scaling, I don’t see why AoC in all it’s fancy graphical glory does not have such a system.
th15, no no. That’s not it. >< Again the problem again isn’t the individual animations, it’s the combo strings. There is in fact animation scaling and they’ve got a bunch of people doing what they call BCC who’s job it is simply to attach animations to things and set their timings so the artist isn’t needed here very much, but when you’ve got a string of 4 or 5 animations in a row, and you need to normalize the string and not the individual animations, it’s a much more complicated issue. And you need to set the timings by hand, not redo the animation. The issue is entirely scale. There’s an awful lot of animations that need to be tweaked.
However, on the front of QA there’s been a distinct lack of “quality” in the assurance. Only recently they were forced to revert a change almost immediately after it going out because it was causing the servers to crash. I’m not quite sure how you miss a server crash bug that results from a loot drop. Seems like it should be pretty obvious if you’ve actually tested it at all. The test server going boom is kinda hard to miss.
This is both a failure of game design and software engineering. If this starts showing up elsewhere, it would imply that the game is not maintainable.
1. The problem should not have happened in the first place if the game play system was properly separated from the game presentation.
2. That it is going to take weeks to fix implies that animations drive game play which may open up some hideous exploits (swap or spoof an animation to change the game performance which is clearly authoritative on the client).
Hindsight is 20/20 obviously. But the combos should have their timings scripted into them, rather than relying on animations for timing. They are not using the animations for collision detection, so speeding up an attack when it’s too long to fit into the combo may look silly, but it wont break anything.
I disagree entirely. DPS is derived from damage over time, so it is a function of damage and time. Unless you’re confident enough to think that you’ve balanced your abilities to the point they need no more tweaking you need the ability to change both.
Take the following situation, with 3 fictitious abilities. One is a 5 second stun. One is a combo, which completes for a male in 4.5 seconds. One is a combo which completes in 1 second for a male and has a 10 second cooldown.
If you attempt to balance the slower animated character through scaling it doesn’t work out the same.
Say the male 4.5 second combo hits for 450 damage, the female equivalent hits for 585 damage in 5.85 seconds. You might say, that’s the same, they both do 100 dps. But a key difference is the male character can finish the combo within the duration of the stun, while the female character cannot, which allows the target to get out of range before the combo completes.
For the 1 second combo, the male does 500 damage, the female takes 1.3 seconds and does 650 damage. Again, the DPS looks the same on the surface, but because of the cooldown the male, using the combo every cooldown is doing 50 dps from the combo, and the female is doing 65.
Finally, assume the fix all of the female animations, so that they are equal with the male’s. Then three months down the line behavior emerges to show that being able to complete the 4.5 second combo within the time of the 5 second stun is unbalanced. Though the DPS of the 4.5 second combo, and the duration of the stun on their own are balanced, the fact that you can do them in combination by yourself is too strong. Without the ability to easily change the speed of the combo, how do you retain the DPS of the 4.5 second combo, the length of the 5 second stun, and disallow the combo to fully execute with no way for the victim to avoid it?
If you could modify the timings, you could extend the length of the combo to 5.5 seconds, and cause it to do 550 damage. DPS is maintained, victim can escape if he’s quick or if the aggressor is slow in executing. Stun duration doesn’t have to change (potentially unbalancing other attack combos)
You can not have full control over the balance of your abilities if you can only change the scale and not the cast times.
Then have them stop playing and slide into a regular death animation/run animation. Losing control of your character while the devs show off “We can be really gory” is cool for the first couple of times you see it. But after that you just want to be able to use your character when you want it.
Why does this sort of exposition dictate the flow of the game? Most players Just Don’t Care. Especially in an MMO where they’ve done it 1,000 times before.
^ yes, though for the regular old solo game (which seems to be what the game is intended to be), I’d let them play out, but having a choice would be nice. They are really cool looking š
Fatality animations also grant an invulnerability state because of this removal of the players control. This often can tip the scales on a large AE battle as the big packs beat on the invulnerable guy who gets a head chop, while the fellows in the back blast them down with lightning.
[…] Raph’s Website Ā» Game grammar in action: AOCās DPS bug Apart from confusing things by making a counterintuitive distinction between experience design and game design, this is a good example of how mechanics and ‘dressing’ interact in games. (tags: raphkoster gamegrammar games gaming play gamedesign mechanics dynamics aesthetics experiencedesign) […]
@Rob,
Er, the design fault is that there isn’t a difference between “script timings” and “animation timings”; the animations, again, are capable of having their timings tweaked. You could look at this particular situation as them not having a single variable for male and female combo strings, but instead having 2. Animation speeds scale based on the timings that are set by the BCC people not by hard coded artist values (I have a good friend who’s doing BCC for them right now, so I know how that part of the system works). And they’re doing those scaling tweaks to fix the issue, but because they combined combat with the animation system they allowed themselves into a setup where they needed those two variables instead of just one, and with the number of combat actions involved, it becomes unmanagable very quickly..
Understand, combos can be interrupted in the middle, even play players just not hitting the next button they need to move further into the sequence, and they still cause damage based on the number of hits that have *already* landed, which limits the ability to implement a single variable per combo instead of using the aggregate of the individual attacks in the combo. It’s not impossible mind, but it’s more involved. You’d basically need to have the strikes of the combos be dot-like ticks to allow the interruption to occur. This is probably a better way to do it, but they probably did’t think the other way would end up being an issue and the dot-like stuff likely would neet more work to be done.
This would be true in most games, but it’s not very useful in AoC. The reason why there’s a DPS problem with females is because it’s out of sync with males, but the difference is not sufficient enough that it would become a balancing issue in terms of standard pve or class balance. Honestly, I’m mostly willing to bet that no one would have noticed without parsing the combat logs, at least in shorter duration combats. A difference of a second can add up fast if your duration goes into several minutes though, but non-raid AoC combat doesn’t really do that much.
AoC combat is entirely too fast for the individual attack durations to be a feature you can tweak for balance. You don’t hit a combo button, watch a casting bar go off, and wait for the cool down. You hit the combo button and then activate another 2-5 attacks within the span of about 5-10 seconds give or take. You can’t speed up combo by reducing the space between those attacks too much, or the player loses active control of them going off, which is what the whole system is based around to begin with, and you can’t slow them down too much either without ruining the gameplay’s feel. If you want to make a combo slower you need to add attacks into it, but that inherently raises the DPS because combo strikes always do more damage than normal ones, and having an extra strike in the middle means you’re getitng one more buffed up hit. And if you want to make a combo significantly *faster*, well, you can’t really (you can make some tweaks to shave off half a second here or there, and that does add up long term, but that’s all the play you get.). You can only lower the cooldown so that the combo can be used more often, and this has some issues, since most combos also generate status effects beyond simply hurting the opponent, and increasing how often those can be applied is tricky.
So the only effective way to tweak DPS in AoC (at least for melee classes, which is the only place the animation stuff comes into play) is to adjust the damage values, or, if the situation warrants it, the cooldown values. There are no casting times to adjust; combo animations are roughly normalized and rarely last even a second, though because they coupled the timings into the animations rather than putting them in the actual combat code there is some variance, and in this case, that variance is actually problematic.
Eh. Fatality animations last at most 10 seconds, you’re invincible for the duration, and you get a pretty nifty buff afterwards for about 5 seconds. They’re rare enough that they’re not going off every 2 seconds, and you only really see them frequently when you’re fighing much lower level enemies than your character. Being able to move and lose the invincibility timer is actually tactically unsound in terms of being an effective combatant. As Kharza says above, the invincibility duration can make a big difference in some cases, but beyond that, you also recover some of your cooldowns, as most combo cooldowns are measured in the tens of seconds (or less) rather than in the minutes. I really don’t see how fatalities are any more annoying than stuns or knockdowns, and at least you get something out of it. Hell, a WoW Mage will spend as long or longer on most casting bars as you spend locked into a fatality animation. Now granted, the Mages move if absolutely necessary, but not if they want to actually play the game.
Again though, this is perhaps me thinking too much in terms of mechanics and less in terms of surface appearance. Being able to move whenever you want to move is very much a “feeling” based consideration, honestly, as allowing unrestricted player movement cuts off more interesting gameplay than it gives back, but that’s a discussion for elsewhere, since AoC barely does anything with it.
But that’s sort of the whole point of AoC. This is the same game where the stated stats didn’t work until a few weeks after launch.
Funcom was trying to make a twitch melee game, not just windowdress the usual underlying RPG with the usual veneer of animations useless to the combat itself.
They did “fail”, but not because they got things backwards. They meant to get things backwards š They failed because they didn’t adequately plan their animation pipeline to account for the stats that were intended to overlay them.
This particular event is just one of a few results.
One might argue that one could easily design a twitch game by separating the model from the view. This becomes especially important in a context like an MMO which is vowed to evolve continuously, unlike the usual console fighting game where bad design doesn’t matter so much after release. Not to mention the implications of backend logic needing to be aware of animation assets…
The downside to WoW’s approach, at least as of when I left, was a huge disparity between melee and ranged classes because of the degree of server authority and client trust. When a melee character pursued a ranged character, activating a skill would activate the ability’s cooldown. The server, however, would reject the ability firing as it was ‘out of range’. The client trusted the server to tell it the enemy was in range, in the interim between the cooldown beginning and being interupted (as the ability had not actually fired, though in many cases misfires consumed the entire cooldown anyways with no result), an attack may have actually been possible, but the client would not allow it, believing the ability was on cooldown. This issue could technically affect ranged classes with an enemy at the extreme edge of their effective distance. However, melee classes aren’t as effective at dealing damage while running away, so you’d typically find it hampering a melee character being kited than a ranged class trying to finish off a runner (ranged classes seemed to have the superior CC skills, anyways).
There’s nightmares in every system, I suppose.
I’m a big fan of Raph and what he did for pre cu swg so my opinion is probably bias but I think Funcom should of had him on board for some of his ideas on making mmo’s more socially entertaining by adding in a few of his ideas from swg.. I realize this blog is in regards to a female dps bug but I’m kind of curious what Mr.Koster thinks of the overall game.. If he’s played it.. I think AoC is a pretty big failure imo.. There wasn’t a single thing aside from fatalities that I liked about the game.. People were/are gawking at the graphics like its the second coming but even those were bland looking in my opinion since all the gear looked exactly the same for 80 levels.
The stats and procs were marginally exciting at best. The melee combat is horrifically primed for 360 its almost a joke to even play it on a pc.. Chances are the game won’t make it to 360 either at the current rate its losing subs.. And the mature crowd is no doubt going on summer vacation soon so it feel like a dead and barren wasteland which will cause more to leave in droves.. If diablo 3 comes out with in the next few months like i suspect AoC is done for.. Failing an early d3 launch War will kill AoC. 5 years this is the best funcom can do? /sigh.. Another 30 million dollar failure.. Aoc fans got the long stiff arm of Mr.Mcquaid.. dry…
Travis wrote:
Where’d you get that idea? Here? The author retracted his analysis.
I haven’t played it yet… Metaplace eats all my time!
@Morgan, man, even *I* get whiplash just reading those two articles. š
@Travis, You’re a bit nuts if you think D3 is coming out anytime in the next 12 months. Blizzard is notoriously slow; I wouldn’t be surprised if we don’t see it launch for another 2 years. Heck, SC2 isn’t even out yet, or dated even (they’re not even projecting the year!), which means there’s likely at least another 6 months to a year on that one, and that was announced what, a year ago? After being in development for what, 4 years already?
And WAR won’t kill AoC either, the games are much too dissimilar. If anything, WAR is likely to have a harder time building up a userbase than AoC was; too many of the systems are very similar to WoW. It’s going to be much harder for WAR to stand out. I really hope it does do well though, because it really looks like a pretty big improvement on several of WoW’s core gameplay systems… but we’ll have to see how that goes. I do worry for it.
And no, the AoC sub numbers aren’t dropping off horribly. AoC will die when Funcom decides to pull the plug, not before; it’s reached sufficient critical mass to be self sustating forever at this point. Hell, Anarchy Online is still around, despite one of the most godawful launches in the history of MMOGs, and AoC is doing an order of magnitude or two better than it ever has. It’s not likely going to surpass WoW though, WoW’s too entrenched.
@Raph, any word when you’ll be letting more people into Alpha or opening beta? I want metaplace to eat all my time too š
You know, I played AoC for a short time on a friend’s system, and I was struck by how similar I found it to EQ2. Sure, the combat system’s much more interactive and responsive, but the classes are… easy to draw lines between, overall.
Hard to say, in that case, if that’s helped it or hurt it, since EQ2 is hardly the competitor that WoW would be.
Not sure if I agree with the ideal of salad and dressing… Surely the ideal for many would be to have the presentation be the representation. We canĀ“t do that because computers are too slow… Still, if your game is a physics based simulation you canĀ“t fully separate presentation from representation.
Think of sports, surely what you can see and how you see it affects your performance? Even the audience and their cheering or booing matters. Most computer-games are dumbed down… too much. We arenĀ“t required to make dumb games, are we?
“Game grammars” never represent the live game when human player(s) are involved. They are abstractions. The rules are abstractions as well. Just think about fun, whether the player thinks the game is exciting or not affects how he plays the game, it belongs to the game proper IMO, hence the presentation does too.
It sounds like FuncomĀ“s mistake was in not making enough of a distinction between female and male characters in the simulation, thus not having the proper “knobs” for balancing. Stupid mistake for a Conan title.
Raph, thanks for the links. Andean Bird is good stuff. What language/what development environment/toolset did you use to make it? I imagine the music is yours too?
I use BlitzBasic for my prototyping and private projects — same language that Rod Humble’s “The Marriage” was done in. In fact, the little tools I have released are done in that too.
And yes, the music is mine, the graphics… everything. I pretty much do everything on those. š
I, and many others, reported this shortly after female avatars became available to beta testers. I found links on the beta forums mentioning it around Feb 2008, female characters weren’t available much before that. So spreadsheet or no, there’s at least 4 months gone before they announce plans for a fix. Now, I’m starting to understand why everything they do seems to take so long, that the design is such that changing or tweaking anything means redoing everything, explains why so little changed between the first announced delay last year and their final release client.
[…] Game grammar in action: AOC’s DPS bug July 3rd, 2008 (Visited 1181 times) Tags: age of conan, game design, game grammar […]
[…] that weāre actually using the animation itself to make the attack; promoting the swing animation from dressing to salad. By now, you are probably thinking of the infamous female DPS bug in Age of Conan. This is […]
[…] […]