Posts on Current Forum | Archived Posts
|
|
|
||
|
||||
|
||||
|
||||
| Frequently Asked Forum Questions | |
|||
|
||||
|
|
|||
| ||||
| ||||
| ||||
| Search Older Posts on This Forum: Posts on Current Forum | Archived Posts | ||||
| ||||
| ||||
| ||||
| So you guys want specifics on game design eh... | |
| Posted By: Hawaiian Pig | Date: 11/12/09 11:34 p.m. |
|
Ok, I was going to post this in response to this and this, but it ended up being much longer than I expected it to be, and I'd prefer it not get swept away to the bottom of the forum. I've decided instead to format it for easier reading and to post it separately.
The casual/hardcore divide has always had pretty contentious debates, and one aspect of it that I haven't seen explored well enough from "my" side of the affair are the specifics regarding how we view game design. There's always an accusation of ambiguity when one says "make something with depth!" Seemingly empty statements like "make it require more skill" or "make it harder" cloud the issue and many of my peers are guilty of this. Hopefully this post will serve to demystify the vague nature of these discussions. That being said, this is a topic where being specific is incredibly difficult unless you have specific examples to consider. Even at that point, there is often some manner of value judgment involved (hello BXR), and we won't settle that here or anywhere ever. With that in mind, perhaps this may help you better understand how some of us make those value judgments. Now I don't speak for everyone and this is not an exact science; just some theorycrafting I was doing after having a few discussions with others about it on IRC. The concepts I come up with are not from any School of Video Game Design and are simply derivative of a few discussions on the topic. Furthermore, I think this is less of a "Hey this is what developers MUST DO" and more of a "Hey this is how it seems high-level games work, maybe we should consider this in game design." Since it's all the rage, I'll add a tl;dr version at the bottom. Let's get what "Fun" is out of the way. Everyone "against" us "damned ivory tower pros" keep asking us for clarification of our terms. So here's a slice of what came to mind as I was reading things here. This discussion at this point has strayed into quantifying "whose fun is more intense," and I don't like that. I will agree with the idea that satisfying a higher-skilled player is tougher to accomplish than satisfying a lower-skilled player (I also admit that this may be a result of my skewed perspective). This doesn't mean that one's satisfaction is more intense than the other (I've noticed Cody trying to say this, and that's just plain silly; you can't quantify what's more "fun"). I could derive fun from taking a hammer to my own balls (some here seem to think that this is my approach to games anyway), to each his own. Who do you design for? That said, I feel certain design imperatives can benefit one side and be inconsequential to the other. This goes both ways. In the same manner that tweaking the game's mechanics for competitive play doesn't affect the lower-skilled player as much, neither does the development of a compelling narrative (for example) affect the higher skilled player. I think my major gripe is that I see more games gloss over the former and focus on things similar to the latter. Nearly true to life graphics (this includes "omfg look at that explosion") and cinematic style story telling, to name a few, have become a huge part of games. Additionally, it seems that a precarious assumption is often made, where it is decided that in order to reach these non-gameplay-centric audiences, games need to become extremely accessible. Am I saying developers ought to design only for the hardcore-pro-guys? No, but I feel the parts of games we find to be enriching can benefit everyone. Let's be sure about one thing though, accessibility is great, most great games are highly accessible; the problem is that, lately, accessibility is often implemented in ways that compromise the complexity of the game (complexity being an issue apparently only matters to a small niche of players). "Sufficient complexity" is always something I stress when it comes to game design. It's toeing the line between being overly-complex and stupidly-simplistic. I really feel that a growing trend in game design is to operate with the assumption that the average gamer is an idiot. When I run into concessions in games that are borne out of this assumption, I find it almost insulting. So why design for this seemingly smaller niche of players? Why not just make interactive movies with big explosions? It would seem to make good business sense. It works for the movie industry. Well, as lame as this sounds, I feel that there's something that sets games apart from the rest of the entertainment industry. This issue has actually been the crux of the ODST discussion earlier this week: the replayability and longevity of a game. Sufficient Complexity and Replayability When I look for examples of games that have stuck around long enough to come to mind, I find that they tend to be games which strike a good balance in terms of complexity. Sufficient Complexity in Single Player Games Replayable single player games hinge on this. The games many of us go back to are not cakewalks. They provide scenarios and encounters that provide the player with a sufficient challenge; enough for them to want to head back and beat it again. Alternatively, they provide the player with a sufficient amount of "things to do," which in itself is a form of complexity. This isn't to say that a great story won't get you to come back and revisit the world again, but I'd imagine this happens much less when the game is neither challenging nor rewarding. I don't think I picked up Bioshock after my first run through, despite the fact that I very much enjoyed the story and atmosphere. I will admit that, in the realm of single player games, complexity of a game doesn't matter as much. When working in a co-operative or single player environment, challenge and variety of activities aren't of central importance (re: casual games). Sufficient Complexity in Multiplayer Games Now, when competitive multiplayer games strike this balance, they not only keep their userbase longer, this balance of complexity is most often conducive to the development of a "high skill ceiling." When a game is too complex it's inaccessible, when it's not complex enough it's too easy to master (and often hits the Nash equilibrium at its highest possible level). So how do you achieve this? How do you address this vague term of "Sufficient Complexity?" How do you increase the height of your "Skill Ceiling?" Developing the large Skill Gradient
It seems to come up constantly in these threads. "What do you MEAN by making something complex," "how DO you increase the 'skill gradient,'" "what IS 'awesome?'" I feel that it has to do with striking that balance of complexity, and there are a few aspects, variables I suppose, that you can fiddle with to effect change in this respect. I haven't explored this as deeply as I should have for Halo, but I have thought of it in terms of Starcraft. When I look at high-level Starcraft play (from my lowly D rank on ICCUP), I think that I can break things down into concepts that apply across many types of games. Fundamentally, you have the game engine and content which comprise (and I hate to use this word) the "sandbox," above this, you have three concepts that directly involve the player: strategy, execution, and available information. The depth of each higher level concept, and the amount of emphasis placed on each, can result in varying types of games. Those games which strike an interesting balance of these concepts can result in something that is both accessible and rich in gameplay. There is no single way to do this; different mixes of each variable will result in vastly different games, each capable of being too thin, too complex, or well balanced. Now while I hold these three variables to be chiefly involved in ensuring a game has a high skill ceiling, they are rooted in lower level concepts. We need to first understand these before we can understand the implications they have on the higher level concepts. Retroactively Inserted Image:  Retroactively Inserted Text: In a nutshell, there are "lower level" concepts here that one can diddle and tinker with to effect change in the "higher level" aspects that are fundamental to broadening the skill gradient. Please do not confuse the use of the word "level" here with connotations of casual/hardcore. The use is simply hierarchical. Lower Level Concepts: Sandbox, Game Engine and Game Content The three higher level variables are at play within a "sandbox," which is comprised of the game's engine and what populates it. Sufficient complexity must be found here, and I'd argue that most successful competitive games actually keep this simple. The sandbox that these variables operate within has the potential to over-complicate things; as its own complexity exacerbates each above variables' complexity. In Halo terms, you can think of this as the reason why MLG players loved the idea of a single hand-cannon to rule them all; the sandbox was thin, but it enabled the higher level variables to manifest. Over-complicate the sandbox (throw in too many weapons in an FPS, too many units and counters in an RTS), and you end up preventing the development of strategies and proficiencies in execution. Games that do this suffer from having an abundance of... stuff... and you end up with gimmicky gameplay. (Which isn't necessarily bad, Grand Theft Auto games are chiefly involved with developing a huge sandbox and lots of gimmicky gameplay; the game derives all of its complexity at this level, and that has proven sufficient in keeping people playing. In fact, you'll notice that the major complaints with GTA IV lie in the removal of "fun" items from the sandbox. Less items, less complex gameplay, fewer "things to do.") This being said, the "stuff" in a game, the game content, is actually the easiest thing to tweak. Testing a game and patches tend to fix these things (the vanilla version of Starcraft that people bought in 1999 differs greatly, now, in 2009 after countless patches and an expansion pack). Indeed, greater issues tend to reside in a part of the game that I see a lot of developers compartmentalize and delegate to a "programmer" team: the game engine. I'm not sure how much feedback goes on between design and programmer teams, but I feel as though they should be extremely tightly bound. I recall reading in Bungie Updates about these two basically fighting with one another over "cool stuff" and "stuff that isn't possible." Still, it is this aspect of the sandbox that I feel has the greatest effect on the higher level variables. At an even more basic level, these lower level concepts directly relate to player input and interaction with the game world; in crafting a game with a high skill ceiling, the higher level concepts ought to be considered when developing the engine. Lower Level Implications at the Higher Level Execution I think I'll use this point as an excuse to hop back up to the top of the hierarchy so we can demystify some of these variables. There is a fairly hot topic regarding Starcraft 2's development at the moment that has to do with player input at the lowest level and how it affects a higher level variable. This one relates to the concept of execution. Allow me to elaborate on this concept first. Execution is a higher level variable that I think encompasses the player's physical dexterity and ability to multi-task. It's very common for people to dismiss this aspect of higher level play as robotic or as "twitch" responses. Games such as Quake place a heavy emphasis on this variable. As I said earlier, a different mix of each variable can result in wholly different kinds of games. Quake (and many other "pro" shooters) is a style of game that has incredible depth in terms of execution of tactics. Sufficient complexity in player movement gives players something to master rather than just holding forward (this includes bunny hopping, strafe jumping, and the like). Think of it like having to dribble the ball in basketball. The game would be much different, much more simple, and undeniably easier to master, if you were allowed to carry. Without considering the concept of execution, and realizing its importance, games threaten to become too simple. In Starcraft's current iteration, one of the things that set apart the highest level of pro players is their incredible multi-task. Simplistically, Starcraft requires copious amounts of concern regarding the management of one's base ("macro-management:" related to procuring and securing resources, transferring these resources into an army through production buildings, and all things related to spending your income as it comes in), and the management of one's army or "micro-management" (as Starcraft's unit AI isn't self-sufficient, many strategies rely on careful and skillful control of your units). Together, these two aspects compose the majority of the game's "execution" feature. In its current state, there is considerable depth here and I feel that it is a great example of sufficient complexity. At all levels of play, two players with equal skills regarding execution will often be well matched (barring differences in the other variables). Cody, Eiii, HunterX11, and I have intensely close matches when we play Starcraft, and yet we all get stomped by goatrope, who himself gets stomped by better players. At the moment, if Korean pros are any indication, it's arguably humanly impossible to achieve perfect control in the game and the skill gradient increases greatly as a result. Some design choices in Starcraft 2 threaten to disrupt this balance regarding player management. Something as little as changing the way players select their production buildings has raised considerable uproar from the Starcraft community. Part of one's macro-management involves producing units, and Blizzard has decided to simplify this process by allowing you to hotkey multiple buildings; this is a fundamental aspect of player input that stems from a lower level concept, in this case: the game engine. In doing so, rather than having to select each production building individually and build units accordingly, one can hotkey groups of buildings and mass produce units with a few keystrokes. This significantly reduces the amount of complexity involved in one's execution and threatens to lower the skill ceiling (a huge part of Starcraft's multitask is concerned with converting your economic gains into a well composed ground army; "Multiple Building Selection" is similar to removing dribbling from basketball). At the moment the community is out to lunch on this one, no one is entirely sure how heavily it will impact competitive play. There are two ways to look at this design choice: one is that it's intuitive and makes the game more accessible, and the other is that it's disruptive to the existing balance and makes the game easier (reducing complexity). I would argue that this particular change probably won't affect the lower skill levels in either case. Proper use of hotkeys are already a slightly-above-average-skilled aspect of the game. This is clearly a design choice fueled by the impetus for greater accessibility that I feel developers are embracing a fair bit too much. Strategy This being said, I'm absolutely terrible with hotkeys. Between those I play with, I probably use them the least, and as a result my execution suffers. That said, I can still compete because I feel I employ a greater emphasis on another concept: strategy. Starcraft, and many other games (video or not), provide the player with a variety of options of which they can choose from based on whats available to them and what they know about the opponent (more on the latter in a bit). As with execution, there must be sufficient complexity involved in these available strategies lest the game become "solvable." In the Halo Wars demo, the Prophet Beam rush a few of us devised demonstrates this to some extent. Granted this only occured in the demo which had only a slice of the full "sandbox" available, it became clear that a specific aspect of the game content (hello, lower-level concept) greatly disrupted the available strategies involved. The solvability (makin' up words all over here) of games is fundamentally entwined with its replayability and longevity. Whereas execution is closely related to the lower level concept of the game engine, strategy is more closely related to the lower level concept of the game content. To find sufficient complexity with regard to strategy, we must consider the capability of a game's system to be solved. In game theory, if Player A is making the best decision he can, taking into account Player B's decision, and Player B is making the best decision he can, taking into account Player A's decision, then the players are in "Nash equilibrium." In this case, no player can do better by unilaterally changing their strategy. Basically, if you give everyone M6Ds, no one will use anything but M6Ds. There are two ways to alleviate this issue, one is to provide the "sandbox" with enough options, whereby no single option dominates entirely, or to appeal to the third concept and limit information availability. This being said, I'd argue that the reason Halo 1 was not considered "solved" in spite of this seemingly apparent equilibrium is the emphasis that it, as with many FPS games, placed upon the execution element of play. Granted, there was certainly room for depth of strategy regarding map control and dominance, I'm attempting to keep things simple by only considering the depth of the available options of weaponry. Perhaps admirably, Bungie attempted to address this with Halo 2 and beyond by decreasing the effectiveness of this one strategy (effectively transforming the pistol into the BR) and attempting to provide alternative options for dominant strategy (equipment and power weapons). The efficacy of these changes are highly debatable, and some argue that this has encroached on the previously developed emphasis on the concept of execution. Ultimately, I feel that the concept of strategy is relatively easier to develop and cultivate because it relies heavily on the highly adjustable lower level concept of game content. However, most often, the problems I see arise have to do with concessions made with the intention of accessibility. It would seem that many of them are made without considering the actual impact upon strategy. For example, Assault Rifle spawns, initially conceived to keep "headshottable" weapons out of the player's hand from the start (to apparently makes things easier for new players) actually serves to make the game less accessible. While a dominant strategy exists (the BR) the available strategies to the player are limited until the first few kills. At this point, whichever team can assert map dominance from the start is able to procure an advantage, and the availability of the more dominant strategy becomes skewed in favour of that team. While on one hand you could call AR starts hyper-pro mode, in that whoever manages to assert map dominance first becomes the rightful winner, on the other hand you have the availability of the dominant strategy being left up to AR battles; an encounter that many players consider to have a low skill ceiling due to the ease of execution regarding an AR kill. Information Availability I mentioned another option for addressing the solvablity of a game earlier, and that involves the regulation of information available to the player. This is a very interesting concept, however I find it has limited applicability to the FPS genre (going maybe only as far as "what I might encounter around the corner"). That being said, while I list this as separate from the above two concepts, it directly informs each of them. I'm gonna fall back on Starcraft again for this one: In Starcraft there are 3 races (factions which include a wide array of units, where no specific unit is truly analogous to units of the other factions [translation: very asymmetrical]). Of the three races there exist six possible matchups. Each matchup plays quite differently, and for a player of a given race, one must learn at least 3 of them. The available options for strategy can be overwhelming. To simplify things, we'll assume for a moment that there is only one matchup. Within an individual matchup there have emerged (after about 10 years of play) a wide variety of strategies that one may employ. Some strategies are particularly dominant against others. On a very very basic level, Starcraft strategy often works in two ways, one goes for either economic play (attempting to procure more resources quicker than your opponent) or tactical play (where one attempts to convert their resources quickly for a larger army or "tech" advantage). In reality, most strategies treat the two elements of play (economic and tactical) as part of a zero-sum relationship. For the purposes of this post, what this means is that there exist a wide array of strategies, some more dominant than others, but no single truly dominant style of play. Essentially, you have a very large game of Rock Paper Scissors (huge oversimplification). Now as we all know, Rock Paper Scissors isn't particularly deep. This is a result of the game's complete lack of available information. You can never truly know what strategy your opponent is going to employ, and you make your decisions based solely on chance. While a balance of dominant strategies exist, you can never know which one to choose. Starcraft, on the other hand is different. There is a great wealth of strategy to choose from, and each strategy can dominate or succumb to other strategies (barring variations in your execution). What's quite interesting here, and what draws me to this game (could you tell?), is that it's a game of imperfect knowledge. The information about your opponent is determinable only through active scouting, and what you see may not entirely reveal the choices your opponent is making. In this way, we can avoid the strategically thinning effects of the Nash equilibrium by never having truly perfect knowledge of your opponent; Player A can't exactly make the best decision he can if he is not entirely sure on what decision Player B is making. As a result, you end up with a game of on-the-fly calculated risks, utilizing out-of-game developed strategy, and in-game execution. Delicious. To be sure, imperfect knowledge is not the key to solving issues of strategy, I just find it quite interesting and the level of knowledge available to a player greatly defines the type of game you're playing. To this end, the availability of information regarding your opponent CAN be perfect and you can still avoid equilibrium by compensating with... sufficient complexity! Chess is a game of perfect knowledge, however the available strategies are so complex that it's hard for one to truly dominate. Twitch shooters such as Quake (and many sports) may involve perfect (or close to perfect) knowledge, however the complexity involved in execution can be so great that the competition remains rich. So what the hell does this mean? Well what do we know now? The sandbox (comprised of the game engine [physics, AI, controls, etc.] and the game content [weapons, units, spells, cherries that enable you to eat ghosts, and what have you]) directly affects the above mentioned higher variables of strategy and execution. In addition to these, a third concept exists, regarding the availability of in-game information, that directly affects the other two (what you know of your opponent will dictate your strategy and how you execute it). These three concepts (and perhaps a fourth could be conceived in consideration of games like rock paper scissors: chance), directly influence the level of difficulty involved in mastering a game, the accessibility of the game, and, indirectly the longevity and replayability of the game. If a game can continually evolve and shift (as many pro games that have stood the test of time do today), then it can gain wide appeal as its "freshness" (hi ODST) does not diminish greatly with time. Perhaps now, when someone says "make the game harder," you may think in terms of these concepts. For example, I have grown tired of ODST's Firefight, because it lacks complexity regarding strategy (it's nearly solvable), the execution of these strategies is fairly simple, and I know exactly what to expect from my (computer AI) opponents. The only redeeming qualities are found outside of the competitive aspect of Firefight. It wasn't a bad way to kick back with some buddies and blow shit up. Thing is, we moved on to other games to do that. Does that make ODST a good game? Does that make it stand out? What am I enjoying here, ODST or the atmosphere Xbox Live provides? I personally feel that this train of thought is the reasoning behind statements such as "ODST is dull." To this end, those who make claims such as these ought to clarify things. What aspect of the game do they feel requires greater complexity? Do we want Halo to be all about bunny hopping and strafe jumping? Should we increase the available dominant strategies for complication and depth in strategic play? (The latter is damned hard to pull off, as the availability of a particular strategy may not be equal among players [re: AR starts, BXR]) The answers to these questions are often rooted in value judgments regarding what style of game we prefer. As I said, the balance of these three variables can result in many varying types of games. I hope at this point the stigma of "oh those assholes just want to dong on noobs" has been abated to some extent. I feel that some of us profess a definite passion regarding the potential depth one can explore in a game. Whether you play bullet-hell shooters like Ikaruga or engage in competitive Starcraft play, there's fun to be had on this side of the fence too. Maybe it's just not as visible with Bungie and most other developers I follow, but I don't feel that any consideration for these complicated issues, such as my concept of "sufficient complexity," is truly given. We may be in the vocal minority, but I think these issues are extremely pervasive and affect the game at every level. That being said, by no means am I saying that what I have outlined above is all one should consider important to a game. Such is the power of the medium; one could likely write up an equally large piece on the depth of a video game narrative. Simply put, the above is concerned with but a slice of the pie that I feel doesn't get enough love. -hp tl;dr version:
- Blah blah blah, everyone has fun in different ways. - I feel that designing with the "hardcore" in mind is less about being a selfish asshole (hi narc) and more about adding depth and replayability to the game - For multiplayer games, adding this depth usually means raising the skill ceiling - To do this I think there are three things to consider: strategy, execution and available information - Strategy has to do with options available to the player - Execution has to do with being able to perform options described above - Available (in game) information directly affects how you make decisions for the above two points - These three variables operate within a game's "Sandbox" which consists of the game's engine and content - Making adjustments at the level of the sandbox has great implications on the three variables I describe - These adjustments ultimately have the ability to increase the skill ceiling, flesh out the skill gradient, provide replayability and add longevity to a game - I think that's pretty important - goatrope is a cool guy, hi goatrope (hey cow, fine, you can get a namedrop too) tl;dr version of the tl;dr version: I feel that designing for higher level play can greatly increase the quality of a game if replaybility and longevity is the goal. Still, it seems developers are very reluctant to do so (perhaps due to expediency and development cost). | |
|
| Replies: |
The HBO Forum Archive is maintained with WebBBS 4.33. |
