Archived

This topic is now archived and is closed to further replies.

Diodor

Constraints; granularity; tactics

Recommended Posts

Diodor    517
The problem of passing rubber objects through rubber holes is an exercise of choosing an approximate matching size between the hole and the object and applying enough force. But if both the hole and the object are rigid, the problem becomes a lot more complicated, the stuff mathematicians work on for hours. A million particles, each moving in a random way in random directions, colliding and interacting in chaos, are nothing more than a bit of inert gas engaged in pointless Brownian motion, which can be described with a few numbers like pressure, temperature and volume. In contrast, an extremely limited system such as a chess board, featuring but a few dozens of pieces, which can find themselves in but a few dozens states is most complex and thought provoking. The point I''m trying to make is that freedom and chaos doesn''t equal complexity and interesting gameplay. Just because the number of possible states in a Starcraft games is staggering Starcraft isn''t nearly as complex as chess. Limiting the freedom, reducing the number of game states, constraining the possible evolutions of the game world _can_ create very good gameplay. Don''t allow your player to build as many Nazguls as he can afford. 9 are more than enough. And a couple of wizards and about four hobbits. Don''t make starship design a matter of filling a satchel with goodies. Divide the ship in sections and only allow one class of weaponry in each section. Only use small numbers to describe anything in the game. Even the satchel problem is interesting if the granularity of the objects is large enough.

Share this post


Link to post
Share on other sites
Sage13    122
Totally off topic, but the chaos thing made me think of it.

"Life is a slippery thing to define, but it consists of two very different skills: the ability to replicate, and the ability to create order. Living things produce approximate copies of themselves: rabbits produce rabbits, dandelions make dandelions. But rabbits do more than that. They eat grass, transform it into rabbit flesh and somehow build bodies of order and complexity from the random chaos of the world."

-Genome: Life by Matt Ridley



[Sig deleted for size/space -- Kylotan]

[edited by - Kylotan on June 29, 2003 6:12:48 PM]

Share this post


Link to post
Share on other sites
MadKeithV    992
Wow Diodor, that''s one of the most profound insight-posts I''ve read here in a long time. You are right! A limited, constrained system can be much more interesting than open systems. Constraints are obstacles, and obstacles are what we want to conquer in our gaming experience.

The way the knight in chess is completely different from the other pieces, and often considered useless, until in the most unexpected situation they turn out to be the only piece with a solution.

Much like the hobbits in Lord of the Rings.

Share this post


Link to post
Share on other sites
RTF    127
To some extent it''s a matter of simulation. Chess is abstracted enough that every move will matter, wheras in Starcraft you have many similar moves whose value varies only slightly. For the purpose of carefully formulating a strategy like those of Chess, where forming position is critical, abstraction in this regard is good.

However, having detail around can really matter in a video game. If Starcraft WERE more like Chess, we would first make the sides and maps mirror images because that''s obviously simpler and easier to learn. Then we could reduce all the units and buildings combined to maybe a total of 20 or 25, with a maximum of maybe 30 units. But guess what - you just destroyed strategy in the name of strategy! In an RTS game following typical design patterns, positioning units automatically takes a backseat to production options, so you actually WANT that complexity!

Certainly you could make a game more like Myth, which is perhaps a closer comparison to Chess, with a more limited range of units, fixed quantities of them, and an emphasis on position, but guess what, they''re two different games.

I would compare it to the design of a vehicle like a car, boat or aircraft. When you choose a given purpose for your vehicle(speed, capacity, comfort), you must balance it so that it functions best in that role. But you don''t try for everything - 18-wheeler trucks don''t enter auto races, and passenger jets don''t carry troops into combat.

Share this post


Link to post
Share on other sites
Ketchaval    186
Slightly off-topic:

Granularity and location size.
Sometimes it may be better to have fewer locations, that vary significantly than to have hundreds of locations that are almost the same. Ie. It is more killer than filler.

For instance consider the old text adventure games, the locations were very granular (?), you moved from one location to another by choosing where to go next. Ie. You type east and move eastwards from the cave to "outside the cave". (and in effect you teleport to the next location) .As opposed to games such as Daggerfall where you have to literally walk everywhere.

Share this post


Link to post
Share on other sites
Inmate2993    222
This argument raises the issue of Strategy versus Tactics.

Strategey
1 a (1) : the science and art of employing the political, economic, psychological, and military forces of a nation or group of nations to afford the maximum support to adopted policies in peace or war (2) : the science and art of military command exercised to meet the enemy in combat under advantageous conditions b : a variety of or instance of the use of strategy
2 a : a careful plan or method : a clever stratagem b : the art of devising or employing plans or stratagems toward a goal

Tactic
1 : a device for accomplishing an end
2 : a method of employing forces in combat

These words are often used interchangably, which is where the confusion arises. A strategy in Starcraft, or for that matter any real-time whatever game is the abstract set of steps leading to victory, and the tactics are the actual actions made. However, with the chaos of starcraft, a strategy can have just a few tactics and be effective. For example, the early-rush, which seems to be greatly effective against players that don''t prepare for it, is just ONE tactics, who''se preparation is a bit complex, accurate timing, troop positioning, on so on.

I''m not saying that stracraft isn''t fun, but strategizing doesn''t work unless they''re really open ended strategies that take into account failure of a few tactics... and try to remember your strategy when a platoon of Dragoons just broke through the choke. In the end, you really have to be a dynamic player who can deal with each situation as it comes, which is more of a subconscious, procedural-memory strategy.

Turn-Based games tend to be unforgiving when it comes to misplacement, or wasted resources. The "granularity" demands exact tactics, meaning a much deeper strategy. Chess, if you just play to not lose any pieces, makes games that end in stalemate. The only way to win chess is to find a potential hole in the other''s strategy and drive them into it. No finding an unprotected hole or hiding the Ghost well, its more of making a list of moves that ends in victory. Try to make a list of moves that''ll end in victory with starcraft and you''ll find yourself laughed at by even the newbies.

Point is, don''t accuse turn-based games of being inferior, because the gameplay is subtley fine-tuned into a mind-game. And personally, it''s always fun screwing with someone''s head.

Share this post


Link to post
Share on other sites
Oluseyi    2103
quote:
Original post by RTF
However, having detail around can really matter in a video game. If Starcraft WERE more like Chess, we would first make the sides and maps mirror images because that''s obviously simpler and easier to learn. Then we could reduce all the units and buildings combined to maybe a total of 20 or 25, with a maximum of maybe 30 units. But guess what - you just destroyed strategy in the name of strategy! In an RTS game following typical design patterns, positioning units automatically takes a backseat to production options, so you actually WANT that complexity!
I disagree completely.

I think the RTS genre has substituted quantity for quality; for games that are supposedly conflict-/combat-oriented, they are encumbered with a baffling amount of domesticity. The need for micromanagement also mires the genre is side-tasks that serve the purpose of lending credence to boxcover claims like "Over 100 hours of gameplay!" but don''t really add to the core game.
  • Why do RTSes place you in obviously hostile/foreign territory with no base, no troops and no fortifications and require you to spend a significant amount of time procuring those? In all the wars that have occured throughout history, the forces are assembled at home and then transported - by foot, on horse, by plane, by ship - to the combat arena. For protracted campaigns such as city sieges, food, water and other resources are brought from home, and scavaging only occurs when resources run out.

    In the real world, armies are always either located at the point of conflict - which tends to be reasonably close to civilization and a source of some supplies - or in transit towards one. I can make exceptions for Starcraft in this one instance because you''re in space, but other than that it falls prey to the contextual errors of the genre.

    Furthermore, real-world armies are supported by countries who have reserve resources - at least at the commencement of engagements - that they can commit to the conflict as the need arises. ie, reinforcements.

    Summary Eliminate resource-gathering as a prime component of gameplay.


  • Why do RTS games present you with virtually nothing in the way of intel? Even in the middle ages, a spy or scout did some recon to gain enough knowledge to plan an assault; in modern times satellite imagery, recon infantry and airplanes/drones and the like provide reasonably detailed information about the lay of the land and the locations of undisguised/unconcealed opposing fortifications and targets. This allows you to formulate a battle plan and deploy your forces as efficiently as you can, reducing the probability (with good planning) of wasteful consumption.

    Fog of War is such a silly idea. Rather than fog of war, what RTS games need is a tactical map with a means of indicating the quality of information about an area. Quality decays with time, so if a recon unit takes two hours to return to base, the information there is two hours old (duh) and may be represented by a color change. If a camera is on location and connected to base via satellite (in a modern or futuristic setting), then information quality will remain perpetually high unless the camera is destroyed.

    Summary Intelligence information!


  • Why do RTS games make you a disembodied entity? This can actually be realistic if you are the Grand Commander or Field Marshall, but in that case you''re not likely to be setting objectives and giving orders to individual units; you''ll work with aggregations and delegate specifics. Either do that or give the player a physical location and eliminate the free roaming camera.

    In either case, delegation is necessary; you can''t physically be everywhere, and you can''t logically track every single action on the map. The "Your base is under attack" warning in Starcraft was such an irritation to me; why can''t I instruct a set number of units to patrol an area and only bother me if the situation is dire? Why can''t I instruct a unit to scout an area and then do A if condition C, B if condition D or E if condition F? I know Starcraft isn''t the state-of-the-art in RTSes anymore, but it is the classic example and most referred-to title.

    Summary Delegation!
Those are just three of my personal gripes, but consider that if you effect them you "simplify" the game without eliminating gameplay. If anything, you increase gameplay and make it an even better RTS - provided they are properly implemented.

Share this post


Link to post
Share on other sites
RTF    127
quote:

Those are just three of my personal gripes, but consider that if you effect them you "simplify" the game without eliminating gameplay. If anything, you increase gameplay and make it an even better RTS - provided they are properly implemented.


Those are valid complaints; but what I''m insisting, and it took me more than a few tries to say this right, is that we have a choice of paths to go down, a very broad potential variety of RTS games.

When we simplify, we reduce our choices until we are left with one. The important thing is to recognize that we have alternatives that can each have strengths and weaknesses and be popular in their own right; this is why we have such a wide variety of games in any form - even if you were to look at "two-player board games involving capture," you have a wide choice between games such as Nine Men''s Morris(Six Men''s Morris, Three Men''s Morris), Checkers(and variants), Chess(and variants), Fox and Geese(and variants)...

But we would never say that one of these games is the perfect one. In the same way, we should look ahead and know that we''ll end up with a similar level of variety in RTS games, and know that none of our changes is a guarenteed improvement, only a likely one and one of many.

Share this post


Link to post
Share on other sites
Diodor    517
quote:

Original post by MadKeithV

Wow Diodor, that''s one of the most profound insight-posts I''ve read here in a long time. You are right! A limited, constrained system can be much more interesting than open systems. Constraints are obstacles, and obstacles are what we want to conquer in our gaming experience.

The way the knight in chess is completely different from the other pieces, and often considered useless, until in the most unexpected situation they turn out to be the only piece with a solution.

Much like the hobbits in Lord of the Rings.



Limiting the number of choices at each moment doesn''t hurt the overall complexity of the game - which, in the long run, is still huge even for the most constrained of games (like chess) - but it inherently allows for planning. Each choice gains meaning as a step in a possible long-term direction.

quote:

Original post by Ketchaval

Slightly off-topic:
Granularity and location size.
Sometimes it may be better to have fewer locations, that vary significantly than to have hundreds of locations that are almost the same. I.e. It is more killer than filler.

For instance consider the old text adventure games, the locations were very granular (?), you moved from one location to another by choosing where to go next. I.e. You type east and move eastwards from the cave to "outside the cave". (and in effect you teleport to the next location) .As opposed to games such as Daggerfall where you have to literally walk everywhere.



I agree, and it''s not off-topic at all. Granularity doesn''t need to limit itself to space - a granularity of time can work just as well. VGA Planets for instance is a turn based game based on a real-time engine. Players give orders for a period of time - which is then simulated - then the players receive the results of the simulation and can give orders again.

Increasing granularity is the easiest tool to limit the number of choices, making each choice a lot more important.

quote:

Original post by Inmate2993
I''m not saying that stracraft isn''t fun, but strategizing doesn''t work unless they''re really open ended strategies that take into account failure of a few tactics... and try to remember your strategy when a platoon of Dragoons just broke through the choke. In the end, you really have to be a dynamic player who can deal with each situation as it comes, which is more of a subconscious, procedural-memory strategy.



The number of choices at each moment in Starcraft is so large and the game situations changes so fast that one can only make short term plans - pushing Starcraft to the reactive side of gaming.


quote:

Original post by Oluseyi

Summary Eliminate resource-gathering as a prime component of gameplay.



What I dislike about the resource-gathering gameplay is that it is such a single-player form of gameplay. Multiplayer Starcraft starts out as multiplayer as playing Pinball in turns to get the highest score. It is a competition, but it has no interaction at all, yet it determines to a large extent the outcome of the rest of the game.

quote:

Summary Intelligence information!



It is very interesting how in Diplomacy, a board game where all the game information is public (map layout, army locations), intelligence is paramount. Even if the game information is completely free, knowing what each player plans for the future is vital for success. Information and disinformation are key to winning Diplomacy - and each game move demands a careful analysis in the light of how the others will interpret it.

quote:

Summary Delegation!



I think increasing granularity can be a supplement of delegation. In Panzer General you do command individual units, but each unit is division size.

Share this post


Link to post
Share on other sites
RolandofGilead    100
Diodor(1st post), I see what you''re talking about but this is sort of a non-issue, at least with the examples you started out with. I think this goes back to top-down versus bottom-up. There are nearly infinite *positions* in Starcraft, but fewer true *states* than chess. It''s like the second word in your post, granularity, do you start your description at the molecule of silicon dioxide, the grain of sand, or the beach? Even in chess, if we were to count the positions a piece is in between moves, there are infinite combinations, but we are already so familiar with its mechanics that we know not to consider them.

So, back on topic and going with the rubber holes, is it better to guide the player to the states that work, let them discover which states work, or disallow extraneous states from existing(rigid holes and between the squares)?

Share this post


Link to post
Share on other sites
Ketchaval    186
The granularity of time ''units''.

Consider a turn-based game that only gave you 10 turns before the players'' scores are worked out. The fewer moves there are, the more important the player''s actions in each move become.

Share this post


Link to post
Share on other sites
RTF    127
I was just reminded of an example of where adding detail works, and it isn''t an RTS game but I think the concept would apply:

GTA3.

The game is an evolution, according to an interview quote, of Pac-Man.

There is a game that is a fascinating stepping stone between Pac-Man and the original GTA: Getaway! for the Atari 800 series of 8-bit computers. This is Getaway!''s design:

-The Pac-Man maze, turned into a large scrolling city setting where the corridors are one-way streets.

-The ghosts and fruit become three police cars and a white diamond truck.

-Instead of dots, cash is distributed randomly through the city, with the total amount kept at a fixed level.

-The player has a hideout that he starts the game at. He must return there with his cash for it to be counted towards his end-of-game total. Running around with large amounts of cash causes the police cars to follow the player more closely. While stopped at the hideout the player is invincible, but his protection goes away if he stays too long.

-Gas stations are placed at fixed locations on the map. The player may stop and refuel for free up to a limit of 99 units, but he must wait and risk getting caught while refuling, and if he runs out he moves dangerously slowly.

-The diamond truck gives the player some cash, but also sets in motion a critical part of Getaway!''s play; after getting caught the first time, three diamonds are placed randomly through the city. Once the three diamonds are collected, catching the truck again releases a new set of three items. Each set of items that is collected increases in value, and additionally increases the value of the generic cash symbols.

-The game''s difficulty increases by having the police follow you more aggressively as your stash(as well as cash on-hand) increases, and by having traps appear on the map that temporarily slow you or cut away a portion of your gas.

-Dying resets you to your hideout in the daytime, but otherwise the game runs continuously, with day and night and no levels or other changes(AFAIK)

The reason I lay out the whole game here is because it has so many obvious elements that do not actually simplify the game, but add to it nontheless. The jump to GTA1, which does away with the maze abstraction, is a fairly clear one, and GTA3 revamps the concept again, bringing it nearer to simulation.

I get the impression, though, that every form of the "criminal" game is correct in its own right, but they act differently on the player, where the simpler, more granular game is at the same time easier to throw away as a diversion but also more enticing to improve at because of its simplicity, wheras the complex one becomes an full experience, but not one that a player is likely to actively focus his energies towards improving his skill at.

RTS games could easily have the same concept applied to them.

Share this post


Link to post
Share on other sites
RolandofGilead    100
quote:
Original post by RTF
The reason I lay out the whole game here is because it has so many obvious elements that do not actually simplify the game, but add to it nontheless. The jump to GTA1, which does away with the maze abstraction, is a fairly clear one, and GTA3 revamps the concept again, bringing it nearer to simulation.

I get the impression, though, that every form of the "criminal" game is correct in its own right, but they act differently on the player, where the simpler, more granular game is at the same time easier to throw away as a diversion but also more enticing to improve at because of its simplicity, wheras the complex one becomes an full experience, but not one that a player is likely to actively focus his energies towards improving his skill at.

RTS games could easily have the same concept applied to them.

For us morons, how?

It's not that we necessarily have to limit freedom, we just have to make sure that when a freedom is added it introduces new choices? We have to control the player's freedom, not limit it? Whoa, I think I made sense.

On the experience part, what of something like the old 2d Mario games, perhaps they're not mutually exclusive or is that an example of something else.

--edit: grammar

[edited by - RolandofGilead on June 28, 2003 7:49:33 PM]

Share this post


Link to post
Share on other sites
Inmate2993    222
Confined to a strategy game, the granularity thing is the subject of ths thread, but lets consider an example of a limit to a game with fairly free movement.

MegamanX5 (or RockmanX5)
2d Side Scrolling platformer and action game. Lots of weapons and moves, basically a lot of freedom to the game.
EXCEPT, you have 15 turns before the game ended to accomplish a goal. Each turn was spent when you entered one of the Levels. So, in a weird sense it was like have 7 continues (8 levels, 7 extra turns).

In fact, I think almost all of the old games have a 3-lives rule. 3 chances to screw-up. So, even then there was a really rigid difficulty.

Share this post


Link to post
Share on other sites
RTF    127
Re: RolandofGilead

You can boil it down to a set of layers. The bottommost layers are the critical parts that you''d always see - the "real-time" in RTS, and units, but maybe only representative of a larger group or abstracted like a board game. In higher layers you''d add resources and start differentiating units as individuals or with distinct types. The highest layers would approach simulating the "real world," or more likely the most realistic version of the fantasy world in the case of an RTS.

One thing that I think is important to consider when looking at it this way, is that you can''t really start at the highest layers, but you have to build up from the bottom every time you make a game. At best you can say, "we''ll steal the groundwork from these games," which is how we get genres in the first place, but you may want to rip out parts at the bottom anyway and rebuild them differently(like the many RTS games that use indirect unit control as opposed to lasso and conquer).

And if you keep your design at the lowest levels, you end up with a Tetris or a Pac-Man or a Chess - games that are probably going to stay around forever, but wouldn''t match up, in a financial sense, to your usual A-list title when put side-by-side on the retail shelf. The progression we''ve had in game design is to push the boundaries towards ever-more-sophisticated simulations, in almost a mirror of our trends towards photorealistic graphics and high-fidelity sound.

On Mario etc.: Games which move you through a story or through themed settings are amusement park rides with challenges thrown in - much of the development work goes into building the player''s excitement and then letting it out with things like harder levels, boss fights, new gadgets to collect, experience and levels to gain, etc. When gameplay can be divorced from the story/settings(like with Elite/GTA/sports games/etc.), then the game becomes a virtual world, with storylines acting as a ride within the world that can cover any sort of ground you want; a far more powerful attraction than your typical rollercoaster, I''d say.

The same ideas regarding layering still apply to the "game-as-ride," only they can also be changed around at different points with things like sub-games, different controls and settings(swimming, flying, driving) to serve the ultimate purpose of making the player happy.

Share this post


Link to post
Share on other sites
MadKeithV    992
quote:
Original post by RTF
... Chess - games that are probably going to stay around forever, but wouldn''t match up, in a financial sense, to your usual A-list title when put side-by-side on the retail shelf.



First, I think that over the several thousands of years that chess has been around, it''s made more money (albeit for different people) than any computer game, but that''s not entirely my point.

My point is that the retail shelf doesn''t really come into this discussion. It doesn''t make business-sense to simplify, because it''s easier to catch a broad group of players by having a (seemingly) diverse game. It _does_ make game sense though.

The one thing you should take away from this thread is that the perceived added complexity of a million different units in fact does the opposite. It completely blands out the playing field to a homogeneous mass, since every combination of strength and weakness is present, and the game is almost reduced to "finding the unit that isn''t quite so balanced as the others".

People should really start playing chess and diplomacy a bit more. Diplomacy is the epitome of incredibly simple gameplay. It doesn''t have any randomness, if you have more armies attacking than the "enemy" has defending, you win. However, planning your moves is the entire game. You talk to the other players, make alliances with them to take over certain territories, lie, cheat, but do it in such a way that it won''t bite you back in a few turns, unless you can mop up the game in a few turns.

Something like this should be just as possible in an RTS, or TBS. For instance, imagine if you start off with a base that''s already there, defended by really heavy cannons, spaced quite wide apart. These cannons will destroy any unit they hit, no question about it. However, they have weaknesses, such as a slow firing rate, or a very limited field-of-view (only in front).
A frontal assault on a base like this with a few really heavy troops would be suicide. A rush with a whole lot of little guys might get you a few onto the walls of the base (with heavy casualties). But then a single sapper can place a charge and blow up a cannon. Just eliminating a single cannon could breach the defenses enough to have a small corridor through which you can send more troops inside the base, to eliminate more cannon.

This all sounds really interesting, and you might say "you can do exactly this in most RTS games".
But now imagine you add cheap aerial units - like in most RTS games. The aerial units are highly effective against infantry, but not against much else (after all, the designer said, there has to be some balance). Combined, the fortress with enough cannon and aerial units is practically impenetrable. You''ve covered too many options with your unit choices, and winning starts coming down to who can take advantage of his opponent''s misclicks the most, or who is the most efficient at gathering resources and getting enough of a material advantage over the other to just rush the living daylights out of him.

Nomatter how different the individual units seem, if there are enough units, and enough variation, the end result is still a very grey mass with little differentiation. Avoiding the grey mass is key to getting interesting interplays. This can be done either through granularity in the number of units (if you only have 16 units on each side, you are going to notice the difference between them), through imbalance in the ability of units (one side is all cheap, mass-produced infantry, the other can slowly build expensive, heavy machinery), granularity in the gameplay (a decision you make lasts for quite a long time, deciding you will put your resources to building infantry is going to have you gasping for breath if you turn out to need artillery in a few minutes), or combinations of those. There are probably other ways of improving the experience, and there are probably also a bunch of AA/AAA games out there that incorporate some of these ideas.

Share this post


Link to post
Share on other sites
Jenison    216
quote:
Original post by MadKeithV
Wow Diodor, that''s one of the most profound insight-posts I''ve read here in a long time. You are right! A limited, constrained system can be much more interesting than open systems. Constraints are obstacles, and obstacles are what we want to conquer in our gaming experience.

The way the knight in chess is completely different from the other pieces, and often considered useless, until in the most unexpected situation they turn out to be the only piece with a solution.

Much like the hobbits in Lord of the Rings.



constraints can also be too much ... while the concept of a game is to do tasks and not just hit one button and your done ...you want the player to always feel like they are making progress of some kind. Keeping them hanging for too long will make them lose interest.

Take a look at the ultimate game ... Tetris. Its so beautifully simplistic with just about 0 flexibility. Sure you could be a little creative in your ordering of the tokens but you can only do that for so long. Then the tokens just drop too fast.

One of the all-time blunders was just had by SOE and Everquest. If you dont know the story ... Planes of Power expansion was released. End game was not completed. SOE specifically prevented players from advancing to end game content by making the task impossible to defeat. This went on for MANY weeks. Finally SOE let the players into the end game content. But in the process pissed off a lot of players ... and caused many too leave.

The point being ... the feeling of progress and accomplishment is needed in ANY game...ridged or complete freedom. The GREAT thing about total freedom games is you can just screw around and have some fun that maybe wasn’t intended. I mean ... we all have played GTA3 and said ok...how many of these hippies can I run over before my car is blown up. Then trying to beat your record

Share this post


Link to post
Share on other sites
Sandman    2210
Excellent post Diodor, you''ve managed to elucidate a lot of things that I''ve been thinking lately, but without quite being able to put my finger on. (although some might take issue with your choice of Starcraft as an example: I think that of all the popular ''generic'' RTS games it suffers from Grey Mass Syndrome the least)

One thing I''ve been striving for in my ruleset is simplicity with emergent complexity. I''ve just completely thrown out my combat rules because I decided they were too complicated and not interesting enough. It''s a tricky process, since you have to decide which rules are necessary for interesting gameplay and which ones are just unnecessary complications.

Another thing I''d like to do is to avoid hardcoded restrictions, but to ensure that certain restrictions are emergent from the simple game rules. For example: Rather than saying the player can only have 9 nazguls, say he can have as many as he likes, but due to the gameplay rules, more than 9 nazguls would be generally a Bad Idea. I prefer this because it lets the player learn where to draw the lines, and there may also be certain cases where the common rule does not apply.

Of course, the problem with this is making those restrictions arise from simple rules, rather than being fudged in order to balance a strong unit. The latter increases the complexity of the ruleset, and results in more ''lawyer play'' and less intuitive play.

Share this post


Link to post
Share on other sites
MadKeithV    992
quote:
Original post by Sandman
Another thing I''d like to do is to avoid hardcoded restrictions, but to ensure that certain restrictions are emergent from the simple game rules.



I agree with this, to a point. It depends on what kinds of "hardcoding" you use. A hardcoded restriction of 9 nazgul is an invisible limit to someone who hasn''t read the books. All of a sudden, you can''t make any more, though you have all the required resources. That is annoying to the player.
Same with the Everquest expansion mentioned. There is no apparant reason why all of a sudden something can''t be done, where it seemed to be possible before.


There are ways you can make it work though. Consider the two "lakes" in stratego. They are there from the start of the game, so you accept them being there. The game doesn''t really need them, and they break an otherwise very regular board, but they do add to the game. Because the constraints have been established a-priori, and not hidden from the player, there is no real issue.

Same with the odd movement pattern of the knight in chess. It''s a weird piece, but it''s there, and the constraints are made clear to the players.

It should be the same for games. No hidden tricks, everything up front. How unfair would it be if all of a sudden all infantry pieces would turn out to be invulnerable to the biggest enemy weapon. Sure, the designer can shout "SURPRISE! There''s a trick you can do with infantry", but I doubt the player who just built the huge gun would be amused.

The reason is that you have to be able to plan your strategy, and the only surprises that you should get are things you overlooked, scenarios you thought were improbable, not that all of a sudden someone can or can''t do something they could or couldn''t earlier.

If something goes wrong, you should think "damn, I should have know that would happen!", and not "hey, you can''t do that, can you?"(of course, assuming the player does know all of the common rules!)



Share this post


Link to post
Share on other sites
Ketchaval    186
quote:
Original post by MadKeithV

My point is that the retail shelf doesn''t really come into this discussion. It doesn''t make business-sense to simplify, because it''s easier to catch a broad group of players by having a (seemingly) diverse game. It _does_ make game sense though.



Off the topic of granularity:
As you say Chess ''seems'' to be less diverse than other games. Is this because it doesn''t have ''levels'', what if it came with pre-set levels that provided ''diverse'' challenges. (Ie. Starting off mid game).

The second thing is that Chess has a steep learning curve, and is probably too complex for many people. It is not clear what sort of strategy is more likely to succeed.
Imagine "Chess for Dummies" where it starts off with easy scenarios, and moves on to steadily harder ones. Ie. You outnumber the computer opponent in the first scenarios..

Of course this approach wouldn''t really work with the rules of Chess as they are currently.. but if there was kind of a simplified variant, this kind of approach could work.

Share this post


Link to post
Share on other sites
MadKeithV    992
quote:
Original post by Ketchaval
quote:

Imagine "Chess for Dummies" where it starts off with easy scenarios, and moves on to steadily harder ones. Ie. You outnumber the computer opponent in the first scenarios..



Actually, that method is used to teach people how to play chess in various situations. You start off with easily resolvable situations, going to progressively harder tasks, until you can play full games and apply your learned knowledge.


Share this post


Link to post
Share on other sites
Sakuranbo    122
I didn''t read the entire thread, so this may have been mentioned already, but I have one beef with the proposed limits described in the OP; they''re an arbitrary black box.

Games like chess are abstracted enough that they can have precise, unjustified limits, but most computer games strive towards a higher degree of realism. The player of the Lord of the Rings RTS will be asking, "Why can''t I have that many wizards and hobbits? Does something bad happen if I do? Even if it does, why can''t I try it anyway?"

Properly explained limits- like the fact that only nine Nazguls exist anyway- aren''t bad, but disadvantages are usually better. Maybe the combined magical energy of your three wizards goes wild and turns that lone orc sneaking up behind your base into a good-sized fire drake. Maybe your fifth hobbit doesn''t get along with Frodo and defects to the enemy, taking a sack of your gold with him. All are more satisfying then being told by a mindless computer, "You can''t have that many, because I said so."

Share this post


Link to post
Share on other sites
Sandman    2210
quote:
Original post by MadKeithV
If something goes wrong, you should think "damn, I should have know that would happen!", and not "hey, you can't do that, can you?"(of course, assuming the player does know all of the common rules!)



Yeah, that's pretty much the feeling I want to get. However, the 'common rules' need to be simple enough that a player can understand them very quickly (qualitatively, at least) or the is somewhat lost.

For example: there are many cases where these 'Doh! I should have known that particular strategy won't work' moments occur in Starcraft. However, because the game relies heavily on the very complex interaction between all the different units, a beginner playing the game tends to experience more of the 'holy crap, I didn't know you could do that' type moments. As a result, the game can be incredibly frustrating for a non-hardcore player or newbie, and it might well put people off getting into the game further.

Chess on the other hand, is quite newbie friendly. You can write all the basic rules on one side of A4 paper, even including some of the weirder special cases like the en passant rule. That's not to say a newbie will necessarily be any good at it of course - but at least he will feel that he has been beaten fairly.

quote:

It should be the same for games. No hidden tricks, everything up front. How unfair would it be if all of a sudden all infantry pieces would turn out to be invulnerable to the biggest enemy weapon. Sure, the designer can shout "SURPRISE! There's a trick you can do with infantry", but I doubt the player who just built the huge gun would be amused.



This reminds me of one of the first games of Red Alert I played. I spotted my (more experience) opponent massing a huge bunch of units not far from my base. Deciding to make him pay for his foolishness, I built a nuke and launched it at his foolish men, only to discover that the nukes in Red Alert were totally stupid and apparently only affect buildings. Thus, my nuke was wasted, and his huge swarm of mammoth tanks came in and wiped out my base shortly after. That was a definite "WTF??" moment.

[edited by - Sandman on June 30, 2003 1:12:13 PM]

Share this post


Link to post
Share on other sites
Guest Anonymous Poster   
Guest Anonymous Poster
Nice Topic Diodor

Anyway i think hard-coded ''unreal'' limits and constraints should be few and well chosen, unless it is made clear to the player that the game is very abstracted (like chess).

Even a chess newbie wouldnt ask _why_ the ''units'' move in the way they do.
If a Starcraft Marine could only move upwards and not back or two tanks would fit in a dropship, but a single tank would not, people would be confused, if there wasnt a realistic/logic (to the world the game plays in) explaination for that.

The reason for that would be that Starcraft tries to be realistic to some level, the player can kinda identify with the action in the game.

Hard-coded rules should come into effect, when the reason for them is purely game-design and not logical (f.e. a unit limit, the godlike unit-control and mapview).

If you want your game-world to be believable, dont tell the player "You can only build 9 Nazguls."
Atleast tell him "You can only build 9 Nazguls, because there are only 9 rings, which have the power to turn a human into a Nazgul" (Yes i know, maybe you should provide a reason for only 9 rings existing too )

The player knows the world we live in, he expects that things fall to the ground, that people eat, that when he goes outside it might rain.

If now in your game you dont have gravity and substitute rain with meteors which are hitting the earth even without gravity, dont expect him to accept this without providing background information.

Unless ofc you keep the game rules simple enough to make him not ask.

Still i agree in general limiting freedom and realism in general can produce the best game-designs.

I have to disagree with the points Oluseyi made aswell tho.
You can move to the realism style of the RTS games he proposed, but i feel this is _not_ necessary to create RTS games which are fun to play.

quote:

Why do RTSes place you in obviously hostile/foreign territory with no base, no troops and no fortifications and require you to spend a significant amount of time procuring those? In all the wars that have occured throughout history, the forces are assembled at home and then transported - by foot, on horse, by plane, by ship - to the combat arena. For protracted campaigns such as city sieges, food, water and other resources are brought from home, and scavaging only occurs when resources run out.

In the real world, armies are always either located at the point of conflict - which tends to be reasonably close to civilization and a source of some supplies - or in transit towards one. I can make exceptions for Starcraft in this one instance because you''re in space, but other than that it falls prey to the contextual errors of the genre.

Furthermore, real-world armies are supported by countries who have reserve resources - at least at the commencement of engagements - that they can commit to the conflict as the need arises. ie, reinforcements.

Summary Eliminate resource-gathering as a prime component of gameplay.



In a fast paced (multiplayer) game, on a relatively small map, where the obvious goal is to destroy your opponent(s). Considering you take out (prdocution-)bases and resource gathering, why would you want to attack your enemy ? In defense he will most likely have a bonus of some kind... bunkers, turrets, higher ground, a river you have to cross. Maybe i am just uncreative, but i cant think of many ways to make a player attack the other under such conditions.

Bases and resource gathering can give a game new strategic possibilities. If your army is smaller than his, without the option to attack him somewhere where he might not have many units
or the option to out-produce him, your chances of a comeback are relatively small.

quote:

Why do RTS games present you with virtually nothing in the way of intel? Even in the middle ages, a spy or scout did some recon to gain enough knowledge to plan an assault; in modern times satellite imagery, recon infantry and airplanes/drones and the like provide reasonably detailed information about the lay of the land and the locations of undisguised/unconcealed opposing fortifications and targets. This allows you to formulate a battle plan and deploy your forces as efficiently as you can, reducing the probability (with good planning) of wasteful consumption.

Fog of War is such a silly idea. Rather than fog of war, what RTS games need is a tactical map with a means of indicating the quality of information about an area. Quality decays with time, so if a recon unit takes two hours to return to base, the information there is two hours old (duh) and may be represented by a color change. If a camera is on location and connected to base via satellite (in a modern or futuristic setting), then information quality will remain perpetually high unless the camera is destroyed.

Summary Intelligence information!



Refering to Starcraft again, scouting is vital in the game, if you dont scout regularly you are pretty dead if your enemy knows what he''s doing.
Scouting options are pretty limited (some suicide unit, a fast flying unit hoping he lacks air-defense, observer, scan, overlord) but adding more units with a scouting purpose would create a grey-mass of scouting options not really different to each other.

Point taken tho, many games lack in that area

Fog of War:
In a classical style RTS, where you can overlook the entire map in a matter of seconds i''m all for keeping it the way it is.
Leave it to the player to decide how useful and current the information he gained is. If the map is huge and span of a game is more than 30-60 mins, such a feature would surely be helpful.

(Off-Topic: Implementing such a ''map'' as base for a scouting AI sounds good to me tho)

quote:

Why do RTS games make you a disembodied entity? This can actually be realistic if you are the Grand Commander or Field Marshall, but in that case you''re not likely to be setting objectives and giving orders to individual units; you''ll work with aggregations and delegate specifics. Either do that or give the player a physical location and eliminate the free roaming camera.

In either case, delegation is necessary; you can''t physically be everywhere, and you can''t logically track every single action on the map. The "Your base is under attack" warning in Starcraft was such an irritation to me; why can''t I instruct a set number of units to patrol an area and only bother me if the situation is dire? Why can''t I instruct a unit to scout an area and then do A if condition C, B if condition D or E if condition F? I know Starcraft isn''t the state-of-the-art in RTSes anymore, but it is the classic example and most referred-to title.

Summary Delegation!



Very dependant on personal preference i''d say, but i actually like doing something all the time, instead of watching AI agents execute my orders and still have the whole map in reach with 1 or 2 clicks.

If the AI was good enough to replace you in your absence, why play at all ?

quote:

I know Starcraft isn''t the state-of-the-art in RTSes anymore, but it is the classic example and most referred-to title.



Worst sentence in your whole post :-)
Starcraft is still the ''best''/most popular/most played RTS game around.

Now for those claiming Starcraft lacks strategic depth or you cant plan long-term strategies. Lets just say you should actually play the game above newbie level. Perhaps tomorrow i''ll write more about Strategy and Starcraft, but imo this is a bit off topic.

And I want to go and get some sleep now, hope to see some criticism tomorrow.

Force

Share this post


Link to post
Share on other sites