• Advertisement

Archived

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

Shameless Self Plug (Play Balance Article)

This topic is 5810 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

Hi All, Gamedev just posted my game balance article. You can see it on the news page. The link is: http://www.gamedev.net/reference/design/features/balance/ Let me know what you think! Tom Edited by - Zileas on February 22, 2002 11:51:44 AM

Share this post


Link to post
Share on other sites
Advertisement
I printed it out and read it right away.

Thought it was a great article. Lots of actual facts and details and references (instead of the usually vague comments people like me make).

Share this post


Link to post
Share on other sites
I thought it was a good article, although I disagreed strongly with the Fragmentation formula. I''ve always found many small units to beat a single larger unit, since (a) most games do indeed model some sort of attrition on damaged units, by slowing them down or so on, and (b) many small units present a harder target to hit than a single large one.

Apart from that, it all made perfect sense.

[ MSVC Fixes | STL | SDL | Game AI | Sockets | C++ Faq Lite | Boost ]

Share this post


Link to post
Share on other sites
very nice article, I''m glad that you brought up some points that I''ve tried to point out before, which seemingly have fallen on many deaf ears (for example the concept of macrobalancing before microbalancing...in my posts I called this going from the large to the small rather than vice versa)

However, there is one point I''d like make. Namely that I don''t think balancing is not such a crucial element as people may think. I had a post in here awhile back that questioned the very purpose of balance. If balance is done solely to try to create a 50/50 chance of winning all things being equal, then I believe that balancing as a design consideration is somewhat superfluous, and perhaps even unnecessary.

I think that some games require balancing...the sort of competition style games. But I think there are other styles of gaming where the point of balancing (to ensure equal chances of victory) can actually be self-defeating. In my post, people came up with the idea of "victory conditions", but I think though this can cover some cases, it''s still missing something of the point. It''s a rather philosophical point, but I think it can introduce another element of gameplay.

We tend to equate games with the goal of "winning". Balancing gives us a sense of confidence that it was the player''s skill that allowed this victory. But I think there are other measuring sticks and other goals that can be pursued other than winning. In these styles of games, balance is not only necessary, but perhaps undesirable.

I know it sounds crazy to consider making a game where the odds are stacked against you, and more than likely the player will lose...but think of many of the greatest movies of all time. They were about sacrifice and duty, no matter the cost. I''m a firm believer you can''t teach a player the same feeling of sacrifice and duty as can be felt in movies or books when you have a good chance of victory.

I just thought it would be important to point out that balancing should not be a necessary requisite of games...and not do to poor planning, but a conscious design consideration

Share this post


Link to post
Share on other sites
Hey Kylotan,

The fragmentation formula is a mathematical truth within the parameters defined. Your are bringing in arbitrary additional game rules which of course make it "not work?

The things you mention which can negate its effect, namely damaged units degenerating, and smaller targets being harder to hit, are properties of specific games. I''m talking about a very abstract mathematical truth It applies to most situations, if adapted properly.

Thanks for the feedback Glad you liked it overall

Tom

Share this post


Link to post
Share on other sites
Dauntless:

I believe balance is absolutely crucial, and i see most of it as bringing out the gameplay you want to be seen. Its not about 50/50 fairness most of the time, but merely, abut game elements all being relatively useful, or else not there at all.

It is a fairness thing with multiplayer, but thats just one application.

Zileas

Share this post


Link to post
Share on other sites
Sorry Zileas, but I still disagree with you about balncing being crucial, since it really depends on the kind of game that you want to make. If you want to make a competitive style game, then balance IS crucial. But if what the game designer is looking for is an emotional experience, then I think balancing has to be thought of in different terms.

Now when I say "balanced", I don''t necessarily mean 50/50 balance, but this is the most common. I also refer to "balance" as a verb. In other words there is the adjective "balance", as in "all things are in balance" (in this sense balance and equilibrium or equality are synonyms), then there is the verb balance as in, "balance the game so that it''s impossible for a player not to have a freedom of choice".

See the difference? I disagree with the adjective form of balance....to have things in "balance". But I do agree with the verb form in that there must be some forethought and planning done so that you don''t wind up with a game that you can only play 10 minutes before realizing you can proceed no further.

If you mean balancing in the verb form to make sure that things have a rational and consistent meaning and/or purpose, then I agree entirely. To me, balancing is not about making things fair, but about making sure that there is a logical consistency between all the various elements...in your terminology, the macro world has to make sense, though even here it is not crucial for things to be "even". As I pointed out in the other post, you can take several examples from history in which the side that everything going against it won. To make a game "even" is only one kind of balancing.

But, everything I mention here is for what I coin "subjective games". If you are trying to design a game with quantifiable goals, and are designing a competitive game (I tend to call these games more "sports" than games, since it conveys their purpose more clearly). So I think your article is superb for these kinds of games, but I categorically believe that balancing (in the adjective sense) is not necessary for all styles of games.

Share this post


Link to post
Share on other sites
Great article - however you don''t note the caveat involved with ''game element modularity'' in that it can easily lead to an issue with what you call ''complexity control''.

If the designers of SC had in fact put mutalisks in their own class, separated their attack types etc, then the results could be quite unintuitive. In most cases, designers will be calling on common knowledge to help players understand the game (eg. broodling only affects organic units, which is a heck of a lot easier to grasp than a list of units). But then you have to stay within the boundaries of common knowledge if the game is complex. You can''t just ''tweak the levers'' willy-nilly.

Share this post


Link to post
Share on other sites
quote:
Original post by Zileas
The fragmentation formula is a mathematical truth within the parameters defined.

1 == 2, for sufficiently large values of 1.
quote:
Your are bringing in arbitrary additional game rules which of course make it "not work?

The things you mention which can negate its effect, namely damaged units degenerating, and smaller targets being harder to hit, are properties of specific games. I''m talking about a very abstract mathematical truth It applies to most situations, if adapted properly.

Let''s just say that I believe the parameters were insufficient since most games I''ve encountered model one or both of those factors. I feel maybe this should have been pointed out, as using mathematical equations tend to carry a lot of weight and may bias designers to make the smaller units cheaper, when this could result in their game getting even less balanced due to the presence of one or both of the above factors.

[ MSVC Fixes | STL | SDL | Game AI | Sockets | C++ Faq Lite | Boost ]

Share this post


Link to post
Share on other sites
quote:
Original post by Dauntless
... so that you don''t wind up with a game that you can only play 10 minutes before realizing you can proceed no further.



Or that you''re not interested in proceeding further because the challenge level does not suit you.

I wonder though what sort of game you have in mind as opposed to a "competitive game". Single player games most certainly do stack the odds against the player, so I presume you don''t mean them.

Share this post


Link to post
Share on other sites
Fragmentation formula: nice one, very useful. Although in real virtual life small units can get very disadvantaged compared to large ones because it''s a lot harder for them to keep cohesion and to concentrate their fire. Some very strong units may destroy the small units as fast as they enter the fray, reducing the smaller unit''s efficiency to a flat zero. In the formula it is assumed all the units fire all the time.

Also, many games offer some form of healing/repairing. If small units engage in a battle they win due to superior numbers, chances are they lost some of their forces. But if large units do this, chances are all they need to restore to full power is a visit to the local battery (protoss archon).

And large units can be better micro-managed too. Often the player can withdraw damaged large units to the back and save them while still using their firepower.


About the need for balance: while the main weapons/units need to be balanced, the game can add some weird units that are too expensive/generally not worth it, but have an unique ability that makes them interesting.

Examples: Half the weapons in CounterStrike. Heavy Machinegun is an underdog and very expensive too. It''s a cool weapon to have tho, I don''t wish they killed it.

Dragons in Warcraft 2. Useless. They costed 2500, took a bunch to build and could be killed by a single mage with full mana (costing 1200). Their special ability - the only unit that could fly was interesting, but overall they were just not worth it.

Share this post


Link to post
Share on other sites
Dauntless: I think balance is needed even if you are not necessarily playing a balanced game.

Consider a game involving two different races. Lets also imagine that you just made all the stats up on the spot. Without going through at least some balancing procedures, either of the following circumstances could arise...

Even if you are using a role based design, you still need some balance. If force A has unusually good artillery, then heavy use of artillery for force A is a dominant tactic. Worse, if your multirole units are too good, then you destroy the whole game balance and the game turns back into a mindless tank rush.

You have no idea that the two forces are even balanced at all. Force A could have better units overall than force B, which gives force A an obvious advantage. While some players may relish the challenge, you should at least give them the opportunity to play a balanced game. Otherwise a simple dominant strategy is 'always be Force A'. Even if you try and give force A players some sort of handicap, this is kind of shutting the door after the horse has bolted - it is just guesswork, and could still be completely wrong.

In other words, I think you need to know whether a fight is fair or not. You don't necessarily have to have fair fights all the time, but you still need to be able to know this.

Edited by - Sandman on February 25, 2002 6:49:02 AM

Share this post


Link to post
Share on other sites
Sandman:

I know what you mean. That''s why I think there''s a difference between using the word balance as in, "both sides are balanced" (think see-saw here, where each side has equilibrium), and saying, "we need to balance our units out so no one unit is too strong" (in Zileas terms, I guess this would be microbalancing). So I''m all for balancing in the sense that you have to ensure that no one strategy or unit will dominate, since in real life, this rarely if ever happens.

So the trick becomes as you said, knowing if the gameplay itself will be balanced. I think if you have microbalancing, then you don''t really need worry about macrobalancing, although even this isn''t always the case. For example look at the Vietnam War. The Americans had superior firepower and quality. The Vietnamese had a couple of advantages of their own of course, namely surprise, a patriotic fervor, and no politcal handicapping. But if you look at a pure microbalancing scale, there''s no way in the world they could have won.

So how do you make sure that no one unit, tactic or strategy becomes unbalanced? I think in some ways, the game has to be open-ended enough so that it is actually a part of the gameplay. In other words, just like in real life, unit capabilities and military doctrine should be allowed to change and to be adaptable to a certain degree.

What most people do is design a rigid unit creation system. This in turn leads to players discovering what combinations of units become most effective against what other types of units. But what if you don''t know the capabilities of the opposing player''s units? What if you don''t know how good his leaders and intelligence capabilities are? What if, like in the real world, you can advance your units capabilities and change their strategies (much as the American Navy did in WWII). As an example, at Pearl Harbor, the thinking was that battleships were still the king of the sea. For some reason though, the Japanese...the very people that had inflicted such grievous damage with their own carriers, did not yet fully realize their power. After Pearl Harbor, that changed...and thanks to the surviving Carriers of the Pacific fleet, the Americans and the rest of the world discovered their potential too. Also, submarines were not given the healthy respect that they deserved before the beginning of the war, but the Germans showed the folly in that thinking as well. Indeed submariner casualties were almost as high (in some reports higher) than those faced by the Bomber crews of the mighty 8th. But the americans own use of submarines in the Pacfic showed much the same damage that the U-boats did.

So how would I do this for a game design? Create an open ended unit creation system. And I don''t mean just a technology tree where you know what kinds of units will be created. Instead, you have to have almost a sub-game itself devoted to the engineering principles of designing new units. When you have the capacity to create units then the enemy will always be on their toes, because they don''t know what the capabilities are. A very well documented case of this was the beginnings of electronic warfare during bombing runs between the Luftwaffe and the RAF. Each side would come up with a new capability and the other side had to figure out how to counter it (in one brilliant case, the British discovered that German bombers were riding on radio waves to guide them toward their target, so they figured out a way to mislead them with their own radio waves).

I can see people arguing that you can create a unit creation rules system that can create uber-units. This is possible, but I have a hunch that it would be far easier to create a patch to kill this form of "cheating" than by having fixed units. I think the unit creation system also has another strategical element to it to add to gameplay. Much as the Germans discovered to their dismay, even though their tanks were superlative, they often faced lots of little bugs. The Germans were so keen on creating new types or modifications, that they had a lot of logistical and maintenace nightmares with their tanks. The Sherman or T-34 tanks, while perhaps not the equal of Panzers, were nonetheless incredibly reliable and easier to maintain. So a player who tries to adopt a strategy of always tinkering with his units may run into these same problems. I do think there is a potential disadvantage to an open system however; the focus of attention on designing new units rather than on the strategy itself. I saw this time and time again in battleTech and Carwas (games that have open-ended unit creation rules), so this is a definite potential pitfall.

But, it is a good question. I think that balancing does one or both of two things. It limits gameplay/creativity or it gives a false sense of certainty. I think it''s far more interesting to play in a world where information is hard fought for and won only through guts and ingenuity. In todays games, too much is taken for granted as to the knowledge of your opponents forces. Even if you don''t know what his forces are comprised of, you know what he is capable of. And this is a HUGE advantage that real world commanders can not afford. Heck, I think players have too much knowledge and control of their own forces.

I think the reason that people have a hard time swallowing alot of my ideas is that they are so radical from the norm, that people just can''t conceive of them. We are so used to games being in a certain format and paradigm, that when something truly breaks the mold, people go, "huh? what the hell is that?". Just look at Battlezone for an interesting game concept that no one really played because it didn''t fit their notions of what a game SHOULD be, not of what a game COULD be. I perfectly realize that my game if it ever came to light probably would be very unpopular But for a few hardcore grognards, I think they would be very interested (I hope). I freely admit that I look at things more from a simulation perspective rather than a game perspective, so that''s probably another reason I think have different ideas.

Share this post


Link to post
Share on other sites
Guest Anonymous Poster
Plasmadog:

I was thinking either against computer opponents, or even multiplayer games but if you played a side with a disadvantage or handicap. Like I said before, I''m more of a simulations kind of guy, but creating "balanced" games would simply be impossible to do justice to portraying certain real world situations (Vietnam War, American Revolution, 6day war, Soviet-Afghan War, etc etc.).

I made a post back here a long time ago about having historical FPS set in pre WWII times. For example, any one of the American Indian wars, the Boxer Rebellion, American Civil War, the English Colony wars of the late 1800''s, etc could have been used as the backdrop. Some people pointed out that usually one side was at a tremendous disadvantage. But even here, I think that you could have some interesting game possibilities (imagine playing a Kit Carson scout hunting down your own people, or being an apache warrior trying to stop them).

In other words, balance to me is a limitation of creativity and imagination....usually in the interests of "fairness". But what''s "fairness"? The chance to be able to win? And therein lies the rub. What''d more important to having fun, "winning", or having an intense and interesting experience?

If you are confused by what I mean, let me take a couple of movies as examples. Look at Braveheart for example. Here, the protagonist of the story dies a horrific death (his real death was actually more gruesome than was portrayed btw) and yet we the audience were carried to an amazing place despite his loss (true the Scots won, but stupidly married right back into the english family again not much later). How about Glory? Here we see the amazing struggles of the most famous African American Regiment in the Civil War, and yet their leader (and the majority of the regiment) gets blasted at the end. How about Saving Private Ryan, where the audience is shown the true carnage of war and the insane conditions that the "greatest generation" had to face...and the main protagonist dies.

So what REALLY are we doing with games? If you look at the definition of a game, I suppose the ultimate object is to win, but computer "games" can be alot more than that. They can be interactive movies. In which case the object of balance is superfluos, and as i said unnecessary. So like I said, if you create competitive style "games", then yes, balance is necessary to ensure fairness, but in many other styles of games (namely games that try to reach an emotive experience) balancing need not be such a hugely overriding concern. As I have pointed out, you do have to do enough balancing to ensure that you can not reah an unplayable point, but being at a severe handicap is not out of the question and may actually be desireable.

Share this post


Link to post
Share on other sites
Yeah, I figured that was you Dauntless. And I see what you mean, but in the scenarios you mention, balance is still important. It''s just that in those situations you are either going for a deliberate imbalance, which entails exactly the same process as if you are creating an equal balance, or you are simply balancing something other than "unit strengths" or other conventional metrics.
This, in very general terms, is how I view the issue of balance:
When you play a game, there is always something that you want to achieve. As discussed in previous threads, this victory condition need not be a conventional win, such as defeating an opponent, but like you say, achieving an emotional experience, having fun, whatever. I think you''ll agree that achieving this victory condition should not be a matter of chance, it should come down to the decisions the player makes (as Sid Meier says, "a game is a series of interesting choices"). Good decisions lead to victory, bad decisions don''t, but may lead to a better understanding for next time. Balance, as a verb, is the process of making your game comply with that statement.

Share this post


Link to post
Share on other sites

  • Advertisement