Okay, let me try to clear this up once and for all. RPS is a very poor element for determining game balance and I beleive actually detracts from gameplay. Once you have units set up that purposefully have built in "nemesis" units, then you run into a situation where you have formulaic armies. Hmm, my opponent has two Orc war wagons with 5 wolf riders and 10 Uruk Splinter Throwers? Ha, fool...I will easily counter that with my 2 Elven War Eagles, 8 Elite Shardmen and 4 Elven Gladiators!! Theonly trick is in having enough resources to create your counter-unit. Trouble is, a units effectiveness is not a simplistic formulaic determination. UnitA will not always beat UnitB depending on a great number of factors. These factors can include but are in no means limited to: 1) morale 2) unit discipline/quality 3) leadership 4) environmental modifiers 5) effective combined arms warfare and unit cooperation So, lets take another example. SideA has 4 rifle squads and an AT squad vs Side B''s one Infantry Fighting Vehicle and two light infantry squads. The first side is apparently at a disadvantage. The IFV is very good at dealing with infantry, and it is supported by two light infantry units. But let''s throw in some other factors now. They are fighting town to town, and SideA has a very talented 1st lieutenant in charge. Also, SideB has just been cut off from the rest of it''s fighting force, causing it to drop in morale, as it is losing contact with the main force. Under normal circumstances in open terrain, SideB would have been able to beat SideA 80% of the time. Under these new circumstances the chances of victory now go SideA. But the most important part is...it''s still not a given. There are random variable factors at play which will cause the outcome to be an unknown. In RPS style of balancing, everything is formulaic...when group A fights Group B, you''ll know who will win. some of these calculations may be very complex when you factor in multiple units, but it is still determinable to a savvy player. What I''m trying to say is, the effectiveness of a unit is only partially innate. It is modified by many external factors and even on what kinds of units they will be facing (MBT''s really aren''t that good against infantry even though they cost a fortune). I hope that game designers will start seeing other means of balancing, or better yet, not worry about balancing that much at all I will shortly within the next week post up my game design ideas on my homepage that everyone is free to look at and use. It is incredibly spartan at the moment, but hopefully I will get some feedback.

Yeah, and CDs are an innefective way of storing data...

What I meant is that you should also explain implementation. This is like giving a blind hungry man food but not telling him it''s food.

You are using the word "formulaic" as if it is a bad thing. This is not necessarily so. While it is a major no-no to my mind when it rears its head in the story line, I see it as a good thing in the strategic side of things. Players need to know how to counter their opponents. Innovation is great. It is really important...in moderation. If you make the learning curve for your game too steep no one will play.

In your example, by the time I know my enemy consists of "two Orc war wagons with 5 wolf riders and 10 Uruk Splinter Throwers", it should be too late for me to build something new to counter it. I will have to fight it with what is already on hand. It MIGHT give me an idea of what he is building so I can counter future waves.

With the complicating factors you cited, none of them change the initial usefulness of the RPS approach. RPS is determined from a tactically neutral stance...or rather without tactics in mind. Further testing in tactical situations will result in tweaks to AI, damage, Hit Points, whatever.

I could see, however, making a unit that sucks by itself, but if you have it in numbers it is unstoppable. Deviating from the RPS road might work for a few specific instances, but for the bulk of the game I don''t think you will find anything else that has balancing issues built in.

Look at MechWarrior 4. I know it isn''t RTS, but the weapons in the game are a great example of balancing done WRONG. Before the expansion came along, you saw exactly 4 weapon systems in multiplayer. Do you want to spend your blood-sweat-tears making a game only to find players are using 1/10 of your effort?

ShadeStorm, the Day_Glo Fish

You are talking yourself in circles here.

First you say RPS is flawed, then you suggest that unit mixing breaks it. I suggest that unit mixing is exactly what makes RPS redeemed. The rest of the things you mention are features you can CHOOSE to add, and you would only add if they enhanced your gameplay. Why would you add them if they broke your core strategic model?

Of course, RPS is a simplification that isnt meant to be used in its "purest" form -- its an academic thing, merely used to explain by simple example the concept of counterability.

A unit counterability circle, or counterability matrix, with lots of interesting strengths and weaknesses to units, CAN be broken down to A beats B, but in an actual game, you are never in the situation (or almost never) where you can say "he has X Y and Z, so I will now build the exact counter". You have to hedge bets, because as the game moves on, you have to respond to what they are doing, in almost every strategy game. That includes changes to your opponents unit composition... so contingencies become a part.

Even with say 5 or 6 simple unit types, and a basic production model, timing alone adds a dimension along which this becomes extremely difficult to master even though the basic RPS system is easy to grasp. Any RTS uses RPS as a building block, and using complementary parts, ends up with a great game (well sometimes anywya -- depends on the complementary parts).

I balance by unit counterability, and it works great. In fact, the most useful tool I''ve found in balancing a game is to confirm that units that are supposed to beat other units do indeed work as expected. Sure you have to consider the whole game when balancing the whole game, but RPS is a great tool for players to think about the game with, and it generally will hold true if you make it so. Things that are easy to think about, are also easy to balance, since its easy to identify when things arent working properly.

I would suggest that too many extra effects, which you can choose to add (the ones you mention, actually muddle the gameplay and make it harder to undersatnd, and thus, strategize about. Formulaic or not, it works, it produces great games, and the burden of evidence is against you

Zileas

My criticism is this: you speak as if the methodology does not work in games, but you back this up by referring to real life. I therefore think your reasoning is flawed, as games do not model real life closely enough.

Even though I agree with your hypothesis.

The Command and Conquer series is one example where the developers have constantly strived to balance their games using the RPS technique. They create a cool unit and balance it by creating an accompanying nemesis on the other side. (I think ''nemesis unit'' should become standard RTS terminology, btw.)

While this sounds ok on the surface, what it leads to is a simple reactive game where your tactics are reduced to choosing your units: "Build some of A to destroy X, some Bs to destroy their Ys and then C to take out the base." Boring.

So, as you implied, the answer is to move some of the elements that dictate victory or defeat out of the unit and into the hands of the player. Make a unit''s position, proximity to friends, altitude and so on affect the formulae. Force players to think about how they use the units rather than which units to use.

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

Rock-Paper-Scissorcs...
(hm, made that typo by mistake, but will leave it as it seems appropriate )

I''m probably the greatest opponent to balancing. To me, playing a game where units have been balanced with the RPS system really does make me feel like I''m playing Rock-Paper-Scissors with someone. No matter what decision I make, if I win or not depends on the choice my opponent makes.

Sure, RPS helps to ensure that at the start of the game both players have an equal chance at winning, based on purely the way they play the game. I personally feel that it simplifies the design by nature. Because if you do use the RPS system, there are many options that can not be included (because they might not fit into the RPS system).

The above listed factors ''morale, leadership, discipline, etc'' would throw off the precariously created balance too much.

What do the players want though? What does the game design require?

I think for a game like C&C, the RPS functions fine. Kylotan mentioned the nemesis unit and I agree that this is something that at least could/should be changed. The nemesis unit is the ultimate RPS unit. And game designers will be happy because it''s easy programming. But that''s just it, it''s the easiest way out. And I don''t like it

I really think RPS will work for specific games. Current RTS games which are still mainly about peon-pumping and massive assaults don''t require much more than the RPS system that''s being used. Sure, morale and leadership etc would be a welcome addition, but those factors just don''t seem to fit into these games.

The RTS games of the future though, will they be satisfied with RPS? Again, it completely depends on the game design. But, I think that RPS should be something of a last resort. It will not be the on the top of my list when I try to come up with a balancing tool.

quote: Formulaic or not, it works, it produces great games

I guess this is true for those that think those games are indeed great. I personally disagree, so for me it doesn''t work. I guess I''m on Dauntless'' side on this one. But, I also realize that IF my taste were just a little different (I have played certain RTS games that I slightly liked, like Warlords Battlecry) I would be on the opposite side of the fence, rooting for RPS.

It''s all about personal taste.
It''s all about the individual game design.

If RPS fits in a game design, use it.
If RPS doesn''t fit in a game design, don''t use it.
If you haven''t decided on a balancing system yet, keep RPS in the back of your mind as a last resort...

I thought it made C+C/Red Alert boring (although it was great compared to what was around at the time). Basically, whatever the opponent threw at me, I knew exactly what to build and what to do. It was just a game of reacting to what you saw.

I think more advanced factors would fit into these games if we took the RPS element out. RPS on its own is a barrier to installing other tactics, as these other tactics would have to be extremely good to be worth pursuing when you will just encounter the nemesis unit anyway. Take out nemesis units, and then winning a game will depend more on how you use the units rather than which ones you construct. (Of course, go too far the other way, and pretty much every unit appears to do exactly the same thing. See also: Total Annihilation.)

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

Of course, the first step towards a system without RPS gameplay is removing build-during-war gameplay (construct buildings, train troops).

That way you take away the ''I''ll build just what I need to counter what I see''.

Instead it becomes ''I have to move my forces around in a smart way so that I can counter enemy forces with units that are capable of dealing with them''.

I just completed C&C Gold (GDI) yesterday - and I disagree with the ''unit nemesis'' thing. NOD have better units - stealth buggies, SAM sites, etc. which don''t have matching nemesis units on the GDI side. However, playing as NOD is hard because the missions themselves are hard - for example, having to airlift 5 people at a time across a river with only one plane, and then to take out half a mapful with them.

Superior Human intelligence is meant to counter for the flaws in the computer AI. Which may be why it took me so long...

Superpig
- saving pigs from untimely fates
- sleeps in a ham-mock at www.thebinaryrefinery.cjb.net

First things first

Andrew, I think that while RPS is not a good system to model balancing I think that in essence, one shouldn''t worry about balncing that much. Too much emphasis is placed on unit design and the units themselves, and not enough is focused on the external factors that control units. For example, what''s better, a Tank out in the open vs. an AT infantry squad, or the same units but in a town. If you answered the tank in both roles, then you don''t know the power of infantry in towns or rough terrain. So my counter to RPS is, don''t worry about balancing and unit design so much. Instead, design units like an engineer would, not like a game designer would. In real life, everything has a trade off, whether it be cost or raw materials.

Let me see if I can put it another way. Most designers see balancing in the terms of Cost and Capability. I agree in the Cost factor, I disagree in the Capability factor. Units should be balanced only in the sense of how much it "costs" for a country to produce them. When you do this, everything else should naturally balance themselves out. When you try to factor in the capability of an object, i.e. rock beats scissors, paper beats rock, etc., you neglect a whole slew of other possibilities. As I made mention before, capabilities of a unit are highly variable on many factors...cost on the other hand is only slightly affected by certain factors.

Shade & Zileas
What you are suggesting here is that what makes strategy "strategy" is guessing beforehand what the other player will build. However, most games due to their Unit-centric design essentially become formulaic plans on what''s the best combination of units to build. Some players like specialized armies just for the heck of it, but these are the exception rather than the rule. What I dislike about formulaic tendencies is that they become something like this...

"when faced with X + Y units, my best counter is with A + B units" True, combinations of units are more tricky to figure out, but on the other hand, due to the lack of fine control with RTS games, effectively managing a combined arms group becomes daunting, and I think the advantages gained through a combination are lost due to the inability to effectively manage a Combined Arms unit.

while it may take some trial and error to learn the best combinations and permutations, eventually they will be discovered. What I like about chess is that any unit can defeat any other unit at any time depending entirely on maneuver. This is not true in Strategy games. You will never see a 10 "cost" unit beat a 100 "cost" unit. And this is my major gripe with formulaic strategies. There is very little sense of unknown other than the composition of your enemy''s forces, which can be taken care of with prudent intelligence. So RPS strategy games boil down to having as much resources as possible to create your desired army as quickly as possible