The idea of counters and how they work (now about armor)

I'll make a post on the topic of unit balance then.

I think that defining units by their roles rather than hard counters is a good idea. In supcom you basically had three roles: antiland, antiair, antinaval. With longer range usually winning out. In TA you had a complete mess of basically random units that while more organic didn't really know their lot in life and there was an insane amount of role overlap where the benefit to one slightly different unit over another wasn't always clear.

Neither is a good system, one is boring the other is just plain confusing.

I think a compromise is best. You stick with the basic roles of supcom, providing a solid base of counters to build on. Essentially you get 3 categories with 3 basic units. naval aa, naval ship killer, naval shore bombardment, bombers, fighters, torpedo bombers, tanks, aa tanks and well... supcom didn't really ever fill the role of land based antiship.

From there you can branch out with more generalized roles. Say a laser or machinegun based weapon would be able to hit both ground targets and air targets easily. And so you can distinguish between flexible units and specialized units. And establish a balance where specialized units are more effective for the cost at what they do, but also harder to manage and more vulnerable to things they aren't specialized against.

So say a more general purpose land vehicle would hold its own against a more general purpose aircraft but be destroyed by a dedicated bomber, which in turn would be destroyed by dedicated antiair and so on. It gives you the basic RPS setup but branches out in a logical manner and gives you a lot of flexibility in what units to choose without being confusing.

In turn you can distinguish between AOE, precision, rapid fire, slow fire, fast movement, slow movement, special types of movement such as amphibious or fixed or rotary wing or whatnot. Once all these categories of movement, specialization and fire rate vs firepower has been clearly established you can start to create a huge amount of varied units with specialized and nonspecialized roles that have no particular role overlap and are all useful in a logical and easy to understand way.

At the end of the day, this fixation on unit counters and composition is killing the RTS genre. Bad RTS games are resulting in decreased sales, which make publishers think RTS games are bad investments, which results in bad RTS games. But so many players have no understanding that anything else is possible, and they actively ask for unit counters and composition-centric gameplay.

Chess gameplay is not about "counters." A knight doesn't counter a bishop, and a rook doesn't counter a queen. They play differently, have different uses, and different mechanics, but you don't need to use unit A to counter unit B.

Go does an even better job of illustrating how positional play and emergent behavior result from simple rules and interesting interaction, even though there is literally only a single "unit" which is 'a stone.' These games have yet to be surpassed for elegance, and for depth of strategy while being simple.

Fixating on unit counters is just so incredibly boring. If they have unit A, you make unit B. It's just dull, it doesn't require much learning, strategy, or even thought, and once you've figured it out, it's just boring.

This.

Agree. Just wanna add, that in both chess and go there are counter strategies. So should be in RTS games instead of direct unit hardcounters.

Most RTS games are about combined arms encounters though. You can't simply ignore unit composition because you think a far more simplistic board game has depth. Even go can't possibly compete with the potential depth of a fully simulated RTS. Which means it's all the more important to fully define the role of each unit in gameplay. And some units are going to be better at some things than others. That's just how it is.

The point is that an RTS game should be able to capture the positional aspects of (for example) Go, but also have a variety of pieces and a complex simulation. It's a more complex game, which makes it considerably easier to have a great deal of depth.

Go is an incredibly deep game, and functionally has two rules; score 1 point for surrounding a piece, and score 1 point for encircling each intersection on the board. (with some minor rules like black goes first, etc.) If you are allowed to have a huge diversity of pieces with different properties, and are allowed to make many moves simultaneously, and so on for all the reasons why RTS is more complicated, it should make the game even more intricate.

Could you make an interesting RTS game with just one unit type? Why not include that unit in PA, such that a game using only that unit by both players would be interesting to play? And make all the other units add still more options and depth.

I don't think PA should be made more simplistic like a board game. I think that PA should strive to be as elegant as Chess or Go (probably quixotic), and aim for a much higher complexity threshold.

Bmb you are thinking of a unit's role in terms of what units it is strong against, and what units it is weak against. That is a boring system. Look at it in terms of real military units- what is an A-10 Thunderbolt II "Warthog" close air support plane? Does it "counter" specific units? It is a tank killer, after all. Or is it more productive to think of the military as a set of tasks which need to be performed, and certain units have roles based on the types of tasks they can perform?

If you looked at units purely as "counter" relationships, you might erroneously conclude that units as different as the M1A1 Abrams, the A-10 Warthog, and the M47 Dragon ATGM are redundant. They kill tanks- it's what they do. But one is a battle tank, one is a close air support bird, and one is an infantry guided missile launcher. You make a grave mistake if you think these units behave similarly because they are effective against similar units. Their roles are extremely different, and indeed it would be inaccurate to say they even fight tanks in the same way.

And you're not thinking functionally. If you have a bunch of tanks you need killing, then you're going to pick the best unit for the job. If the best unit for the job is hard to find then you have a problem.

In terms of unit roles you must think in terms of what they do for the player and which units overlap with that function. If you have a unit that does a lot of functions well then that unit will be OP. Clearly categorizing what functions a unit does well and does not do well helps avoid such simplistic balance issues.

In this case a warthog is fast, and is invulnerable to its target, but is very vulnerable against AA. It's good for picking off stragglers and lone groups of tanks at range. An abrams is more of a direct fire unit that is effective at short range, but is also vulnerable to other tanks, but not vulnerable to their antiair or other general purpose units. These are the functional values a player will appreciate whether you design for them or not.

And if you don't design for them how can you balance them?

For example they weren't thinking in these terms when they designed the missile launchers, which have long range, high damage and almost perfect precision, making it almost universally useful against any ground target to the point where wasting your resources on other units is a bad idea.

If Go is able to match the strategic depth of all of these other more complex games.... isn't all that design work wasted?

Go isn't an RTS.

That's nice.

There might be several units suited for the same job but only one is the best counter in 1 situations. That is good. It means there is depth:

Which is the best antitank unit? That depends on the situation. Good thing.

Even if a unit is does alot of things there might be other units that can do some of the things better.
This might cause redundancy rather than imbalances.

What missile launchers? In what game?

Counters are kinda lame. I would rather have some units that just do different things and then have them fight.. granted some will counter others better but i don't want to see RPS play

Counters are inevitable. They will exist in any game no matter what. Whether they are intricate/situational or explicit/obvious is a matter of how creative you can get.

The more clever things get, the more difficult it is to keep track of everything. However, the potential for long term interest and metagame evolution is much higher.

Its true, if there were NO counters there would just be one unit that people spam.

its good that some units are better than others and there should be a theoretical maximum betterness (with some exceptions where math breaks like bombers vs tanks). I'd call this a hardness scale, with my personal favourite being around 20-30% (a unit good against another unit is about 30% better after whatever caused the difference). Thats just a number i plucked out of my head though. 100% would be units can't hurt their counters which is silly and 0% would be the all units are equal, which is also silly.

Tanks tend to be better against the lower armoured maneuverable types in straight battle, but they are generally good against all ground targets, even if they loose out to artillery or sniper type units. The concept here is that while a unit might be best against one thing they still have an advantage over some others, are equal to a few more and lose to things that are best against them. The more a unit specialises the better it is against its best opponent (say low damage but fast guided missiles for anti fighter use) the worse it tends to be against other areas (comparatively low health and ground DPS, despite near perfect hit rate).

I think that's a good insight yogurt. Interesting.

Flexibility is a quality too. You might need units that can deal with a large number of unexpected situations. They will not be as good but they will have far better survivability in variable situations than specialized units.
This is what I'm advocating, the more generalized a unit is the less effective it should be against any particular thing. A unit that can reliably attack all three kinds of units land sea and air should never be as good as a unit that can only attack one. This makes sure you don't actually have redundancy because for a dedicated task the dedicated unit will be the best.

Likewise for the roles other than just AA/tank/shipkiller, AOE specialized units should be the best for dealing AOE damage. Fast units should be the most useful for hit and run attacks etc. etc.

Sorry but that is stupid imo, why should there be a "best" unit, why shouldnt there be some relativly equal methods to do the job. Thats what decision is about. Else its not a decision but an memorizing game of counters.

I for this one am on ledarsis side.

^this

BMB gets it.

Units can be comprised of any random set of attributes. Without understanding how those attributes work together to create a unit's value and power, you have total anarchy in unit design.

Strengths and weaknesses do not exist on their own. They are derived by comparing against OTHER units. A unit with 50 speed is meaningless, until you also state that another unit moves at 20 speed. Then it becomes obvious the fast unit has an advantage over the the slow one, thus giving it more power.

You could not be more wrong. The game is boring if there is a single "best" unit for anything. There should be multiple interestingly different options, and the player needs to select between them looking at their current strategic situation. There should be tension between these decisions; resources spent on one are unavailable for another, and investment in one may lock the player into a certain type of action later.

If unit B counters unit A, and the enemy is making A, then by your logic if B is the "best" unit because it is the hardest counter, it is what I and any other rational player must build in order to maximize the probability of winning. Such a small, trivial, boring schema for what is supposed to be a strategy game.

If the enemy has a tank army, you should have various possible responses. Perhaps you think you can manage to take air superiority with a bit of investment in SAM's or fighters, and that consequently using the close air support tank hunter A-10 is optimal in your situation. Maybe the enemy has too much AA for that, or you have no airbases in range, and need an alternative. You could go for a tank army of your own, but you may lack the industry to produce it in the local area quickly enough, or your transports are occupied doom-dropping an enemy base, or whatever other reason. There should be a thought process behind your choices in this regard that looks at the current game situation.

What you propose is to always have the same single response due to the design of the game, which is incredibly dull. And can easily result in a broken game if the enemy finds an efficient way to do X and pre-emptively prepare for X's one counter.


Relativity of stats, bobucles, is being sophomoric. Obviously stats are relative. 100 damage is meaningless unless we know the approximate value of HP because they are defined relative to each other. It's a silly tautological argument from definitions, and proves nothing synthetically.

And half the time, take out your own ground units whilst attempting it :lol: