I have an issue with the "Skilled players can repel rushes" thinking. The psychology of it turns the entire game into a rush-or-be-rushed game.

For example, lets take starcraft. Player A picks Terrans, Player B picks Protoss. B immidiately starts building a rush force of Dragoons. Player A immidiately builds a rush force of Firebats. Firebats beat dragoons, player A has enough of a force left over to take out the pylons and probes, player A wins.

BUT, Lets change the situation around a bit. B starts and focuses his resources on Zealots, A still does Firebats. Zealots beat Firebats, B successfully rushes and wins.

AND AGAIN. B decided not to rush, and works on trying to get Reavers, but A builds a rush force of Firebats. B has no opposing Rush force and gets rushed.

The only situation where we have an actual strategy game is where both players decide not to make a rush force. But that won''t happen because both players fear getting rushed. Its the classic Libertarian Mutual Advantage argument. Both players would have to reveal their bases to the other to assure that Rush forces aren''t being made.

The way I thought of tackling the "rush" tactic is not to look at the units, look at the logistical support the player gets in order to make the rush.

Maybe a credit based system is the wrong way to look as the sole method of limiting construction and force size, more cash = more units/more structures, seems fundamentally flawed, and would a synonymous to the situation of making a Yugo 25 faster by adding a bigger engine, you''re trying to codge a fix together for something that appears to be already ''broken''.

Methods of approach to finding an alternative solution ''should'' be investigating by looking at the logistical draw backs of getting a large army into the battle zone(we will ignore rush tactics in real life for now ). Many problems spring to mind, being ammo/fuel/medical/hardware/water/food supplies to name just a few. How do these get into the battlefield? Answer.. a convoy/supply chain.

How this would affect the gathering of a rush team..

Troops need food which needs to be supplied, and this doesn''t come quickly. Armaments need to be driven into the battlefield, and wouldn''t ''appear'' from some building which at first also ''appeared'' out of the ground. So it all lies in the supply line, which of course, needs to be defended.

Also... since when does an army start without a pre-built base of some description? command center, comms tower, tent-barracks..

{any grammatical errors/typos can be ignored }

After reading other thoughts and having some time to think about it, I''ve decided that really, the problem with allowing rushes is that it chokes off the potential complexity of any strategy by making the time given to develop and position yourself before the first attack non-existent. Inmate''s Starcraft example shows that pretty obviously; if, to play for higher-level units is rendered a guessing game or mutual agreement over the "will the other guy rush" question, then the people who have no interest in innovating their strategies can consistently win, stagnating the lasting complexity of the game.

The best thing to do, I think, is to take a page from probably the best strategy game ever, Go. Both sides are forced to develop equally in Go, with few quick victories to be found; it is how they use their resources that count. Therefore, anyone''s strategy has plenty of room to let itself develop and prove its worth on an equal basis, and the crude shoving, eye-gouging, or backstabbing that rushing conjures images of cannot exist.

To build an RTS following the play of Go does not mean to eliminate resources or base-building, however; it should instead eliminate balancing economic development against military development. This turns the game away from the simplistic "strategy" of getting weak units now or strong ones later, opening the gates for lots of other concepts that involve using what you have. An exact implementation of a no-resource-juggling RTS could vary, but it''s certainly not impossible.

Making the world furry one post at a time

I will now defend:

The first rule of war is to know yourself and know your enemy. If you do not know what your opponent is doing, or you do not know how to use your own troops, then you can never hope to win. This is why I say,

"Only low-skilled players are effected by rush-tactics".

If you know what your opponent is doing (building an air/water/land rush), then you can counter that attack. Scouting out the enemy is a vital tactic, and costs you practically nothing (normally just a single peon). This is why games like starcraft and warcraft give you the ability to choose a Random race... so that you force your opponent to send a scout to figure out what kind of strategy to use against you.

Secondly, I never said that the game should have no defensive structures (ir2 that post about terran bunkers in Starcraft). Choosing to spend resources on defensive structures or a defensive force is all part of the strategy.

Thirdly, a rush-tactic may be part of a larger "economic denial" strategy (where you consistently attack your opponent to prevent them from effective expansion/teching), so disabling it or heavilly stacking the game against it doesn''t just stop one form of strategy, but two. If you deny rushes you may as well just give everyone full economy and tech trees from the start. And that may well be the most realistic way to set up a RTS, but... there''s a reason why we have to expand and tech in these games, it''s so people can develop critically timed attacks, yet another form of strategy.

And as for that comment about all games desolving into an attack or be attacked mess... If you want to build empires, then play a nice CIV game instead of a combat game. The point of the game is destroy your opponent, so ultimately the game has to come down to attack or be attacked. Whether you can do it successfully with low-tech units or later on with the sexy big beasties is really kind of unimportant.

Lastly, every attack is won or lost based on an imbalance in the two forces'' strengths. This can happen at any time in the game (because of better expansion techniques, well chosen teching, opportunistic attacks at points of weakness, whatever). If a player can get his/her build orders snapped down in the first few minutes of a game to get a single unit out faster than the opponent, then why deny them the ability to use it? And, if you can successfully defend a rush, you have two choices. You can either take advantage of the possibly weaker economy of the rusher and attack, or you can further your defenses and try and bed-down for the long, economic battle.

I see rush attacks as being very similar to the three move check-mate in chess. They''re cheap and nasty, they do the job, and ultimately they require the co-operation of your opponent. You don''t hear anyone saying, "How do I break the rules of chess so that the three move check-mate is not valid."

Last thing to go through a tortoise''s mind before it hits the ground at 40mph? "D3DXVec3Add (Tortoise.Pos, Tortoise.Delta, Tortoise.OldPos"

quote:

---

Original post by Inmate2993
I have an issue with the "Skilled players can repel rushes" thinking. The psychology of it turns the entire game into a rush-or-be-rushed game.

---

The flaw in your thinking is that skilled players tend to build varied forces, rather than spamming one particular unit.

Also, defensive forces tend to have a slight advantage, particularly since in the time it takes the agressor to find the enemy, the defender has had a bit of extra time to build units. This means that the 'rushing' player has to either be much better at the early stages of the game, or very lucky to succeed.

That's not to say you shouldn't be agressive, but mindless rushing tends not to work - at the very least you need to find a weak spot in his defense before sending your forces in.

Of course, none of this seems to deter those morons who like to zergling rush you, and then disconnect if their attack gets repelled.

[edited by - Sandman on May 6, 2003 6:03:43 AM]

quote:

---

Original post by SoaringTortoise
The first rule of war is to know yourself and know your enemy. If you do not know what your opponent is doing, or you do not know how to use your own troops, then you can never hope to win. This is why I say,

"Only low-skilled players are effected by rush-tactics".

If you know what your opponent is doing (building an air/water/land rush), then you can counter that attack. Scouting out the enemy is a vital tactic, and costs you practically nothing (normally just a single peon).

---

From what I''ve seen of most RTS games, you rarely get much of a chance to scout if you''re concentrating on not being beaten in the arms race, which often requires micromanagement of order queues and resource gathering units.

quote:

---

Secondly, I never said that the game should have no defensive structures (ir2 that post about terran bunkers in Starcraft). Choosing to spend resources on defensive structures or a defensive force is all part of the strategy.

---

My original point is mainly that such defensive structures should probably be cheaper, especially early in the game.

quote:

---

If you deny rushes you may as well just give everyone full economy and tech trees from the start.

---

I don''t agree with that, to be honest. I know different RTS games can take different lengths of time, but a tech-tree shouldn''t be so quick to traverse that eliminating rushes is equivalent to giving out the whole tree in advance. I know I''ve played Age of Empires games lasting well over an hour where none of us were near having the whole tree. Sure, we weren''t great players, but any real rush attack worthy of the term would have come long before we reached the final age.

quote:

---

And as for that comment about all games desolving into an attack or be attacked mess... If you want to build empires, then play a nice CIV game instead of a combat game. The point of the game is destroy your opponent, so ultimately the game has to come down to attack or be attacked. Whether you can do it successfully with low-tech units or later on with the sexy big beasties is really kind of unimportant.

---

If you have tech-trees, you have development and construction. That requires the time to be able to make some interesting choices. If the choices are forced upon you by a dominant strategy or a lack of time to respond, you lose much of the strategy element. "Attack or be attacked" is fine, but these are games of construction and economics, which means that they''re supposed to be longer-term.

quote:

---

Lastly, every attack is won or lost based on an imbalance in the two forces'' strengths. This can happen at any time in the game (because of better expansion techniques, well chosen teching, opportunistic attacks at points of weakness, whatever). If a player can get his/her build orders snapped down in the first few minutes of a game to get a single unit out faster than the opponent, then why deny them the ability to use it?

---

Because it''s rewarding faster mouse clicks over better strategy. Some people might like that, but it doesn''t allow "diverse and emergent strategy" as I initially called for.

quote:

---

I see rush attacks as being very similar to the three move check-mate in chess. They''re cheap and nasty, they do the job, and ultimately they require the co-operation of your opponent.

---

If you execute a Fool''s Mate on a chess player, they can defend against it from the next game, and without having to sacrifice a wide variety of potential strategies. You simply don''t move those 2 pawns that way. But with many RTS games, being the victim of a rush does not teach you how to defend against it since no part of your defeat was tied to a very specific mistake on your part. Additionally, to defend against a rush often limits your strategies to either out-rushing them or digging in. This is why a far higher proportion of players fall prey to RTS rush tactics than chess players who fall to 3 or 4 move checkmates.

Again, the truth of the matter here comes down to whether the game is sufficiently biased in favour of defence or not. In chess, you start off with a large defensive force which you then can deploy as attackers, but in RTS games you are usually trying to develop both at once, so a minor advantage is often amplified with little chance of recovery.

[ MSVC Fixes | STL Docs | SDL | Game AI | Sockets | C++ Faq Lite | Boost
Asking Questions | Organising code files | My stuff | Tiny XML | STLPort]

Rushing can be a viable part of RTS. It depends on the design of the game. If a player loses the technology battle because they defended against the rush, then the game design should be reworked to allow for both play styles to be effective.

Like another poster said, if scouting is a necessary way to find out what the other player is doing, then so be it. Perhaps you can tell from a quick look at their base what they are planning on doing, and build your defenses up in the time it takes for them to travel across the map to get to you (which hopefully is long enough of a distance for you to get your defenses up).

Ultimately, though, I think it comes down to game design. Rushing is a reality of online play. The best way to avoid it is to build in startup defenses in your initial buildings that will stop the quick rush. The opponent will have to build something bigger in order to get through it, which would give you time to build up your defenses or a proper counterattack. But that would also take a dimension away from the game, I think. Players enjoy starting with the bare minimum instead of being assisted. Maybe it could be a map option.

I''m not a rusher myself, and I hate rushers because I never really developed the desire to micromanage enough to stop them. But I understand it as a reality in RTS, and it''s up to the design of the game to keep it from breaking the game, and keep it from turning into a "play method 1 or method 2, don''t choose wrong or you lose" paradigm.

I see the begin game as one of a few possibilities:
1. Prepare a rush attack
2. Defend against the rush attack (sacrifice some initial tech buildup, and no rush for you!)
3. Go all technology (sacrifice defense against rush - risky!)
4. Middle ground between rush defense and technology buildup (survive rush but no counterattack until later)
5. Prepare a rush attack with some defense to somewhat counter their rush (big gamble - micromanage battles to win)

I respect designers that can come up with game designs that serve all these possibilities well without making one single tactic the best. StarCraft must have been a BEAR to balance between three such distinct races.

To answer the original question, one first has to define the term "rush".

Many people both in games and in this thread seem to have differing opinions on what constitutes a "rush".

Common usage indicates that a "rush" is simply an attack on one''s opponent(s) relatively early in the game. This might be a WC2 grunt rush, or a C&C Generals gatling tank rush, or a Kohan skel rush. But if we simply removed all early threats, isn''t the optimal strategy then simply to build one''s economy until such point as there may or may not be threats?

At this point one then has the same dilemma - there is always an earliest point at which the application of military force can end the game. So given this view of the term "rush", there is always an effective rush in any standard RTS. I consider that a reductio ad absurdum of holding this view simultaneously to holding the view that rushes are to be avoided.

The argument Kylotan seems to be making is that rushing reduces strategic diversity. But only dominant rush strategies produce this effect, and few games have these (perhaps the pre-patch SC 6-pool?). Note that removing the possibility of rushes also reduces strategic diversity.

From Kylotan :

quote:

---

From what I''ve seen of most RTS games, you rarely get much of a chance to scout if you''re concentrating on not being beaten in the arms race, which often requires micromanagement of order queues and resource gathering units.

---

Firstly, almost all top players in RTS games scout regularly, whether it''s with a zergling in SC (cheap cost), a wisp in WC3 (can produce resources from any tree on the map), or a short shade in Kohan (fast). Secondly, micromanagement is simply strategy in the small - a big factor in almost any RTS you might care to name. Effective "build orders" are widely available from any site, and are present either explicitly or implicitly in pretty much every strategy game where one has decisions on what to produce. The good ones are the culmination of much strategic thought, and good strategies are often copied.

You are of course at a significant disadvantage if you choose not to take into account the opinions of others in developing your own strategy - I typically take a nice build order and modify it to suit my purposes. Or do you expect to be able to build willy-nilly and have the same output as one who executes a carefully planned build order?

I tend to agree with you that faster mouse-clicks should not win out over better strategy. Fortunately that isn''t often the case. Fast decision-making is typically the deciding factor. On being rushed, a newbie is typically overwhelmed in seconds, still thinking about what they should do. Afterwards, they then have time to think about how to counter such a strategy, or get advice from other players on how to counter it, and on encountering it the next time will fare much better. And so on. In some games you need to be lightning fast. But in most RTS games you do not. At top level some measure of speed is required, but then of course it is an RTS game and so speed must eventually have some impact.

Most strategy games offer some advantage to the defender early on, enough to force strategy to get into more of the game (as you suggest). All RTS games that I know of have distance - you start some distance from your enemy and the time it takes to find/reach your enemy is a barrier to an effective rush. WC3 has various features - enabling workers to turn into half-decent fighters for a short time for the humans, allowing orc workers to jump into their "farms" and shoot out at attackers, and simply putting a strong barrier over most of the night elf workers (which additionally can detonate on an opponent). Kohan has town militia and outposts. C&C Generals has only towers, and is quite open - there is at least one very strong rush in that game (on smaller maps at least).

In short, I think most of your concerns have been met by the current crop of RTS games without removing the possibility of rushing. If you choose not to build any defence and instead put those resources into your economy you take a risk. And that''s just a strategic decision.

quote:

---

Original post by Argus
To answer the original question, one first has to define the term "rush".

Many people both in games and in this thread seem to have differing opinions on what constitutes a "rush".

---

I would define a ''rush'' as any strategy specifically geared towards attacking your opponent before he has had time to build a basic defense - essentially, while they are helpless.

Although from a purely strategic point of view, it is a perfectly valid strategy, it isn''t a terribly interesting one, nor is it a particularly fun one. Personally, I like my games to last slightly longer than 3 minutes, and I like them to involve more than memorizing a rush/anti-rush build order and repeating it over and over again, without ever playing long enough to get into the more interesting areas of the game.

Yet, as I mentioned before, some people seem to love this strategy so much, they''ll happily put all their eggs into one basket and just throw everything into a rush - if it succeeds, they claim victory, if it fails, they''ll disconnect rather than take the loss. This isn''t really fun for anyone, it''s just annoying. The only real way to prevent this sort of play is to make it clear from the start that an early rush will never succeed.

quote:

---

Original post by Argus
But if we simply removed all early threats, isn''t the optimal strategy then simply to build one''s economy until such point as there may or may not be threats?

---

That''s assuming "build one''s economy" is a single strategy. I would say that if it ever was considered a single strategy, then something about the game is sorely lacking. There should be so many different ways to develop your force and use them that you could ban all combat for the first 15 minutes and still have very different games occuring.

quote:

---

At this point one then has the same dilemma - there is always an earliest point at which the application of military force can end the game. So given this view of the term "rush", there is always an effective rush in any standard RTS.

---

There''s a continuum from one extreme to the other, and "rush" is an abstract term, so you can''t expect to get a precise definition.

quote:

---

The argument Kylotan seems to be making is that rushing reduces strategic diversity. But only dominant rush strategies produce this effect, and few games have these (perhaps the pre-patch SC 6-pool?).

---

My chess comparison earlier is the best way I can explain this. A rush is generally a tactic used by someone who is good at micromanagement and/or who knows the best ratio of resource gathering to unit production. Often it involves learning a predetermined build queue. In this sense, it''s no different from learning a quick-victory opening in Chess. The difference is, when you''re beaten by a rush, there often is no defence at all unless you know how your opponent achieved it. All you can do is go back to lesser-skilled opponents and experiment like crazy to try and work it out. Whereas in chess, you can easily block the attack. A skilled chess player has to win by drawing from a depth of strategies, whereas many RTS players win by being effectively able to micromanage one or two of them.

I don''t claim to refer to Starcraft since I never got into the game. (I found it pretty tedious, frankly.) However several other RTS games I''ve played, including the Age Of Empires ones that this thread was originally based on, do suffer from this, despite having great potential for a wide breadth of strategies. My original post was more of a theory than a criticism of the genre.

[ MSVC Fixes | STL Docs | SDL | Game AI | Sockets | C++ Faq Lite | Boost
Asking Questions | Organising code files | My stuff | Tiny XML | STLPort]