StarCraft

[Getcha cheats here]

StuartFraser, who is currently playing through the Zerg campaign just to get at the Protoss in the hope that they're better missions, heartily disagrees with this statement. Not only is the StarCraft story banal, annoying, fails to explain remotely how the player manages to change sides twice, and inferior to those in the CommandAndConquer series in every respect, but it also imposes the arbitrary limitation of forcing you to play the campaigns in "order". I don't particulalry want to since I hate the Zerg but am quite happy to play with Terran and (probably) Protoss forces, and strongly resent this. As stories go, Homeworld is obviously the best RTS for them; Blizzard have yet to demonstrate any more talent for originality in their writing than they have in their games.

AlexChurchill doesn't disagree with most of what's said here.  Obviously "banal and annoying" is subjective.  The changing sides is a very odd objection... clearly no character in the story is changing sides, it's just that the point of view of the story shifts.  Surely it's fair enough to see a story at different times from the POVs of a Terran adjucant, a Zerg cerebrate and a Protoss executor?  I don't think any of the chapters other than ch5 (Brood War Terran) actually name the character who you are actually playing.  And if they're wanting to tell one coherent story, then surely they needed to enfore an order of the three sections?  You can't tell a story's end, then its middle, then its start, and keep the audience's attention [*].  There are plot developments in the Zerg storyline that will let the Protoss one, um, actually make sense.

[*]: with the exception of the rare stories that are crafted to be deliberately told end, middle, beginning.  Even then, the suspense is only relieved once the beginning is discovered - in particular, you can't let people choose the segments in any order!

There are plot developments in the Zerg storyline? That *weren't* obviously implied during the Terran one? I don't think the existence of Dark Templars and their ability to kill Zerg and have them stay dead couldn't have stood alone in the Protoss storyline, if actually written properly.

Just to be picky, I'll point out that truly non-linear stories can be crafted.  Doing so is difficult, of course.  EtherLords (2) has a particularly fine example of how to do it technically well.  The story itself is banal and badly told, though that may be mainly due to the "Your kung fu is weak!  Fight me!" nature of the game.  Humm.  Thinking about it, the major trick to such a non-linear game is that every action the player takes is cleaning up after a mess.  Then later sections can happily show you how that side gets into that mess, and the two things work in either order, which they would not do if you had any way to prevent the mess in the first place.  --Vitenka (Uh, we now return you to your regularly scheduled conversation about StarCraft - oh, while we're at it, if you really hate the missions, protoss1 at the chat prompt should get you out of there)

I have, without cheating, got as far as Zerg 10, due to the nonchallenging nature of the majority of Zerg missions probably more than anything else.

What is the RTS startegy named?  Zerg rush.  Can we think why?  --Vitenka ;)

Since the final Zerg mission seems to potentially be hard, I may have to skip it, thanks for the cheat. And yeah, I was moderately impressed by the non-linearity of the Etherlords plot; Dune 2000 tries a similar thing but doesn't quite pull it off, due to the divergent nature of the final missions.

Once again, I'd stress that I don't think the Blizzard storylines are anything special.  And all the way thru the WarCraft 3 one I was laughing at the hugely over-hammed-up dialogue and voice acting.  But I think with some of the criticisms you make, they couldn't really do anything else withing the constraints of a three-race RTS game?  --AlexChurchill

I would be perfectly happy for all three storylines to be concurrent rather than consecutive. I'm about to reify another page which will go some way to explaining how I think non-linear RTS plot should be written, in the context of an entirely different game...so, I may see you over there. --SF

Pah.  I gave that as an example which sprang to mind having frequently used it myself.  But every key makes perfect sense - "W" is for Wraith.  If you wanted "B"attlecruisers, "T"anks, "M"arines, the keys are similarly obvious.  I happen to choose to assign my factories to key 5 and 6, but that's completely customisable.  All the hotkeys are instantly visible and equivalent to clicking appropriate buttons - it's just that it's faster and less distracting to tap a few keys.  For the same reason that giving keyboard shortcuts to menu items in a WordProcessor? is a sign of a good UserInterface?, providing hotkeys to buttons in RTS games is also a sign of a good UserInterface?.  Surely it's better to be able to press "www" rather than type "wraithwraithwraith", or "buildmeawraithplease<ENTER>buildmeawraithplease<ENTER>buildmeawraithplease<ENTER>" ??  --AlexChurchill

As for hotkeys - I don't see 5wwwbtwww6wwwbtww etc. etc. as actually, you know, 'short'.  Ok, it's better than choosing from all those menus - but it's still horiffic.  I'd suggest some sort of scripting language as being nice, but then the initial learning curve would be event worse.

Again I say, Pah. :)  That's an example of something that'll only work late-game anyway, as you have to be able to afford 10 Wraiths for that keysequence to work.  The more normal version is, "Tank factory, build me another tank or two."  Which translates, under the way I usually assign my hotkeys, to "6TT".  I don't see how that could be much shorter.

Scripting language would be a gooood idea.  You can't argue usability issues against it if the scripting language is optional, as then the initial learning curve doesn't have to feature it until the player's ready.  You can't coddle people forever - and hopefully, if there's any point to an RTS game being RTS and not TBS, they'll have to learn to think fast and play fast at some point.  --AlexChurchill

Well, if a player using the scripting language always beats one that doens't, then it would be only optional to players willin to lose.  And I don't doubt that would happen.  The ability to think fast and react quickly is good - the fact that that reaction has to include wrestling with the interface isn't.  Scripting lets you get a little bit closer to 'do what I mean - right now' but probably not close enough.  And of course it lets people fire off 'win the game, strategy 2' scripts...  Turn based is a whole different argument.  Basically for me it devolves into micromanagement and takes so long that you forget what your initial plan was.  That's definitely just me though - I hate long term strategy and prefer tactics (which rts cater for better, sorta)

Sorry, Vitenka, just to clarify: do you still view "5WWWWW6WWWWW" as a translation of "factory 5 and 6 each build 5 wraiths" as an example of a UserInterfaceFromHell?  Given my defence above?  It's not clear from your latest response whether you retract that or not.

Yes, it's hellish.  Of course, clicking on the factory and then each tank in turn is worse...

How on earth would you prefer to order each of factories 5 and 6 to produce 5 wraiths, then?? :)  Given that you might also want to tell them to produce science vessels, battlecruisers, or some mixture of the three?

Something telepathic perhaps?  Or maybe just hit the 'tanks' key then a number?

"Only optional for players willing to lose" - I disagree.  I'm sure if any of us were to play the world champions at our RTS game of choice we'd lose horribly.  Does that mean there are things we've still to learn about the game?  Probably.  Does that mean we're bad players?  Not necessarily.

Does it mean that we could beat a good RTS player when he is using an RTS interface and we are using a TBS interface?  Almost certainly.  Not certainly (I'd never beat a grandmaster at chess, no matter how short his timeclock) but certainly with a great advantage.

Sorry, sorry, typo.  I meant RTS.  Delete this comment and reply to what I meant to say. --AlexChurchill

Uh - same answer still applies.  An RTS player, handicapped by the real-time part, vs someone with all the time in the world (as provided by an insanely impossible but never mind that for now) set of scripts should lose.  Otherwise the 'realtime' bit isn't a handicap and most of the 'skill' argument goes away.

"Does that mean there's something "compulsory" which we haven't learnt?  I doubt it.  It could be near the end of things which the hypothetical-new-player learnt.  They'd still be able to beat medium AI opponents, and some human opponents, without using it.  I have no doubt that there'd be a minority who stubbornly refused to use the thing at all - just like there are some who refuse to use keyboard while playing StarCraft at all, playing it all with the mouse and still doing quite well.

A "I win the game strategy 2" should be utterly impossible in any game worth its salt.  Not only is there no unbeatable top-level strategy, but if the script can adapt fluidly to opponents' actions as well as a human can, that's no script - that's an impressive AI. :)

But, I would argue, most of these games don't really warrant a top AI - they only have decent game programming to make the computer opponent feel human.  To just win, by - for example, flat out producing unit type 3 and swamping the enemy with them every 14 seconds - or any of the other strategies which a vast array of TypingMonkeys peck out on a daily basis - seems much easier.  Clearly you wouldn't be able to progress into the top echelons of gaming (although having a vast array of 'top strategies' at your fingertips would help you get there) but you could flatten a sufficient number of players to kill off the game (or force them to use the same scripts - and the number of players capable of coming up with good ideas for scripts is limited)  FPS? games definitely have this problem.  The accusation of 'no skill bot' runs rife..

Well, sure. But my point is, in StarCraft any such single-minded strategy can be defended against, and quite easily.  I played against a few people who used a Marine rush in a row once... by the third, I was quite happily slaughtering them as they approached.  As it happened, he didn't have much other strategies planned, so I crushed him with aircraft.  Now as I understand it, you say that if a new player comes up against endless players who're X, then they won't want to play any more.  Could you clarify which of these is X: "hugely better than them", "all using the same simple-minded strategy", ... or what?  In the first case, well, the Net games often say "New players only!" for this kind of reason.  In every gaming field, people gravitate to those who aren't too far from their own level.  In the second case, I assert that in StarCraft, it's easy enough to defend against any one strategy that this doesn't become a problem.  If someone has an array of six strategies which they switch smoothly between based on their opponent's actions, then sure they're going to be hard to beat; but I'd say such a player is officially Good. :) 

Just to be clear, I'm asserting that StarCraft doesn't suffer from the "produce unit 3 and attack every 14 seconds" affliction, because (I assert) any one strategy can be defeated, and ones as simple as that can be defeated quite simply.  --AlexChurchill

Um.  From what I've seen, the opponents both set up their strategies (almost in isolation, although taking some time to find out what your opponent is up to is a strategy in itself) and then let rip and see who wins.  Now, admittedly, I'm not talking top level competitive play here - but if that level of play is 'good enough' then I'll never get to SEE top level play. And most people would never even figure out it existed.

I simply must disagree, I've played several hundred games on battle.net and my experience is that if the opposition fails to adapt their strategy to what I'm doing they will lose, and visa versa.  There are only two exceptions I can think of: 1) A player has already lost due to earlier gross mistakes, it doesn't matter that the strategy the other player is using is suboptimal, they have 20 times the resources and production, the other player is going to lose. 2) One player is insanely good and running a near flawless tank rush (by which I mean a combination of vultures, mines, tanks, turrets, engineering bay scouting, etc. working in perfect unison, with production running full steam, before you're 6 minutes into the game.  I've tried it and I just don't have the skill to pull it off.  Just tanks loses), the other player is not damn near perfect, but that's a case of 'I'm sufficently good not to have to adapt to you, because I plan on winning before whatever you're doing has time to come into effect.  I think it's also worth noting that I've gotten lazy in the past and stopped using keyboard shortcuts leaving slower than my opponent.  I've done quite well in such conditions: they may have been doing things faster but I was doing the right thing. --NR

Do you know if any games writers have considered hotkeying the hotkeys? Cos then you could just type 7 and have it enter '5WWWWW6WWWWW' or something. Would be quite handy and very flexible. - CorkScrew

I'm not atall convinced as to its handiness, how often do you want to build 10 wraiths?  How often do you want to build *more* than that?  Hotkeying the hotkeys sounds cool (it is indeed cool), but I suspect that it's too lacking in quick flexibility to be useful.  That said I have found myself with Zerg wanting a "build hydras/mutas everywhere!", as that is the race that wants to swarm.  In fact since it would be useful to Zerg and I don't like Zerg let's not have it.  Cute concept tho.  (All this said I will admit that SC isn't perfect, for example why does the disruption web targetting track flying units?  Like my own interceptors for example ... grr!).  That said I realise that what you're propsing is a neat solution to what's been raised.  Perhaps what I'm more disagreeing with is that you end up in situations where you want that kind of block construction, perhaps Alex is used to rather different games than me. --NR

Wow, that was an incredibly involved, well-thought-out response to my behavious pattern of "bored now, let's come up with a dumb idea for the wiki" - CorkScrew :)

The RTS vs TBS discussion should be moved to one of those two pages, I think.  --AlexChurchill

Well, obviously not QUITE so simple - but damn near.  The mileage that even a relatively unskilled player can get out of following a script (which other people worked out) as quickly as possible is sufficient to prevent a new payer from ever learning any nuance of the game.  It really is a classic meta-game shutout, and every RTS seems to suffer from it.  I believe the particular on I am whining about this instant is the 'x-rush' (tank-rush, blue-blobby-piglet-rush etc.)  The fact that 'go faster than your opponent' is a basic winning strategy in ANY conflict is a pretty big reason for this - especially when modified by arbitary game limits.  (Such as the aforementioned and much hated 'only 99 units') - there is no disadvantage to having the maximum number of units, and possibly a large risk if you don't - so you pick which unit to develop to (low enough that you get there before your opponent, as high as possible otherwise) - build 99 of them and flood the enemy (who you hope hasn't also got to 99)  You'd think computers would be fast enough to remove such arbitary limits by now.

The trouble with Rushing is that it loses to Defending (which, in turn loses to TechTreeClimbing?, which loses to Rushing).  A successful strategy is a mixture, of course.  This is not to say that I don't agree on the 99 unit limit (200-foo limit in StarCraft, but... shrugs).  You have to remember in that context, of course, that StarCraft was designed for P120s...

Also remember the 200-unit limit is arbitrary.  StarCraft in fact includes a cheat to allow you to bypass it, and achieve numbers of units limited only by memory and processing power.  This cheat is made public on [Blizzard's site].  But the limit is included by default for multiplayer games, to prevent anyone accumulating /excessive/ amounts of units.

Side swipe - if the limit could be removed but is not, then the limit stands.  Yes, it's better than WarCraft - but not much.

While initially set against the limit I think it does actually improve the game (it used to be slightly too restrictive, they've since tweaked the top units down from 8 to 6 since when it's never bothered me).  Why?  Because if the game is going well you never actually hit that limit: 200 is more than enough to deploy your strategy and either your stuff is getting blown up, or you've won before hitting 200.  If the game isn't going well: i.e. both players are just staying clear of one another, not engaging, just building then the unit cap forces yo to say "well I can't build any more, guess I might as well go and try an attack" so it gets otherwise stand off games moving.  (We could observe that it also stops Zerg just winning insane resource maps except that 1) insane resource maps are bad and wrong anyway 2) Zerg does just win these anyway). --NR

I always found Protoss horribly effective on insane-resource maps; getting 12 stargates up and sending waves of carriers against the foe Just Worked.  This may, however, have been because we were all on dial-up.  The swarms of interceptors confused the automatic targetting horribly, and the resulting slideshow kept the human from intervening.  I'm not sure how good it would be on a decent connection. --NT, who hasn't played since sixth form

I would certainly say that Terran were the least effective because their units take more micromanagement so don't scale as well, but on insane resource maps it's possible to build so many cannons/colonies/turrets that they eat all the interceptors from a fleet of carriers before they can significantly damage them.  A sound Zerg player can get up an absurd number of colonies (each giving him access to yet more resources) and then produce 150 hydras in a shot, they reach you with the knowledge that if they all die then another wave of the same size will be along in 30 seconds.  Protoss just don't have the build speed to compete - each carrier just takes too long to make. (I will conceed that I haven't played silly resource for a very long time because it annoys me). --NR

And Rushing, like any other strategy I've seen in StarCraft, can be fought successfully - as Senji says, often by Defending.  That's one of the things that gives StarCraft real staying power as a multiplayer game - the fact that any strategy has an effective counterstrategy.  --AlexChurchill

Have you ever faced the tank rush, and how would you beat it? --NR

Dark Templars and Scouts --Angoel

Ah, I think you misunderstand me.  By tank rush I mean the strategy which uses tanks to do the damange but primarily consists of vultures, mines and missile turrets.  Scouts get slaughtered by the turrets, and mines track DTs.  do you mean send your scouts and DTs in to their base in the hope of doing damage faster than they can? --NR

I probably am misunderstanding, yes.  How exactly does this work as a rush tactic?  Aren't turrets and mines inherently immobile?  --Angoel

They run vultures up to the edge (well, nearly) of your base, deploy a cache of mines so that you can't get out to the SCV presently building a turret.  As it finsihes a tank or two deploys next to it, and an engineering bay floats forward to expose the view.  Your base is shelled, the mines and turrets continue to be deployed working forward into your base.  It's not a rush in the usual sense of 'charge in', it is an early attack, but I rather slow attack.  It's fiendishly hard to conduct, but I've yet to find a truely effective non-Terran defence.  (To be fair I use Zerg very little, so almost all that I've tried is Protoss based). - NR

Dark Templars can get to the SCV provided they're built before turret's finished, and once they deal with the SCV, they can probably deal with the rest of it.  Guardians can deal with everything since they can attack from outside turret range (although you probably can't get them early enough to solve the problem).  A spotter and units with strong ranged attack can take out the mines before they explode.  Indeed, you could probably take out the mines by sending zerglings, one at a time, into the minefield, before attacking in force with your hyrdolisks.  That said, I haven't played any multiplayer starcraft in years (and when I did I was hopeless), so all of these probably fail for one reason or another.  --Angoel

My problem isn't that rushing is an impossibly effective tactic (although in many RTS it does seem to be overly effective) - it is that it is sufficiently effective to prevent a new player from entering and enjoying the game.

Hmmm... how is this hypothetical new player learning?  From a friend?  (If they're actually a friend, they'll explain or show this kind of thing to their acolyte.)  By playing online?  I don't know how they're meant to become a good player except by either someone showing them, or playing the thing on their own enough to learn.  And certainly the StarCraft AI teaches people to defend against rushes.  Even if it's not quite the strength that utterly single-minded human opponents can use, they'll be familiar after playing the AI with the kind of defence concepts they'll need.  --AlexChurchill

Agree with learning whilst sat next to an experienced player, although such teaching is likely to be of a poor quality (most players of anything tend to forget what it was like to not know things, and to base explanations upon too high level constructs.  but that's a quality of teaching thing, not intrinsic to the game)  By playing online, was, of course, what I meant.  I disagree that concepts learnt against the AI teach you the concepts you need.  At best it teaches "So, I need to get x billion amount of defence up, within y nanoseconds?  How is that possible?" Follwed by either a speed race (hardly desirable) or disillusion and one less player in the game.  Working out what to do is supposed to be the interesting and fun part of these games, not convincing the computer to do it quickly enough.  I'm not even talking about expert level play - in general expert players tend to be nice enough to explain things - the problem is moderate level players, who are effective only because they can follow someone elses script (with or without using macro programs) and who tend to be, on the whole, catasses.  And the game encourages them because 'not too clueful but rapid formulaic play' wins often enough against a large enough section of the population of players.  This is, to be fair, a problem with the genre, not with this particular game.

I agree it's a problem.  But I think it's rather wider than just this genre... most beat-em-ups and FirstPersonShooters? also suffer from the same affliction!  Are there any RTS which don't??  --AlexChurchill

Well, no.  But damnit, there should be.  The closer a game gets to this 'no-meta' ideal, the happier I shall be.  I'd pimp my columns here, but I can't figure out wiki enough ;)

I think you misunderstand where meta arises from.  It is a property of games with hidden information.  Whenever there is hidden information there comes the question of 'what is my opponent likely to be doing?' and hence 'how can I counter it', especially if the game has the 'every strategy has a counter strategy', which we hope it does.  Now you can avoid a meta game by going to games with no hidden information, like Chess or Go.  But aside from that you will have a meta game. - NR

Ahh... now this I agree with you on.  I think.  Obviously Go is a much better game for encouraging genuinely thoughtful play, and not just following formulae :)

Well, obviously.  I have a whole different set of problems with go (which can be breifly summed up as "I suck at it")  Let's try that column link then:  [This one springs to mind]