Some say men think about girls every five seconds. For me, this is also a good approximation of the rate I think about Starcraft. So when I re-read the book "Ender's Game" last week, I couldn't help but relate the story to our beloved RTS game.
If you're one of the two people who has never read or heard of "Ender's Game" (read it), it's about a boy military genius, Ender, and his struggle to win simulated battles. Even when the odds were unfavorable, Ender beat his fellow trainees through adaptability and creativity... elements I say is lacking in modern Starcraft.
"But that's blasphemous! Haven't you even heard of Boxer!? OMFG BBQ!"
Let me explain: Clearly, early Starcraft was very different. Players like Boxer used new strategies to catch their opponents off guard. But as expected with all games with static rules, the obvious strategies were soon figured out, and overtime, new ideas appeared less frequently.
To be sure, the game is still evolving (i.e. mech-build vs Zerg), but these new ideas appear very, very infrequently and once the dust settles, the new quickly becomes the mechanical. Forge fast expand, for example, wasn’t always the norm for PvZ.
The result? Modern games follows a repetitive pattern of player A doing an opening and player B reacting by choosing a known counter (I see you 9 pool, so I’ll put down 2 canons before nexus). So rather than a game of "who can out think the other", the weight has shifted to "who can click the fastest".
"So what? Clicking requires skill, and even if the strategies are repetitive, it’s fun to watch players macro and micro to those set patterns!”
But ask yourself this, why do so many people still like Boxer these many years after his prime? And why are so many others from the old days never heard of again? Simply, Boxer did the ridiculous, and his creativity is exciting to watch, and we prefer innovative players over the mechanical ones. It’s the players who innovate – Savior (use of swarm), Bisu (Bisu build), Flash (Dual Armory build), Fantasy (Fantasy build, mech build), July Zerg (Mutalisk Stacking), Jaedong (2-hatch muta, queen usage) etc… we remember and like to watch.
But modern players face a serious problem when it comes to creative play. The game has been so played out, that viable strategies are already known by everyone. Even Battlenet noobs like myself know the basic openings for each race. But who can blame the players? Repetitive strategies are repetitive because they work. Deviate from known counters, and you’re sure to lose. So in a sense, the players have become slaves to overused, over-studied strategies. Strategies which, I say, ultimately have their roots in the maps.
The map is a huge influence, if not the sole culprit in shaping strategies. To give an extreme example, Zergling rush works EVERYTIME on the old map Blood Bath. And in modern games, some maps heavily favor certain strategies. For example, we only see ZvZ on Battle Royal. No team would send P or T players because Battle Royal favors Zerg players so much that no P or T strategies works. Besides these extreme examples, modern maps all have very similar layouts: A main with a ramp, a natural expo or two, and the size of the maps are also very alike. So expectedly, strategies remain largely the same from map to map.
Maps affecting strategies is expected, but it becomes detrimental to creativity when players enter a match fully aware of the terrain. They no longer have any reason to deviate from the tried and true build orders. So even before the match begins, creativity has already become a minor element during gameplay.
The solution? Randomize the terrain for each game, and force the players to adapt. Force them to scout aggressively (not only the other player, but also the battlefield), force them to analyze the terrain for the best build order, force the players to decide where to position their units. No longer will players enter a match with a playbook. Instead, they’ll have to outsmart the enemy with novel strategies to best suit these novel battlefields.
Only then will each match will be truly unique. We’ll see victories resulting from maneuvers and adaptation rather than macro and static build orders. Gone will be the days when commentators use the first few minutes of the VOD to give shutouts. Important decisions will begin from the start.
Here's my point: include map generator in Starcraft 2, and make adaptability and creativity a bigger part of the game!
Edit: On Balance of generated maps:
Something to keep in mind is that maps we already have are imbalanced and they're made by PEOPLE. So nothing less should be expected from generated maps.
In non-mirrior matchups, it's inevitable that some imba will occur. But ultimately this is a problem of small sample size. In a best of, there is a better chance that players will get lucky or unlucky equally. Additionally, the unique nature of each map also means that players won't immediately know if a map favors their race, much less the specific things they can do to take advantage of the terrain. In other words, a imbalanced map does not automatically grant a win or loss because the mystery of the battlefield gives extra room for the underdog to win - a balancing element that repetitive maps lack.
Finally, imba can also add excitement to the game. It’s admirable when a player wins, but much more so (and more entertaining to watch) if he wins despite the odds.
I get what you're saying, and I agree with the concept you're talking about because I do think that creativity and adaptability are some of the most interesting parts of any competitive play. There always seems to be a tug-of-war between game balance and randomness, though, that makes it an understandably difficult concept to implement. For example, look at Left 4 Dead. Used to be that a team had no idea what was ahead of them, and this definitely made it more fun because you didn't know if you had a tank around the next corner or a normal zombie. When it came to competitive play, though, you got to the issue where one team might have a much easier time than the other because of the randomness of what they were going up against. One team might have to fight six or seven tanks, while another team might have to fight three, and this gave the latter what's considered an "unfair" advantage. A while after Valve changed it so that both teams would face the same bosses in the same places, which made it more balanced but deprived it of the excitement over facing an unknown challenge.
It's just a trait of competitive play, eventually everything will be boiled down to the optimal solution. There may be options to pick, but the entire point of this strategies is that they find the most efficient solution, and at a point it gets to where you can't do much more in terms of creativity. So while watching adaptability and creativity is extremely fun, it's significantly less fun if you're playing and you're punished for doing so. Procedurally generated maps would have the same problem, eventually all the possibilities would be memorized and planned for, or it would feel like it was giving too much of an advantage to one side or the other. Games with a high level of random factor can be fun, but usually not when you're playing competitive. Look at the fail of World of Warcraft's RNG-fest in level 70 arenas, you'd have situations where random factors would end up screwing one team over and making skill unnecessary.
The way competition works, there's just not much you can do other than explore the few new creative opporitunities when they come around.
Ender's game is my favorite book of all time. I enjoy reading, and whenever someone asks me for a recommendation, that usually is in the top 3. Beyond that, I like the idea. Whenever I think of the end when Ender is battling the buggers, it reminds me of fighting off the swarm - indeed there are many similarities between the two. But one thing is certain, Ender never entered any of those battles knowing the terrain beforehand - he had to constantly analyze the situation and adapt his strategy accordingly.
In fact, as you will recall, in the battleroom, once it became apparent that Ender had completely mastered the standard play environment, the odds began to be stacked against him. Graff started to change the rules, which accordingly, caused Ender to have to struggle and to grow - he couldn't use the same proverbial "build orders" and counters.
I think the idea of having a random map generator is a fun idea, but the practicalities of actually implementing it would be difficult, and as stated previously, could possibly turn out rather generic. However, one thing that I think they could easily do is make it so you can make a map pool, and then have it randomly choose a map from within that pool. So maybe you take all the iccup maps of the week, throw them into the list, and then the game will randomly choose one of those maps, and you won't know which one until the game actually comes up on your screen and you're having to get your workers moving on the mineral lines. That way, it's kind of like playing random - you don't know what race/strategies/counters you're going to have to implement until the game is on.
I think it's a way fun idea though - good thought!
Something to keep in mind is that maps we already have are imbalanced and they're made by PEOPLE. So nothing less should be expected from randomly generated maps.
In non-mirrior matchups, it's inevitable that some imba will occur. But ultimately this is a problem of small sample size. In a best of, the players will get lucky or unlucky equally.
Imba can also add excitement to the game. It’s admirable when a player wins, but much more so (and more entertaining to watch) if he wins despite the odds.
I've thought of this idea, and I think it is workable, especially now that you can see the terrain as the game begins.
But in order to be a truly workable solution, it needs to have civ style configurable settings for things like map style, size, and such. Inevitably it'd generate maps no one likes so it would require an approval system before the game starts.
I think that if it could be done right, and it is a very tall order by the way, but if it could be done, it would be an amazing thing.
But then for competitive play you end up with another problem right off the bat. Right now when you watch a match, the map is obviously reasonably balanced. With a map generator, were you really better than the other guy? Or was it a "map win"?
That said since the sc2 team apparently has some time on their hands, I'd love to see them develop it just to see if it turns into something. There were plenty of game modes in the original that never stuck, this one certainly deserves a chance.
Good post, I definitely agree. I've pretty much stopped playing SC because it's just boring to do the same exact thing that 100,000 other people have done before. The fun in a game, to me, is in innovation. I expect to do my best in SC2.
I'd love to see a decent random map generator. It wouldn't become a standard in standard tournaments, but I could see random map tourneys and the like, or even just for messing around.
On August 08 2009 07:45 zazen wrote: Not only this would be very, VERY hard to implement, the maps would also be completly imbalanced... This could never truly work in SC2... Sorry.
It could work, I'm sure that the actual coding a random map generator isn't that hard. What is hard though, is making a decent algorithm that would generate even half decent maps.
It would be sooooo luck based because you would have to build some of your buildings and units before you actually get to take a good look on whole map which would put one player in better position or everyone would just start with some safe build every time which would be extremely boring. There is already luck factor now in cases when one player takes shorter scout route, and other takes longer. Imagine what would it be if terrain was random.
Problem is not in maps. Every sport has pitch which never changes but players and their moves do. What should be done is more versatile units, buildings and overall engine to allow more things.
I think map generator would be good thing, but not in competitive play.
Age of empires has a random map generator (versions 2 and 3). The thing is, Age also has fewer cliffs on maps, resources are everywhere around you. In either case, the random maps were often fun to play. I doubt pros will use them because they could be imbalanced, but if someone thinks it could be made into SC, that might be pretty fun. There's also this random map setting in Age, where a premade map in a list is randomly picked. It might offer more creativity...and very little coding. No one will practice different BO's for 48 different maps.
On August 08 2009 08:19 MannerMan wrote: I think it's funny when you say in a bestof series the luck will be equal... maybe best of 101.
Keep in mind that map balance is unlike flipping a coin. When you flip a coin, there are two clear outcomes. But when there is an imbalanced map, the outcome is much more ambiguous. Players won't automatically know if the map favors them, or the things that needs to be done to take advantage of the terrain. In other words, a imbalanced map does not automatically grant a win or lose, the mystery gives extra room for the underdog to win.
Randomized maps is a terrible idea. In every way, shape, and form.
Seriously.
If you really want to spice up the Starcraft scene with something overly-radical, add more map features through the use of map settings. For example, doing this you could have a 2v2 with shared bases. This could lead to much more interesting 2v2's, with higher tech units and more wins based on the teamwork of players as opposed to their individual skill levels.
On August 08 2009 08:54 KhaosKreator wrote: Randomized maps is a terrible idea. In every way, shape, and form.
Why are random maps so bad? I love random maps on Age of Empires, random maps define the game, perhaps for the pro-scene, random maps may not be used, but for single player or casual gameplay, random maps are awesome.
I enjoy preparing and watching progamers preparations for certain games on certain maps. Even prepared cheeses that abuses certain dynamics of the maps can be creative.
Just add some new maps every few months and there will be variety and the map pool will get bigger and better over time.
On August 08 2009 08:54 KhaosKreator wrote: Randomized maps is a terrible idea. In every way, shape, and form.
Why are random maps so bad? I love random maps on Age of Empires, random maps define the game, perhaps for the pro-scene, random maps may not be used, but for single player or casual gameplay, random maps are awesome.
I'm thinking random maps won't work for starcraft.Just think of it you get a main base and 1 or 2 nat expansion with every map. But with a random map generator you could have like 6 minerals and no gas in your main and then 1 mineral spread everywhere on the map, and your opponent could have like 4 minerals and 1 gas. In terms of terrain I don't really see it as too much of an advantage as you are not sure what to take advantage of in a map. Maybe for building turrets in the middle of maps is the only obvious thing.
I sometimes remember Ender's Game, in connection with Starcraft, when I'm watching Jaedong or Flash play. When I see those guys being pushed beyond their limits in terms of physical and emotional energy, I can't help but draw the comparison.
Obviously their strict training regimen and match-pressure are no way near in the same league as the cruelty that was inflicted upon the child-geniuses in Ender's Game, but I'm sure if you've read the book you'll know what I mean.
On August 08 2009 08:54 KhaosKreator wrote: Randomized maps is a terrible idea. In every way, shape, and form.
Why are random maps so bad? I love random maps on Age of Empires, random maps define the game, perhaps for the pro-scene, random maps may not be used, but for single player or casual gameplay, random maps are awesome.
I'm thinking random maps won't work for starcraft.Just think of it you get a main base and 1 or 2 nat expansion with every map. But with a random map generator you could have like 6 minerals and no gas in your main and then 1 mineral spread everywhere on the map, and your opponent could have like 4 minerals and 1 gas. In terms of terrain I don't really see it as too much of an advantage as you are not sure what to take advantage of in a map. Maybe for building turrets in the middle of maps is the only obvious thing.
Hm I guess. Maybe Starcraft isn't the type of game that would go well with a random map, because the 3 races are radically different. In Age of Empires though, the randomizer allotted resources to each person together, so it was like each person got x sheep, x gold mines, etc.
I think for random maps to be used in the proleague, they would have to be restricted in their randomness. Parameters could be set to make sure that players get a main base and natural expansion with adequate resources, and so that the terrain is suitable enough for a game with further expansions and paths from base to base. I think it would be pretty cool and the only imbalance would really be the players' ability to adapt, which is what we want to see! BRING ON THE RANDOMS
On August 08 2009 09:00 D10 wrote: Not every good idea is feasible for implementation.
Sorry lad, it will never be done with this level of technology.
You understand technology as much as I understand why human beings post stupid shit they have no idea of.
If you think random maps in Starcraft need to be any different from simply having players play on a different human made map they never played before on every game. Then you have no idea what procedural content is.
You are such an arrogant guy VIB, you didnt do anything other than insult me, without disproving my point.
If you think Its only because of our technology hasant reached that level yet you are deply wrong, and didnt understand my point at all.
All I meant to say is its not possible to make it in a way that will work without being completely cumbersome/having tons of misgivings
Blizzard knows it wouldnt becost effective in the effect it will provide compared to all investiment it will take take to make one that doesnt suck ass, and the end result is so intangible compared to just having good balanced maps that it doesnt make sense.
You cant simulate weeks of testing and tweeking like we have right now with pro bw maps into a program, if we had a mind blowingly better AI technollogy capable of thinking like a human and running simulations of all kinds of abusive (and would would have to program what are abusive parameters) situations on one map, then I would say that there is nothing stopping us from doing it, but we dont, and therefore doesnt deliver
There's a lot of problems w/ random maps in competitive play imo. First of all, the game is at such a fast paced in which gamers wouldn't be able to analyze the map nearly enough to make any sort of creative strategic play on the fly. Ender had the luxury of just sitting there, and analyze the entire battleroom right from his view. People playing on a randomized map will not have that luxury. Standard openings will always be standard because they are the most flexible of openings that can deal with almost every strategy. A safe 1 or 2 rax cc can deal with any strategy a zerg throws, be it 4 pool, 2 hatch build, 3 hatch, etc, for instance.
IMO, this how a random map will boil down to in competitive play: Either cheese, or open with a safe build (2 rax, 2 gate, 9 pool). Obviously this will end up favoring zerg a bit. Go for the standard build path that is safe, scout and adapt accordingly, and abuse terrain if luck is on your side. Not too much different from normal games, it'll just be a lot sloppier and we won't see any super-refined build orders prepared for the match.
D10, YOU are being extremely arrogant for assuming that you have any idea when you clearly don't. I personally don't know shit about rocket fuel, so I don't go saying that NASA should use engines moved by fart to get to Saturn, that would be arrogant and stupid by my part. You don't have a slightest clue of what technology and AI are. So don't be arrogant and stop.
For me, another problem with randomly generated maps is that, if that's all we were playing, there would be no 'classic' maps that everyone likes to watch/play games on.
On August 08 2009 10:42 D10 wrote: Then you care to explain where im wrong sire ?
No. Every single day I get dozens of newbies asking me stuff they would take less time googling then typing a question. I just wasted my whole fucking work day with an idiot who didn't read the Docs and told me my program had a bug when it was doing exactly what the Docs said. YOU are the same type of person, who could have searched how procedural content is generated in less that time that it took you to write your crappy posts in these threads. I'm fed up with lazy bums like you. So look it up yourself.
I'm thinking random maps won't work for starcraft.Just think of it you get a main base and 1 or 2 nat expansion with every map. But with a random map generator you could have like 6 minerals and no gas in your main and then 1 mineral spread everywhere on the map, and your opponent could have like 4 minerals and 1 gas.
Not if you make the random generator (gasp) not do stuff like that!
I played AoE2; note that random maps there always give you 4+2+2 sheep and a couple trees near your town hall and some gold/stone fairly close and some bushes and.... Or go look at the good random map scripts for Heroes of Might and Magic 3--those generate fairly balanced maps since they place stuff by "zones," and starting zones have the same sort of stuff. I'm not saying that maps in those games are the same as maps in Starcraft, but the ability to random maps with the same amount of resources per player is definitely out there.
I don't see anything wrong with adding the option to play on random maps to the game. It will never become the standard, but I'd be surprised if a good random map generator wouldn't lead to a decent amount of play.
I agree with OP as well as with VIB, in that random maps could be interesting, also could be really fun for skirmish vs computer, you could set up maps like 2 human vs 5 comp with a narrow choke and it is generated, it all depends on how advanced they would build it, and i cant see how it would be that hard to implement a random map generator for blizzard, it has ben done so many times before. And i have a lot of knowledge about programming and game development.
On August 08 2009 10:42 D10 wrote: Then you care to explain where im wrong sire ?
No. Every single day I get dozens of newbies asking me stuff they would take less time googling then typing a question. I just wasted my whole fucking work day with an idiot who didn't read the Docs and told me my program had a bug when it was doing exactly what the Docs said. YOU are the same type of person, who could have searched how procedural content is generated in less that time that it took you to write your crappy posts in these threads. I'm fed up with lazy bums like you. So look it up yourself.
you do understand the sense of a forum dont you? as a pm this post probably would have made a lil sense, but not as a forum post. You do realize that more people than D10 are reading your posts? Even if he looks anything up, what about all the others interested? Yep, these lazy bums can look it up too! Cus hey, the combined time of 20 people looking shit up on the internet sure isnt worth a damn compared to your 10 mins to explain it for all of them.... You fail at grasping the concept of a constructive forum.
On August 08 2009 11:42 djdolber wrote: it all depends on how advanced they would build it
You're smart, I like you.
There is technically no reason why a randomly generated map would be ANY different from a human made map. The code that generates the map was made by a human after all. Random maps doesn't need to be any different from playing on a different human made map that you don't know on each game.
From Blizzard's point of view it's simply a matter of managing human resources. Buildings a robust content generator = more resources in software engineers but less resources in level designers and artists. While building many static maps instead would be spending resources in the opposite. The game developing industry is slowly moving towards more and more procedural content every year. But Blizzard is a "complacent giant" like any other big monopoly. So don't expect them to be too innovative with this.
I think we need to clear something up, some people are talking about randomly generated maps (randomly made and put together by a computer) and some people are talking about going into a match and not knowing what map out of the pool they will be facing.
I would be all for the second option, random map choosing, but against the first, the randomly generated maps, they would be retarded and deeply affect balance.
People say that men like to score every five seconds. So do baseball teams. So when I reread the book "The Homerun Kid" by Dan Gutman, I couldn't help but think of our beloved national pastime.
If you don't live in the US, Japan, Korea, or the Dominican Republic, you may know little about baseball. It's a game which involves a ball, nine players per team, and four bases. Baseball teams beat their opponents by running around the bases more times... elements that are sometimes lacking in modern baseball.
"But that's blasphemous! Hitting has gone up in the last decade! RBI, HR!"
Let me explain:
***
Every sport/game/competition works in one particular way the majority of the time! I mean, should baseball move the fences around every inning? Throw beachballs instead of baseballs? I'm not sure you're watching the right thing, just in the same way that the bottom of the ninth is interesting in many different baseball games, starcraft is interesting, because no two games are ever alike, no matter what builds are used! And depending on the situation, the drama is still there! Oy!
The extra patch of land in original LT that let siege tanks kill gas on a really privileged position come to mind, how to prevent these minor glitches from completely imbalancing the game ?
Its not that constant maps remove the adaptability and creativity skill, they just add new ones, like map knowledge, specific BOs, etc...
Random maps only are imbalanced if the maps have time to be examined.
This requires on the fly adaptability to win instead a boring textbook rush where the advantage of knowing the route is nullified. Especially if the map turns out to be an island map.
I'd find it useful for making large maps when I don't have the time to carve out the land, but nothing else. Most maps are so small that there wouldn't be much randomness at all, seeing as there are certain qualities needed to make the map even remotely playable (much less "balanced" for pro and ladder). Plus, it sounds like the map editor will have tools powerful enough to make randomization practically obsolete. Randomization works best for obnoxiously huge maps where you can build virtually anywhere, like in Civilization. Starcraft maps are small and there are tons of places where and lots of terrain types upon which you cannot build.
It is a cool feature, but it would be virtually pointless for Starcraft II. What we really need is a dedicated community of map-makers who can whip up insanely creative maps with the click of a button. We also need to take the time to play on larger maps than usual, so there is more room for such creative map development -- the larger the painter's canvas, the greater the brush's freedom.
On August 08 2009 13:43 D10 wrote: The extra patch of land in original LT that let siege tanks kill gas on a really privileged position come to mind, how to prevent these minor glitches from completely imbalancing the game ?
Its not that constant maps remove the adaptability and creativity skill, they just add new ones, like map knowledge, specific BOs, etc...
Something to keep in mind is that maps we already have are imbalanced and they're made by PEOPLE. So nothing less should be expected from randomly generated maps.
In non-mirrior matchups, it's inevitable that some imba will occur. But ultimately this is a problem of small sample size. In a best of, chances are that players will get lucky or unlucky equally.
And because the maps are unique every time, players won't automatically know if a map favors them, much less the things they can do to take advantage of the terrain. In other words, a imbalanced map does not automatically grant a win or lose, the mystery of the battlefield gives extra room for the underdog to win.
On August 08 2009 14:39 Aerox wrote: Random maps only are imbalanced if the maps have time to be examined.
This requires on the fly adaptability to win instead a boring textbook rush where the advantage of knowing the route is nullified. Especially if the map turns out to be an island map.
That is some fucked up logic. Why are maps not imbalanced if you can't analyze them? Often times maps aren't imbalanced because of 1 abusive, calculated strategy, but because the terrain is naturally in favor of one race. No amount of analysis will change that and the players don't have to do anything to take advantage of it. If there are tons of walls around, protoss is gonna have a hell of a time flanking a tank push, and it doesn't take a week's prep with a coach and some b team practice slaves to figure that out.
What random maps require the most to win is luck, and that's why they shouldn't be adopted by any competitive SC2 groups. I mean the feature would be tons of fun, but I and many others would be very disappointed if preparing for specific maps was taking out of SC2. Plus no one wants an OSL final to be determined by Jaedong getting God's Garden vs Flash game 5.
On August 08 2009 15:27 Benthum wrote: In non-mirrior matchups, it's inevitable that some imba will occur. But ultimately this is a problem of small sample size. In a best of, chances are that players will get lucky or unlucky equally.
The players taking turns getting lucky maps wouldn't make for a good series (well it might, see bisu v jaedong in GOM, but it wouldn't make for good games or a truly epic series.) A series where Protoss is 80% to win games 1&3, then Terran is 80% to win games 2&4 isn't nearly as fun to watch as a series where the maps are balanced the whole way through. On average each player will have 2 wins going into game 5 but the series with balanced maps has more room for variety, the games have more standalone value and viewers stay more interested.
Plus once a player gets a gift from the map gods, odds are it's not gonna even out. We don't have the sample size to ignore the effects of any 1 lucky map, so it could be devastating to series play as we know it.
On August 08 2009 15:27 Benthum wrote: In non-mirrior matchups, it's inevitable that some imba will occur. But ultimately this is a problem of small sample size. In a best of, chances are that players will get lucky or unlucky equally.
The players taking turns getting lucky maps wouldn't make for a good series (well it might, see bisu v jaedong in GOM, but it wouldn't make for good games or a truly epic series.) A series where Protoss is 80% to win games 1&3, then Terran is 80% to win games 2&4 isn't nearly as fun to watch as a series where the maps are balanced the whole way through. On average each player will have 2 wins going into game 5 but the series with balanced maps has more room for variety, the games have more standalone value and viewers stay more interested.
Plus once a player gets a gift from the map gods, odds are it's not gonna even out. We don't have the sample size to ignore the effects of any 1 lucky map, so it could be devastating to series play as we know it.
The uniqueness of each map means that players won't automatically know if a map favors their race, much less the specific things they can do to take advantage of the terrain. In other words, a imbalanced map does not automatically grant a win or loss because the mystery of the battlefield gives extra room for the underdog to win - a balancing element that repetitive maps lack.
The mystery of the battlefield doesnt even make sense from a lore point of view, in real life you have lots of time to study the terrain (even if a few hours) and make a strategy before hand.
On August 08 2009 15:41 huameng wrote: The players taking turns getting lucky maps wouldn't make for a good series (well it might, see bisu v jaedong in GOM, but it wouldn't make for good games or a truly epic series.) A series where Protoss is 80% to win games 1&3, then Terran is 80% to win games 2&4 isn't nearly as fun to watch as a series where the maps are balanced the whole way through. On average each player will have 2 wins going into game 5 but the series with balanced maps has more room for variety, the games have more standalone value and viewers stay more interested.
The maps would be random each time. And I doubt you could be 80% to win in any map you see for the first time, simply because there are so many unknown factors to consider...
Plus, the main challenge in making a random map generator algorithm isn't having it create reasonably balanced maps... that would most likely be fairly easy. The hard part would be having having random maps that are both balanced *and* interesting - allowing a high enough degree of variation in the maps that it doesn't feel like there are just a couple of maps with different "skins". It would be interesting if they managed to pull it off.
However, i doubt random maps would really be used in the pro scene, and that's a shame, really...
On August 08 2009 16:10 D10 wrote: The mystery of the battlefield doesnt even make sense from a lore point of view, in real life you have lots of time to study the terrain (even if a few hours) and make a strategy before hand.
In starcraft you have time to study the terrain too before the battle - that's why you send workers / overlords out to scout. It might not be "lots" of time, but hey, in real life training one soldier takes months / years compared to a few seconds in starcraft.
Random maps is a good idea. If nothing else have them mirror so both players have to deal with the same stuff.
Let there be parameters... same minerals in each main, same amount of expos near the base, same base layout, etc etc.
Though in the pro scene? That's pretty dicey. Thats adding more RNG which is the exact reason WoW is an awful e-sport. It would make for great special matches though
If SC2 plays out similarly to how SC does, then, IMO, this would be imbalanced: certain races are easier to adapt with than others; for me, Protoss takes no time to learn a map -- you can basically learn a map on the fly. On the contrary, I have never been able to win my first game on a map when using Terran: Terran is so terrain dependent that if you have not mastered the layout, then your play will suffer greatly, and psychologically you will feel that you are going to lose since you are not comfortable with something that is critical to your race's success.
The result? Modern games follows a repetitive pattern of player A doing an opening and player B reacting by choosing a known counter (I see you 9 pool, so I’ll put down 2 canons before nexus). So rather than a game of "who can out think the other", the weight has shifted to "who can click the fastest"
I hate it when people combine to statements with some kind of causal word; putting a causal word between two statements does not create an causal relation ship.
1) people don't have to think that much on the fly so 2) it becomes a click festD
why? Chess openers are written out untill move 20 or smt. Does that make chess into "who can move his pieces the fastest?"
I for one hope Blizzard does NOT implement random maps for the ladder.
Having played all three Age of Empires games competitively, I've seen my fair share of games being decided by the random map outcome. In fact the biggest Rise of Rome tourney (Age of Empires I expansion) got decided by random maps, as one of the finalists got berry-fucked twice in a row (difficult access to starting resources).
But this is not even my main complaint against random maps, as I'm sure it is possible to come up with a decent algorithm with sufficient restrictions for random maps. The thing is random maps often makes for a very blurry game where there's no clear execution of build orders or strategies. Players often explore the map while they build up trying to get the edge. You often find yourself going with a mainstream strategy that works well overall, and then adapt the strategy to the map as you go along. It basically defeats the purpose of having an ultra fine tuned gameplay such as that in SC1. Random map aliminates so many aggressive strategies.
If they include it for custom games, then cool, but not for ladder.
I tend to agree with that the game has become too much of a click fest. Years ago, there were pro gamers who could excel with what we would call "low apm" today, and probably even considered low then (was before my time).
In today's game, if someone has 240 apm and they can fare decently as a pro, then we marvel at it as it is some sort of great feat. I feel that 240 apm in any game should be enough...
I think having to play the game at a fast pace is great, but... when you make a strategy game, I think you want to find a delicate balance where strategy will remain at the forefront, and solid mechanics and apm will be in a more auxiliary role than it is now. I feel hand-speed is starting to overshadow strategy to a certain degree.
Some people have wrong logic here. People say that Boxer is great because he did ridicoulus strategies etc, but that doesn't mean that there is randomness in his play. Read his autobiography and you'll find out how many hours did he spent on creating strategies etc. There is no way that someone can pull great strategy on the fly. It can only be completely luck based.
imho it could be fun in custom games but terrible in ladder games.
On August 08 2009 07:45 zazen wrote: Not only this would be very, VERY hard to implement, the maps would also be completly imbalanced... This could never truly work in SC2... Sorry.
Has everyone forgotten about age of empires? Random map generation worked perfectly in these games, although the scale of the map was larger than in Starcraft. Random map generation would be amazing and would definitely add quite a bit of unpredictability and fun to a game that after 11 years is just starting to get kind of boring.
I think you are misjudging the situation. It's not like there was almost no innovation nowadays, and in fact you even gave us a list of important developments in SC that are spread somewhat equally over that time frame. But those developments could never have been made by confronting people with random maps. You could never make the jump from 3 hatch muta to 2 hatch muta on a random map because it is such a fragile build that it would fail in 99% of the games, and thus could never be refined enough to be useful. You also couldn't create innovative builds because in SC a great part of your build has to be decided before you scout your enemy, and certainly before you have the time to scout and analyze the whole map. So you would always go for some cookie cutter builds or all-ins. Depending on the frequency of cliffs over mineral lines, terrans would either tech early drops or not. But usually you don't get to scout that before you have to put down your gas.
All I could see coming from random maps is standard play, until someone spots a terrain feature he can abuse and win the game. There is no time to adapt to a random map, simply because of the timings in a fast-paced game like SC.
Besides, if you don't see innovative play in SC that often, you've got to open up your eyes for the details. In almost every progamer match I can see some creative or unexpected move. It's just that innovation in SC today is not as easily spotted as it was in boxers time. A good example that also shows the importance of non-random maps: Recently there have been several zvzs where one zerg makes 6 lings and hides them from the opponents overlord, then he makes some more lings, shows them and attacks with all lings at once, surprising the enemy with a larger force than expected. I am not sure who invented that strategy (I somehow got effort as a name in my head, but don't know for sure), but it would not have been possible if that player had not known the timings and patterns of overlord scouting, as well as places where he could hide his lings. You could not do that on a random map because that build is decided long before you can find out all the stuff you need to know to make this work.
To conclude this: IMO random maps don't give us anything desireable in competitive SC. Maybe for fun every now and then. They only disrupt balance and take away many strategic options.
Oh and about AoE: in AoE all the "races" are about the same. As long as the map has similar starting positions everything will be fine. This does not work for SC at all, though. Just because a map is mirrored it does not mean it is balanced at all.
And one more important thing. SC1 is 11 years old, it is normal that there are no more big changes, but I really don't care what will happen with sc2 in 11 years. I want to play it for 4-5 years at most and after that I expect tecnhology to advance, gief holographs and simulators like in star trek (:
On August 08 2009 16:24 HuskyTheHusky wrote: Random maps is a good idea. If nothing else have them mirror so both players have to deal with the same stuff.
Let there be parameters... same minerals in each main, same amount of expos near the base, same base layout, etc etc.
Though in the pro scene? That's pretty dicey. Thats adding more RNG which is the exact reason WoW is an awful e-sport. It would make for great special matches though
RNG in mmo makes you adapt to the situation so you cant fight a preplanned battle i dont see why this would be bad in rts.
Randomized map? You know how damm hard it is to make even a semi balanced map in the map-editor? I'm postingthis without reading the rest of the posts but this just seems ridiculous to me. In SC2, even though technology has progressed, I highly doubt that randomly made maps by the computer will have a sheer semblance of balance. If the computer is made to have certain parameters in the map, then each map will have a similar layout. Conceptually new maps will not be made such as Troy or Outsider. Also, there is still much creative play in StarCraft. Look at Fantasy. He makes great use of dropships TvP, something that hasn't been done for a long time.
'Random map' being selectable when choosing a map so that neither player knows which one it will be before the game begins is a solid idea (at least in ladders or other situations with a set map pool). Randomly generated maps on the other hand sounds beyond bad.
I would be interesting, however I dont actually think it would increase creativity to have randomly generated maps. Your not going to do a cheese build on a map you dont know anything about, neither are you going to go for risky fast expansion. Build orders will actually suffer as people will have extreme difficulty breaking a standard safe build without time to sit down and spending time creating something specifically designed to beat it.
Also, for spectators who will have absolutely no map knowledge either, it will be very hard to follow the game on TV as you wont recognise any parts of the map.
The way to keep starcraft fresh with adaptability and creativity is to change the map pool often, not random maps.
I'm not too sure, but I think it would be impossible to have a random map generator. The paramaters would either be so generic that the maps would be inevitably the same (clones), or they would be so complex that having unique maps would be impossible to do - and be playable that is.
On August 08 2009 16:24 HuskyTheHusky wrote: Random maps is a good idea. If nothing else have them mirror so both players have to deal with the same stuff.
Let there be parameters... same minerals in each main, same amount of expos near the base, same base layout, etc etc.
Though in the pro scene? That's pretty dicey. Thats adding more RNG which is the exact reason WoW is an awful e-sport. It would make for great special matches though
RNG in mmo makes you adapt to the situation so you cant fight a preplanned battle i dont see why this would be bad in rts.
HAAHAHAHAH, what's your experience with WoW PvP? My guess is none. RNG is fucking terrible. It makes you do the exact same thing with either a chance of double damage or no damage. It gives your stuns a chance to hit or miss. You're still going to do exactly the same actions, except it's either going to kill the opponent, or it's going to do nothing.
I think that having a computer make the map would'nt be a very good idea. That would cause for some imbalances. But as a few have said before, it would be fun if you didn't know what map you're playing on beforehand. That would be awesome. You'd have to really know your stuff . Another thing, in pro games they should change the maps every... week? That way gamers have to adapt quickly, a true test of skill.
On August 08 2009 16:24 HuskyTheHusky wrote: Random maps is a good idea. If nothing else have them mirror so both players have to deal with the same stuff.
Let there be parameters... same minerals in each main, same amount of expos near the base, same base layout, etc etc.
Though in the pro scene? That's pretty dicey. Thats adding more RNG which is the exact reason WoW is an awful e-sport. It would make for great special matches though
RNG in mmo makes you adapt to the situation so you cant fight a preplanned battle i dont see why this would be bad in rts.
Rofl, just... wow. RNG in any game makes it so that games are won and lost based on a dice roll. That is completely terrible for any competitive environment. Having a game lost because the RNG decided to roll "Resist" or a heartbeat resist that breaks a cc is completely awful.
yea rng can make you win or lose a battle if you are very lucky or unlucky but in most fights its average for both player in those fights it takes alot more adaptability to win than if there was no rng at all.
besides in some games you can try to escape those unlucky fights.
On August 09 2009 03:51 LuckyOne wrote: not talking about wow,
yea rng can make you win or lose a battle if you are very lucky or unlucky but in most fights its average for both player in those fights it takes alot more adaptability to win than if there was no rng at all.
besides in some games you can try to escape those unlucky fights.
You might say you're not talking about wow, but name one other MMO played on a professional competitive level. Massive requirements of adaptability are not the pinnacle of a competitive game. BW still requires adaptability and creative thinking from the players, though not nearly to the extent that it used to. Look at July, he's still kicking it because he not only has kept up on the mechanical side, his creativity is amazing. Introducing RNG into a competitive environment in any form is a bad idea because the result is that the number generator is deciding games, not the skill level of the players. The RNG doesn't care if this is day1 or the grand finals. Over the entire season you might get an average of lucky and unlucky rolls, but that doesn't help a poor roll of the dice costing you the one game that matters at this moment. When a match is decided by individual games, the overall average of the season means nothing. Supreme adaptability does nothing in the face of a horrible streak of luck, and a competitive game should not be based on luck. The negative consequences of RNG are far greater than the benefits gained by it.
Idle wrote: Rofl, just... wow. RNG in any game makes it so that games are won and lost based on a dice roll. That is completely terrible for any competitive environment.
A random map generator can be made very intelligent. A computer can calculate rushing distances (in time, obviously), mineral line to perimeter distance (for tank and muta harass defence), choke point size and positions, number of routes, resource amounts...
It is possible to determine the defensive strength of a position by calculating ground and air distances from a base, and check building room, cliff direction, path importance..
Of course, a human might spot opportunities for abuse that a computer can't see. Maybe the random map generator can be enhanced to parametrize these settings in a way so that balance is assured. But when players spot opportunities for abuse, these players should be rewarded for it.
In terms of competitive starcraft, random maps might be an issue. It would be possible that maps are randomly generated beforehand and approved by a commission, keeping the map secret from the players. Just in case the map generator causes too often an imbalance, which might especially happen easily on higher level play as imbalances could be magnified by good micro.
Someone mentioned the risk of cheesing on random maps, making it never viable. Good point I'd say, how about players get a set amount of time to prepare a strategy before the match, say 10 minutes? Maybe 5 for a nice fat commercial brake. The casters could look at the map at the same time and predict strategies.. that'd be fun. Hey, combined with the knowledge that the map generator put into map making, it could present the main features that it tried to get into the map, that'd be rich
The OP is phrased as a single argument, but it should be split in two. One point made is that starcraft's best times are when new strategies unfold. The other is that random maps allow for 'better dynamics in available strategies' (my own words, they might not match 100%). I'll leave discussion on the truth on either count out of my post. It is a very difficult discussion. The only thing I want to add here is some ideas of the possibilities.
You guys don't understand one big thing. ALL of the best tactics which we have seen in last few years weren't made during the game. Best tactics are always made before, with alot of testing, practicing etc. Players can excecute some really great moves but it is impossible to be creative and think of some good tactics while maintaining good macro etc. That would just decrease overall game level and games would become worse. Everyone would just be using same safest build all over again because they wouldn't want to risk anyhing. This is so terrible idea for competitive play because it comes from wrong logic. This wouldn't increase creativity because game does not alow it. You can't just make one build, and then scout, find enemy units and figure out that you want to be creative and then change your entire build. That is not possible during the game, that's why best tactics are always made after long time of thinking and experimenting while practicing. Only thing which can make things more exciting is more units and more balanced units because nowadays some units are really rarely used which is limits the number of possible tactics etc.
On August 09 2009 06:09 crate wrote: Also maybe the best game I've ever played imo is Magic: the Gathering.
Actually, MtG is, IMO, an example of how randomness can create a very unfortunate situation for a competitive game. While the randomness can create a wide variety of game-states, it also makes it hard to see players consistently do well. A single bad draw can make a player who's top 8-ing Pro Tours and Grand Prix fail to make day 2 at an event. From the view of a spectator and fan, it's hard to really be a fan of a player or team when one wins so inconsistently.
Also: Civilization series > Wesnoth. Even in multiplayer.
On August 09 2009 06:55 Badjas wrote: A random map generator can be made very intelligent. A computer can calculate rushing distances (in time, obviously), mineral line to perimeter distance (for tank and muta harass defence), choke point size and positions, number of routes, resource amounts...
The problem with this is that the generator has to preemptively evaluate factors. How can a computer assess the balance of a map without knowing how relevant a given feature will be? How could it know that the positional imbalance on an LT-style map is huge without a large sample set of games on similar maps?
Imbalance comes from contemplation of the map before hand. The comments about imbalance are effectively white noise, and I think a random map generator could be integrated into multiplayer. That is not to say the whole map would be randomly generated. What you get then is a uniformly indistinct maps that are just boring. But what if you only randomised some of the map?
Imagine Python in your head. Now, instead of a barren centre, imagine the centre had randomly placed strategic features. Perhaps a hill somewhere. Mini forests. Blockages. Patches of buildable or unbuildable terrain.
But what if you get hills right in front of naturals? Wouldn't Terran rip that shit up? The answer is algorithms of course. Every 'randomly' generated feature could have it's position determined in a fashion that's the reverse of how the 'clouds' of probability for electrons are calculated.
Pros: Novel map features without the imbalance cons of a random map
Cons: Imbalance could pop up, but honestly in a well coded system it'd be squashed in any series Much more work would go into making a map, and much more testing would be requried Plays that require a static map may be made more difficult, or (very unlikely) impossible.
On August 09 2009 06:09 crate wrote: Also maybe the best game I've ever played imo is Magic: the Gathering.
Actually, MtG is, IMO, an example of how randomness can create a very unfortunate situation for a competitive game. While the randomness can create a wide variety of game-states, it also makes it hard to see players consistently do well. A single bad draw can make a player who's top 8-ing Pro Tours and Grand Prix fail to make day 2 at an event. From the view of a spectator and fan, it's hard to really be a fan of a player or team when one wins so inconsistently.
Also: Civilization series > Wesnoth. Even in multiplayer.
On August 09 2009 06:55 Badjas wrote: A random map generator can be made very intelligent. A computer can calculate rushing distances (in time, obviously), mineral line to perimeter distance (for tank and muta harass defence), choke point size and positions, number of routes, resource amounts...
The problem with this is that the generator has to preemptively evaluate factors. How can a computer assess the balance of a map without knowing how relevant a given feature will be? How could it know that the positional imbalance on an LT-style map is huge without a large sample set of games on similar maps?
Could not have said that better myself. Magic is a great game, but the randomness does not add to the competitive aspect, it detracts from it. One lucky draw can be the difference between complete domination of a game and losing miserably.
On August 09 2009 06:55 Badjas wrote: A random map generator can be made very intelligent. A computer can calculate rushing distances (in time, obviously), mineral line to perimeter distance (for tank and muta harass defence), choke point size and positions, number of routes, resource amounts...
The problem with this is that the generator has to preemptively evaluate factors. How can a computer assess the balance of a map without knowing how relevant a given feature will be? How could it know that the positional imbalance on an LT-style map is huge without a large sample set of games on similar maps?
Starcraft has been patched many times to enhance balance after it was released. A random map generator can also be patched.
From my post: Of course, a human might spot opportunities for abuse that a computer can't see. Maybe the random map generator can be enhanced to parametrize these settings in a way so that balance is assured. But when players spot opportunities for abuse, these players should be rewarded for it.
Not claiming that random map making is the solution to strategic creativity. I do feel convinced that a random map generator can produce quality maps. You mention that LT style maps are imbalanced. I'm pretty sure a map generator can calculate what a human can see (with regards to strategy of course).
The best way to find out what a random map generator can do and what it can do for casual / non-pro play, is to create one for starcraft (sc1 / bw).
Have fun with coding one for SC1. Getting somewhat proper terrain done is a ton of work already. Especially if you don't even have the ISOM Algorithm, and as far as I know neither Blizzard nor SuicidalInsanity have released anything. Add all the the stuff you'd have to hardcode (like custom ramps because they basically use random tiles) and you're quite occupied. Then you'd have to develop all the algorithms, gather all the parameters through replay analysis or whatever, and even work in the metagame.
That might be possible if you got a team of people who are experienced in those fields and get paid for that, but otherwise this will take quite a while.
Magic is a great game, but the randomness does not add to the competitive aspect, it detracts from it.
I'm going to just say I don't agree. Magic without randomness would be an awful game. Where's the fun in a deterministic card game? That's what it would be.
No point in even playing if you take out the randomness. Just build decks and compare the decklists and you're done.
Also maybe the best game I've ever played imo is Magic: the Gathering.
Both have large elements of luck in them and are still fanastic competitive games.
I agree that Starcraft is not the place for randomness in a competitive setting, but I think your comment here is ridiculous.
Oh, what's up, my leader red mage missed all 4 shots on a red hp melee character and died the next turn to 6/6 direct hits (and consequently missed all his staff hits) while on a mountain?
Wesnoth is a HUGE RNG gamefuck. There's only so much you can do when you're -30EV, not to mention that getting any significant -EV during the first major fight screws you pretty majorly. If the game had more units and experience wasn't so powerful, the RNG in wesnoth wouldn't be terrible for competitive play, but it is. There are very few player in the wesnoth competitive community specifically because most people correctly realize how fucked the game is.
Randomly generated maps would bring a whole new level to starcraft IMO, because instead of memorizing maps from a map-pool beforehand, both players have to go through the random map together not knowing anything about it. The concept probably sounds silly to most people, and I doubt there will be any way to make a randomly generated map for SC that is actually close to balanced, or doesn't look shitty.
Magic is a great game, but the randomness does not add to the competitive aspect, it detracts from it.
I'm going to just say I don't agree. Magic without randomness would be an awful game. Where's the fun in a deterministic card game? That's what it would be.
No point in even playing if you take out the randomness. Just build decks and compare the decklists and you're done.
You're right, Magic without randomness would be an awful game. That doesn't mean that randomness adds more to the competitive aspect of the game. Magic is a TCG, comparing magic to SC as a competitive skill based game is like saying blackjack is as competitive or skill based as chess. The game is not designed around showing player skill, its designed around selling trading cards.
But to be honest, it's the progamers that dictate what kind of gameplay occurs in Starcraft. If one can defend against a 1base rush while getting 2bases and win because of that economic advantage in the end, why not?
Here's my analogy for you guys. Starcraft has often been compared to chess, so this might be an appropriate comparision.
Chess is a game whose board and rules never change. The only variable is player choice. The result of this are dozens of "standard openings", not unlike the kinds we see in professional Starcraft. Each standard opening has a rather standard response, though deeper into chess games you can see how creativity can affect the game. However, some chess games can be played rather mechanically, as well, several moves deep into a game.
There is a chess variant called Fischer Random Chess, or Chess 960, which randomizes the starting position of the nearest row of pieces (the non-pawns) except for the position of the king (must be between two rooks), and the two bishops must start on opposite colored squares. This results in 960 different possible starting positions of those 8 pieces (mirrored on the other side, but notably not reversed). It's named after Bobby Fischer because he prosposed it, as it allowed an individual's maneuvering and tactical prowess in the game to shine, rather than standard play (or more specifically, standard play is what the smart people created and everyone else copied when it was proven to work). A true depth of chess understanding is required to know how to do the best with your units and their starting positions, rather than knowing exactly how you're going to open weeks beforehand (as I'm sure Fantasy knew exactly what he was going to do to Jaedong on a map like Neo Medusa well ahead of time).
There's no reason this can't be true of Starcraft II maps, as well. No one is saying the map should be completely randomized (just as Chess 960 is not). What should be randomized in this scenario are distances between starting positions, arrangement of expansions, presence of islands, inter-base terrain features (chokes, high ground, watch towers). The computer can of course be told to generate these maps based on simple formulas like "starting base requires 9 mineral patches and 2 gas, with adjacent natural expansion with 9 mineral patches and 2 gas", and what is random is the way they open onto the map, the elevation and the choke (defensibility), and destructible features around the map.
What if the map has randomly generated a vulnerable backdoor cliff that a colossus or reapers could exploit? That means all of the bases have it. Scouting and adapting to the map as play unfolds requires skill, and exploiting it requires on-the-fly creativity. These are aspects of competition that should be more rewarded than running a build you've practiced thousands of times. Sure, after the match you could say it was imbalanced, but you can also ask yourself "What could the loser have done to prevent the exploit of these features?"
Anyway, random maps would of course be mirrored and modular. That is to say if a base spawns with a nat and mineral-only third, all behind the same choke, then so do the other starting positions on the map. Does this provide a distinct advantage to one race? Maybe. It's your job to figure out how to crack it before the other opponent can figure it out of exploit it. This can only increase the level of competition in Starcraft.
Taking away standard maps at least allows for some players to prove and win based on their adaptability. I'm kind of surprised that KESPA doesn't introduce "surprise" maps, one that the players don't know about ahead of time, but are familiar with and have practiced on. Would Fantasy have done as well against Jaedong if he had not known the map would be Neo Medusa for the ace? What if it had been hidden from the players until game time? Would've been interesting.
Magic is a great game, but the randomness does not add to the competitive aspect, it detracts from it.
I'm going to just say I don't agree. Magic without randomness would be an awful game. Where's the fun in a deterministic card game? That's what it would be.
No point in even playing if you take out the randomness. Just build decks and compare the decklists and you're done.
You miss the point. No one said that Magic without randomness would be good. But the fact of the matter is that while randomness adds to the GAME, it detracts from making it competitive, because of how luck can so often allow a worse player to beat a better one.
The thing about randomized maps is that they don't add significantly to the game experience, while they do significantly detract from the format's competitive viability.
Idle wrote: Rofl, just... wow. RNG in any game makes it so that games are won and lost based on a dice roll. That is completely terrible for any competitive environment.
Also maybe the best game I've ever played imo is Magic: the Gathering.
Both have large elements of luck in them and are still fanastic competitive games.
I agree that Starcraft is not the place for randomness in a competitive setting, but I think your comment here is ridiculous.
It's a subtle point, but these are two entirely different ideas of randomness. In one case (randomly generated map), an entire game could be decided at one instant with a single 'die roll'. You could have a PvZ where the map happens to have 4 easy gasses and favors the flanking style of the Zerg and has a tiny main base etc. etc.
In the case of Wesnoth, though, the randomness is something that you manage. Is there a chance that an important unit could die if you attack now? If yes, is it a high chance (for those that don't know, the game calculates all probabilities for you and displays the result before any action is taken)? If yes, then you risk a major decisive blow against yourself - it's probably best not to. This act of playing conservatively over the course of dozens of random encounters represents good "risk management". The winner is usually the player with the best "betting" skills - that is, he bets on the smart money. He buys low and sells high and isn't taken in by excessive greed.
If this kind of randomness was in Starcraft, it probably wouldn't be that big a deal - this kind of randomness depends on the scale. Wesnoth is probably at 'medium' on this scale, meaning that there is some not-insignificant chance that luck could make or break you. WoW is at 'high' on this scale, meaning that a significant chance exists that luck will determine the outcome. However, if Wesnoth were a game where the map was randomly generated, you bet you could kiss balance goodbye. The locations of villages (similar to expansion locations) and the terrain surrounding them would determine the winner.
I think randomized maps would be an interesting feature, but I would be really unhappy to see it in progaming. Some of the most exciting games are ones with well-planned strategies and timing. Why should we reward on-the-fly creativity above all? Aren't execution, preparedness, and a sense of existing game balance good too? I personally prefer games where crazy strategies are performed based on heavily calculated build orders, building locations, scout timings and positions and so on. I think those demonstrate an incredible depth of understanding.
On the other hand, I fear that having randomized maps would just force the game even more towards mechanics than it already is. Builds would no longer be tailor-made, and I fear that the finesse they sometimes require would be gone. Builds would become more like general concepts based on map allowances, and the player with more EAPM would be able to carry out their general build the best.
I realize my two objections are sort of mutually exclusive. They are more "arguments against other arguments" than whatever I actually feel. What I do feel is that random maps have no place in progaming.
Some of you guys write waaay too many words to make a simple point. =P
Anyway, would a random map generator actually be useful? I doubt it. Example: how many people have played on every: - Progaming map? - Blizzard map? - User-made map on brood war maps?
If these options haven't been exhausted yet, then I don't see how having even more maps would be worth the effort.
On August 09 2009 18:31 spinesheath wrote: Have fun with coding one for SC1. Getting somewhat proper terrain done is a ton of work already. Especially if you don't even have the ISOM Algorithm, and as far as I know neither Blizzard nor SuicidalInsanity have released anything. Add all the the stuff you'd have to hardcode (like custom ramps because they basically use random tiles) and you're quite occupied. Then you'd have to develop all the algorithms, gather all the parameters through replay analysis or whatever, and even work in the metagame.
That might be possible if you got a team of people who are experienced in those fields and get paid for that, but otherwise this will take quite a while.
No one is asking for the random maps to look good, afaict. That part can always be taken care of by a human, anyway, if the map turns out to be fun.
As a starting point, in SC1, the algorithm could draw out a main base + nat + 3rd base in a corner of the map using a handful of square tiles, then rotate it to make the other 3 corners.
I'd say it's do-able by a highly-motivated community member.
On August 10 2009 12:26 Bill307 wrote: Some of you guys write waaay too many words to make a simple point. =P
Anyway, would a random map generator actually be useful? I doubt it. Example: how many people have played on every: - Progaming map? - Blizzard map? - User-made map on brood war maps?
If these options haven't been exhausted yet, then I don't see how having even more maps would be worth the effort.
Quoted for truth. Some of the posts in here are huge considering what you want to say is "Randomness can be ok, but not always lol!"
The OP is inherently flawed, because he suggests that everything pro-level players do is mapped out. Yea, they might go in with a build order, but whether or not they:
-expand (They've weakened their opponent but not enough to finish them) -move in for a kill (weakened opponent enough to finish) -turtle up (expecting a finisher and is trying to counter) -or move out (catching opponent while expanding) -Tech decisions -Upgrade decisions -Multitude of other decisions
These are all things that must be that can be decided in reaction to the opponent.
Even though SC is getting more and more mapped out, there are so many variables involved in a game that no game is the other one alike. Plus, it's still evolving.
Maps are fine. Gimmicks like this aren't required to promote a vibrant and exciting scene.
To those who think that random maps will be mysterious and their imbalances not readily perceivable: A progamer does not need to be some strategic savant to recognize an insanely short overlord/rush/air distance ala Battle Royal, a plethora of gas per defensive point ala God's Garden, or simply too much map center clutter helping the mech push ala Lost Temple.
This is not a good idea. If you don't like what SC has developed into, that's just fine. I think there's plenty of variety in gameplay, and I don't need to see Boxer-esque play every game to be entertained/satisfied. It's hard enough learning to properly execute the "standard" build order, and I don't think you appreciate the meticulous preparation that progamers put into every televised match in order to counter what they expect their opponent to do.
Beyond that...randomized terrain? The positional imbalances would be ridiculous. How would you guarantee that the main only has one entrance? How would you guarantee there's a natural? How would you guarantee the bases are distributed fairly?
Not many seem to understand this, but even mirror matchups can suffer from map imbalances if they're positional and terrain-based ones, just like on old Medusa at the 6 o'clock base with the main gas being tankable from the back ridge.
There'd be no point in preparing build orders anymore if you had no idea what the fuck the map you were going to play on would be. Actually...really, how the fuck would you practice in the first place?
I can't believe how many people actually LIKE this idea.
It’s the players who innovate – Savior (use of swarm), Bisu (Bisu build), Flash (Dual Armory build), Fantasy (Fantasy build, mech build), July Zerg (Mutalisk Stacking), Jaedong (2-hatch muta, queen usage) etc… we remember and like to watch.
Yeah you basically wrote down people who have been dominant with these things because its not true that they innovated them (at least for the most of them), but were good enough (strong mehanics) to implement them successfully in proscene.
On August 11 2009 04:54 PH wrote: There'd be no point in preparing build orders anymore if you had no idea what the fuck the map you were going to play on would be. Actually...really, how the fuck would you practice in the first place?
thats the point you wouldnt need to practice and prepare just play and pwn or be pwnt.
Here's my analogy for you guys. Starcraft has often been compared to chess, so this might be an appropriate comparision.
Chess is a game whose board and rules never change. The only variable is player choice. The result of this are dozens of "standard openings", not unlike the kinds we see in professional Starcraft. Each standard opening has a rather standard response, though deeper into chess games you can see how creativity can affect the game. However, some chess games can be played rather mechanically, as well, several moves deep into a game.
There is a chess variant called Fischer Random Chess, or Chess 960, which randomizes the starting position of the nearest row of pieces (the non-pawns) except for the position of the king (must be between two rooks), and the two bishops must start on opposite colored squares. This results in 960 different possible starting positions of those 8 pieces (mirrored on the other side, but notably not reversed). It's named after Bobby Fischer because he prosposed it, as it allowed an individual's maneuvering and tactical prowess in the game to shine, rather than standard play (or more specifically, standard play is what the smart people created and everyone else copied when it was proven to work). A true depth of chess understanding is required to know how to do the best with your units and their starting positions, rather than knowing exactly how you're going to open weeks beforehand (as I'm sure Fantasy knew exactly what he was going to do to Jaedong on a map like Neo Medusa well ahead of time).
There's no reason this can't be true of Starcraft II maps, as well. No one is saying the map should be completely randomized (just as Chess 960 is not). What should be randomized in this scenario are distances between starting positions, arrangement of expansions, presence of islands, inter-base terrain features (chokes, high ground, watch towers). The computer can of course be told to generate these maps based on simple formulas like "starting base requires 9 mineral patches and 2 gas, with adjacent natural expansion with 9 mineral patches and 2 gas", and what is random is the way they open onto the map, the elevation and the choke (defensibility), and destructible features around the map.
What if the map has randomly generated a vulnerable backdoor cliff that a colossus or reapers could exploit? That means all of the bases have it. Scouting and adapting to the map as play unfolds requires skill, and exploiting it requires on-the-fly creativity. These are aspects of competition that should be more rewarded than running a build you've practiced thousands of times. Sure, after the match you could say it was imbalanced, but you can also ask yourself "What could the loser have done to prevent the exploit of these features?"
Anyway, random maps would of course be mirrored and modular. That is to say if a base spawns with a nat and mineral-only third, all behind the same choke, then so do the other starting positions on the map. Does this provide a distinct advantage to one race? Maybe. It's your job to figure out how to crack it before the other opponent can figure it out of exploit it. This can only increase the level of competition in Starcraft.
Taking away standard maps at least allows for some players to prove and win based on their adaptability. I'm kind of surprised that KESPA doesn't introduce "surprise" maps, one that the players don't know about ahead of time, but are familiar with and have practiced on. Would Fantasy have done as well against Jaedong if he had not known the map would be Neo Medusa for the ace? What if it had been hidden from the players until game time? Would've been interesting.
I completely agree with that. As a chess player as well - I love this kind of stuff. We would play games where we had to choose our pieces based on points (9 for queen, 5 for rook, etc), up to a maximum, so we could use what we wanted. We would play ``blind`` chess, where we could not see our opponents pieces. We would play 4 way chess. There are so many things you can do, if you take your head out of your ass and look at all the possibilities.
I am all for random maps - as long as they have proper coding. There are a number of ways to help reduce or eliminate positional and map imbalances this way, without having to actually make a map.
They don`t need to look incredibly pretty - that`s not the focus of a map. The focus is on the game.
On August 11 2009 04:54 PH wrote: This is not a good idea. If you don't like what SC has developed into, that's just fine. I think there's plenty of variety in gameplay, and I don't need to see Boxer-esque play every game to be entertained/satisfied. It's hard enough learning to properly execute the "standard" build order, and I don't think you appreciate the meticulous preparation that progamers put into every televised match in order to counter what they expect their opponent to do.
Beyond that...randomized terrain? The positional imbalances would be ridiculous. How would you guarantee that the main only has one entrance? How would you guarantee there's a natural? How would you guarantee the bases are distributed fairly?
Not many seem to understand this, but even mirror matchups can suffer from map imbalances if they're positional and terrain-based ones, just like on old Medusa at the 6 o'clock base with the main gas being tankable from the back ridge.
There'd be no point in preparing build orders anymore if you had no idea what the fuck the map you were going to play on would be. Actually...really, how the fuck would you practice in the first place?
I can't believe how many people actually LIKE this idea.
You may not understand how a "random map" is nothing but just a variation of a certain map design. Have you ever played Age of Empires 2/3 or Age of Mythology? If so, you would know there are plenty of maps in the game. For example, in Age of Mythology there is a map called Midgard. The basic design of the map is 2 settlements near each player's starting spawn, and a number of hunting that spawns at equal distances from the starting spawns. It also produces a number of fishing spots in the water. The map layout is generally the same as well (one huge island with water all around). Every time the map is "randomly generated," it will produce the same design. The only differences are the terrain being different. In this case, the map is not symmetrical.
Oh, and there were build orders for the map that would ALWAYS apply. The scout in this game would be used early on to find resources (but in SC2, your scout will probably find the natural on most maps if there was a random map generator).
A random map generator could be sophisticated enough to make a symmetrical map with an equal number of expansions, mineral patches, the distances to and from, all the time. The only things that would change are the terrain, but it is possible to keep cliffs a certain distance away, ramps, etc and of course make it symmetrical.
On August 11 2009 04:54 PH wrote: There'd be no point in preparing build orders anymore if you had no idea what the fuck the map you were going to play on would be. Actually...really, how the fuck would you practice in the first place?
thats the point you wouldnt need to practice and prepare just play and pwn or be pwnt.
On August 11 2009 04:54 PH wrote: This is not a good idea. If you don't like what SC has developed into, that's just fine. I think there's plenty of variety in gameplay, and I don't need to see Boxer-esque play every game to be entertained/satisfied. It's hard enough learning to properly execute the "standard" build order, and I don't think you appreciate the meticulous preparation that progamers put into every televised match in order to counter what they expect their opponent to do.
Beyond that...randomized terrain? The positional imbalances would be ridiculous. How would you guarantee that the main only has one entrance? How would you guarantee there's a natural? How would you guarantee the bases are distributed fairly?
Not many seem to understand this, but even mirror matchups can suffer from map imbalances if they're positional and terrain-based ones, just like on old Medusa at the 6 o'clock base with the main gas being tankable from the back ridge.
There'd be no point in preparing build orders anymore if you had no idea what the fuck the map you were going to play on would be. Actually...really, how the fuck would you practice in the first place?
I can't believe how many people actually LIKE this idea.
You may not understand how a "random map" is nothing but just a variation of a certain map design. Have you ever played Age of Empires 2/3 or Age of Mythology? If so, you would know there are plenty of maps in the game. For example, in Age of Mythology there is a map called Midgard. The basic design of the map is 2 settlements near each player's starting spawn, and a number of hunting that spawns at equal distances from the starting spawns. It also produces a number of fishing spots in the water. The map layout is generally the same as well (one huge island with water all around). Every time the map is "randomly generated," it will produce the same design. The only differences are the terrain being different. In this case, the map is not symmetrical.
Oh, and there were build orders for the map that would ALWAYS apply. The scout in this game would be used early on to find resources (but in SC2, your scout will probably find the natural on most maps if there was a random map generator).
A random map generator could be sophisticated enough to make a symmetrical map with an equal number of expansions, mineral patches, the distances to and from, all the time. The only things that would change are the terrain, but it is possible to keep cliffs a certain distance away, ramps, etc and of course make it symmetrical.
That does make it sound better. I haven't dealt with RTS random map generators since the old C&C ones, and I've never played AoE or AoM.
However, I really don't think this would work out for any major league. Sure, I guess it'd be fun to have a small Heritage League-sized tournament where they use these kinds of things.
However, with how much a map's balance can be altered with even the most minute changes to the terrain, expo placement, even mineral and gas placement, I still think this would kill competition more than it would refresh it.
We could just have humans make the maps but not release them prior to tournaments. (Maybe 2 games in a Bo3 are played on pre-released maps, the decider is a surprise? Something like that...)
Map selection could be randomized on ladder play.
These preserve the basic idea without the potential technical challenge of a map generator.
As a side note, I really sort of miss maps that DO have terrain imbalances, like older versions of LT (creating opportunities for T to push between bases) or Jungle Story or Gorky Island.
On August 11 2009 12:35 PH wrote: However, with how much a map's balance can be altered with even the most minute changes to the terrain, expo placement, even mineral and gas placement, I still think this would kill competition more than it would refresh it.
I think one of the main points of the OP is this:
The ONLY reason a maps balance is so radically changed with even minute changes in the map is exactly BECAUSE the maps get played so much that all those little flaws are discovered. You have so many thousands of Koreans playing those same maps for so many hours every day, that even a small imperfection in the map is going to get spotted by someone, and then it has to be fixed.
However, in a real-time setting, this would be impossible, no one would no of any of the maps weaknesses ahead of time, many that did exist would probably never be discovered. And anyone who could discover minute weaknesses in a map and properly take advantage of them would be a legitimate genius in the mold of Ender Wiggin and would be lauded as the Ultimate Bonjwa. I'm not counting on this happening very often if you had a good map generator.
My idea for the generator would be set bases/nats that could maybe be moved around the map, and then all the terrain in between the bases could be randomized different ways. Seems like it would be easy to set a certain distance around each main as "uncliffable" by tanks and just include it in the "movable base" section. Everyone would have the same access to starting resources, but after that, you have to discover on your own. I do think this would lead to some very safe builds becoming more prevalent, but you'd also see some real risk taking and innovation as well. I don't think there'd be a lot of "map wins" though.
IMO, this would not supplant the current map structure, but would be an awesome addition to the community and one that would make for both a lot of fun gameplay and fun viewing. I doubt it will ever happen, but I personally think it would be really cool.
On August 11 2009 11:26 lMPERVlOUS wrote: I completely agree with that. As a chess player as well - I love this kind of stuff. We would play games where we had to choose our pieces based on points (9 for queen, 5 for rook, etc), up to a maximum, so we could use what we wanted. We would play ``blind`` chess, where we could not see our opponents pieces. We would play 4 way chess. There are so many things you can do, if you take your head out of your ass and look at all the possibilities.
The thing is, none of those things introduce randomness into the game. They introduce variation. Which adds the point that there are a ton of ways to introduce variation into the game without introducing randomness.
Plus, there's another consideration. Fischer Random Chess is fun, but would you really want it to be played where it affects the ranking of the top players in the world? Implementing random maps for casual play is fine. Forcing them into progaming isn't.
On August 11 2009 04:54 PH wrote: There'd be no point in preparing build orders anymore if you had no idea what the fuck the map you were going to play on would be. Actually...really, how the fuck would you practice in the first place?
thats the point you wouldnt need to practice and prepare just play and pwn or be pwnt.
On August 11 2009 04:54 PH wrote: There'd be no point in preparing build orders anymore if you had no idea what the fuck the map you were going to play on would be. Actually...really, how the fuck would you practice in the first place?
thats the point you wouldnt need to practice and prepare just play and pwn or be pwnt.
And you think that's a good idea?
yea, then it would be more skill based than time.
Utter bollocks. By all accounts of what the word skill constitutes, investing time into exercising a faculty is invariably related to one's skill in leveraging the faculty. The vast majority of human skills display variance in ability related to the time and effort invested. If the results of competitors in an activity are wholly unrelated to their efforts, luck and talent (in a specific sense of the word) remain as the determining factors.
Furthermore, in order for a game to be a viable e-Sport, wherein a full-time career as an athlete is an option, organized, full-time effort must generate materially superior results to freelance, part-time effort. If skill fails to vary based on time (a hyperbole that will almost certainly never happen), all pro-gaming institutions as we know them lose their raison d'être. The world's best competitor can be literally anyone with sufficient talent. He is no longer a wonder of the world since he is no longer a testament to both human genius and dedication, but to the former alone. All the top players lose half of their mystique since we the audience could easily invest equal effort.
On August 11 2009 04:54 PH wrote: There'd be no point in preparing build orders anymore if you had no idea what the fuck the map you were going to play on would be. Actually...really, how the fuck would you practice in the first place?
thats the point you wouldnt need to practice and prepare just play and pwn or be pwnt.
And you think that's a good idea?
yea, then it would be more skill based than time.
Utter bollocks. By all accounts of what the word skill constitutes, investing time into exercising a faculty is invariably related to one's skill in leveraging the faculty. The vast majority of human skills display variance in ability related to the time and effort invested. If the results of competitors in an activity are wholly unrelated to their efforts, luck and talent (in a specific sense of the word) remain as the determining factors.
Furthermore, in order for a game to be a viable e-Sport, wherein a full-time career as an athlete is an option, organized, full-time effort must generate materially superior results to freelance, part-time effort. If skill fails to vary based on time (a hyperbole that will almost certainly never happen), all pro-gaming institutions as we know them lose their raison d'être. The world's best competitor can be literally anyone with sufficient talent. He is no longer a wonder of the world since he is no longer a testament to both human genius and dedication, but to the former alone. All the top players lose half of their mystique since we the audience could easily invest equal effort.
it wouldnt eliminate effort since playing more ppl would make you better, just remove the boring parts of the effort like practicing build orders (since the optimal build would vary depending on the generated map, ideally)
Are you promoting the suggestion that randomized maps be the competitive norm on the basis that the relevant time spent practicing will include less boredom? If not try to relate what exactly you're supporting and how.
I don't want to answer you based on my own assumptions, but I'll note that pro-gamers may practice builds for other reasons than rote memorization. Obviously I don't know first-hand, but from interviews and related discussion, I've read that learning and formulating answers to gameflow permutations, maximizing one's own play strengths towards the map, matchup, and opponent, and weighing various contingencies to opponent moves by ease and efficacy... are all crucial considerations for a pro-gamer working on his build. You might agree with me that they're intellectually stimulating enough to stay boredom.
On August 12 2009 03:29 EchOne wrote: Are you promoting the suggestion that randomized maps be the competitive norm on the basis that the relevant time spent practicing will include less boredom? If not try to relate what exactly you're supporting and how.
I don't want to answer you based on my own assumptions, but I'll note that pro-gamers may practice builds for other reasons than rote memorization. Obviously I don't know first-hand, but from interviews and related discussion, I've read that learning and formulating answers to gameflow permutations, maximizing one's own play strengths towards the map, matchup, and opponent, and weighing various contingencies to opponent moves by ease and efficacy... are all crucial considerations for a pro-gamer working on his build. You might agree with me that they're intellectually stimulating enough to stay boredom.
well if im not wrong they spend alot of time fine tuning builds and strategies for specific maps and then play alot of ppl to see how they react to it (kind of like writing a book of how they should play with all possible cases),
then on serious matches they try to execute what's on the book flawlessly.
even with good mechanics and game sense you would never be able to compete with them because they spend 16h a day writing their book.
the randomness would ideally eliminate its use because there would be too many different cases, leaving mechanics and game sense.
On August 12 2009 03:14 LuckyOne wrote: it wouldnt eliminate effort since playing more ppl would make you better, just remove the boring parts of the effort like practicing build orders (since the optimal build would vary depending on the generated map, ideally)
Wrong. Reducing a part of the game that requires effort and practice to fine-tune reduces the overall effort that needs to be spent. Just because progamers spend less time perfecting builds does not mean that 100% of that time is suddenly shifted into other practice scenarios, because of the diminishing returns on additional practice after a certain point. Its not like if Jaedong didn't have to spend 8 hours preparing a new build for a map, he'd just spend all that time practicing muta micro instead, because the marginal gain is no longer worth the payoff. Players only keep practicing as far as the practice is worth their time, and if you reduce the gain from planning a good strategy on a know map, you reduce the benefits gained from practice.
On August 12 2009 03:53 LuckyOne wrote: the randomness would ideally eliminate its use because there would be too many different cases, leaving mechanics and game sense.
So you want to emphasize mechanics and on-the-fly thinking over planning. That's fine, but I don't see how that actually benefits the game. It's not *worse* than the current way of doing things, but you can't really say its better either.
The thing is, the potential imbalances and luck-based factors introduced by randomized maps outweigh the potential benefits. Emphasizing mechanics and game sense is DIFFERENT from emphasizing planned strategy, but not objectively BETTER.
even with good mechanics and game sense you would never be able to compete with them because they spend 16h a day writing their book.
Is this sort of status quo lamentable or in need of improvement? No reputable competitive activity allows competitors that have exhausted the limits of human effort to be bested by those who have only dedicated a small fraction of their potential.
To be honest, if I were able to regularly beat players who've invested superior effort at a game, it would tell me that the game is quite shallow and a waste of my time, since evidence suggests that, in the game, hard work does not pay off.
The reality is that although practice can benefit certain faculties (as you cite, mechanics and game sense), planning is the one department where both time and creativity can reap limitless yet measurable (one's progress in game sense is largely amorphous while one can actually record data on different situations, contingencies, tendencies exhibited during planning) benefits due to the number of permutations that could spawn in a game of Starcraft.
throwing out a weird idea that popped into my head what if maps were generated at random, and a certain amount of time was allotted for each player to examine the map before the game even started, like say 2 minutes (arbitrarily picked). this would allow for some room for creativity, as players aren't playing completely blind, while at the same time squashing down a bit on the current methodology of playing a map to death.
not sure if it would work for starcraft 2, but who knows
On August 12 2009 16:00 alphafuzard wrote: throwing out a weird idea that popped into my head what if maps were generated at random, and a certain amount of time was allotted for each player to examine the map before the game even started, like say 2 minutes (arbitrarily picked). this would allow for some room for creativity, as players aren't playing completely blind, while at the same time squashing down a bit on the current methodology of playing a map to death.
not sure if it would work for starcraft 2, but who knows
The thing is, this doesn't help the issue of map imbalances. It actually increases the problem, because with that time, players are more likely to discover those imbalances.
On the other hand, the imbalance could be reduced if the system generated 3 maps, and each player were given 15 minutes to vote one out, which would leave the least imbalanced of the 3 to be played on.
Honestly though, I don't see why people dislike the current method of playing a map to death. As EchOne mentioned, why should it be possible for a noticeably less prepared player to beat a more prepared player? Map-specific practice drastically improves the chances of winning for a prepared player, because spending a day perfecting a strategy on a map yields much more visible results than a day practicing general macro/micro, especially at the high level of play which progamers are at. You may not *like* the system, because it means that an obscene amount of time has to be devoted to seemingly dry practice, but it creates entertaining games (arguably more than the alternative because practice allows a greater array of viable ingenious strategies).
Ironically, one player who actually has shown this kind of on-the-fly ingenuity in current competitive Starcraft, but has not used varied prepared builds is written off by the general fanbase of competitive Starcraft as simply being able to 1a2a3a. Best sealing his own front door on Troy against Nada in response to Nada's bio build in the EVER2008 OSL, and using a shuttle to ferry zealots out after having his door sealed by Luxury on the same map show instances of on the fly responses to dire situations, yet most people dismiss Best as only being able to macro, and not showing any creativity.
On August 12 2009 18:14 dcttr66 wrote: too much arguing. extra game mode= extra fun. do it.
Once again, the discussion isn't directed toward the idea of implementing it in the game, but at the heavy implication in the OP that it should be used in competitive play.
Well even if it wasn't used in competitions, it would be a good feature for casual play- I find my friends don't like playing the maps I practice too much on.
Rather than doing a complete switch to all random generated maps. Would it not work to have throw them in as a variant- maybe the rubber match on best of 3 or else on comps with best of 5. It would require more flexibility for the players who could still memorize build orders, but then were also forced to figure out the best use of an unknown map. It could showcase the versatility of certain players without throwing out the training program for known maps.
I love this idea...being able to think on your feet should be the basis of starcraft. I, too, have gotten real sick of the repetitiveness...heck, I'll admit I rarely stray from playing on maps that I know. Flip up the terrain, and make it randomized...
Of course I think the odds of this actually being implemented are like 1 in 1000. People are too stuck on their perfectly balanced maps, guaranteed natural, exact number of gases, single ramp, etc...
Uh, idk if this has been mentioned cus I cba reading it all :<. but Red Alert command and conquer 2 had a random map generator, where you could make it random a map and then watch it before start? That looks like it could work out without too much trouble.
I think the bottom line and ultimate counter to this argument is, why?
Why do you need random maps when pre-made ones already have zero-chance of undetermined imbalance and are familiar so you can learn them after playing a little bit?
A diverse amount of creative strats already are created for pre-existing maps.
Maps evolve, and an algorithm that makes code doesn't. When SC1 first came out, there were maps with no gas in expos, a gas geyser that requires 8 workers, no natural expansions, imbalanced spawn points, etc. But as the maps evolve, we now have maps that have an easily securable natural and usually 3rd, a nat gas, only 3-4 workers required, we have maps like holy world with the dwebs and neutral ccs, and maps like neo medusa with the blocked temples. IMO if a random map generator was made, it wouldn't be able to keep up with the innovations that human mapmakers are making.
I agree with this thread. Unfortunately, creativity only goes so far in this game. Creativity loses against good mechanics and trite strategies and therefore, in order to compete and secure your victory, I'm afraid you can only counter using better mechanics - hence the dilemma of the game favoring boring strategies with good mechanics than creative strategies with mediocre mechanics. Randomized maps cannot be solution because that would result in many players winning from sheer luck or imbalances of the map rather than adapting to the environment. I agree that this can allow more creative strategies but realistically, it's not the best solution to tackle this matter. Because, for example, if the randomly generated map is an island map, zerg would make mutalisks which would force terran or protoss to make valkyries/corsairs and then counter with drop etc. which wouldn't change one bit from the Starcraft you talk about. And to tell you the truth, professional Starcraft isn't always like that. I mean, there are innovative players and always will be, and we will definitely see more of brilliant strategies in the future that revolutionizes the style of play. This is also the reason why Starcraft is loved by so many for such a long time.
dhe95 you are exaggerating...i think it was the players that were putting 8 workers on gas, not the geysers. it was probably something like a gas only expansion and you probably needed more than the 8 workers for it... but yeah there definitely were quite a few maps that you would need 5 workers for the geyser. i think random map generation would be great for like a game mode for ladder that people don't really play much like warcraft 3's free for all. i don't see a problem with it.
i feel like they would want different map generation directions instead of different maps in the ladder when they do this. so basically if you made a custom game with random map generation you would have to pick your map generation directions, or, they could set a randomize the directions feature...which would make the map layout even more random, in other words, it wouldn't let you choose that you don't or do want an island map, or it wouldn't let you choose that you want or don't want yellow minerals or whatever...it would randomize that. i think it's a good idea.
EPT
