|
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.
|
On August 08 2009 06:52 Benthum wrote:
To give an extreme example, Zergling rush works EVERYTIME on the old map Blood Bath.
This is a ridiculous and untrue statement.
|
Randomized maps would probably look very generic and not be very fun to play on.
|
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!
|
Edit: On Balance of randomly 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 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.
|
Well, to appease the nazis, they could make a random mirror-map generator...
|
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.
|
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. 
|
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 work with WC3 where the races are so much alike but no way to balance this in SC.
|
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.
|
I think it's funny when you say in a bestof series the luck will be equal... maybe best of 101.
|
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.
|
Not every good idea is feasible for implementation.
Sorry lad, it will never be done with this level of technology.
|
On August 08 2009 07:01 Bosu wrote:
Randomized maps would probably look very generic and not be very fun to play on.
I laugh at the ignorance of this statement.
|
|
|
|
|
|
EPT
