|
CIRCLE SYNDROME
Local Resource Density, Central Map Width, Map Control, and the Baseline 2-Spawn Map Template
(Army Positioning Gameplay)
INTRODUCTION
Almost two months ago I came up with a theory I named "Circle Syndrome". It's a pretty catchy name, but it doesn't do much in the way of actually describing the problem.
My original posts about it are as follows:
http://www.teamliquid.net/forum/viewmessage.php?topic_id=288965#17
http://www.teamliquid.net/forum/viewmessage.php?topic_id=288965¤tpage=2#33
I have come to a much fuller understanding since writing these; this new OP is designed to bring it more to light.
WHAT IS CIRCLE SYNDROME?
Circle Syndrome is the bane of a macro-oriented map. I cannot actually describe what it is (as it has many levels and many forms), but I can describe how you make it and what it does.
First and foremost, Circle Syndrome reduces the complexity of map control by rendering it far too difficult to attain and therefore largely useless and uninteresting.
Circle Syndrome stunts the players' ability to utilize a map to its full potential. Even though your map has 12 or 14 or 16 bases, Circle Syndrome makes it difficult for the players to hold more than 40-70% of them simultaneously.
I look at the past three paragraphs and I don't see many fouls (explained or not)... But the fouls that are there are enormous.
By the way, the two main factors contributing to Circle Syndrome are Distances and Local Resource Density.
As you might have guessed, Circle Syndrome is incredibly Complex. Please bear with me as I start with one piece at a time.
LOCAL RESOURCE DENSITY
I first wrote about Resource Saturation here:
http://www.teamliquid.net/forum/viewmessage.php?topic_id=273780
In there I talked about map-wide, overall, "Global" Resource Saturation. But here I want to talk about Local Resource Saturation.
First a brief science lesson:
"Gravitation, or gravity, is a natural phenomenon by which physical bodies attract with a force proportional to their mass." - wikipedia
Think of each base like a "physical body", and consider how there is a sort of gravity between them. The gravity represents how easily one base flows into the next one. If you have one of these bases, how easy is it to take the other one?
It should be clear that bases that are on completely different sides of the map don't flow very well into each other. It should also be clear then, by contrast, that bases that are only 15 spaces away from each other (instead of say 150 spaces) flow much more easily into each other. There is a much stronger connection between bases that are closer together.
Now realize that every base on a map flows, or has a gravity, towards every other base on the map. Again, the strength between each connection largely depends on how close or how far the two bases in question are. The terrain of the path between them is also a large factor.
Not all bases are created equal. Some are "heavier" than others. Indeed, the place in which you originate (your main) is the heaviest. The base right next to it (natural) is almost as heavy. Somewhat lighter than that is the (potential) third(s). Etc.
Furthermore, some bases have less resources and are thus lighter. Some bases are high-yield bases and are generally heavier.
Some bases (think tertiary bases) are very easy to hold - heavier - because they have only a single fairly small choke. Some bases are extremely wide open making them much lighter.
Even if you did not fully realize this stuff before, you have surely felt some of it.
CENTRAL MAP WIDTH
The classic case of Circle Syndrome is when there is a ring of bases around the perimeter of a square(ish) map. What happens is these "lighter" bases complete a full circuit of a ring of bases around the map (this is where the "circle" in circle syndrome comes from).
The number one way to break this circuit is by manipulating something I call "central map width".
To find a map's 'central map width' you must first identify the bases that will always belong to a single player when they both still have the main/natural. This must be done on a case-by-case basis and can potentially be ambiguous or even dynamic. You have your side of the map, I have my side of the map. Which bases are truly mine and which ones are truly yours? (Can you easily deny one of the ones that can only belong to me?)
Loki II seems like a good map to help illustrate this. I have identified the bases that will pretty much always belong to a single player (player one in red; player two in blue):
+ Show Spoiler +
Now you should draw a (imaginary) line from the farthest ground entrance from one side to the farthest ground entrance to the other side for both players, like so:
+ Show Spoiler +
The central map width will always be between these two lines.
The central map width has two parts. The first part is the shortest single line of pathable area you can make between them, which in this case happens to be like so:
+ Show Spoiler +
The central map width is also pretty much all of the pathable area between these lines.
Note: This pathable area can be, often is and sometimes should be separated into multiple lanes. You must reach a balance between raw width (including air space) and actual pathable width - (with an ideal in mind) they should be roughly inversely proportional.
-
There is also something I will now be calling the "central map length". It's probably pretty much what you'd imagine:
+ Show Spoiler +
Note: The more bases a map has (or the more macro-oriented it is) the bigger it's central width and length should be (not proportional).
MAP CONTROL
The central map width and length are absolutely integral to the idea of map control and circle syndrome.
To go back to what I said at the top,
First and foremost, Circle Syndrome reduces the complexity of map control by rendering it far too difficult to attain and therefore largely useless and uninteresting.
This sounds like an opinion on the surface, but there is something truly tangible to be had here. Fully understanding what this is and why it is important probably takes a lot of experience with SC; for the sake of space I am going to leave much unsaid here (we can discuss later in thread if anyone wants).
So how do the central map width and length make map control too difficult to attain?
In a very broad sense, it's actually very simple. When a map's central length is significantly shorter than it's central width, map control becomes extremely (almost exponentially) more difficult to attain.
It has a lot to do with whether or not you can sneak an army across the map without them scouting you. It's largely about scouting and actually being able to intercept armies while simultaneously defending your bases and possibly attacking theirs. Obviously you don't want to make it too easy, but making it too hard has dire consequences.
If the central width is longer than the central length (or just very long period), and there are ~4 bases on the edges of the central map width, then it becomes nearly impossible to hold many of these bases; both players could probably hold one, but neither player can hold more than one for very long.
Whether or not this is a good thing is a matter of opinion, but it is objectively less complex (and therefore tends to be less interesting pending other factors).
Simply put, I personally think that the central length should almost never be smaller than the central width. I wanna throw some numbers out there, but damnit don't follow them exactly:
On an average 2p map, the Central Map Width should be something like 60-90 spaces wide (not necessarily including multiple lanes)
On an average 2p map, the Central Map Length should be something like 70-100 spaces long.
Just to give an idea. In other words: the center of your map should be significantly smaller than you would normally do unless you're aware of Circle Syndrome.
-
It is not enough to simply make this area and identify it. Interesting map control is dynamic, and the only way to make it dynamic is to put incentives (like bases or high ground area) along the area of map control.
It is more than possible to have 4 bases in the map control area without causing circle syndrome if you are careful about it. And do not forget about keeping racial balance; these central bases are GENERALLY more for terran and protoss while the high saturation on each player's side of the map is more for zerg (they can all be for everyone).
This part is massively important and largely unexplored; I leave it up to your imagination to come up with ways to fill it.
BASELINE 2-SPAWN MAP TEMPLATE
You might (probably should) be thinking "wtf?! How can a standard map template be a good thing?". Just hear me out.
This template does not really tell you how to do specific map features or structures. All of the terrain is up to you. The size of the map is up to you.
In return, roughly following this template is pretty much guaranteed to produce a map without too much circle syndrome. Furthermore, it helps serve as a 'baseline for comparison'; think of this template as like the controlled variables in a scientific experiment.
While you do get to choose the size of the map, I recommend using a rectangular base as it is the easiest to create a rectangular(ish) central map width with. Honestly pretty much any size is fine (if you know what you're doing), from 110x90 to 170x140 and everything in between and probably more.
![[image loading]](http://i.imgur.com/MQI8s.jpg)
green = high resource saturation (many bases)
yellow = medium resource saturation
red = low resource saturation
black = unpathable
For a map with no circle syndrome at all, the main/nat (and at least one third) should not get too close to the central map width. Remember that these are the "heaviest" types of bases, and putting them right next to the shortest central map width can single-handedly complete the circuit even if your central map width is plenty small.
![[image loading]](http://i.imgur.com/ZSbsB.jpg)
I mean, you can do whatever the hell you want with the map control in the center really. That's arguably the most flexible part of this template.
The word "hourglass" describes this template well, but it doesn't necessarily have to be like that. Make it interesting. It can be a single wide line or multiple thinner lanes, or a combination of both. I assure you there are many possibilities. But yes, the idea is definitely to generally concentrate bases on two far halves of the rectangle.
Put incentives (bases, high ground, etc) along the map control area to give reason to move along it dynamically (in a slow and methodical fashion).
PLEASE Twist this template around (see Loki II for how to fit it into a square map) and skew whatever you want with it in a controlled manner. Maybe start with only a single map that tries to follow it completely, but then branch off and try to make maps that are somewhere in between zero circle syndrome and massive circle syndrome.
BACK TO CIRCLE SYNDROME
Circle Syndrome is not necessarily a bad thing. It is usually a bad thing when there is too much of it, but if every map had exactly zero circle syndrome then it would get pretty boring before too long.
And if you happen to make or see a map with circle Syndrome, PLEASE do not just say "that map has cs" and call it a day. Identify WHY it has CS. Try to understand to WHAT DEGREE it has CS. And if necessary, try to identify what could change to lower the CS.
Some aggressive maps even WANT to have CS. Like on Deception, it can create a dynamic flow of bases that should be regarded as expendable (or not taken in the first place). Circle Syndrome is the ideal box. It's more than reasonable to stay inside the box (we certainly haven't fully explored it), but it is also more than acceptable to think outside of the box as well (if you know what you're doing).
EXAMPLE MAPS
There are three maps in particular that helped me understand Circle Syndrome.
If you don't already know or haven't looked at/memorized the maps in Ragoo's post: one of these maps has circle syndrome, which one is it?
Shakuras Plateau + Show Spoiler +
Daybreak + Show Spoiler +
Dual Sight + Show Spoiler +
+ Show Spoiler [ANSWER] +Dual Sight has heavy Circle Syndrome
Check that out; the other two fit my template's idea quite well (or maybe the other way around ^^).
---
I hope I have cleared some things up.
Thanks for reading,
- Barrin
|
Yes, first on this epic thread. Honestly a lot of what you said was extremely confusing even though I understand what you're saying. It would help if you expanded on some of your ideas. Just write all of your thoughts and don't worry about it being too long or having incomplete sentences, then go back and fix it up.
|
*Reads Barrin's post*
*Looks at current map project*
"Crap, my map width is bigger than my map length. FFFFFUUUUUU..."
But honestly, great post. Definitely can look at my maps more critically with these concepts.
Edit: Also for Dr. Who fans: http://i.imgur.com/y002R.png
|
While the post was a little confusing because thoughts seemed to start and trail off into nothing, I understand (more or less) of what you are saying. It sounds a bit 'foreign-y' in style (I mean that it seems to be a little stilted, as if someone who is ESL (such as myself) wrote it), but apologies if it is not.
Apart from that, I LOVE the fact that this is presented in this way. I feel that there needs to be much more discussion over the nature of maps because it will prevent a stagnation in design, repetition, or falling into patterns. So all in all, huge props for writing this!
|
I cannot actually describe what it is
Hmm, this seems to be an ongoing theme with 'Circle Syndrome'. Can somebody please just provide a concise and coherent definition of Circle Syndrome? My current understanding is that Circle Syndrome is something like this: Maps in which most expansions (excluding main and natural) are ambiguous, i.e., approximately equidistant from both players' naturals. Is that correct?
|
On January 20 2012 12:40 whatthefat wrote:Hmm, this seems to be an ongoing theme with 'Circle Syndrome'. Can somebody please just provide a concise and coherent definition of Circle Syndrome? My current understanding is that Circle Syndrome is something like this: Maps in which most expansions (excluding main and natural) are ambiguous, i.e., approximately equidistant from both players' naturals. Is that correct?
That's the most prominent case of circle syndome, as in if that occurse there is likely circle syndrome (and vice versa). But like Barrin said, CS is too complicated to really boil down to one factor, although in my opinion what you listed is the major factor on most maps with circle syndrome.
|
On January 20 2012 12:42 RumbleBadger wrote:Show nested quote +On January 20 2012 12:40 whatthefat wrote:I cannot actually describe what it is Hmm, this seems to be an ongoing theme with 'Circle Syndrome'. Can somebody please just provide a concise and coherent definition of Circle Syndrome? My current understanding is that Circle Syndrome is something like this: Maps in which most expansions (excluding main and natural) are ambiguous, i.e., approximately equidistant from both players' naturals. Is that correct? That's the most prominent case of circle syndome, as in if that occurse there is likely circle syndrome (and vice versa). But like Barrin said, CS is too complicated to really boil down to one factor, although in my opinion what you listed is the major factor on most maps with circle syndrome.
Okay, thank you. But if it's not possible to define it, I would argue that it isn't a useful term.
|
On January 20 2012 12:45 whatthefat wrote:Show nested quote +On January 20 2012 12:42 RumbleBadger wrote:On January 20 2012 12:40 whatthefat wrote:I cannot actually describe what it is Hmm, this seems to be an ongoing theme with 'Circle Syndrome'. Can somebody please just provide a concise and coherent definition of Circle Syndrome? My current understanding is that Circle Syndrome is something like this: Maps in which most expansions (excluding main and natural) are ambiguous, i.e., approximately equidistant from both players' naturals. Is that correct? That's the most prominent case of circle syndome, as in if that occurse there is likely circle syndrome (and vice versa). But like Barrin said, CS is too complicated to really boil down to one factor, although in my opinion what you listed is the major factor on most maps with circle syndrome. Okay, thank you. But if it's not possible to define it, I would argue that it isn't a useful term.
Actually I wrote a long thread about it, with a definition. Check it out.
Noun: Circle Syndrome [sərkəl ˈsinˌdrōm] - a two player map expansion layout where each player can expand clockwise or counterclockwise around the map which results in the last base being in close proximity to the opponent
|
On January 20 2012 12:47 monitor wrote:Show nested quote +On January 20 2012 12:45 whatthefat wrote:On January 20 2012 12:42 RumbleBadger wrote:On January 20 2012 12:40 whatthefat wrote:I cannot actually describe what it is Hmm, this seems to be an ongoing theme with 'Circle Syndrome'. Can somebody please just provide a concise and coherent definition of Circle Syndrome? My current understanding is that Circle Syndrome is something like this: Maps in which most expansions (excluding main and natural) are ambiguous, i.e., approximately equidistant from both players' naturals. Is that correct? That's the most prominent case of circle syndome, as in if that occurse there is likely circle syndrome (and vice versa). But like Barrin said, CS is too complicated to really boil down to one factor, although in my opinion what you listed is the major factor on most maps with circle syndrome. Okay, thank you. But if it's not possible to define it, I would argue that it isn't a useful term. Actually I wrote a long thread about it, with a definition. Check it out. Noun: Circle Syndrome [sərkəl ˈsinˌdrōm] - a two player map expansion layout where each player can expand clockwise or counterclockwise around the map which results in the last base being in close proximity to the opponent
As I understood it from Barrin's examples when he first discussed the idea, it's not just a matter of the last base being close to the opponent, it's all the potential expansions. I'm open to being corrected on that though.
|
You are actually correct, whatthefat.
|
circle syndrome is a problem but not entirely because of bad design or bad designers. the problem manifests itself as a result of the way the game itself is played. that is, deathball syndrome, warpgate abuse, overly strong units.
when i design my maps, my mind is ablaze with visions of BW. armies moving back and fourth, here and there, just generally all out mobile trying to have true map control (as defined as what your army actually covers, rather than how far your expos reach). as a result my maps have many 'paths' or 'corridors' to mimic this.
BW maps are also full of expos all across the map, in the far reaches and corners, which can allow for any number of map designs; this is because 1) there are no warpgates - the reinforcement distance and time is larger compared to sc2 and 2) units are not as powerful or abusive, so these two things allow expos to stay alive longer and allow you to get reinforcements to try save that expo.
in sc2, mappers must have a coherent understanding of distances and timings otherwise any number of sc2-specific aspects can cause severe upsets in the base layout and balance. thusly the easiest solution to this circle syndrome, having a progressive series of bases to take to allow for quick reinforcements for defense. take for example: in bw, cross map expos are a prevalent tactic for zerg to 1) hide expos 2) force long attack path for enemy, but do you see this much in sc2? why not? there are so many ways to abuse a lone zerg expo due to the design of the game, its just not worth it in the eyes of many players.
the question is, can a paradigm shift in the way the game is played solve the above issues? or is the game's design too locked down to allow for these wild and varying maps?
|
Interesting thoughts, that this is a gameplay issue not a mapping issue. I guess I agree. I do stay quite positive in believing that the players can improve all these things, these little flaws in SC2, with whatever help from Blizzard we get. And HotS looks promising.
I remember when I watched the old GSL open 1 finals, Fruit Dealer vs Rainbow, and on the first map seeing Fruit Dealer expanding in opposite sides of the map, with good success. It's a shame this sort of play hasn't stuck around, and I wonder why. Have players gotten locked into a certain mindset or is this the best way to play the game?
I guess as much as I think the players can make a change, they sometimes need a push. I don't think players think that much about making the game better like this, but they just do what it takes to win. Because of this, sometimes we have to use maps to force them to think in new ways. Crossfire is my go-to example, a map that is pretty non-standard, but the players learned how to play it and now it's pretty decent. Maybe what we need to do, as mappers, is not say "the games on this map turned out poorly, we need to scrap the idea," but rather be more stubborn and force the players to figure the map out so we can get the game to a better place. I think this might be one of the problems with the map making community (at least foreigners,) in general. That we are too soft and won't try to get the play to change to be better.
Actually, the play seems to be slowly gravitating to be more a more like BW, and the maps as well. And I've actually noticed the similarities between the SC2 maps and the very old BW maps. This gives me confidence about the future.
|
On January 20 2012 14:59 a176 wrote:
circle syndrome is a problem but not entirely because of bad design or bad designers. the problem manifests itself as a result of the way the game itself is played. that is, deathball syndrome, warpgate abuse, overly strong units.
when i design my maps, my mind is ablaze with visions of BW. armies moving back and fourth, here and there, just generally all out mobile trying to have true map control (as defined as what your army actually covers, rather than how far your expos reach). as a result my maps have many 'paths' or 'corridors' to mimic this.
BW maps are also full of expos all across the map, in the far reaches and corners, which can allow for any number of map designs; this is because 1) there are no warpgates - the reinforcement distance and time is larger compared to sc2 and 2) units are not as powerful or abusive, so these two things allow expos to stay alive longer and allow you to get reinforcements to try save that expo.
in sc2, mappers must have a coherent understanding of distances and timings otherwise any number of sc2-specific aspects can cause severe upsets in the base layout and balance. thusly the easiest solution to this circle syndrome, having a progressive series of bases to take to allow for quick reinforcements for defense. take for example: in bw, cross map expos are a prevalent tactic for zerg to 1) hide expos 2) force long attack path for enemy, but do you see this much in sc2? why not? there are so many ways to abuse a lone zerg expo due to the design of the game, its just not worth it in the eyes of many players.
the question is, can a paradigm shift in the way the game is played solve the above issues? or is the game's design too locked down to allow for these wild and varying maps?
It will change, and in the meantime we should hold everyone's hands by making maps that are conducive to map control -- Daybreak is a decent example of a current map in popular competitive use that does this -- that gently nudge play in the direction of the paradigm shift you envision.
SC2 is in a phase where there are lots of dangerous timings. As standard play is honed, all the windows will get smaller and smaller and players will be able to handle themselves with a high degree of security (actual and virtual map control -- safety) without overscouting, which is what you need to do now. Right now you have to have nearly perfect information of the enemy build or the enemy army position(s) to be safe. A partial mix is usually pretty unsafe.
This will change for two reasons. First, players will learn what they must do to hold off which attacks. Second, attackers will begin to respect that the chance of committing serious damage is low, and there will be less flailing attacks.
When you watch BW games, you see a lot of situations where the aggressor immediately pulls back after probing the enemy resistance, even if the defender looks weak. This is because the aggressor isn't attacking, he is just pressuring to obtain information and create space for himself. This sort of play is only exhibited in a very basic way in current SC2, but it's improving every day.
The paradigm shift you envision will open up possibilities for maps, but not for a while, like years--also there are expansions coming out that throw all the standard knowledge out the window.
edit: Also, there is a limit to the amount of acceptable circle syndrome, regardless, to use the en vogue terminology.
<3 barrin
|
A lot of the explanations are unclear to me, but I (like Barrin, like others) have an intuitive grasp of the problems Barrin is pointing out. The map-width stuff greatly confused me, but I have my own ways of thinking about maps which seem to roughly line up with Barrin's conclusions. When considering expansion paths, I often think in terms of "front lines" and "harrassment paths"; in the case of maps with "circle syndrome," each expansion greatly increases both front lines and harrassment paths, making expoing on these more difficult with every expo taken.
Anyway, though some of this is puzzling, I am very happy to see an anlysis like this. If the mapping community can progress to more concrete terms for describing the way a map functions and potential problems, maps will definitely improve. I look forward to seeing Barrin expand and clarify his thoughts on resource "gravity," map control, and his template.
|
Although I am very inclined to answer direct questions, in a general sense I feel that I have explained it well enough. Perhaps there are important parts somewhat hidden that should be bolded. I will try to come up with a better way to explain this gravity thing (I truly feel it is an excellent analogy anyway).
For the most part, I was relying on everyone's "intuitive grasp" to fill in the gaps (many of us know what happens when I try to fill in all the gaps >.<). I am also relying on time; I already can tell you guys are getting it, it just needs some time to permeate.
I definitely recommend reading multiple times.
|
btw
On January 20 2012 15:15 Gfire wrote:
I guess as much as I think the players can make a change, they sometimes need a push.
boom, headshot
circle syndrome hinges upon this being true
|
First of all thank you so much for writing this incredibly informative article.
After reading it I did an analysis of my own map:
![[image loading]](http://i.imgur.com/lX7UC.jpg)
as you can see the length is a bit longer than the width by what feel like just the right amount in game. But when I played and got feedback from GM players I had one player mention that once a terran moves out (in zvt) it just feels like their army disappears. This was probably due to the lack of XWTs on the map. I think the length / width can be expanded to include the inclusion of XWTs since they effectively lengthen the map by proving an earlier warning for attacks.
The more interesting analysis comes from the gravity of the map's expansions. From taking the standard third the corner base feels like a natural progression if you don't factor in the natural:
![[image loading]](http://i.imgur.com/4Gtzl.jpg)
But once you do the centre base feels like a viable option for the races that are better at holding ground and are less mobile:
![[image loading]](http://i.imgur.com/dtyDw.jpg)
The red and blue lines represent how far an army must move to deal with attacks on the extremities of the bases. Comparing these distances to the distances that the opponent has to travel to attack the extremities might be a more accurate measurement than just length and width when discussing how difficult it is to hold expansions although it takes significantly more time.
![[image loading]](http://i.imgur.com/PmV5q.jpg)
In this picture the pink is the defender's path to be able to deal with attack on both sides while red represents the attacker's path.
![[image loading]](http://i.imgur.com/0APXk.jpg)
Notice how taking the central base changes these proportions completely. The defender has to move much less to deal with the attacker's threats except the central base has to take the brunt of an attack for a little longer since the attacker has a relatively faster way to attack that expansion than the other ones. This would explain why terrans and protoss favor the central expansions: lack of mobility and stronger static defenses. While Zergs tend to favor the corner expansion: more mobility and weaker static defenses.
Overall I don't feel it suffers from Circle Syndrome due to the somewhat easy to hold third but I'm having a hard time deciding on this one since there are so many variables that are up in the air due to what races are playing what races.
Of course the analysis can be taken much further when you add factors such as destructible rocks and XWTs. Both lengthen the length (make it easier for defenders) in most cases but in must be taken on a case by case basis. I think the mapping community is coming close to being able to move beyond simple analysis of size and distance thanks to this post.
Also Barrin please bold the "Back to Circle Syndrome" part to hell and make it bright pink before the map making community becomes incredibly dismissive like this:
http://www.teamliquid.net/forum/viewmessage.php?topic_id=305363#2
Sure it's easy to say circle syndrome but now a days that's all I see when I run into a new map. People say circle syndrome and dismiss it without even comparing the merits of the map.
|
So if you understand Circle Syndrome (especially the part about "gravity"), then an easy way to understand Breadth of Gameplay in SC2 is simply the following:
I think each base is too "heavy" with 8m2g.
|
|
|
|
|
|
EPT![[image loading]](http://i.imgur.com/W50nD.jpg)
![[image loading]](http://i.imgur.com/TxQbt.jpg)
![[image loading]](http://i.imgur.com/YiR5s.jpg)
![[image loading]](http://i.imgur.com/gIz38.jpg)
![[image loading]](http://i.imgur.com/BpIn8.jpg)
![[image loading]](http://i.imgur.com/w06L0.jpg)
![[image loading]](http://i.imgur.com/sLpmA.jpg)
![[image loading]](http://i.imgur.com/7PaJf.jpg)
![[image loading]](http://i.imgur.com/kcNqK.jpg)
![[image loading]](http://i.imgur.com/L4pyy.jpg)
