|
Much of the current balance debate has focused on maps, and specifically map size. Increasing map size has been proposed as a potential fix for several perceived issues:
(1) Increasing rush distances will help stabilize the early game.
(2) Providing more securable expansions will allow Zerg to more safely acquire a third base (something that’s very difficult on some of the current map pool).
(3) Bigger maps should promote a more macro-oriented game, and less 2-base all-in style plays.
While I agree increasing map size and improving map layout could help with some of these issues, there are clues that the problems could be arising at a more fundamental level.
Mineral collection and the need to expand
A superb recent analysis by LaLuSh demonstrated that the basic mechanics of resource collection are fundamentally different between SC2 and BW. You can read the full thread here: http://www.teamliquid.net/forum/viewmessage.php?topic_id=191702
One of the key figures is repeated here:
+ Show Spoiler +
The key points to note are that:
I. When the number of workers is less than or equal to the number of mineral patches (n), mining rate is essentially linearly proportional to number of workers in both SC2 and BW.
II. In BW, this linear growth begins to taper off once the number of workers exceeds n. By contrast, in SC2 each mineral patch can support 2 workers without any interference between workers, meaning linear growth is sustained until the number of workers exceeds 2n.
III. In BW, the benefit of added workers does not saturate until there are over 3n workers. By contrast, in SC2 the benefit of going from 2n workers to 3n workers is small, and complete saturation is essentially achieved before reaching 3n workers.
The ramifications of points II and III are far reaching. First, in BW there is more benefit to be gained from taking a first expansion before reaching 2n workers. In SC2 limiting returns don’t kick in until much later, so continuing to pump workers off one base until exceeding 2n workers is much stronger than in BW.
Second, because each expansion can fully support 2n workers without mineral losses in SC2, there is little incentive to expand beyond 3 bases (3x2x8 = 48 workers on minerals, + 3x2x3 = 18 workers on gas). By contrast, in BW 2 bases with n workers each mined much more efficiently than 1 base with 2n workers, meaning more expansions were favored.
An obvious objection to the above analysis is that it ignores the role of gas. This is especially important to Zerg, who often take 5 bases in the late game purely to meet gas costs. However, if a Protoss or Terran secures 3 fully saturated bases, they have little incentive to expand again. This theoretical prediction is borne out in practice; when a Zerg on 5 or more bases faces a Protoss or Terran on 3 bases, the game is usually fairly balanced from a macro perspective.
![[image loading]](http://imgur.com/bLVCh.jpg)
As LaLuSh concluded:
“In this light, it seems slightly foolish of the community to expect the metagame of SC2 to eventually evolve into something resembling that of Broodwar. You will likely never see players opt for as early of a third base as used to be the norm in Broodwar. Rather players will tend to be bottlenecked on 2 bases for longer (especially on Blizzard-sized maps).”
How can map designers address this?
Map designers are in a relatively privileged position when it comes to dictating game dynamics and balance. Players can lobby for unit changes, ability changes, and race mechanic changes, but they are ultimately at the behest of Blizzard. By contrast, well designed maps are readily taken up by the community, and provide an excellent way of potentially changing the status quo. The current standard in map design is 8 mineral patches and 2 vespene geysers per base/expansion. Exceptions to this template are relatively rare. But does it have to be this way? What would the game be like if Blizzard had instead decided on 7 or 9 mineral patches per base?
![[image loading]](http://imgur.com/wpALg.jpg)
To get some feel for this, imagine the following 2 extreme scenarios:
(1) Suppose all bases/expansions had 2 mineral patches. You would be saturated immediately, so aside from gas, there is no point making more workers until you expand. Expanding would almost immediately double your mineral income.
(2) Suppose all bases/expansions had 30 mineral patches. You would not be saturated until far into the mid-game. Expanding would yield almost no additional mineral gains until you had over 60 workers.
From this thought experiment, I would expect maps with less mineral patches per base to generally encourage earlier expansions. Thinking about this led me to consider a number of potential approaches that map designers could explore.
A. Changing the total resource count on mineral patches
+ Show Spoiler +A relatively simple change would be decreasing the amount of minerals held in each patch. With bases mining out slightly more quickly, there would be a clear incentive to expand more frequently. However, this would not address the saturation issue, i.e., players would not actually need more operational bases at any point in time, they would just need to expand more frequently.
B. Changing the number of mineral patches
+ Show Spoiler +
The simplest change would be to reduce the number of mineral patches to 7 per base/expansion. I’m hopeful that a simple change like this could have a significant effect on overall game dynamics. Alternatively, main bases could have less mineral patches than expansions, although this would obviously lose any effect on gameplay once the main is mined out. Furthermore, if the natural was too much better than the main, there would be the possibility of Terran’s relocating at the beginning of the game.
C. Changing the distribution of mineral patches
+ Show Spoiler +One important difference between BW and SC2 is mineral patch placement. In BW, minerals were more often placed in a straight line, whereas in SC2 they are typically arranged in a concave. Mineral patches that are closer to the base are effectively higher yield (until all patches are fully saturated), and players take advantage of this by mining from those patches first. The fact that this effect was more extreme in BW also promoted quicker expansions to take advantage of the ‘high yield’ patches at a new location. It may thus also be worth investigating the effects of different mineral layouts in SC2. ![[image loading]](http://imgur.com/3ZeEn.jpg) Creating non-uniform mineral fields by mixing gold and blue patches would be one way of achieving extreme differences in patch yield. However, one potential balance flaw in any resource layout where some patches give a much higher yield than others is the ability of Terrans to exploit this with the MULE.
Summary
I hope this provides some fresh ideas for the map making community. The effects of the current 8 mineral patch template have not yet been critically tested, and improved game balance may be achieved through changes in this aspect of map design.
|
There are two things I'm going to immediately comment on, and I'll probably make a later statement once I think a little bit more about this.
1. This was discussed to death early on.
2. The general consensus was, the game is still so young, people don't even understand their races yet, lets not screw with mineral layouts until they do.
3. You are incorrect when you state that mining rate in SC2 is essentially linear, and your graphs proves it. :/
|
On February 11 2011 07:31 iGrok wrote:
There are two things I'm going to immediately comment on, and I'll probably make a later statement once I think a little bit more about this.
1. This was discussed to death early on.
2. The general consensus was, the game is still so young, people don't even understand their races yet, lets not screw with mineral layouts until they do.
Sure, but I've seen no testing of it since game release, and given the current state of the game I think it is now an important issue to reconsider. Not to say that the game won't continue to evolve, but it's no longer a matter of theorycrafting, we now have some clear idea of game dynamics.
3. You are incorrect when you state that mining rate in SC2 is essentially linear, and your graphs proves it. :/
I said it is approximately linear in SC2 up to 2n = 16 workers, which is clearly true from the graph. This is also confirmed by LaLuSh's graph showing minerals mined per worker, which is flat up to 16 workers.
|
On February 11 2011 07:40 whatthefat wrote:Show nested quote +On February 11 2011 07:31 iGrok wrote:
There are two things I'm going to immediately comment on, and I'll probably make a later statement once I think a little bit more about this.
1. This was discussed to death early on.
2. The general consensus was, the game is still so young, people don't even understand their races yet, lets not screw with mineral layouts until they do. Sure, but I've seen no testing of it since game release, and given the current state of the game I think it is now an important issue to reconsider. Not to say that the game won't continue to evolve, but it's no longer a matter of theorycrafting, we now have some clear idea of game dynamics. Show nested quote +
3. You are incorrect when you state that mining rate in SC2 is essentially linear, and your graphs proves it. :/
I said it is approximately linear in SC2 up to 2n = 16 workers, which is clearly true from the graph. This is also confirmed by LaLuSh's graph showing minerals mined per worker, which is flat up to 16 workers.
I had actually edited Lost Temple and made mains have 7 minerals and natural 7 minerals. Would like to experiment more on this.
|
I find it funny that adding 26th probe in BW makes protoss mine LESS minerals than stopping probe production at 25-this makes no sense 
|
i don't think mineral patches are an issue as they are at the moment
i'll go as far as to say having 2 gases per expansion is what limits the expanding opportunities and thus leads to significantly less macro-oriented positional play, and has done so since the get-go. think of toss in bw, you'd often take extra bases just for gas to feed your gas-hungry army. and there's good reasoning to support this
it's usually the extra gas required for teching and later game armies that provides most of the incentive to expand, and having less gas from the get-go slows down the rate at which a game develops, thus encouraging a more cerebral approach to the strategy of the game (aka leads to more depth). sc 'suffers' from all this oftentimes simplistic 2base play because it can be so strong, and it can be so strong because there's all this gas to be had in a relatively confined space that requires little maneuvering to defend
the new maps starting to be used in tourneys address the second part of the issue by being larger, with more expansions and more strategical points to control, but medium-sized maps (medium, as in bigger than say dq, but not as big as say terminus re) should be able to achieve the same result in encouraging strategical play with good scouting and positioning over compositional play with good timings by reducing the number of gases available for expansions
ofcourse, some cheese-favorable maps never hurt, and there's all sorts of fun that could be done with doodad blocks and such to encourage various specifically desired types of play. but for the standard bw-ish dynamic that people tend to appreciate, this could be given a tryout on a larger scale
|
@monitor
Interesting. Did you do any serious play testing?
@Dicetoss
I'd say that's down to measurement error. Although knowing BW worker AI, I couldn't say for sure!
@anatem
That's an interesting perspective. Do you think the game would be balanced with 1 gas per base? My worry is that there would just be even more of the mineral heavy style plays from terran (i.e., marines). Also, as a zerg it's almost impossible to do anything in the midgame without 3 geysers.
|
I assumed they made 2 geysers from BW to SC2 was to prevent the gas stealing technique (or at least they could only block half your gas income, only for a short while)
|
The reason there are two geysers is that units are more gas intensive.
Still thinking about this btw
|
I just made a large post in Lalush's thread. It discusses at length the options introduced by whatthefat in his diagrams. I took my time to post here,, because I wanted to be sure of my analysis. After much thought, I believe that convex mineral lines are the solution to the "problem".
Now, SC2 is a different game, period. Shaping it towards BW based on blind principle is flat wrong. Do we need to augment the expansion incentives with map design? I am inclined to say yes. It seems inconceivable that maps would ever disavow the consecrated arrangement of main + natural. If the "functional macro ceiling" is 3 mining bases, this means that only one other expansion beyond the default available resources comes into play before the late game. That seems shallow, and it certainly under-utilizes the potential of maps.
I want to try playing some games on modified ladder maps. If anyone would like to join in, please hang out in channel MotM on NA. Otherwise you can PM me. I also plan to do some experiments. I'll post the results in Lalush's thread.
|
I can't see convex Mineral Lines ever becoming popular.
If you've noticed, any time mineral lines look a bit "off", players immediately comment on that. Imagine what they'd say to convex lines.
I'm not saying that it wouldn't fix the problem - I don't know one way or another - but it won't solve the problem because players won't take to it.
|
On February 15 2011 07:54 iGrok wrote:
I can't see convex Mineral Lines ever becoming popular.
If you've noticed, any time mineral lines look a bit "off", players immediately comment on that. Imagine what they'd say to convex lines.
I'm not saying that it wouldn't fix the problem - I don't know one way or another - but it won't solve the problem because players won't take to it.
if by "off" you mean that on some maps the mineral/gas layouts are actually different depending on where you spawn, thats unfair.
if everyone is playing with the same setup people wouldnt be making that complaint.
one quick comment about mineral field layouts. although it would be a smaller issue, terrans ability to drop mules on close patches pushes favour to them even if the minerals themselves arent high yield. if thats the goal as the meta game changes thats fine its just something to keep in mind.
one way to possibly encourage faster expanding, via quicker saturation is to move the minerals closer to the base, but to lower the amount you get per trip. currently patches are 4/5 squares away and give 5 minerals. if patches were 3/4 squares away but only gave 4 per trip (same as gas, which just feels "neat" to me) income would be similar but saturation would be sooner as even prior to 2n workers there would be some disruption on the 3 square patches
|
@EatThePath
Great, I'll be happy to participate, and I look forward to hearing the results of your own testing.
At this point, I'm leaning towards changing the number of patches as the simplest way of promoting expansions. I can't seem to find any way around changes in distribution making the MULE ridiculously imba.
|
I think it isn't the minerals that is the biggest problem. The gas seems to be more important. There should only be one geyser per base and more gas in each geyser. People would then take more bases for more gas and build more high tech units. With the current amount of gas in each geyser, it depletes to quickly.
|
On February 15 2011 08:24 turdburgler wrote:Show nested quote +On February 15 2011 07:54 iGrok wrote:
I can't see convex Mineral Lines ever becoming popular.
If you've noticed, any time mineral lines look a bit "off", players immediately comment on that. Imagine what they'd say to convex lines.
I'm not saying that it wouldn't fix the problem - I don't know one way or another - but it won't solve the problem because players won't take to it.
if by "off" you mean that on some maps the mineral/gas layouts are actually different depending on where you spawn, thats unfair. if everyone is playing with the same setup people wouldnt be making that complaint.
Using contextual clues, you should be able to determine that that is not what I meant. Whenever the mineral lines look a bit "off" (for you I'll exchange the word "different from normal ones" so that you don't get confused), players immediately make a comment. Top mappers realize this.
When you play with top players on custom maps, and talk to them about it, I'll listen to your informed opinion.
one quick comment about mineral field layouts. although it would be a smaller issue, terrans ability to drop mules on close patches pushes favour to them even if the minerals themselves arent high yield. if thats the goal as the meta game changes thats fine its just something to keep in mind.
This comment is hard to follow, but I'll assume you mean that Terrans favor expansions with more close mineral patches. This is a very minor detail, and I can guarantee that no pro player will ever take a non-standard expansion purely because there is one more close mineral patch. Also, no pro will ever choose a blue expansion with one or two closer patches over a HY, (assuming all else is equal).
one way to possibly encourage faster expanding, via quicker saturation is to move the minerals closer to the base, but to lower the amount you get per trip. currently patches are 4/5 squares away and give 5 minerals. if patches were 3/4 squares away but only gave 4 per trip (same as gas, which just feels "neat" to me) income would be similar but saturation would be sooner as even prior to 2n workers there would be some disruption on the 3 square patches
Minerals will not be moved closer. This is a stupid suggestion that none of us can do anything about anyways.
The only suggestion in this thread that has any chance of actually being implemented is to reduce the patches/base.
|
On February 15 2011 08:24 turdburgler wrote:Show nested quote +On February 15 2011 07:54 iGrok wrote:
I can't see convex Mineral Lines ever becoming popular.
If you've noticed, any time mineral lines look a bit "off", players immediately comment on that. Imagine what they'd say to convex lines.
I'm not saying that it wouldn't fix the problem - I don't know one way or another - but it won't solve the problem because players won't take to it.
if by "off" you mean that on some maps the mineral/gas layouts are actually different depending on where you spawn, thats unfair. if everyone is playing with the same setup people wouldnt be making that complaint. one quick comment about mineral field layouts. although it would be a smaller issue, terrans ability to drop mules on close patches pushes favour to them even if the minerals themselves arent high yield. if thats the goal as the meta game changes thats fine its just something to keep in mind. one way to possibly encourage faster expanding, via quicker saturation is to move the minerals closer to the base, but to lower the amount you get per trip. currently patches are 4/5 squares away and give 5 minerals. if patches were 3/4 squares away but only gave 4 per trip (same as gas, which just feels "neat" to me) income would be similar but saturation would be sooner as even prior to 2n workers there would be some disruption on the 3 square patches
Mineral patches are 3 or 4 tiles away from the command center.
Forgive me for being frank, but you're talking out of your ass right now.
EDIT: Just to comment on the actual context of the thread so I'm not completely flaming here, I have to disagree with the OP. These kind of discussions were done to death in the beta and as far as I am concerned, I will be sticking to what is considered relatively standard as that is how the game was designed. There are things I am willing to experiment with, but without concrete evidence towards something like a convex mineral line I'll find other ways to make my map unique.
|
Sadly I agree with you guys, there's no way you could get people to buy a convex mineral line, right now. I have a feeling that some of Lalush's hypothesis will be borne out in the upcoming GSL season with the large maps. Perhaps at that point people will be willing to go off the beaten path.
It's been brought up a lot that this was already exhaustively discussed during beta. I think it's obvious we've come a long way since then. I know it must seem repetitive, but the comprehension of the relevant factors has surely increased, meriting a fresh look (if you care to). If not, I am sure SC2 will be just fine with normal or 7 patch expos.
Briefly, I have made some effective mineral placements that don't look awkward, just larger than usual. I don't think they'd put anyone off once they played it once or twice. I have to make an epic thread to present it all with pictures, it will be useless without a comprehensive treatment. So, whenever that happens...
Oh, and about mules. The layouts I'm using only have an aggregate throttle on the mining rate over the 16-24 range, with a slight effect at 12-16. Considering the boost the first mule gives you on normal patches anyway, that seems fine. Subsequent mules will be slightly stronger relative to the mining rates compared to standard mineral layouts. I think this small "balance shift" can be absorbed by the game. Mules are so imba anyway, right? ; )~
|
ETP, can you post a teaser example pic please?
|
Yes, just a sec igrok
[edit]
These are two of the basic shapes I have thoroughly tested so far. I have also tested an inverse of the "pocket", but it's not pictured here. I have tested more organic arrangements as well... but this is just for starters eh.
Pocket
![[image loading]](http://img692.imageshack.us/img692/6033/pocketz.jpg)
An example standard expo in the same configuration, for comparison
![[image loading]](http://img695.imageshack.us/img695/3484/pocketstandardcompare.jpg)
Slash
![[image loading]](http://img262.imageshack.us/img262/7581/slashr.jpg)
An example standard expo in the same configuration, for comparison
![[image loading]](http://img337.imageshack.us/img337/6874/slashstandardcompare.jpg)
I should revise the numbers I mentioned above. These bases start having a different mining profile above 10 workers, noticeable at 14-16. From 16-24, they actually "catch up" to a standard expo because any of these patches is saturated or near-saturated with 3 workers. Hence, the mule effect is only seen during the 1st or 2nd mule.
A mineral patch 6 squares away is ever so slightly undersaturated by 3 workers. So these examples are functionally equivalent to standard mineral lines assuming 24 workers.
A mineral patch 3 squares away is actually significantly oversaturated by 3 workers, which is how these arrangements "catch up".
In fact, the pocket and the slash are remarkably fast at settling workers into triplets on each patch, unlike all the standard expos I've tested on, which have significant bouncing indefinitely. This is why there are arguments about 24 is not saturated, 27 or 30 is, etc. For the pocket and the slash, 24 is saturated, period. Adding workers causes bouncing and can only slightly decrease mining sometimes. More "organic" arrangements cause bouncing, and those are what you would want in a real map, to raise the saturation cap to 30 and give slower mining gain per worker.
My next step will be testing even deeper pockets and more extreme convex shapes with distant patches greater than 6 patches away, and varying ratios of close / far patches. The goal is to create a situation where it's advantageous to expand and put 12 workers on each base as opposed to 24 at home. This is true for the pocket and slash, but I want to test the limits of feasible mineral distance. I think irregular mineral lines that cause bouncing are the best way to raise the saturation cap above 24.
|
Here are some pictures of Lost Temple to give you a sense of the shape and appearance on a real map.
+ Show Spoiler +LT 6 oclock main ![[image loading]](http://img151.imageshack.us/img151/6946/ltpocket.jpg) ![[image loading]](http://img543.imageshack.us/img543/1475/ltpocketstandardcompare.jpg) LT 6 oclock natural ![[image loading]](http://img522.imageshack.us/img522/7399/ltslash.jpg) ![[image loading]](http://img143.imageshack.us/img143/7131/ltslashstandardcompare.jpg) This is just for reference. Slash is pretty ugly arranged like that. Pocket looks okay though.
|
|
|
|
|
|
EPT![[image loading]](http://imgur.com/vG24W.jpg)
