|
Blizz listens
Creep will now spread evenly in all directions. http://us.battle.net/sc2/en/forum/topic/1213111662
Edit: Please +1 my post on the battle.net forums if you think this should be fixed. Hint: it should be fixed. (((Thread is now highly rated thanks!!! I don't think it hurts to keep upping it though ^_^))) http://us.battle.net/sc2/en/forum/topic/981608428#1
The Creep Conundrum
So yeah it's not really a "conundrum". It's just rather catchy IMO ^_^
Creep. We all know what it is. Most of us probably think we know everything about it. But I bet most of you don't yet know what I'm going to show in this post. This might seem small; and well, it is small, but not insignificant to "perfectionist" mapmakers like myself.
So WTF am I talking about? Simple: For Hatcheries(Lairs/Hives), Creep Tumors, and Nydus worms (NOTE: NOT OVERLORDS): More creep spawns on the left(<west<) and bottom(VsouthV) than on the top(^north^) and right(>east>).
BTW The following pictures were taken in the Galaxy Editor, but rest assured that the creep generation is the same in-game.
Do you see it? Me neither (at first). It's hard to see. Let's take a closer look:
Notice how the creep tumor is actually one space closer to the northwest than it is to the southeast (in relation to red lines), HOWEVER, the creep itself is aligned rather perfectly (red lines).
In the above picture, I moved the creep tumor one space left(<west<) and on space down(VsouthV). Take another look at the creep in relation to the red lines.
So what does it take to make the creep perfect?
This picture illustrates the problem quite well.
Here's some proof that the same thing happens to both Hatcheries and Nydus Worms (look closely): + Show Spoiler [Hatchery] + + Show Spoiler [Nydus Worm] +(Woops... Ignore that unplaced nydus worm on the bottom left... too lazy to fix)
Again, remember that Overlord Creep does not suffer this same fate. I don't have pictures for this... I guess you'll have to take my word for it.
On Xel'Naga Caverns there is a slight ZvZ imbalanced based on starting positions. Zelniq explains it here in this thread. Basically, optimally placed spine crawlers on the top position cover a little more of the top of your ramp than optimally placed spine crawlers on the bottom position. I believe this is due to both the "creep conundrum" and the nature of the way terrain itself is kind-of generated into 2x2 boxes.
Just like on Xel'Naga Caverns, mapmakers should at least be aware of the way creep generates itself if they truly care about carefully placing the bases.
That is all. <3 - Barrin
edit:
+ Show Spoiler +NO, not really bad job from Blizzard (we love Blizzard, they gave us Starcraft), its just that there are limits with having a grid system instead of being able to place buildings absolutely everywhere. It's a question of where is the center of the creep-circle.
The problem is that to fix it they're either gonna have to mess with the diameter of creep generated by all buildings (enlarge/shrink so that the building is the same distance from all edges of the circle, so trim off a bit of the bottom left part) or they're going to keep the diameter the same, but make the creep generate so that the circle is equidistant from the centre.
Its like a pylon. Imagine if pylon power originated 1 hex to the west and 1 hex to the south, how do you solve the problem? = same thing I think increasing the range that creep spreads can fix it. Look at the ranges of the placement indicator (funky green thing) very closely. + Show Spoiler +(Btw the white line represents the max... it is indeed uniform... see spoiler in the above experiment example)
What I see is that:
- There is still a little bit of room southwest(<<vv) where if the range that the creep spreads from the origin was increased, this side of the creep would not be changed
- There is an even smaller margin (?) where if the range that the creep spreads was increased, the northeast(>>^^) side would become uniform with the other sides in terms of creep spreading.
--------------
+ Show Spoiler +the main thing is that mappers need to take into consideration how creep spreads, that's the point of this thread. Maps like Scrap Station and XC are a problem for Zerg because the positions aren't equal for creep spreading. The way creep spreads isn't the problem or where the location of the tumor is, the problem is that some maps have favorable spawn locations for optimal creep spread where others do not. Mapping in this editor isn't perfect, but with enough play testing I think some maps out there could be close to perfection. Yeah! You're right... I can't believe I didn't think of this... Experiment examples + Show Spoiler +(keep in mind that this is assuming the creep tumor is the only source of creep to place the next creep tumor on... the actual potential range of the next creep tumor is actually uniform assuming the tumor is surrounded by even more creep than it generates) : Straight <<<< Can be placed 10 spaces away. vvvv Can be placed 10 spaces away. >>>> Can be placed 9 spaces away. ^^^^ Can be placed 9 spaces away
Diagonal <<vv Can be placed 7 spaces away. >>^^ Can be placed 6 spaces away. <<^^ Can be placed 6.5 spaces away. >>vv Can be placed 6.5 spaces away. (6.5 means that it's actually 6 if you go for a pure straight line diagonal. In the <<vv and >>^^ you can't go any further, but in the 6.5's you can actually go non-diagonally to get a tiny bit farther than just 6)
very late edit: uhh 14%? wtf? lol. How did I come up with that... It made sense then I guess... maybe I just don't see it anymore.
Conclusion: Yes, it is safe to say that spreading creep directly southwest is about 14% [spoiler]100 / 7 = 14.2857143[/spoiler] faster than spreading it northeast, assuming you wait until the creep is done spreading itself (it's not really 14% if you keep spreading the creep as soon as the creep tumor cooldown is finished, which is not always what you should be doing because it can be dangerous).
--------------
Oh yeah. Units can be slightly off creep and still get the speed bonus. I was thinking that maybe (just maybe) they got the bonus easier depending on which side of them the creep was on. NOPE. Bonus speed to units is not related to the "creep conundrum".
--------------
edit#2:
Since I'm a map maker I will post here.
You can fix this issue by going into footprints and changing the creep tumor footprint. For some bizzare reason blizzard chose an even number instead of an odd number for the footprint size for creep tumors, which causes the problem.
In the map I made, Hero Attack, the creep tumors do not have this issue because I changed the footprint size. Genius man is genius.
|
Post this on battle.net forums, blizzard have to see this. This problem shouldn't even exist, really bad job from blizzard here
|
NO, not really bad job from Blizzard (we love Blizzard, they gave us Starcraft), its just that there are limits with having a grid system instead of being able to place buildings absolutely everywhere. It's a question of where is the center of the creep-circle.
The problem is that to fix it they're either gonna have to mess with the diameter of creep generated by all buildings (enlarge/shrink so that the building is the same distance from all edges of the circle, so trim off a bit of the bottom left part) or they're going to keep the diameter the same, but make the creep generate so that the circle is equidistant from the centre.
Its like a pylon. Imagine if pylon power originated 1 hex to the west and 1 hex to the south, how do you solve the problem? = same thing
|
On November 06 2010 07:57 bITt.mAN wrote: NO, not really bad job from Blizzard (we love Blizzard, they gave us Starcraft), its just that there are limits with having a grid system instead of being able to place buildings absolutely everywhere. It's a question of where is the center of the creep-circle.
The problem is that to fix it they're either gonna have to mess with the diameter of creep generated by all buildings (enlarge/shrink so that the building is the same distance from all edges of the circle, so trim off a bit of the bottom left part) or they're going to keep the diameter the same, but make the creep generate so that the circle is equidistant from the centre.
Its like a pylon. Imagine if pylon power originated 1 hex to the west and 1 hex to the south, how do you solve the problem? = same thing While this is pretty much /thread material here, Blizzard should still be made aware of the issue. Nothing wrong with slapping this up on the B.Net forums and seeing what they can do to resolve it.
|
main thing is needing 2 tumours for one of the spawn location on scrap station to get to the bottom of the ramp, iirc its the bottom location.
|
I have like a 5 min. break from a serious crunch at work and I jump on TL, scan the Custom Maps sidebar and see a thread titled "The Creep Conundrum" and I thought I'd find a wacky UMS map about playing as non-Zerg and trying to fight off waves of creep, or some ludicrous puzzle game based on creep.
So first of all, I'd like to officially request from a the community a creep-focused UMS map!
Okay, to the thread at hand: Barrin is definitely a perfectionist as he already stated--and if you follow the link to his VARS map symmetry thread I think funcmode's reply is relevant to this discussion also. To paraphrase: there may be a bazillion minor differences between spawns on a map, but devoid of major advantages it is generally OK because the players' actions have much, much more impact on the outcome of the game. I think the creep offset, too, doesn't break any particular spawn layouts.
That said, it is good to know about this issue and so any good map maker should at least think about it. I didn't know about the offset from the creep generator, but as a habit I place a hatchery at every base on my maps to see where the usable creep is, what parts of the base are out of creep sight, etc.
And we know at least two key issues that should be checked that the creep offset can affect: make sure the number of creep tumors to connect the main to the natural is the same for every spawn, and that the opportunity to defend with crawlers is the same (minerals, main chokes, all of that).
The problem with creep tumors is that sometimes the range of the tumor lands in an unpathable area, like the map cells where a cliff transitions up or down, and so even though the distance between the main and nat is almost the same, one square more or less of creep can change the number of tumors needed to "get over" the cliff. So keep this in mind!
@Barrin: you are ridiculous, nice find
|
the main thing is that mappers need to take into consideration how creep spreads, that's the point of this thread. Maps like Scrap Station and XC are a problem for Zerg because the positions aren't equal for creep spreading. The way creep spreads isn't the problem or where the location of the tumor is, the problem is that some maps have favorable spawn locations for optimal creep spread where others do not. Mapping in this editor isn't perfect, but with enough play testing I think some maps out there could be close to perfection.
|
Portions of this post will be put at the end of the OP.
Post this on battle.net forums, blizzard have to see this. This problem shouldn't even exist, really bad job from blizzard here Yes sir!
+ Show Spoiler +NO, not really bad job from Blizzard (we love Blizzard, they gave us Starcraft), its just that there are limits with having a grid system instead of being able to place buildings absolutely everywhere. It's a question of where is the center of the creep-circle.
The problem is that to fix it they're either gonna have to mess with the diameter of creep generated by all buildings (enlarge/shrink so that the building is the same distance from all edges of the circle, so trim off a bit of the bottom left part) or they're going to keep the diameter the same, but make the creep generate so that the circle is equidistant from the centre.
Its like a pylon. Imagine if pylon power originated 1 hex to the west and 1 hex to the south, how do you solve the problem? = same thing I had to read this twice to believe it. I think it can still be fixed; I'll get back to that.
main thing is needing 2 tumours for one of the spawn location on scrap station to get to the bottom of the ramp, iirc its the bottom location. Hmmph
- Blizzard Maps are not the only maps that matter.
- Even if they were, the map pool will continue to change over time.
- I thought it was the top location? I forget.. so rusty.
+ Show Spoiler +To paraphrase: there may be a bazillion minor differences between spawns on a map, but devoid of major advantages it is generally OK because the players' actions have much, much more impact on the outcome of the game. I think the creep offset, too, doesn't break any particular spawn layouts. 99.9% agreed.
+ Show Spoiler +The problem with creep tumors is that sometimes the range of the tumor lands in an unpathable area, like the map cells where a cliff transitions up or down, and so even though the distance between the main and nat is almost the same, one square more or less of creep can change the number of tumors needed to "get over" the cliff. So keep this in mind! Nice nice, very trixy, very trixy.
@Barrin: you are ridiculous, nice find *bananagrin*
+ Show Spoiler +the main thing is that mappers need to take into consideration how creep spreads, that's the point of this thread. Maps like Scrap Station and XC are a problem for Zerg because the positions aren't equal for creep spreading. The way creep spreads isn't the problem or where the location of the tumor is, the problem is that some maps have favorable spawn locations for optimal creep spread where others do not. Mapping in this editor isn't perfect, but with enough play testing I think some maps out there could be close to perfection. Yeah! You're right... I can't believe I didn't think of this... Experiment examples + Show Spoiler +(keep in mind that this is assuming the creep tumor is the only source of creep to place the next creep tumor on... the actual potential range of the next creep tumor is actually uniform assuming the tumor is surrounded by even more creep than it generates) : Straight <<<< Can be placed 10 spaces away. vvvv Can be placed 10 spaces away. >>>> Can be placed 9 spaces away. ^^^^ Can be placed 9 spaces away
Diagonal <<vv Can be placed 7 spaces away. >>^^ Can be placed 6 spaces away. <<^^ Can be placed 6.5 spaces away. >>vv Can be placed 6.5 spaces away. (6.5 means that it's actually 6 if you go for a pure straight line diagonal. In the <<vv and >>^^ you can't go any further, but in the 6.5's you can actually go non-diagonally to get a tiny bit farther than just 6)
Conclusion: Yes, it is safe to say that spreading creep directly southwest is about 14% + Show Spoiler + faster than spreading it northeast, assuming you wait until the creep is done spreading itself (it's not really 14% if you keep spreading the creep as soon as the creep tumor cooldown is finished, which is not always what you should be doing because it can be dangerous).
-------------
Back to what bITt.mAN said: I think increasing the range that creep spreads can fix it. Look at the ranges of the placement indicator (funky green thing) very closely. + Show Spoiler +(Btw the white line represents the max... it is indeed uniform... see spoiler in the above experiment example)
What I see is that:
- There is still a little bit of room southwest(<<vv) where if the range that the creep spreads from the origin was increased, this side of the creep would not be changed
- There is an even smaller margin (?) where if the range that the creep spreads was increased, the northeast(>>^^) side would become uniform with the other sides in terms of creep spreading.
Hot damn... if I wasn't the person who made this thread I would have to read what I just wrote like 10x before I understood it. Sorry if it's confusing, I can try to explain again if necessary.
-------------
Oh yeah. Units can be slightly off creep and still get the speed bonus. I was thinking that maybe (just maybe) they got the bonus easier depending on which side of them the creep was on. NOPE. Bonus speed to units is not related to the "creep conundrum".
|
Since I'm a map maker I will post here.
You can fix this issue by going into footprints and changing the creep tumor footprint. For some bizzare reason blizzard chose an even number instead of an odd number for the footprint size for creep tumors, which causes the problem.
In the map I made, Hero Attack, the creep tumors do not have this issue because I changed the footprint size.
|
Since I'm a map maker I will post here.
You can fix this issue by going into footprints and changing the creep tumor footprint. For some bizzare reason blizzard chose an even number instead of an odd number for the footprint size for creep tumors, which causes the problem.
In the map I made, Hero Attack, the creep tumors do not have this issue because I changed the footprint size. O.O Thanks dude!
Good thing I read this before telling Blizzard.
|
|
done and done, good work swarm
|
"O.O Thanks dude!"
You can thank me by playing hero attack (it's on the fourth page of the north american server).
Also, I just want to say that it's not the footprint that the creep tumor has directly. Rather the creep tumor has a generate creep behaviour and this behaviour has a creep grown footprint. The creep grown footprint is a problem because they made it even rather than odd.
|
Cool discovery, as you stated it doesn't matter that much but i guess it would be better off fixed.
|
e^ipi-+1=0, what you're saying is that you made the creep behavior originate at an odd number (in the center of the tumor)? Or that you enlarged the creep circle by 1?
If it's the former, you effectively decrease the creep radius by 1 because the properly centered circle reaches to a fractional radius.
Either way you change the amount of creep you get. Is it better to have lopsided creep or misadjusted size creep?
|
What I'm saying is that the creep tumor has a behaviour called 'Zerg - Creep Growth Medium'. This behaviour has a footprint called 4x4 creep source grown. Blizzard made creep source grown a 20 by 20 footprint (an even size). If you go to the footprint in galaxy editor you can change the size of the footprint (either to 19 by 19 or 21 by 21). If you go to placement apply and placement check for the footprint you will see a blue circle (well rather a pixelized circle). To make the creep centered around the creep tumor you will have to change the footprint size to an odd number and then adjust this circle.
The only way you can fix this problem is by changing the size of the creep the creep tumor generated (or alternatively you could increase the size of the creep tumor from 1x1 to 2x2). So yes, there is a choice. Do you want the same sized creep or non-lopsided creep?
|
|
meh, its creep, it doesn't need to be perfect. In fact it's more realistic if it isn't perfect
|
That would be a valid argument if the creep was randomly lopsided, which isn't the case. If I understand this thread correctly, a Zerg player spawning on top will spread creep faster than a Zerg player spawning at the bottom of the map.
|
LOL I just made a (closed) thread trying to figure out more about creep spread a couple days ago :p Anyway great find. I liked your post on BNET.
|
|
|
|