Competitive Map Making Discussion
Posted: Tue Mar 14, 2017 12:01 pm
Continuing from the map pool discussion and the 1v1 map discussion and to create a resource for future map makers to draw from.
I thought it'd be an appropriate time to make this post with the map making contest voting period starting a couple of days. That being said, I'd like to ask for people to keep nonconstructive criticism or excessive slander of any particular map out of this thread.
____________________________
Creating a well-designed and a well-liked map is no simple feat. No one can seem to agree on what makes a decent map, and everyone has their own criteria as to what constitutes a good map. The following are just my observations and opinions on the features that make or break maps. For simplicity's sake, I'd like to focus on 1v1 maps and try to categorize the maps by 3 standards: size, money, and design.
Size : Large or Small. There is no well-defined "medium" sized map that people can agree on right now
It's clearly obvious you can change the size of the map by altering the initial dimensions and the amount of traversable area. However, I think one aspect map makers overlook are the spawn distances. Here are the spawn-to-spawn distances for a few maps:
Orelord: 43.84c
Sidestep: 51.7c
Sirroco: 63.32c
River Crossing: 70.0c
Pitfight: 80c
Shadowfiend: 87.81c
Behind the Veil: 89.87c
A lot of the more modern maps are using very long spawn distances, which generally favors greedier, eco-heavy style builds. At this stage, I think any spawn distances smaller than Orelord is too small, and anything bigger than Behind the Veil is way too big.
Money : High-eco or Low-eco
This is probably the most difficult aspect of creating a well-designed map. There are quite a few factors that affect if a map is high eco or low eco:
• Starting Resources: 10k starting resources + growth is barely enough to cover a 2ref SD opener with 2 harvesters...9.3k is used towards infrastructure, leaving around 700-1.3k for combat units until the mines are empty. In the interest of expanding the game rather than forcing players to build in a particular order, I believe that the minimum amount of starting ore should be no less than 11k for each player. Also, be wary of positional imbalances with the starting ore mines; for instance, a spawn with ore mines above it will be at a disadvantage versus a player who has ore mines south of their MCV.
In general, the lowest total resources tends to be around 50-65k with 10-12 mines, and the highest total resources is around 70-85k, with at least 12 ore mines. Several maps have been experimenting with 16 or 18 mines, which provides many options to expand to, but also runs the risk of diminishing the importance of economy management. 1v1s on Tournament Island and Roadkings are good examples of what happens when there is too much money available on the map.
• Ore mine: There is a lot of room for creativity here, and I think it'd be easier to list what I've found not to work as intended as opposed to listing where to place ore mines:
-Put the edge of the ore patch at least 20c away from the next closest ore patch. If the ore patches are too close, the seeking of the harvesters will cause them to wander when the ore mine is empty. It also makes it so that the harvesters more likely to go to a different refinery if its overcrowded.
-Middle, shared, and vulnerable ore mines tend to lead to snowball and stalemate situations. Too many of these will make the game one dimensional. Of particular note, the vulnerability of every ore mine should be controlled with intention. There are many maps that I've played on where the mine is completely vulnerable to any kind of attack, and I suspect several of these were not intended to be like that. I've found that these types of ore mines are either extremely dangerous and require the player to secure it immediately, or they're not worth the effort to mine from (easier and more effective to put a tesla/turret next to it rather than a refinery).
-Ore mines with little to no ore. Single ore mines with less than 5k+growth are not worth the time and money that goes into placing a refinery at that location. A single harvester can harvest 5k's worth of resources in around 2m. At this eco level, I've found it to be better to invest that 5k into more cost-efficient units such as artillery, aircraft, and static defense. These types of ore mines are also often placed strategically to make a particular location more attractive, but using too many of these ore mines creates a similar problem with the vulnerable/shared/middle ore mines.
Design: Circular, Mid-centric, Creative
Welp...if there are a lot of places that you could put the ore mines, there are even more ways you can design a map. Rexy's thread discussed a lot about the lane theory already, so I'm not going to go too in-depth about that. Basically, maps require at least 3 main lanes in order to allow for enough maneuverability and options for strategical movement. Circular maps with only 2 lanes and Mid-centric maps with only 1 lane don't offer a lot of strategical variety. In my opinion, some of these maps would include Winter Storm, Singles, Northwest Passage, Mountain Ridge, and Crucible. All of these maps are either too mid-centric or have no middle at all.
Personally, I believe Sidestep was a great map because it was Circular and Mid-centric at the same time. This allowed for a lot of different strategical options, and sometimes, we got to see players using 2 different strategies against each other. From a glance at the Map Contest, I can see that Showdown , Mountain Pass, River Crossing, State Highway 95, and Trapped sharing a lot of the same qualities as Sidestep. Mountain pass and Trapped, in particular, are interesting because they have different economic situations. Trapped has 16 ore mines and a lot more available eco, which makes all of the routes extremely viable, while Mountain Pass has extremely low and vulnerable eco. The former seems to have so much eco that it is nearly impossible to shut down a player's eco completely, and the latter seems to have the exact opposite problem. In my opinion, for a combo mid centric / circular map, it should have slightly safer eco than what Mountain Pass offers, and it should have less total resources than what Trapped offers. River Crossing seems to have an interesting map design with porous passageways to the middle, which is something Mountain Pass also features.
People are also making bigger and larger maps to accomodate for the current tank/infantry blob meta. Such maps would include Frontier Duel, Sirroco
, and Ó Ceallaigh. These maps are interesting in the sense that it highlights the pros and cons of having multiple armies versus 1 giant blob. In my opinion, I think Sirroco is the biggest a 1v1 map should be, in terms of ore mine to spawn distance. I don't really know how these maps are going to play out in the long run, but I would venture a guess that it would favor the multitasker since it forces players to move their screens.
Several new maps have implemented creative ways to prevent or to inhibit base pushing with neutral buildings/civilians debris, random ore tiles, and beach tiles. However, all of these aforementioned methods are also detrimental to a mobile army as well. Civilians, trees, and buildings force infantry to stack on top of one another, which makes A-moving through these types of territory ill-advised. Debris, ore tiles, and beach tiles slow down mobile armies, which means that infantry stays prone for a lot longer, meaning rifles/rockets are more likely to die before they start firing back. In short, trying to prevent an MCV from deploying at a particular location will also affect the mobility of an army. Of particular note, Apocalypse Now uses beach tiles to inhibit MCV movements and basepushing, but it also makes it much harder for a normal army to move. In essence, this completely negates the advantages of a tank/inf army and increases the power of a v2/arty based army.
--- As a footnote to this particular paragraph, I think implementing high ground/low ground vision or accuracy modifiers could be a potential avenue for map makers to explore if it were ever to get added, as well as larger bridges, roads, and the beach tile usage from Apocalypse Now.
Lastly, I have to give Monty Hall it's own category because it breaks every single rule of what 1v1 has been thus far. It has numerous yaml changes and the starting positions are on an island with broken bridges, not to mention the creeps and extra money being added in the latest revision. Is this type of map, one with pseudo-islands, yaml changes, and creeps, viable for a competitive format?
People describe it as "unique," but that does not necessarily mean it's a good for the RAGL. Much like how Apocalypse now has too many beach tiles, I think Monty Hall breaks too many barriers. The addition of the fenced-in creeps is pretty creative because a normal fence is too easy to break through, but a concrete wall is extremely annoying to break through. The creep basically functions as a health bar for the entire wall, which I thought was a unique addition. Plus, it follows the theme of goats behind the wall, which is appropriate I guess. However, this map might set a "bad" precedent for future maps if it kept the creeps and was used in the RAGL. The moral of the story though, is that this is not the only map-breaking change there is in this map, and maybe the map should be modified so it doesn't break all these barriers.
As for the capturable gem mine, civilian buildings, and tech center, I think mapmakers should start using these yaml changes more in their own maps. it creates new objectives in the early game and creates a different type of barrier to the late game. For instance, it might be wise to unlock the gem mines earlier so you have more resources later on, but your opponent may see this and decide to steal that spot from you before you can get there.
As a parting note, I think it's unfortunate that most players are practicing for the RAGL rather than going through the map contest maps...perhaps we could line up the schedules for next year?
I thought it'd be an appropriate time to make this post with the map making contest voting period starting a couple of days. That being said, I'd like to ask for people to keep nonconstructive criticism or excessive slander of any particular map out of this thread.
____________________________
Creating a well-designed and a well-liked map is no simple feat. No one can seem to agree on what makes a decent map, and everyone has their own criteria as to what constitutes a good map. The following are just my observations and opinions on the features that make or break maps. For simplicity's sake, I'd like to focus on 1v1 maps and try to categorize the maps by 3 standards: size, money, and design.
Size : Large or Small. There is no well-defined "medium" sized map that people can agree on right now
It's clearly obvious you can change the size of the map by altering the initial dimensions and the amount of traversable area. However, I think one aspect map makers overlook are the spawn distances. Here are the spawn-to-spawn distances for a few maps:
Orelord: 43.84c
Sidestep: 51.7c
Sirroco: 63.32c
River Crossing: 70.0c
Pitfight: 80c
Shadowfiend: 87.81c
Behind the Veil: 89.87c
A lot of the more modern maps are using very long spawn distances, which generally favors greedier, eco-heavy style builds. At this stage, I think any spawn distances smaller than Orelord is too small, and anything bigger than Behind the Veil is way too big.
Money : High-eco or Low-eco
This is probably the most difficult aspect of creating a well-designed map. There are quite a few factors that affect if a map is high eco or low eco:
• Starting Resources: 10k starting resources + growth is barely enough to cover a 2ref SD opener with 2 harvesters...9.3k is used towards infrastructure, leaving around 700-1.3k for combat units until the mines are empty. In the interest of expanding the game rather than forcing players to build in a particular order, I believe that the minimum amount of starting ore should be no less than 11k for each player. Also, be wary of positional imbalances with the starting ore mines; for instance, a spawn with ore mines above it will be at a disadvantage versus a player who has ore mines south of their MCV.
In general, the lowest total resources tends to be around 50-65k with 10-12 mines, and the highest total resources is around 70-85k, with at least 12 ore mines. Several maps have been experimenting with 16 or 18 mines, which provides many options to expand to, but also runs the risk of diminishing the importance of economy management. 1v1s on Tournament Island and Roadkings are good examples of what happens when there is too much money available on the map.
• Ore mine: There is a lot of room for creativity here, and I think it'd be easier to list what I've found not to work as intended as opposed to listing where to place ore mines:
-Put the edge of the ore patch at least 20c away from the next closest ore patch. If the ore patches are too close, the seeking of the harvesters will cause them to wander when the ore mine is empty. It also makes it so that the harvesters more likely to go to a different refinery if its overcrowded.
-Middle, shared, and vulnerable ore mines tend to lead to snowball and stalemate situations. Too many of these will make the game one dimensional. Of particular note, the vulnerability of every ore mine should be controlled with intention. There are many maps that I've played on where the mine is completely vulnerable to any kind of attack, and I suspect several of these were not intended to be like that. I've found that these types of ore mines are either extremely dangerous and require the player to secure it immediately, or they're not worth the effort to mine from (easier and more effective to put a tesla/turret next to it rather than a refinery).
-Ore mines with little to no ore. Single ore mines with less than 5k+growth are not worth the time and money that goes into placing a refinery at that location. A single harvester can harvest 5k's worth of resources in around 2m. At this eco level, I've found it to be better to invest that 5k into more cost-efficient units such as artillery, aircraft, and static defense. These types of ore mines are also often placed strategically to make a particular location more attractive, but using too many of these ore mines creates a similar problem with the vulnerable/shared/middle ore mines.
Design: Circular, Mid-centric, Creative
Welp...if there are a lot of places that you could put the ore mines, there are even more ways you can design a map. Rexy's thread discussed a lot about the lane theory already, so I'm not going to go too in-depth about that. Basically, maps require at least 3 main lanes in order to allow for enough maneuverability and options for strategical movement. Circular maps with only 2 lanes and Mid-centric maps with only 1 lane don't offer a lot of strategical variety. In my opinion, some of these maps would include Winter Storm, Singles, Northwest Passage, Mountain Ridge, and Crucible. All of these maps are either too mid-centric or have no middle at all.
Personally, I believe Sidestep was a great map because it was Circular and Mid-centric at the same time. This allowed for a lot of different strategical options, and sometimes, we got to see players using 2 different strategies against each other. From a glance at the Map Contest, I can see that Showdown , Mountain Pass, River Crossing, State Highway 95, and Trapped sharing a lot of the same qualities as Sidestep. Mountain pass and Trapped, in particular, are interesting because they have different economic situations. Trapped has 16 ore mines and a lot more available eco, which makes all of the routes extremely viable, while Mountain Pass has extremely low and vulnerable eco. The former seems to have so much eco that it is nearly impossible to shut down a player's eco completely, and the latter seems to have the exact opposite problem. In my opinion, for a combo mid centric / circular map, it should have slightly safer eco than what Mountain Pass offers, and it should have less total resources than what Trapped offers. River Crossing seems to have an interesting map design with porous passageways to the middle, which is something Mountain Pass also features.
People are also making bigger and larger maps to accomodate for the current tank/infantry blob meta. Such maps would include Frontier Duel, Sirroco
, and Ó Ceallaigh. These maps are interesting in the sense that it highlights the pros and cons of having multiple armies versus 1 giant blob. In my opinion, I think Sirroco is the biggest a 1v1 map should be, in terms of ore mine to spawn distance. I don't really know how these maps are going to play out in the long run, but I would venture a guess that it would favor the multitasker since it forces players to move their screens.
Several new maps have implemented creative ways to prevent or to inhibit base pushing with neutral buildings/civilians debris, random ore tiles, and beach tiles. However, all of these aforementioned methods are also detrimental to a mobile army as well. Civilians, trees, and buildings force infantry to stack on top of one another, which makes A-moving through these types of territory ill-advised. Debris, ore tiles, and beach tiles slow down mobile armies, which means that infantry stays prone for a lot longer, meaning rifles/rockets are more likely to die before they start firing back. In short, trying to prevent an MCV from deploying at a particular location will also affect the mobility of an army. Of particular note, Apocalypse Now uses beach tiles to inhibit MCV movements and basepushing, but it also makes it much harder for a normal army to move. In essence, this completely negates the advantages of a tank/inf army and increases the power of a v2/arty based army.
--- As a footnote to this particular paragraph, I think implementing high ground/low ground vision or accuracy modifiers could be a potential avenue for map makers to explore if it were ever to get added, as well as larger bridges, roads, and the beach tile usage from Apocalypse Now.
Lastly, I have to give Monty Hall it's own category because it breaks every single rule of what 1v1 has been thus far. It has numerous yaml changes and the starting positions are on an island with broken bridges, not to mention the creeps and extra money being added in the latest revision. Is this type of map, one with pseudo-islands, yaml changes, and creeps, viable for a competitive format?
People describe it as "unique," but that does not necessarily mean it's a good for the RAGL. Much like how Apocalypse now has too many beach tiles, I think Monty Hall breaks too many barriers. The addition of the fenced-in creeps is pretty creative because a normal fence is too easy to break through, but a concrete wall is extremely annoying to break through. The creep basically functions as a health bar for the entire wall, which I thought was a unique addition. Plus, it follows the theme of goats behind the wall, which is appropriate I guess. However, this map might set a "bad" precedent for future maps if it kept the creeps and was used in the RAGL. The moral of the story though, is that this is not the only map-breaking change there is in this map, and maybe the map should be modified so it doesn't break all these barriers.
As for the capturable gem mine, civilian buildings, and tech center, I think mapmakers should start using these yaml changes more in their own maps. it creates new objectives in the early game and creates a different type of barrier to the late game. For instance, it might be wise to unlock the gem mines earlier so you have more resources later on, but your opponent may see this and decide to steal that spot from you before you can get there.
As a parting note, I think it's unfortunate that most players are practicing for the RAGL rather than going through the map contest maps...perhaps we could line up the schedules for next year?