I thought it was time to have an exhaustive discussion about maps in SC2, their role in the game, their effect on balance/design and what the community wants going forward. While there has been some discussion from Pros (http://www.thescoreesports.com/sc2/news/6508) and Blizzard in community feedback updates reaching up to 7 months back since September, neither have gone in depth into its history or its aims.

Debunking the Myth of
Same Maps = Same Strategy

Here are some direct quotes from Blizzard regarding some of their thoughts on the past map pool.

September 11
We’ve clearly seen how stale the game becomes both in terms of playing and watching when we’ve had map pools that everyone agrees is ‘good.’ The matches are all very standard and similar in terms of playstyle, and we want to clearly avoid this from ever happening again.

With that said, because we are constantly exploring new things that can potentially be cool for the game, obviously there is a higher chance of making a mistake. Maps such as Daedalus Point are examples of something that we tried that didn’t work out. However, we believe the positives that we gain from pushing map diversity outweigh the negatives. If necessary, it’s easy to remove a map that doesn’t work out mid-season, and we’ve seen from experience this doesn’t happen on a regular basis.

October 29
We know that there are many of you out there including top tier pros who prefer the same, standard maps every time. As we’ve discussed many times over and over again, it’s much more exciting in terms of playing and watching to have map diversity which leads to strategic diversity as well.

December 11
The general idea here is the same as it always has been: we want to push map diversity so that the game is more interesting to play and watch due to not every single map having the exact same build orders, attack timings, and strategies per matchup.

February 4
Obviously, map diversity is something we must push for the game, because we’ve seen in the past that when all 7 maps in the pool were basically the same, we were only seeing 1 timing/strategy/build order per matchup and the game became stale really quickly.

March 8
We’ve also seen this same sort of thing when all the maps in the map pool were of the same type: you play the exact same strategy capitalizing on the exact same timings on every map, so every game felt too similar.



The reaction from the Pros is varied. The only thing they all agree on is that Dusk Towers and Orbital Shipyard seem to be the best maps in the pool. Incidentally they are the most standard. (source: http://www.thescoreesports.com/sc2/news/6508) Some statements directly contradict Blizzard’s own statements about the interesting and diverse gameplay (Pengwin calls the map pool abysmal and says that these unique diverse maps have pigeonholed pros into playing 1-2 strategies per map, a direct contradiction to Blizzard’s own views).Some say it’s good and almost all of them question balance on multiple maps.

Of course that should be taken with a grain of salt as pros in general like to play on standard maps that have little variance at least in the early game. But it’s interesting to note that there isn’t a consensus among pros beyond them all liking Dusk Towers.

The biggest point to me in the Blizzard community updates is that a majority of it talks about how old map pools were too much of the same type and therefore they led to too many similar games. While maps did play a part in the staleness of metas in WoL and HotS, it was also in large part due to what the best play styles were at the time.

For example, if we reverted back to the end of Wings of Liberty, broodlord infestor would still reign supreme regardless of the map pool since Protoss and Terran would still not have answers to that strategy. We’ve seen BL/infestor work on 1 base, 2 base, 3 base+. Even on massive maps where you could outmaneuver them (as Ryung did against DRG on Whirlwind), Zergs would later just uproot 20 spines/spores and take the basetrade knowing that they would win anyway.

A second example is mech at the end of HotS. Mech was always a viable composition in HotS and players like Reality and SuperNova proved its viability early on. However it wasn’t used until the end of HotS, because INnoVation and Flash proved that bio/mine was a more effective strategy and much more fun to play with. It wasn’t until Zergs like soO, Life and ByuL started to close the door on the effectiveness of bio/mine did mech start becoming the prominent strategy. Both eras had turtle maps that were suited to turtle mech, but it was only used at the end because Terrans understood that playing bio vs Zerg was close to a near impossibility and the only Terran that did so successfully in the last year of play was Dream who hit the apex of his skill in his two run-ins with Life in SSL.

Maps do play a part in both balance and design, sometimes a very large part. But the meta will be dictated by whatever strategy or unit composition is the best at the time.

Looking Back Into 6 Years of Maps

In this section I go over all of the maps from WoL to the end of HotS to show that they were not all similar types of maps leading to the similar types of gameplay as the Blizzard posts say. It goes over the 6 years of maps and breaks down every map pool into standard and non-standard maps and how they all have distinct qualities that let them play out differently from each other.

For the sake of simplicity, I’ve only included maps from the GSL/OSL/SSL over the past 6 years. They start from 2010 GSL Open 1 to now. I break down the map pools into sections. Keep in mind this is a very simple overview as each map could have an entire article written about its complexities in every matchup. The first is from GSL Open 1 to the end of GSL January 2011. Also I use my own judgement for what a standard/non-standard map is and will later discuss the problems with that label.

GSL Open 1 to GSL January 2011

• Blistering Sands
  
• Desert Oasis
  
• Scrap Station
  
• Steppes of War
  
• Xel'Naga Caverns
  
• Kulas Ravine
  
• Lost Temple
  
• Metalopolis
  
• Jungle Basin
  
• Delta Quadrant
  
• Shakuras Plateau
  
• The Shattered Temple

The best thing about the first map pool was that it convinced everyone that tournaments needed mapmakers. There were multiple terrible non-standard maps with almost no redeemable games: Blistering Sands, Kulas Ravine, Lost Temple, Jungle Basin and Delta Quadrant, Desert Oasis. The only good game that used a feature of the map was JulyZerg’s crazy all-in on Delta. The opponent scouted his building natural and left. Upon leaving, JulyZerg canceled his natural and built a baneling nest on the remaining creep. So when the Terran would later scan his main, he would find nothing then die to the bane bust. The only other notable games on those maps were NonY vs Painuser, Fruitdealer vs TOP and Jinro vs Socke.

The only non-standard map that worked was Scrap Station. It created multiple good games because of the map features: Polt vs MC, MC vs Jinro, NesTea vs Choya and NesTea vs Zenio.

The maps that were considered standard during this time were Metalopolis, Xel'Naga Caverns, Shakuras Plateau (after it was fixed) and Shattered Temple (an altered Lost Temple that took out the abuse-able cliffs). They had a few things in common: Fast 2 base play, multiple spawning positions, and multiple attack paths. However they also had important differences that made them distinct from each other.

Metalopolis had a gold in the center, only 2 attack paths by ground and 1 by air unless you spawned close positions. The expansions were formed in such a way that you either were expanding in sync towards the same clockwise direction or expanding towards your opponent. Xel'Naga was a 2 player map with 4 possible attacks by ground and 1 by air. It was very congested and rocks were everywhere making it a very aggressive map. Shattered Temple had the same features as Metalopolis without the gold in the center, but it had 3 main attack paths by ground and possibly 1 by air if you spawned close air. The center was large and wide making it possible to sneak in ambushes and whoever took center control also had more paths of attack by air. It also had gold bases blocked by rocks and semi-island bases. Shakuras Plateau (the later version) had only 1 main attack path by ground and 2 mini-paths. It was a turtle map, though rarely played as such.



There are too many good games to count for those 3 maps, so I’ll just list a few: Mvp vs MKP, MKP vs Kyrix, NesTea vs Mvp, Mvp vs IdrA, NaDa vs Leenock, Clide vs Leenock, Jinro vs Ensnare, Losira vs TLO, Losira vs MMA, Polt vs MMA, July vs Mvp, Sheth vs Puzzle, MKP vs NaDa , DRG vs MMA, Genius vs qxc, SlayerSBoxeR vs FruitDealer, SlayerSBoxeR vs Hyperdub, TLO vs Hyperdub.

Total non-standard maps during this period: 8
Total standard maps: 4

These were the added maps from GSL March to GSL November:

• Crossfire
  
• Terminus
  
• Tal'Darim Altar
  
• Crevasse
  
• Xel'Naga Fortress
  
• Bel'Shir Beach
  
• Dual Sight
  
• Antiga Shipyard
  
• Daybreak
  
• Calm Before the Storm

This part of WoL’s map pool was when we started to slowly move away from the idea of 2 base being the standard meta to 3 base. This was an explicit directive as the community at the time wanted games with more bases. To do that, mapmakers started to include more maps with easier to take 3rd bases and queen range and infestors were buffed. The non-standard maps that didn't work were Terminus, Crevasse and Calm Before the Storm. Terminus gave 3 bases too easily. Because of how WoL economics worked, that meant generally slow turtle games. Crevasse was a weird map as it gave one rich vespene geyser in the natural, but constricted movement too much in the mid-to-late game with the rocks and importance of the center. Finally, Calm Before the Storm combined both of the problems from the previous 2 maps. Too easy to get 3 bases, too constricted to do much in the late game. The only notable games were Nestea vs July, Jjakji vs Leenock and Jjakji doing a 2-2-2 with the bunker capacity upgrade to beat a Protoss once.

The non-standard maps that worked were Crossfire, Xel'naga Fortress and Dual Sight. Crossfire was closer to the old standard set in early 2011 with fast 2 bases, but it had only two attack paths on the map leading to very distinctive games. Xel'naga Fortress was a 3 player map that had a hard 3rd to take and a vision tower that exploded after a certain time limit. What made it weird was as you went into the mid-late game there were only 2 attack paths by ground, one by air and the players' last expansions were always uncomfortably close, leading to some interesting games. Dual Sight was a smaller map with 2 bases, but it had a huge amount of freedom of space. Both players could get a gold, but the gold was in a hard to defend position for both players. Additionally, it had a multitude of attack paths that increased as more expansions were taken by both players.

Interesting games from these maps include: Bomber vs ByuN, TOP vs NesTea (both Crossfire and Xel'naga Fortress), NesTea vs sC, Mvp vs MMA, SuperNova vs MKP, Mvp vs Leenock, NesTea vs anypro. All of which were in some way influenced by the map design.

The standard maps from this pool include: Tal'darim Altar (TDA), Bel'shir Beach, Antiga Shipyard and Daybreak. TDA gave players either easy or hard to get 2nd/3rd bases depending on spawn location and opponent race. The longer it went into a game the more attack paths opened and attack paths changed depending on spawn. Despite being a macro map, the map openness and size made it hard to turtle or split the map as each expansion was harder to get along the way. This compelled players to start splitting their forces or go for an all out attack before it reached that point. Bel'shir Beach was an easy 2 base with a moderately hard 3rd to take. It had 2 attack paths, the harder one of which gave you easy air access to your opponent’s main and covered your expansions, but the shorter of which got you to your opponent's expansions faster. Despite its simple design, the attack paths gave players a subtle way to express their own styles of play.



Antiga and Daybreak are both known as standard maps and both gave birth to what we now call the BL/infestor age. A fast 3 base map with 3 attacks paths, constrained map movement and easy to turtle. The biggest difference between them is that Antiga had multiple spawns and should have played out differently depending on where you spawned, but the prevalent meta of BL/infestor made it so all Zerg matchups played out similarly. Outside of that Antiga created a lot of interesting dynamics as the center and extra expansions forced a lot of multi-tasking in non Zerg matchups, particularly in TvT and less so in TvP. Daybreak was very constricted to a 3 base meta, but because Protoss had a superior late game army with Collosus/HT, Terran was forced into constant aggression in both multi-task and attacks and made interesting games (barring SCV pull games). TvT was mostly unaffected by the constrictions of Daybreak as a map as bio/tank bypasses the turtle phase and Mvp proved that mech vs mech games could be broken open with incredible flair..



The interesting games from those maps were: jjakji vs Leenock (Bel'shir Beach] and TDA), ThorZaIN vs MC, EffOrt vs DRG, Mvp vs Bomber, Sage vs Miya, Sen vs HerO, DRG vs MMA, Bomber vs Seal, NesTea vs NaDa, GuMiho vs MMA, TaeJa vs YongHwa, DRG vs SuperNova, TOP vs Mvp, Polt vs TaeJa, Polt vs Ryung, Flash vs PartinG, NesTea vs DRG, Genius vs sC, Freaky vs Life, SuperNova vs Trap.

Total non-standard maps added: 6
Total standard maps added: 4
Grand Totals: 14 non standard - 8 standard

These were the maps added through 2012

• ClouD Kingdom
  
• Entombed Valley
  
• Metropolis
  
• Atlantis Spaceship SE
  
• Ohana
  
• Whirlwind
  
• Abyssal City
  
• Bel'Shir Vestige

In the year of 2012, Stephano would make two revolutions of play. The 3 hatch vs Protoss and the BL/infestor composition. We now remember this era (and early 2013) as the BL/infestor era. While standard maps like Daybreak and Antiga Shipyard made the strategy more viable, the design of the game made it so no map could break BL/infestor as we saw it sometimes win even on 1 or 2 bases against an opponent with many more. At this point Daybreak was the definitive standard of maps: 2 player spawn, fast 3 bases, 3 attack paths.

The non-standard maps added at this time were: Metropolis, Atlantis Spaceship, Ohana and Abyssal City. Metropolis was a weird map in that it was massive, had 2 attack paths and gave each player 5 bases if they could secure their chokes and protect themselves from harassment. Generally considered a failure of a map as it gave players incentive to turtle in a turtle meta. Atlantis Spaceship was interesting in that it was also a split map but it gave a lot of different attack paths. The problem was it was too constricted and there was no incentive in the game to go beyond 3-4 bases, especially against BL/infestor. Ohana was a smaller map that led to a lot of aggression, but because it was possible to get 3 bases easily, it was also dictated by BL/infestor. Abyssal City was the last non-standard map, but it didn’t have a long lifespan. Again it was suffocated by BL/infestor. In other matchups, it had the problem of being too open by air and too closed by ground.

Interesting games from those maps include: Mvp vs NesTea, Mvp vs VortiX, Mvp vs Squirtle, Seed vs ByuN, Bomber vs Freaky, YugiOh vs PartinG, Squirtle vs BBoongBBoong, TaeJa vs Seed, Dream vs Rain, Seed vs Sniper, CranK vs Freaky, Leenock vs Oz, Leenock vs Bomber.

The standard maps that were included during this time were: Cloud Kingdom, Entombed Valley, Whirlwind, Bel'shir Vestige. They were closer in line with the Daybreak model of fast 3 base, but due to how the late game worked out, were less likely to end in stale turtle matches (with one exception). Cloud Kingdom is an interesting map in that it was pretty clearly a turtle map and there was a lot of jockeying for position in the center, especially the high ground. The difference between it and some of the other maps during this era and why it’s remembered fondly is that it still gave room for the aggressor to outplay the defensive player at points in the game and even sometimes in the late game. Entombed Valley had 3 easy to take bases and a natural 4th, but there were 3 attack paths with the middle wide open. If it had come out during an era where BL/infestor wasn’t the norm, I think it could have given a lot of dynamic and highly aggressive late game scenarios. Even during BL/infestor it did so.



Whirlwind was similar to Entombed Valley, but unlike Entombed, the more expansions that were made the more attack paths were created for both players (both by ground and air). Because it was multi-spawn it also created spawn location specific strategies in the early and late game. It was large enough that before Zergs realized they could uproot and win basetrades vs Terran players, some Terran players could out position and beat BL/infestor with bio/tank play on this map (most notably Ryung). The last standard map to be made during this time was Bel`Shir Vestige. It’s very similar to Daybreak with a little less constriction. The least interesting standard map to be added.

Interesting games from the standard maps include: Grubby vs Nerchio, Grubby vs sLivko, Leenock vs Life, HyuN vs Sniper, jjakji vs PartinG, Mvp vs Rain, GuMiho vs Creator, NesTea vs aLive, Polt vs DRG, MKP vs DRG, MKP vs PartinG, Hack vs RorO, DRG vs GuMiho, Dream vs PartinG, Mvp vs MMA, Flash vs Ryung (twice on WW), Soulkey vs INnoVation (PL), TaeJa vs Rani, Soulkey vs TaeJa, INnoVation vs SuperNova, Ryung vs DRG, INnoVation vs Flash, YugiOh vs Strelok, YugiOh vs LucifroN, GuMiho vs Losira, sOs vs Losira, Seed vs Symbol, TRUE vs FanTaSy, First vs FanTaSy, DRG vs INnoVation, Bomber vs Goswer.

Non standard maps added: 4
Standard maps added: 4
Grand Totals: 18 non standard - 12 standard

These are the maps added through 2013 (now Heart of the Swarm)

• Akilon Wastes
  
• Icarus
  
• Neo Planet S
  
• Atlas
  
• Star Station
  
• Red City
  
• Gwangalli Beach
  
• Anaconda
  
• Newkirk Redevelopment Precinct
  
• Derelict Watcher
  
• Frost
  
• Polar Night
  
• Yeonsu

The non-standard maps in this pool were Icarus, Neo Planet S, Gwangali Beach, Anaconda and Yeonsu. Icarus was defined by its back natural expansion that had 1 gold mineral and rich extractor. It was a timing attack map and most famous for Soul Trains and all-ins. Neo Planet S had the basis of the fast 3 base, but the 3rd wasn’t that easily taken. The pathing in the mid-late game was what made it non-standard as it’s circular and created 4 attack paths by ground. Gwangali Beach was a weird map as both players had to expand towards each other and there is one direct attack path from one player to the other. Anaconda was another really weird map where you could get fast 3 bases, 2 of the 3 were especially vulnerable to air. Yeonsu is the final map with fast 2 base, a harder to get 3rd and a very hard to take 4th. Depending on where the players expanded and how they expanded, it created a lot of dynamic games, but the map feature of having a large blinkable main base made sure that it never lived up to its potential and it was removed because of the Blink Era. An example of a map that adversely affected the balance of the game (in contrast to the BL/infestor era).

Notable games on these maps include: sOs vs INnoVation, Squirtle vs Soulkey, Soulkey vs DIMAGA, Mvp vs Tefel, FanTaSy vs Life, Symbol vs Dear, sOs vs some random Zerg victim on Anaconda (when he busted down his rocks with void rays), soO vs Zest, DRG vs INnoVation, ByuL vs Revival, Polt vs Protoss.

The standard maps in this pool were Atlas, Newkirk Precinct, Akilon Wastes, Star Station, Red City, Derelict Watcher, Frost and Polar Night. In this case a lot of the maps during this time played similarly for two reasons. Either it was during the BL/infestor era which was still dominated by 3-4 base Zerg infestor play which could ignore map features. Or it was being dominated by INnoVation’s 4M parade push which was a very unidirectional kind of strategy and one that only the best players could stop. So a lot of these maps in TvZ played similarly. At the same time a lot of the maps helped create the Blink Era with bases that could be very easily all-ined with blink stalkers (Star Station, Frost, Polar Night, Anaconda, Yeonsu). This would be a bigger problem later. Newkirk had potential, but it was too large so players normally stuck to either turtling or massing armies on this map (the only players who could play it to its potential were TaeJa vs INnoVation). No one was able to make use of Atlas and it was too similar. Star Station was a very weak standard map. Same with Derelict Watcher.



Akilon Wastes was complex for a standard map. The map was split in half (and could be played turtley especially during BL/infestor era) but since both players expanded towards each other it created a lot of openings and aggression on both sides. Red City was a very strange map that was cut up into quadrants and very chokey with a lot of different ways to harass a player. However it didn’t last long though a few games were exceptional. The two best maps to come out during this time were Frost and Polar Night. Frost started off as a standard fast 3 base map, but the situation in the mid-late game changed drastically depending on the matchup and spawn positions. It was possible to turtle, though it wasn’t done too often as there were a lot of different ways to attack. Polar Night was a 2 player map that held the same concept as Crossfire, except there were 4 lanes of movement. The oval shape created a lot of complex and reactive movement in matches.



Notable games from the standard maps of this time were: First vs KeeN, TaeJa vs INnoVation, SuperNova vs Soulkey, Mvp vs INnoVation, GuMiho vs Losira, PartinG vs RorO, PartinG vs Life, HyuN vs HerO, Ryung vs Apocalypse, Dream vs Scarlett, MadBull vs Maru, Dear vs Maru (Frost and Polar Night), Happy vs HyuN, Symbol vs jjakji, Maru vs jjakji, duckdeok vs INnoVation, DRG vs Flash, Curious vs INnoVation, herO vs HyuN, Has vs Jaedong, Polt vs Protoss, First vs jjakji, Maru vs Classic, Classic vs soO

Total non-standard Maps added: 5
Total standard maps added: 8
Grand Totals: 23 non standard - 20 standard

2014 GSL Map Pool:

• Alterzim Stronghold
  
• Daedalus Point
  
• Heavy Rain
  
• Waystation
  
• Merry Go Round
  
• Habitation Station
  
• Catallena
  
• Deadwing
  
• Nimbus
  
• Foxtrot Labs
  
• Overgrowth
  
• King Sejong Station

In 2013, a lot of the standard maps weren’t good. Among the 8 standard maps that were introduced, only 3 had real success in creating dynamic interesting games. Maybe because of that more non-standard maps started to be included. The biggest meta change during this time was the Blink Era which was in part created by the map pool. A great example of how map design adversely affected balance.

The maps I consider non-standard during this iteration were: Alterzim Stronghold, Daedalus Point, Waystation, Habitation Station, Deadwing, Foxtrot Labs. Alterzim and Deadwing were both massive maps meant to create a lot of chaos in the late game, but Alterzim was too straight forward as a map. Deadwing was bit less so as the differing spawns as well as the meta allowed it to still create interesting games (bio vs muta/ling/bling). However it was perfect for mech vs swarmhosts which also was a point against it. Waystation was a map where you either had super long spawns or very short spawns. It was panned for the most part, though it did give a few map specific games that could have only happened here (Bbyong vs Flash and Billowy vs Symbol).

The Billowy vs Symbol match is notable because it is the best example of maps creating interesting gameplay, yet despite all of the non-standard maps made in the last 6 years and the thousands of tournament games, no one player has ever been able to create an entire plan that exploited map features as beautifully as Billowy did in this one game. In this game, he opened phoenixes to kill overlords, take map control and deter mutas. He then made a warp prism and took the island base and transferred workers with a recall. From Symbol’s point of view it looked like a weird 2 base all-in. In actuality, it was a 3 base all-in, and Billowy abused the massive long spawn maps to proxy a hidden nexus in an unscouted part of the map that was close to Symbol’s base. He then recalled his max army there and completely bulldozed Symbol.

Next is Habitation Station. A weird map that created Gangnam Terran. While it had two good games (Scarlett vs Bomber, SuperNova vs TY), it was for most of its career horrific as it was either a bunch of all-ins, mech vs swarm hosts, or SH vs Protoss turtle matches. Foxtrot was another map with a lot of distinct features. However, unlike Habitation Station, it mostly worked as the map features gave a lot of different options and the pathing of that map was decently complex in the late game.

Notable games from these maps include: Life vs First, Scarlett vs Bomber, TY vs SuperNova (Way Station and Habitation Station), Rogue vs Seed, Jaedong vs Patience, Billowy vs Symbol, Bbyong vs Flash, TaeJa vs Zest, soO vs Flash, Maru vs MyuNgSiK, Polt vs Hydra, Ruin vs Daedalus, Polt vs HerO.

The standard maps from this time period were: Heavy Rain, Merry Go Round, Catellena, Nimbus, Overgrowth and King Sejong Station. Heavy Rain had 5 paths, the 2 outer paths were open, but took the longest to traverse. The middle path was the shortest, but had the narrowest chokes and could be closed by rocks. The 2 paths in between were blocked by large rocks. This map didn’t make the mistake of Atlas and throw rocks everywhere, but had them control key paths and expansions that came into contention in the end game. Merry Go Round was a 3 player map akin to Xel'naga Fortress, but it was larger and had expansions to fight for in the middle which created a lot of different ways to expand and change the tempo of the game depending on the type of player. Catellena was a more defensive version of that with no bases to take in the center. Because of its lack of late game complexity, it was never able to create as many good games as Merry Go Round.



Nimbus felt like a throwback to Crevasse, but it fixed the constricting pathing issues and had interesting army paths, movements and differing strategies depending on where you spawned. Overgrowth was the most popular standard map of the time and created a lot of great games for a few reasons. It was balanced, but could be used for early game shenanigans (most notably some specific cannon rushes as well as Zerg gold rushes). The late game forced players into complex positions that forced them to weigh positioning, bases and army types. The only negative was that it was fairly good for SH strategies. King Sejong Station was a map with a lot of distinct map features, especially where the expansions and natural were laid out. It was still considered standard and was a great macro map that also lent itself to unique strategies. It lost its appeal once SH players found out you could siege everything if you took the center base as the 4th base.

Notable games were: jjakji vs HyuN, TaeJa vs Life (Heavy Rain and KSS), Maru vs Dear, Zest vs herO[jOin], Life vs Dream (MGR and Overgrowth), SuperNova vs TRUE, soO vs Flash, PartinG vs Terran (MGR as he had the best micro), TRUE vs TY, TaeJa vs soO, Flash vs Solar, INnoVation vs TaeJa, Leenock vs anyone (Overgrowth), Life vs Flash, TY vs SuperNova, TY vs anyone (KSS), TaeJa vs MC, Polt vs Classic.

Total non standard maps added: 6
Total standard maps added: 5
Grand Totals: 29 non standard - 25 standard

2015 GSL S1

• Iron Fortress
  
• Expedition Lost
  
• Vaani Research Station
  
• Cactus Valley
  
• Coda
  
• Terraform
  
• Echo

The final year of HotS had the least amount of new additions. There were two main meta changes. First there was SH which had started to gain more and more prominence in 2014 as Zergs came to realize that the only way to reliably beat Protoss was to use SH. Though the only player who could win with it consistently at the highest level was Snute. It was both a combination of game design and mapmaking design. Once they nerfed SH, Zergs would spend the rest of HotS trying to win games against Protoss off of surprise gambits, really strong aggression or timing attacks. They were never able to find a way to reliably play vs Protoss in the late game, but it ended up not mattering as HotS ended.

The other thing to note was the death of bio in TvZ and the rise of mech. Dream was the last great BiovZ player. Everyone else gave up on it as the refinement/tactical innovations of Life, ByuL and soO completely dominated that matchup. This brought in a bunch of mech styles in TvZ. Later on in the year, maps were made in such a way to force non-mech play which resulted in catastrophic loss of the top Terran players to Zerg players. This was an example of how map design could force players to play a certain way, but balance wise the game wasn’t there resulting in fairly bad or lopsided games.

The non-standard maps were: Iron Fortress, Expedition Lost and Terraform. Iron Fortress only gave 2 easy to take bases and Protoss had a very hard time taking a 3rd which led to a lot of bad games in general. Expedition Lost was another throwback to 2 base style play with a lot more early aggression. While it created a lot more early strategies, there are very few notable games that came from it. The only one that I can remember is Life vs Maru. I consider Terraform a non-standard map as while it gives you three bases easily, the entire map pathing is incredibly constricted and narrow. The pathing and movements are completely different from the majority of maps both standard and non-standard. It was also a safe haven for mech players.

Notable Games from these maps include: ByuL vs INnoVation, Rogue vs Mech, Maru vs Life, soO vs Flash, MarineKing vs the Minimap



The standard maps added were: Vaani Research Station, Cactus Valley, Coda and Echo. Vaani was an interesting standard map as it gave you the three bases like a standard map. The map layout was very similar, but the gold bases added a couple of weird strategies that could surprise. Despite having that map feature, it very much had the same problem that KSS had in that the central base had too much access to all other parts of the map. This was supposed to be combated by the fact that there were so many open lanes running up and down the map, but SH in the early part of the year and mech in the later part of the year could cover all of that space. When the games weren’t dictated by the stale meta, they were fairly good though. Cactus Valley was a fairly standard map. It was large and avoided having expansions out in the center making it a very base trade heavy sort of map. In terms of the late game, it was too simple. Coda and Echo were both standard maps that had fairly interesting dynamics in the mid and late game, but had the misfortune of also living in the SH and mech eras where it was too easy to split the map.

Notable games from these maps: ByuL vs Terrans, ByuL vs herO, Rogue vs Protoss, Polt vs Rain, soO vs INnoVation, Rogue vs herO.

Non standard maps added: 3
Standard maps added: 4

Total non standard maps: 32
Total standard maps: 29

That is a basic summary of the entire map pool from WoL to the end of HotS. It lists some basic features and distinctions from every map and while some of the maps do have similar features and/or play styles, a vast majority of them do not. It is hard to distinguish between core gameplay and the maps and there are examples of maps influencing the game play (Blink Era), gameplay superseding anything a map can do (BL/infestor) and parts where both contribute (SH and Mech).

Finally there is another contributing factor as to why Blizzard might feel there was a lot of staleness in the game. Basically all pro players spend tens of thousands of games playing and refining thought-out strategies and build orders that deal with a majority of what anyone can throw at them in one way or another. That is just the nature of the beast.

As for standard maps contributing to the staleness of the game: I can see it to some extent, but I think a majority of the great games come from the standard maps. There are close to an even amount of nonstandard to standard maps, yet the majority of notable games lean heavily in the favor of the standard maps. Though I do understand that they want more games like Billowy vs Symbol, the truth is I’ve only seen 1 game that ever used map features to that crazy extent in 6 years. In all likelihood that was Billowy hitting a revelation on a map that coin tossed him into the right positions to use that strategy. It is an incredibly rare thing and not something you can reproduce.

On the other hand you also have to point out that pros will be vetoing the non-standard maps as much as possible so the amount of times a map is played also leads to that, so take all of this with a grain of salt. And finally this is just my opinion on what standard is and it could change drastically depending on how you define standard. Which is another problem when we talk about maps.

Moving Forward

There is no defined standard when it comes to maps.

I've talked to pros, mapmakers and read all of Blizzard’s postings. No one has come out with an actual definition of what they believe the standard to be beyond the map that players like to play the most. "Standard" is an ever changing concept so trying to define what you want as standard is incredibly problematic when trying to design a map as no one has a set definition for it.

While I do not believe that a hard definition is necessary, or that there is a single map that encapsulates "standard" the most because the game constantly changes, it is helpful and possible to define what kind of map we consider the standard currently. In early WoL, a medium sized map with easily accessible 2 bases without abuse-able map features was the standard.

Later on 2 player maps with fast 3 bases and 3 paths of attack (i.e. Daybreak) became the standard. In HotS, maps with 3 bases and an accessible 4th became the standard. Assigning even a temporary standard should help mapmakers and players give feedback as to how to improve the playing (and watching) experience for everyone.

Yet it isn’t enough to go non-standard. We need to break everything.

At the beginning of LotV, Blizzard said that they were willing to do whatever it took to keep the game as fresh as possible. While the goals are admirable, the execution needs some adjustments. There are currently too many restrictions on map pool guidelines that dictate everything from rock health/armor, high yield gas, mineral counts, watchtowers that detonate after a certain time—maybe even maps that used the old HotS economic model.

With free rein, one could even make maps that have alternate spawning positions based on the current matchup so that it can be balanced for TvZ, but in ZvZ they’re forced to just ling/bling knife fight all day. As it stands now the new non-standard maps aren’t really creating a lot of crazy game scenarios and they don’t have the great strategic/tactical movement of the standard maps like Dusk Towers or Orbital Shipyards. An example of this is Prion Terraces, which was said to only be interesting because of gold bases, but why is it better than any other map that has a bunch of gold bases? It doesn’t play that much differently and only skews it for Zerg against other races.

What we’ve learned from 6 years

While I do believe in releasing strict guidelines so that mapmakers can make more diverse maps, I also believe that map features are not the key to success. Rather the opposite. One of the biggest problems with the current map pool is the obsession on map features.

As I’ve shown in the entire history of SC2’s map pool, you can have non-standard maps that do not rely on heavy gimmick features or very strong racial advantage one way or another. Instead of focusing on trying to create action as early as possible, there are other ways to balance the game based around complex mid and late game situations when players need to decide where to expand and how to attack.

I’ve given 6 years of past example including notable games from every map pool for both standard and non-standard maps. Watch those games and take a look at how the maps affect the game, especially the mid to late game.

Here is a basic example from three of the standard maps: Whirlwind, Frost and King Sejong Station.

Whirlwind had 4 spawning locations, so a player had 3 possible scenarios. Either they had favorable by air (meaning their natural faces towards their opponent's main/natural giving them free access to air attack paths), unfavorable by air (which is the opposite of that), or cross spawns.

In the favorable by air case, the players had choices on how they want to expand. There were two choices: either the pocket or the straight line. For players that spawned favorable by air, they chose the pocket as it gave them the best defense against both ground and air attacks. For the player who spawned in unfavorable by air, they opted for the line as the pocket surrendered too many attack paths by air to the opponent. From there a late game ensued where favorable by air would be constantly putting on pressure and had a choice of doubling down by expanding to the closer 4th or getting the farther 4th and trying to keep it safe via pressure. They could even hide bases in the top left or top right corners. Unfavorable by air had a few choices. In terms of spawning location, he was in a slightly worse spot, so they could either go for the all-in, or fight very hard for mid or inner control. If unfavorable by air won mid-control (a lot of factors change depending on who gets mid control), he was in a much better position now than favorable by air as he opened 3 attack by air paths on the pocket 3rd, the main and the natural. He even had double the attack paths by ground. This complexity only increased as both players expanded to 4ths and 5ths throughout the game.



Cross spawns was much more complex as expanding patterns were based on what type of strategy and army comp the player had in mind, what the opponent did in the opening stages and how the mid to late game played out. I will say that it was a very hard map to turtle on and even during BL/infestor, Ryung proved you could win on this map in a positional game against the greatest imbalance the game had ever seen (before Zergs realized they could uproot spines and have them walk with the unstoppable army anyway).

Frost worked completely differently. Like Whirlwind it was a 4 player map. There were 3 possible ways to spawn. Air to air, 2 paths by ground or cross-spawn. Air to air had both players' mains exposed to each other via air attack paths. While the pocket third was generally considered safer and protected the air weakness of the main, it put one's army in a position where it must commit to the defense of that area, which opened up the other paths to run bys, especially at the natural and later the open 4th. If a player took the open 3rd above, there was the risk of being harassed in the main as the army had to commit to taking center control to protect the natural and third. The small path at the bottom still gave potential for run-bys. Both players had to deal with these complexities and as the game went on further, more and more attack paths opened up and players were forced to think about how they wanted to break up and use their units to secure superior economic and tactical positions across the map. It became even more complex as both players expanded towards the rest of the map as the new expansions gave you money, but all of the old infrastructure was still in the main creating multiple areas of strategic and tactical value. The 2 paths by ground scenario were much more stable as both players could take the 3rd without taking a notable vulnerability. But mid control was even more important as whoever controlled mid forced the enemy to react to their army and how they set up the attacks.

Frost cross spawns worked in a similar way, but after the initial 3-4 bases there wasn't a clear place to take your 5th as all expansions essentially expanded closer towards enemy territory for both players and each succeeding expansion created more attack paths. (This was all moot for certain types of A-move strategies like mech, or long range map controlling strategies like SH)



A third standard map was King Sejong Station. It was a 2 player map with the natural jutting out towards the opponent by air. The natural had two entrances, one open, one closed by rocks. Players had a natural expanding pattern, though taking the center 4th was a risk in non SH/mech compositions as it was hard to control without falling apart to harassment. Similarly as the game drags on both players had to weigh aggression with defense as there were a possible 6 attack paths by both ground and air and the number increased as more expansions were taken.



This is just a small example of how these standard maps were excellent at creating varied, engaging and interesting gameplay in the mid and late games. Despite being grouped as ‘the same standard macro maps”, they all played differently in their respective eras.

There has been some hoo-hah about having more varied unit compositions in each matchup. But for my money, I consider WoL TvT the best matchup to spectate because of its inherent value in positioning and its large potential for aggression. Similarly, TvZ with bio-mine vs mutalingbling (HotS) or bio-tank vs mutalingbling (WoL) were both very compelling as they had a lot of action that forced a lot of army movement and strategic jostling throughout the game. TvP was the same when the meta was bio-ghost vs mass templar strategies. Though the current iteration of TvP in LotV is very good too and has yet to be played out. The matchups that are most prone to what people call "maxing out armies and A-moving" are PvZ, ZvZ, and PvP. In that respect, not much has changed for any of those matchups in LotV. You can do some harassment and all-ins, but it is still a butting of heads as the game has always snowballed too hard in those particular matchups.

So, I don’t think more varied unit compositions is the answer. Games can vary widely based on tactical execution and economic expansion. The ways strategies are played out can be very different and compelling even when we’ve seen the same type of unit composition before. And even standard maps could be open to incredibly creative cheeses like Polar Night where Has 7 pylon rushed Jaedong—back when pylons didn't even shoot lasers.

Essentially maps need to have a larger overarching look at what they’re trying to accomplish when it comes to trying to make interesting gameplay.

For instance here is a comment from Theo about modifying the current map pool to make it more playable:




What is strategic diversity and interesting gameplay?

The justification for including more non-standard maps is to increase strategic diversity and make games more interesting. I cannot speak for the community or the audience. I can only speak from my own experiences . I don’t know what that statement means.

As I’ve shown, maps do have a large impact on the way the game is played, but you cannot force strategic innovation. No amount of maps you throw at a pro is going to make him discover a paradigm shift that will forever alter the way future pros play a certain matchup. A zerg will keep going roach/ravager. Protoss are pigeonholed in what they can open vs Zergs. Terrans will go bio and pray to god they get to mass liberators against anyone but themselves. If they mean tactical execution, army movement, economic decision making then the only maps that I’ve seen it work out on are Dusk Towers, Orbital Shipyard and a little bit on Prion. On Orbital, the price to pay for 5 convenient bases is a very exposed layout. They can be easy to defend with enough guile and good positioning, but they are easily exposed by multipronged attacks.



As for interesting gameplay? I think interesting players make for interesting gameplay. Players like Mvp, sOs, SuperNova, Bomber, Polt, Classic, PartinG, Has, ByuL, soO, Life, GuMiho, TaeJa, Rogue, Maru, FanTaSy and TY are going to be interesting no matter what map you throw them on. As for the rest of the pros, can strange maps force them to play to even 1/10th the level of interesting as GuMiho does in his standard games? Some of these players have been around since the beginning, and we've seen that regardless of map pool or expansion, creativity is still a part of their gameplay. It is a personality and a mindset that players apply regardless of the maps, and it is something that can't be forced.

I am for both standard and non-standard maps as long as they have depth to what they’re trying to accomplish in all parts of the game.

Changing the Map Pool

I think map pools should be changed. One of the key factors that is holding SC2 back is that they don’t want to add on another layer of complexity to an already very complex game to the casual player. In my view, it is time to give this attitude up.

SC2 is already known to be one of the hardest 1v1 competitive games in the world. It is time we embraced that hardcore aspect, like how Dark Souls markets itself. I don’t think it’s wrong to try to appeal to more casuals and bring them into the game, but the subtleties of the map pool and how they play into strategy is at the very end of the road.

So my proposed solution is to have multiple map pools. One for the lower leagues and another for Masters and above. From a casual standpoint, they may feel that they are bad at this game, but this game was never going to be easy to begin with. It’s hard, it’s competitive, it drives you insane, but the payoff is larger because of it.

An alternate solution is to make two different map pools for non-ranked and ranked. This methord won’t have any weird psychological effects on non-ranked players as they’re playing for fun anyways. And it gives a space for mapmakers to beta test the maps with a larger sample size. Finally it allows for more freedom in the actual competitive map pool.

Tournament Map Pools

It is not difficult to see that Proleague has had some of the strangest maps ever—remember Arkanoid?—, but the league also exhibits some of the most exquisite strategies we've ever seen. Many examples of revolutionary or refreshing strategies occur in Proleague games. One possible reason for this is that players are given their maps and opponents a few days before their official matches. This gives them (and their teammates and coaches) the time and freedom to come up with strategies specifically for their opponent and map.

In most tournaments, a player's group or opponent is usually already predetermined. However, the maps aren't revealed until the games are about to be played. While this does have advantages, it also means that players often resort to standard strategies regardless of the maps. If players were given a few days to prepare knowing the maps they are set to play on, then we could see more strategic variety in tournaments with predetermined matches. We often praise "preparation tournaments" more than weekenders, yet players still resort to standard play more often than not in these instances. Maps play a large part in preparation, and it would encourage more creativity and cleverness if maps were announced ahead of time.

On Map Rotation

The current map rotation works like this: There are 7 maps, 4 of the 7 are rotated out based on age, so it is always the older maps. This means that 6 of the 7 maps will always have a guaranteed lifespan of 6 months no matter how shitty or great the map is. This is way too strict as we should be trying to create the best map pool possible.

And while I can see why they would want to do it like this to avoid upsetting schedules for tournaments, the entire system has changed in 2016. WCS as a long running league is dead. It is now a bunch of weekender events so putting it between those is not a problem. GSL/SSL do their own thing and there are only 4 of them anyway. Because of their elongated scheduling they can pretty easily justify changing the map pool between the ro32/ro16/ro8 without a problem. There is no reason to not change the way maps are rotated in or out.