SPL '13-14
Finals

MAP ANALYSIS
Battlegrounds of the finals

PL Strategy
Primer

Detailed information on
Liquipedia

by TL Strategy

It's no secret that Proleague can produce the most unique games and strategies known to Starcraft, or that this has been the case for a very long time. Proleague is a melting pot of strategy and planning separated from other tournaments by its format. With each team meeting each other several times a year and most teams utilizing a 4-5 player rotation, many teams will have to deal with the same player multiple times, in new scenarios each time. When facing a player that far outmatches any your team can muster, you cannot turn to gameplay alone. A layer of strategy and surprise has to fill the gap. While the amount of surprise can vary from one-off builds (such as Solar's 10-pool spore crawler) to completely nullifying an opponent's playstyle, it is definitely one of the most important aspects of Proleague competition.

Throughout the 2013-2014 SK Telecom Proleague, we have witnessed the rise of several new styles and builds devised by a variety of players, in the most unlikely situations. Mech against Protoss? Done. Hidden gas into speedling all-in with a zergling blocking the opponent from building a wall? That, too, has happened.

While individual leagues may carry more weight in Starcraft II, there is no denying Proleague remains uniquely fitted to creating upsets and shifting the meta, even if it's just for a single game. In this first part of our SPL Finals Preview, TL Strategy have looked closer at some of this season's most strategically diverse games, and the maps on which they took place.

The Maps

by Jer99 and Thaniri




Potentially the craziest map of the lot, Habitation Station is the map to watch for crazy snipe strategies. This map has seen the full gambit of cheese and off-tempo builds, ranging from early pools, proxy barracks, and hardcore cannon rushes to void ray busts and the seemingly unbeatable Gangnam (Terran) style. The map has particularly been harsh on non-Bbyong terrans, with a 44.6% winrate in TvZ and 53.8% winrate in PvT, but don't let that fool you - it does not rule out a potential Terran pick. A player to watch out for in particular would be KT Rolster’s TY. He is known for playing absolutely chaotic games, but he is equally apt at preparation. Having played arguably the most meticulous and calculated snipe build against Cure on Outboxer, a well-prepared TY is a force to be reckoned with.. Any of the lesser known players as well could be used to try and catch a star player off guard with a wild cheese on this map. Players like KT's Miso and SKT's BrAvO - seldom fielded in the regular season - could really make a name for themselves against a top level player. Habitation Station has the potential to produce miracle upsets, as such expect it to be, if not one of the more interesting games of the night, then at least a tipping point.





Overgrowth has many ridges and choke points in it, requiring players to maintain both excellent map awareness and army movement. Getting caught out of position on this map is particularly devastating due to the geography of it. As such, zergs have a tendency to stay on lower tech on this map, as ultralisks are not particular good in the chokes. soO’s heavy mutalisk and corrupter style against protoss could potentially come out on this map. PartinG and TY both have demonstrated strong two base plays on this map against zerg, abusing the relative vulnerability by isolation of the third base in comparison to other maps in the pool such as Frost. Balance wise, this map is the most even in the ZvP matchup, with terrans struggling against zergs and protosses surprisingly against terrans. Given that, and also the fact that hard all-ins are a bit rarer on this map compared to frost, most likely this map will have a ZvP played on it.



Frost is the map that is standard to the hilt. It is relatively uncommon for cheesy strategies to be used on this map given the long rush distances and easily defensible bases. This map has a tendency to play out to at least four bases for each player, lending itself towards sending out players who are mechanically sound in longer games. Players that would immediately jump to mind would be Zest and Flash for KT or the rather daunting SKT1 death gauntlet of PartinG, Rain, soO, Soulkey, and Classic. The star studded lineup of SKT1 will make it difficult for KT Rolster to prepare a sniper for this map, as any one of the their aces could potentially be playing on this map. Expect to see the highest level of play on this map, not to say that the other maps won’t be high level, but that this one in particular has the least potential for trickery to catch players off guard.



With Outboxer never used outside of proleague, it's probably the least figured out map of the bunch. Nevertheless, the map has few features out of the ordinary and standard play is fine. With the main base having a sizeable chunk of open space, this can make for a good blink build, and even a siege tank line can be set up and cut off reinforcements to the natural. The first three bases are very close together, making defending them easy. Like The King Sejong Station, the fourth base will be a big point of contention as it is out of the way of the previous bases and easily sniped. The middle of the map won't be used much as its low ground, so the contour will be the main attack paths. There is plenty of air space making warp prisms, mutalisks and medivac drops more effective. A relatively tight map structure will make surrounds difficult; most engagements will be a straightforward army vs army fight to decide the game. Interesting to note about Outboxer is the high frequency which players are typically fielded on the map. More so than any other, Outboxer is the map of map specialists.



Being the only three player map in the pool, Merry Go Round is the map for unorthodox/cheesy openings meant to take a quick win. Aside from the natural, the structure of the map is very open, resulting in the third base being extremely hard to establish. Because of this, fielding Protoss players on this map can be tantamount to suicide without a well thought-out strategy, and if there is, expect 2 base all-ins. With so many avenues of attack, we will be seeing lots of runbys and counter attacks on the exposed third and fourth bases, and a lot of skirmishes all over the map should the game reach that stage.



As one of KeSPA's Proleague maps, King Sejong Station has a long rush distance at 66 seconds, making cheddar mostly ineffective. But despite the distance, this is the best map in the pool for proxy play; there are many spots to hide buildings at. Setting up the initial three bases will be easy, but like on Outboxer the fourth base is the key location. The 3 potential fourth base locations each have their advantages and drawbacks, so it will depend on the player and their style will determine where they take it.
Generally, Terrans have been taking the middle location as their next base as they are continuously moving around with their army and have map control, while Zergs and Protoss will want to take the outer bases while taking a defensive posture. Doom drops are effective as well due to how long it takes ground units to cover the distance between bases.

Proleague Strategy: Preparation and Planning

by jowj and Corazon

Classic vs Bbyong on Outboxer VOD

In our first example, we have Classic vs Bbyong in an extremely important semifinal match for the coveted Grand Finals spot of Proleague 13-14. Although many games were played over this three-day span, each and every one of them were important in deciding who would go to the Finals and who would have to wait four more months for next season..

Classic spawned in the bottom right of Outboxer, a map that has been a staple of Proleague for the past three rounds. His opponent was Bbyong, the Gangnam Terran, who spawned in the top left. Classic opened up fairly standard with a 1-gate expo, while Bbyong varied slightly from a normal Terran fast expand, building a gas on 15 supply. Both players continued to play out a normal game until Bbyong threw down a fast factory at 4:15. This opener allows for Terran players to be extremely flexible and potent, with strategies ranging from hellion drops to cloaked banshees. Classic continued to play a standard robotics facility strategy that involves getting out a fast Colossus. Classic responded to scouting a 15 extractor (a sign of a delayed tech build) by getting his robotics bay right as soon as possible, delaying his second and third gateways. He knew he could get away with a quick Colossus because Bbyong's opener did not allow him to put on aggression very quickly. Because of this, Classic was able to get a Colossus out by 8:00, just in time to deflect Bbyong's medivac aggression.



Bbyong continued to open up the amount of options available to him by adding tech labs to his factory and starport. He made a siege tank and a raven while throwing down a third command center in his main base. With the Raven, he was able to snipe Classic's observer, only revealing Bbyong's 1-1-1, a few marines, and a tank. With the observer out of the way, Bbyong added two more factories and continued his mech production, unbeknownst to Classic. Bbyong started his simcity on the third base, adding bunkers and supply depots intended to barricade himself on three bases. However, he made a very quick fourth base at the island base on the bottom left side of the map.

As long as Classic did not scout the hidden expansion, Bbyong knew that he would be free to turtle on three bases and not be forced to overextend himself by taking a normal fourth base farther away from his core. Classic had no clue that any of this was going on, as Bbyong's opening and drop harass signaled nothing out of the ordinary. A raven and a siege tank are units that make up mech compositions, but only seeing one of each was not enough for Classic to definitively say that Bbyong was going to be playing mech. Classic continued to make Colossus and Blink until he sent a hallucination across the map and saw Bbyong turtling up with siege tanks.

Classic had neither the tools nor the economy to break the stronghold Bbyong had created. He attempted to attack Bbyong's third as quickly as he could, but he simply did not have the right type of army to go head-to-head with Bbyong's mech force. At no point in the game was Classic in a position to be ahead, instead consistently falling farther and farther behind. After his big attack failed, Classic tried to take a fourth but had a very tough time defending all of his mineral lines from Bbyong's hellion attacks. Every time Classic would move up the left side of the map, Bbyong would respond by sending hellions down the right side and vice-versa. After losing 50 workers and failing to break Bbyong yet again, Classic was forced to tap out after being absolutely stomped by his terran opponent.



So why was Bbyong able to fool Classic? It was a mixture of Bbyong's strategy and Classic's lack of attention to detail. Bbyong opened with a quick factory, which allows a Terran player to do anything in the book. His marine aggression was designed to make Classic think that he made a factory and a starport simply for a quick drop, nothing else. Bbyong made a Raven in order to kill the observer, giving him the time he needed to build two more factories and get his simcity set up. Classic's hallucination scout should have come a few minutes early and not at 12:15. Classic simply fell into the "metagame trap" and expected his opponent to play standard. Once Classic found out that Bbyong wasn't playing as expected, he panicked and decided that attacking Bbyong's army was the best course of action, which it was not. Alternatively, Classic could have tried to play passive and expand while amassing a better army composition versus mech but he still would have been hard-pressed to catch up to Bbyong.

Rain vs Bbyong on Outboxer VOD


Two days after dismantling Classic, Bbyong would yet again face another Protoss on Outboxer. This time, it would be Rain, the Protoss known for playing safe and standard, sparing no expense to survive the midgame and win in the lategame with a superior army.

Bbyong spawned in the same position on Outboxer as the previous game in the top-left corner, opting to go for a reaper expansion this time around. Rain spawned in the bottom right and opened the same way as his teammate Classic did with a 1-gate expansion. The game continued on just like before with Bbyong executing a 1-1-1 with a tech lab on both his factor and starport. Rain made a robotics facility and pumped out three observers. He managed to see that Bbyong made his buildings in the exact same location as before. Along with this, another observer saw Bbyong's medivac heading for a peculiar place; the top right corner of the map. These signs pointed towards Bbyong doing the exact same build as before. Now that Rain knew Bbyong was going mech. How would he respond? By making three stargates and a fleet beacon, of course.



After making his first three carriers(!), Rain headed directly towards Bbyong's ninja expansion in the bottom-left corner, having incapacitated the other one with a warp prism attack. After putting the two ninja expansions out of commission, the game became pretty simple for Rain. Bbyong was forced to take his fourth base earlier than he wanted to, spreading his army extremely thin. This allowed Rain to take as many bases as he wanted, knowing that his economy would be untouched as long as he could defend the hellion harass. While Bbyong did manage to kill at least 20 probes with a rather fortuitous hellion attack, he simply did not have enough air units to deal with the might of the carriers and tempests. Rain was able to stroll through Bbyong's defenses, forcing the gg from the CJ Entus player shortly after the 22-minute mark.



The fact that the same strategy worked against Classic but not Rain demonstrates the nature of Proleague. From his pre-match preparation, Rain knew that Bbyong could potentially do this build, seeing as it was so strong the first time versus Classic. The turtle mech style works really well if not pressured and Rain's reputation as a player who plays as safe as possible convinced Bbyong that it would work again. However, Rain's preparation and excellent scouting tipped him off to Bbyong's build and allowed him to counter it absolutely perfectly.

herO vs TY on Outboxer VOD

While the matchups in Proleague are different due to the format, the maps can also make a difference in strategy choice. In this final example of PvT, TY shows us how playing the map can be just as good, if not better, than playing the matchup.

In this game, both players go for a standard opening. KT's TY went for a command center first build, while CJ's herO went for a 1-gate expansion. It is important to note that TY went for a quick refinery after his first barracks, as opposed to the 2-barracks opening that usually follows. This was risky as herO could have chronoboosted out a stalker/zealot/mothership core and put a lot of pressure on the Terran. However, the calculated risk paid off. TY went for the same 1-1-1 opening as Bbyong but decided to make two quick siege tanks and a medivac, setting up his tanks right below herO's natural mineral line. This attack killed a few probes but also denied one of the gas geysers, crippling the Protoss tech advancement. This went on for a while until herO decided to send his entire army to clean up the siege tanks and the few marines guarding them. This proved to be a huge mistake as TY unleashed a triple-pronged attack towards herO's third, natural, and main that herO had absolutely no way to deal with, losing a crippling amount of probes. Behind all of this TY had made a third command center and possessed a sizable bio army. After consolidating his forces, he simply pulled SCVs and rolled over herO.



By utilizing an unexpected build which takes advantage of the map's layout, TY is able to easily beat herO, never really looking behind at any point. I'm sure that if herO knew that build was coming, he could have found a way to deny the tanks from ever getting close to his mineral line. However, that is the beauty of Proleague. Many builds are designed to only work once because they only need to work once. Bbyong's turtle mech play and TY's quick siege tank drop are prime examples of this type of mentality and strategy. As skilled as one needs to be to come up with builds such as these, one must also be wise in choosing to do these builds. Having faith in your build is one thing, but sometimes misguided faith can cost you a crucial win. TY's build was not used again, while Bbyong used his turtle mech style three times in three days against the same team, losing twice after being successful on the first try. The moral of the story is that the level of innovation is matched by the level of preparation shown in Proleague, which makes for some unique games and for some one-sided stomps, often both.

TY vs Soulkey on Yeonsu — VOD

This is from Round 1 Week 3 of Proleague, KTRolster vs SKT1. These two team league titans clashed and, despite Parting taking out FlaSh for an early SKT1 lead, KT Rolster was able to pull out a 3-1 victory in heavily unfavorable match ups. The key figure in this turnaround was TY. TY vs Soulkey on Yeonsu was the linchpin of this match by preventing the heavy SKT1 lineup from building up momentum into an eventual ace match. In order to secure the lead, TY channeled iloveoov and showed some of the greatest build prep seen yet to date.

The game opened with a proxy 2rax, although a relatively safe one. Due to the nature of Yeonsu, it is easy to proxy barracks at the third and then lift the barracks to your main quite safely over the cliff. TY took advantage of this and began bunkering up the natural cliff into Soulkey’s second base rather than pulling scvs and going for a make or break timing. This ended up working wonderfully, as Soulkey’s natural went down uncontested and TY completed his three bunker wall-in at the front of Soulkey’s base.

Rallying back, Soulkey took down the wall with a baneling bust, and attempted to get aggressive on TY’s side of the map, but TY was more than prepared for this with a solid wall of two barracks and a CC locking him in safely. From here, the game stabilized. Soulkey went up to three bases, and TY sat safely on two bases with triple CC. This was the calm before the storm as TY prepared his next move: a tank/SCV/marine drop on the cliff above Soulkey’s 3rd.


Arming the strategy

Using his tanks on the highground to cover for his ground forces, TY dropped his marines in the mineral line and performed an incredibly micro-heavy strategy with excellent execution, dealing immense damage to the economy of Soulkey, killing several drones and eventually the hatchery as well. The marines were eventually cleaned up, but the SCV that was dropped with the siege tanks had made three turrets which prevented the tanks from being cleaned up by anything but the most determined of muta balls. This forced Soulkey to remake his third base in the alternate, less more exposed position than where he would like it to be, playing to the next step in TY's plan.

The next stage of TY’s build was to begin his push toward Soulkey’s newly located third and siege it from the low ground. To accomplish this, TY sent out a drop first to pull Soulkey’s army back so he could push out in safety. Soulkey was prepared for the drop with mutalisks and a well placed spore crawler to block any drops moving up from the third top the main, but TY’s preparation served him well, and he had a fallback drop point planned out.


Contingency plans

This positioning allowed TY to completely avoid any ground forces from taking out his marines, forcing Soulkey to react with only mutalisks (which in small numbers would take quite a bit of damage), further delaying Soulkey from being out on the map. As the drop was cleaned up TY pushed forward with his mine bio push on the back side of the cliffs below the 3rd of Soulkey.


Abusive as hell

This forced Soulkey to engage from awkward angles around the back of his third base on the lowground with his zerglings and banelings alone while his mutalisks had to protect from a drop into the mineral line itself. This kind of inefficient engagement cost Soulkey his new third, and he was forced to remake his third base in the original location, where the tanks were still waiting. Soulkey was unable to get rid of the tanks at his base due to the three turrets that were still there, so all TY needed to do was reload his strategy with an infusion of fresh marines to kill more drones and eliminate the hatchery.


reloading the strategy

From here the game was all but over. With no third base and TY on four bases, Soulkey performed one last ditch effort to equalize, but was forced to tap out.

What is incredibly interesting about this game is how like a normal standard game it was. It took basically some very common ideas like aggressive openings and widowmine/bio pushes and fused them together by adding stepping stones -- in this case, a tank/turret drop and marine drops -- to get from the early game to the Terran-favored mid game. This level of build prep has yet to be seen commonly in individual leagues due to the differences in mechanics and scouting that is so variable across players. As a result, this level of preparation seen in the Proleague format draws in viewers eager to watch fascinating and well thought-out game plans.

RorO vs Rain on King Sejong Station - VOD

A different sort of preparation can be seen in RorO vs Rain from SKT1 vs Samsung Galaxy in the Round 1 playoffs. This was an incredibly close game in an even closer series. In the end, SK Telecom won 4-3 before going on to get stomped by KT Rolster in the finals. As SKT1 got up to 3-2, Samsung sent out Roro to take on Rain.

Before analyzing the game, it is important to go over RorO's strategy for this match. He knows that Rain is an incredibly robust player that utilizes a heavily defensive style to get to the late game, where control and map awareness becomes king. Taking advantage of this knowledge, RorO opted for an incredibly greedy opening that would transition into an incredibly passive mid game where RorO would be able to set up his own late game plans. He wanted an end game Zerg army that could go toe to toe with a late game Protoss, and that army relies heavily on swarm hosts. What follows is a great example of build innovation and refinement that everyone should watch.

The game opened mostly standard with a 15 pool vs FFE and both players opting to send out scouting workers to block their opponent's natural expansion from going down precisely on time. From there, RorO continued his economic focus into a fast third base, standard play vs a FFE build.

This early game fit perfectly into RorO's game plan. His strategy did not rely on doing early damage and, in essence, ignored this portion of the game, trying to cut directly into the middle and later stages of the match. The mid game started around 9:30 when he threw down a fourth base and a spire. This was quickly followed up by a roach warren, infestation pit, +1 missile and +1 air attacks, roach speed, Enduring Locusts, burrow, and a host of other mid game upgrades. This all exploded out around 11:30. Up until this point, RorO did his level best to not make any units other than a pack of zerglings for map control. However, once he had all the upgrades he needed started, he began producing corruptors and swarm hosts while continuing his upgrade frenzy with +2 missile, +1 carapace, and burrowed movement.


Upgrade ALL the things

By 15:00, Rain's passivity had come to an end and he began posturing around the map. Meanwhile, RorO remained passive with swarm hosts while implementing different pieces of his overall game plan; he was rushing straight towards the late game. He deployed roaches with burrow roach movement all around Rain's territory, doing almost zero damage but buying time and delaying for the next phase of his plan. As hive finished, RorO started yet more upgrades, again throwing up the signal: "Not now. Now is not the time".

At 18:45, RorO started his viper production after Rain secured his fourth base, but RorO was still not ready to begin moving out on the map. That would be far too soon! Instead -- he got more upgrades. With double spire up, he started +2/+1 air while getting melee upgrades to round out his ground composition. He continued to distract Rain with burrowed roaches while defending at home against various phoenix/colossus pushes and mass zealot warp prism harass. He did an admirable job defending, and eventually Rain was pressured into assaulting the forward fifth base of RorO in a desperate attempt to slow down the super tech-heavy style that RorO was using.



Rain had amassed a considerable army by that time of void ray/colossus, but even that was not enough to crack the nut of RorO, trading armies and still keeping his 5th base up. While defending, RorO rebuilt a spire that was sniped during earlier harassment and added on the final missing piece to his composition - the infestor. With 8+ infestors out plus his pre-existing viper/corrupter/swarmhost force, RorO had finally stabilized on four bases with his dream composition and was ready to set his plan into motion, over halfway into the game.

However, Rain had something to say about it. He sniped the fifth base of RorO twice while maintaining pressure to keep RorO at bay and on the other side of the map. This was done at great cost to Rain, and after several skirmishes Rain ends up down in supply and unable to remax, while RorO keeps up his burrowed roach harass and finally begins sending waves of locusts to Rain's sixth base. At 36:00, RorO took an incredible fight using fungal/abduct/corruption, giving him an incredible lead of maxed supply to a 140 supply Rain. This quickly snowballs into cancelled bases and yet more resources lost by Rain, from which Rain never really recovers. He is forced into remaking stalkers to make up the lost supply, which do not trade well with swarm hosts, and the cycle continued to viciously repeat.

Up to this point, RorO played patiently and he wasn't about to throw away his lead by being too aggressive, so he continued his passivity, allowing Rain to posture up and down the map while RorO retained his crucial mining base. From this point, RorO finally began his game-winning siege of the last mining Protoss base and forced the engagement. Rain musters a valiant effort, knowing that this was his last shot. However, Rain's army was torn to pieces once again by RorO's incredible, fully upgraded, caster-heavy army.



The preparation and practice that RorO had to have put into this build is truly mind-boggling. This game went an incredible 50 minutes, and in order to have prepared this build RorO would have to have spent days of man hours just for this one game, on one map against this one opponent, and that's just the actual practice of the finished build. The creation of the build and the refinement would have taken even longer to perfect, even while working with practice partners and coaches.

This is again one of the benefits of the Proleague format, as it forces great focus onto individual matches which leads to astonishingly deep study and performances in an already deep game in order to succeed. The successful players in Proleague work harder than any other players in the Starcraft II scene in order to be able to step up to the plate for their team.

Conclusion:

These games are both examples of the highest level of preparation in all of Starcraft, but they are prepared in different ways. TY spent his time leading up to his match fusing ideas into a never before seen strategy that would catch his opponent off guard while remaining strong against anything thrown at him. He did this by studying Yeonsu and examining what works in the early game (a long-distance proxy 2rax) and what is strong in the mid game (a tank drop abusing the cliffs overlooking the common third base location of Zerg) and developing it into a standard-looking finishing move: a biomine parade push to the alternative third base.

RorO on the other hand, focused much more on the player than the map. Of course, the map was taken into consideration when constructing the strategy on some level, as swarm hosts are heavily map dependent and they were the linchpin that held RorO's build together. But even more than this, he spent an enormous amount of time studying his opponent and learning his tendencies and typical build orders. This knowledge allowed RorO to safely construct one of the greediest and tech-heaviest builds we have ever seen in professional Starcraft II play. His build wasn't a gamble. It was a constructed device meant to punish the typical play out of Rain on that map, and it worked beautifully.

Bbyong's decision to play mech against Classic, Parting and Rain in succession may have been a product of supreme confidence in the style, but it might also have been a decision to play an alternative style in an effort to throw his opponents off with a unique build and manipulation of information. He displayed the full power of this in the first game, while all the weaknesses were put in display in the second and third games.

These games are examples of what you won't usually see in the GSL. Sometimes this level of preparation is shown in individual series, and those are usually some of the most famous games of all time: Leenock vs jjakji on Whirlwind, MVP vs Squirtle, Maru vs Rain. These are good examples of match preparation that led to some of the greatest games of all time. In Proleague, these sort of games are seen every week due. In an individual Bo7 series, it's often better to just drill mechanics and make sure you're in peak condition so that you can play a slugfest series and take the win in a standard fashion. In Proleague, the emphasis is on just one game, on one map, and one opponent every week. This inevitably shifts the focus from purely mechanical play and puts the ball squarely in strategic choice's court. The result is beautiful and, first and foremost, effective.

Writers: jowj, Thaniri, Corazon, Jer99
Complimentary writing: Zealously
Images: KeSPA
Editors: Zealously, SC2John