Note: This is an editorial. The opinions expressed by this article do not reflect the official position of TeamLiquid.net or its staff (other than TheDwf).

After the disaster of WoL's final days, HotS first came as a youth therapy. The mere fact that one could now breathe freely in game without being instantly covered in deadly green fungus, or that we were finally spared the dreadful chorus of "GREAT FUNGALS!" (yawn) promised a new dawn for the game. Starcraft II was anew terra incognita, and that alone made it exciting for people who had long dwelt upon the beaten tracks of the broods/infestor desolation.

Unfortunately, those days of prosperity were not to last. Not learning from their previous mistakes in WoL, Blizzard's misplaced meddling resulted in a sizeable casualty: an entire race is struggling since the autumn of the last year. If by any accident you are knowledgeable in the lore of yesteryear, you may know it as Terran.



The game has come a long way from the Warhounds a-moving to victory in the beta—before being promptly removed—to a third of the spectators being reduced to prayers each time Maru enters the booth in GSL. If you remember Terran at the beginning of HotS, it was definitely a strong race. Some would say it was broken. The LR thread was absolutely shocked when Mvp boosted his Medivac over Grubby's 4 Stalkers or when the King of Wings dropped to death the King of the Orcs in the following game on Cloud Kingdom. Boost Medivacs were broken for sure, there could be no other explanation. Did you see how fast they are? Hell, he even had some units in position yet could do nothing to stop them. Such is the effect of novelty against the unprepared: the winning side looks untouchable, and therefore is OP by necessity.

Following that same pattern came the days of 4M parade pushes: when the armada of Korean Terrans came to conquer the MLG in the first "official" days of HotS, the same triple OC dual EB bio/Mines builds which, a couple of details aside, are still in use as of now were already there; thus Terrans were overall in advance regarding the bulk of 4M play, and Zerg players could of course not come up with instant, perfect answers to the new standard—no more than Protoss could magically understand all the options unlocked by the MSC. Unsurprisingly, most of the Aiur blades were still playing a rather conservative WoL style. It was no coincidence that the Protoss who went the furthest at this tournament, MC, was the one who had immediately incorporated newly implemented stuff (i.e. the Oracle) in his play—on top of all-inning so he did not have to deal with Terran's standard play anyway (1).

Completing the trinity of pain, last came the Hellbat drops; SC2 was set ablaze for a few weeks with Hellbat drops reigning supreme in all match-ups. The first flames seemingly sprang in Europe, with for instance LucifroN carrying the torch in the ATC or the infamous ForGG vs Bunny 30 minutes barbecue mirror on Akilon Wastes in WCS Europe, but after a while Korea ignited too and for a few weeks the traditionally Protoss-dominated Proleague was nothing but Hellbat drops, much to the dismay of Zerg and Protoss (and soon enough also the Terrans themselves as they were forced to play their mirror).

It was logical that Terrans had the upper hand at first (2): it was relatively simple to incorporate the new tools—Hellbats and Mines, on top of Medivacs which were already there—in Terran's standard bio play, while Zerg and Protoss virtually had to rebuild from scratch their "reactionary macro play," taking into account the new threats around to create a modern standard, and learning how to defend the revamped timings and bio agression. Blizzard simply had to monitor whether Zergs and Protoss were catching up in the next few weeks/months or not to see if the reason for Terran success (still confined at the top, mind you (3)) was a temporary advantage in strategies or if some kind of fundamental imbalance was at work. As months went by, it was obvious that Zergs were catching up. Learning how to optimally control lings/banes/mutas, defusing Mines, defending Medivac pushes better, having less troubles securing their fourth, refining transitions, etc. was a slow and demanding process, but progress was being made. As for Protoss, they too were exploring the various new build possibilities offered by HotS; in particular the now viable Stargate tree opened by the addition of the MSC. In short, as the novely effect wore off Zerg and Protoss players were closing the strategical gap by incorporating their new tools. The game would stabilize. Where? That we shall never know.

What is important to note is that Terran outside of Korea was still not shining. Think what you want about the state of Terran during the first months of HotS, but we did not witness a generation of "patchterrans" regularly winning against superior opponents in dubious ways. The attempt at building a "releaseterran" meme did not last. A series like SjoW vs Life was not the upset of the year for nothing; such a thing remained exceedingly rare. The mechanical requirements of the race alone make sure that mediocre players cannot be carried very far by the brute power of the race; Terran requires an already high skill to unlock its potential, which naturally denies heathens entrance to the holy temple.

For instance, regardless of what you thought about Hellbat drops back then, the fact is they required a lot of multitask to be an efficient opening: toasting a bunch of Drones was of little use if you found yourself floating 1.2k minerals with no buildings launched after the action. In this game or this one, you can see Bogus is certainly not lying idle.



In this other example, Bogus neither managed to land the Hellbats nor played a transition requiring to swap add-ons; had it been the case, it would have been much harder to optimally execute the opening (4). Incidentally, notice how sOs had no troubles defending the dreaded Hellbat drops, no more than Jangbi here against Flash. As for 4M vs Zerg, it required a ton of micro on top of a superb regularity in macro cycles, which was the reason behind Bogus and Flash's apparent invincibility with the style, even while the rest of Terrans looked oddly mortal.

Then came the Hellbat nerf. Despite the official justification, it was not (mainly) done for the sake of TvT. Had this truly been the case, the new values would have taken into account the fact SCVs and Marines have 45 hit points instead of 40 for Drones and Probes: the Hellbat's attack against Light targets could have been reduced to 20, 21 or 22, allowing SCVs and Marines to survive two attacks while the other workers would have still been destroyed. In particular, the 20 solution had the advantage to allow Drones to sometimes survive two shots thanks to the Zerg regeneration (Hellbats are not always perfectly synchronized in their attacks), while Probes would have still been destroyed in the dual wave of flames (5). Thus Blizzard had a lot of leeway for the nerf: they could have specifically targeted one match-up or two. The axe blow was directed at the three.

Consequences? The complete extinction of all pressure openings involving Hellbat drops. TvT was thankfully saved from the absolute disgrace that was the mirror Hellbat drops era, but this change came at the cost of one of Terran's staple pressure openings against Protoss. Without any compensation: Photon Overcharge completely killed Banshee harassment, so Terrans had absolutely no use for this "counter-buff" in TvP. What's more, the Hellbat quickly disappeared from TvP bio compositions; teching Blue Flame to get back their original power in midgame is impractical (6), and without it they simply don't deal enough damage to the hordes of Zealots to be worth it.

In TvZ, the cheaper Cloak cost would allow (or rather favor) the return of the Hellions/Banshees openings, but here too the Hellbat nerf left a void: the Hellbat drops into Hellbat pushes (e. g. see here or here) that gave Terran a strong 2-bases pressure into third option were no longer working, as Flash experimented at the Dreamhack Bucharest: without the threat of an early Drone barbecue, neither YuGiOh on Bel'shir Vestige nor Dimaga on Neo Planet S had any problems holding it, remaining decidedly ahead in the macro game that ensued.

Not many tears were shed for the Hellbat nerf. TvT had became so ruled by the HPM (Hellbats Per Minute) metric that the vast majority of players and spectactors were happy to see the match-up being saved from this dark age; you know one match-up is terrible when even the prehistoric 4G mirrors on Tal'darim were more pleasant to watch. Zerg and Protoss could opportunely hide behind the TvT excuse to have this unpleasant threat removed from their own vT match-ups, thus strenghtening their standard builds and even opening a bunch of new possibilities for them (pure lings/queens defences again, room for mass Zealots compositions), possibilities that would become problematic for Terran down the road.

---

Even though Terran was already on a slippery slope, gradually losing ground in both match-ups, the race was not yet doomed. Code S representation was still fine, with 11 Terrans in Code S in RO32, 5 in RO16 and 2 in RO8. In September, various Terrans were still winning Premier Tournaments: Polt, Bomber, MMA won WCS events while TaeJa won Dreamhack against Bogus in a TvT finals. At this phase, it would have been very interesting to simply keep monitoring the game and see what would spontaneously emerge from there. But Blizzard didn't call it quits, and instead came the infamous and uncalled-for patch 2.0.12:


Along this, Bel'shir Vestige, Derelict Watcher and Star Station were replaced with Habitation Station, Heavy Rain and Daedalus Point.

The aftermaths of those changes were devastating. Terran's representation in Code S brutally collapsed, going down from 11 to 3 in the span of one season. Virtually all Code S regulars were exiled to Code B. Protoss maintained a humble 80% winrate vs Terran in Code A, going a modest 31-8 in maps and 15-1 in series (for an element of comparison, keep in mind Zerg was winning "only" 65% of their maps in GSL vs Terran at the end of the famously Zerg-dominated era of WoL). Terran spent 7 months without a Premier win before TaeJa wrote the hyphen between himself... and himself. Since that period, Protoss won 12 Premier tournaments, Zerg 5 and Terran 4. But more than the mere amount of titles claimed by each race, the list of winners speaks for itself:


Enough said.

Even if you consider Major tournaments along Premier ones, do you find more different Terran winners? Nope. To find a triumphant blue eagle again, we have to further downgrade at the Minor events (link). At the table of crowns, Terran eats last.

In 2014, the Terran representation in WCS Premier League is no more brighter:



Compare those figures to the ones from 2013: the supposedly OP Hellbats and Mines resulted…



… in a nearly perfect racial distribution. Go figure.

---

Such is the law of the jungle: where danger roams, falls first the weakest link. The Terran downfall is best illustrated by looking at what happened to some Code S players, in 2013 when Terran was still a good race, and in 2014 when it is massively struggling.


KeeN has never been a championship contender, but he's always been a solid Code S player. You would regularly see him in RO32 or RO16 each Code S season. Invariably, he would drop to Code A but would return to Code S the following season. After the disaster of last fall, KeeN was demoted from Code S regular to Code B regular.


Much has been written about Flash's performances in GSL, both seriously and in jest. In the whole year of 2013, the tragicomedy of Flash vs RO16 would inevitably repeat itself. Flash would dominate his RO32 group, dropping a single map to hyvaa's unscouted bane bust in the span of a whole year, then would be placed in a group of death (Bogus, Life, PartinG; Bogus, Bomber, Bbyong; Maru, LosirA, PartinG) and drop to Code A from there. Still, his 16-1 record in RO32 and Code A was a pretty clear answer to the question of Flash's level. Perhaps moved by his personal tragedy in the RO16, Blizzard's Balance Team made sure no such thing would occur again; Code B has no group of death.



Similar itineraries: players who were swaying from Code A to Code S RO16 were kindly invited to drop their slot to more deserving Zerg and Protoss players.

In this case, the rules in the new jungle were so brutal that the "weakest links" included several players of high caliber; some would even fit in a top10 world Terran ranking. So much for Terran being the "elitist" race: united by the Balance Team's undying love, the scrub and the gosu now share the same cold bench in the Code B void.

Only a few chosen ones survived the upheaval, and "survived" is the adequate word there: they maintained their Code S status, but did not go far before falling.


Probably one of the most underrated Korean Terrans, the WoL veteran SuperNova managed to retain his Code S position. Had you been told Code S would have only 3-4 Terrans left, would you have named him among the survivors? Certainly not, yet SuperNova is there with (or thanks to?) his unique builds and rather unorthodox play. Nonetheless, when you saw him play, you couldn't help but feel the despair in his games. Whatever he would try, the flame of his top6 OSL run was no longer there.


Bbyong improved tremendously in the last months, upgrading from a solid Code S player to one of the best Terrans in Korea, but he didn't have the opportunity to reach the playoffs for all that. When he was in position to do so last season, herO (who had lost 1-2 the first series) wisely built two Councils in the decider's match.

Youngest Yet Oldest

Since his OSL win, Maru has consistently been the last Terran standing in GSL. The last Terran title in Korea—already over a year ago—belongs to him. What makes Maru unique? Master of all styles of Terran, from hyper aggressive bio to the much slower-paced mech, his willingness to violenty gamble with risky cheeses combined with his exceptional micro in macro games, and probably the aura surrounding him in Korea, make him a chaotic force to be reckoned with. But there is more than just that, and here's the link with the current most successful Terran outside of Korea: TaeJa. You could also add Polt there; even if he had some slow-down lately, he fully belongs to the argument.

Aside from their race, what do those 3 players have in common?

Behold the first GSL Open Season… in 2010.

They were already here at the beginning. They started competing at the top four years ago. Of course, they were not immediately succesful, but since that time they gathered a colossal amount of experience. Is it a coincidence that the most successful Terrans during those troubled days are the ones who went through so many trials and eras? After all, they saw GomTvT, they witnessed Lings of Liberty, they experienced Heart of the Storm. Without his broken body, odds are Mvp would probably be here along them.

One question mark has to be written on TaeJa's résumé, though. TaeJa no longer competes in Korea since the beginning of HotS; therefore it remains uncertain how he would fare against all the top players in Korea since he never has the opportunity to meet some of them in tournaments. After all, at the Homestory Cup or Dreamhack, the likes of soO, Soulkey, herO or Zest are not there. There is no reason to believe he would be outclassed, but it is not said either he would dominate week-end international tournaments if there was a higher amount of those top Korean players competing in them. No doubt his recent 17-0 win at the Dreamhack was extremely impressive, but it falls under the same question mark. Besides, let us admit there were harder paths to victory than running a gauntlet of Prebs MorroW Harstem Oz Sjaak Patience Jaedong HerO (7).

The exception to this "longevity rule" (although not fully so, given a BW background) is Bogus. Arriving late in 2012 with the KeSPA switch, Bogus has been a top Terran player since the end of WoL. His strength is the epitome of straightforward play: exceptional, virtually unparalleled mechanics in standard play that would put almost every player to shame. And yet, he could not avoid the pit that is Code B in the first season of 2014, nor was it enough to make him reach the Code S playoffs this one. Bad luck, or is something more needed in these rainy days for Terran?

At any rate, TaeJa's and Maru's extraordinary performances do not mean Terran is fine or healthy, no more than the precedent of Mvp and TaeJa still being successful at the end of WoL meant TvZ was OK. A few players manage to overcome the odds (8), but it cannot be expected that everyone could do the same.

---

So how does one explain Terran's downfall? Grotesque statements such as "Terran players are all slumping" or "Good Terrans all left Korea" will naturally be disregarded with a frown; it makes absolutely no sense that all top players from one race start underperforming at the same time, especially—surprise, surprise—after a heavy nerf occurred. As for the other "argument," only two Code S players left Korea between the critical seasons: Bomber and jjakji left WCS Korea while Hack (of Code S caliber) returned from WCS America. Unless 11 minus 1 (or 2 if you want to dismiss Hack) magically equalizes 3, we have to consider other factors than this mostly meaningless transfer.

The original onslaught came from Shakuras. The first heavy blow to Terran in Korea was indeed dealt during the Code A session of January, which saw only 2 survivors emerge; thanks to that, Terran now holds the award of the lowest amount of Code S players for two consecutive seasons in the whole history of SC2 (previously, Aiur was the keeper of that sad record with 5 Protoss in 3 GSL seasons). Responsible for this disaster was, first and foremost, the infamous Blink attack with all of its variations. Terrans fell unceremoniously under the wrath of the Twilight Council, and the two Terrans who would join a seeded Maru in Code S, Bbyong and SuperNova, did so… by dodging TvP.

The first question we have to ask is: what made Blink attacks (9) dominant? Blizzard and a large part of the community, especially Protoss players, blamed the maps for this. Heavy Rain, Yeonsu, Daedalus were pointed out for having too much vulnerable space. Blink was fine, but not on those maps. Just change the map pool and everything would be OK again.

At first glance, this explanation seems perfectly reasonable. After all, Blink was never a problem on a map like Habitation Station, void of any usable cliff. And yet, if you look deeper into it, this thesis is absurd. Or rather, it's superficial and fails to understand the roots of the Blink issue.

Let us compare the different map pools that were used in the last 2013 WCS Season and the first 2014 one:



Frost, Polar Night and Yeonsu were already there in the previous map pool, so by definition they had no particular effect on the rise of Blink play. Whirlwind and Habitation cancel each other: both maps are bad for Blink play, having essentially only the entrance at the natural and a terribly awkward Blink access from third to main. Neither Alterzim nor Derelict Watcher are particularly hot for Blink; the vulnerable surface is globally similar. Were Heavy Rain and Daedalus Point so much better Blink maps that it made all the difference compared with Akilon Wastes and Bel'shir Vestige? No. Defending Blink on Bel'shir was definitely easier than in Daedalus, but Protoss could perform the same "cast Time Warp on ramp, Blink where he is not" trick on both maps. Additionally, Akilon Wastes had a lot of surface for Blink as well; not as much as Heavy Rain for the main, but the natural was more exposed.

At any rate, if the problem was only contained to the two new maps, then it would have been solved easily: after all, players have vetoes. Just like Zerg always vetoed Yeonsu against Terran or Protoss, Terran players would have simply kept vetoing these problematic maps and played on an essentially identical variant of the map pool (for Blink) used during the third season of WCS 2013, thus preventing the issue from ever hurting them. Right?

No. Blink had suddenly become problematic on Frost, Polar Night and Yeonsu as well. And yet, Terrans had not been massacred by Blink attacks the season before. Why? Because the real reason for the 2-bases Blink dominance does not lie in the maps, but in the following patch notes:


The Mine nerf did not have the sole effect of crippling bio play in TvZ. It is also responsible for the PvT slaughter of January 2014 in Code A.

Think about it that way: what is the particularity of a 2-bases 6-7g Blink build? Before the attack starts, it has about 40 Probes, 3 Stalkers and the MSC. And that's it. From there, Protoss secures a proxy Pylon, starts warping Stalkers near Terran's base, pokes the cliff with his MSC (at this time, keep in mind the MSC still had its wide 14 sight range (10)) and Blink wherever the Terran is not while preventing Terran bio from having a correct formation/position with Time Warp. What generally happens afterwards is unknown to none: barring perfect defence, Terran either loses instantly or falls behind sacrificing too many SCVs to hold. But never mind that; let us rewind.

As previously said, at 7'30 Protoss has 4 units (one MSC and 3 Stalkers), 40 Probes and a few gates in production. What is missing in the picture? Detection. What could happen shortly before Protoss starts warping Stalkers and attacking? A Medivac carrying two Mines hits Protoss, lands one Mine per mineral line and Protoss is suddenly in a world of pain. The problem is that after the Mine nerf, Mine drops became absolutely terrible (11). Even when you managed to land one Mine in each mineral line, the outcome would be laughable: at the first shot, one Probe would die, two if you were lucky, whereas pre-Mine nerf hits were a guaranteed 5-9 if Protoss didn't pull his Probes out of the mined area.

Thus Protoss, previously prevented from attempting detection-free attacks by the threat of Mines landing in their mineral lines (12), were now free to tech quickly to Blink and get up to 7 gates on the back of 4 units and zero detection. Bio builds were the norm, and even in the cases where the Terran was going 1-1-1 after expand, the Mine drops no longer packed the punch required to cripple Protoss.

In short, the 2-bases Blink menace arose not due to maps, but because of the lack of early aggressive options (a role previously held by the Mine) that ultimately forced Terran into the vulnerable bio builds. This is why options, even if they are actually rarely used, are of utmost importance, because they can passively act as a weapon of deterrence for the other side. For instance, imagine that Cloak Banshees were still viable in HotS TvP, even against standard play (like in WoL). That sole threat would have given some pause to Protoss; just like WoL Protoss playing 4g pressure after expand could get a build order loss against a Cloak Banshee because they had no detection at home (13), Protoss trying to go 7g Blink would have needed to take some necessary risks. If Protoss was exploiting bio builds with 2-bases Blink, then Terran could have had the option of reacting with more Cloak Banshee openings, forcing Protoss to deviate from the most powerful 7g Blink to, say, 4g robo Blink, i.e. something way more manageable because it hits later and/or with less Stalkers.

Despite the existence of Blink maps during all of HotS (e. g. Star Station, Akilon Wastes), 2-base Blink was never truly problematic until Terran was robbed of its only counterpart threat, 1-1-1 builds. Teching 7g Blink out of 4 units was absolutely unthinkable at the beginning of HotS: a simple Marines/Mine poke would paralyze your mineral line at the natural for almost two minutes, and you would likely straight up lose to any Mine drop into expand build (14).

To put the final nail in the coffin into the "Blink was broken due to maps" thesis, consider how many of the core WoL maps would suddenly prove to be "broken" if they were used today. Holding HotS Blink on Antiga? A nightmare. Ohana? Too much surface to cover. Shakuras? Not a chance. Abyssal City? More like Hell City. Cloud Kingdom? Rename to Blink Kingdom already. Something else than the terrain architecture should be considered when suddenly half of the maps ever used in SC2 competition would become problematic in one way or another regarding a single PvT strategy.

Last but not least, what about now? The recent scarcity of Blink attacks may have given you the illusion that the issue has been fixed. This is not the case. If we consider the current map pool…


… we can notice that there are still 3 maps out of 7 (or 8 if you consider the two spawns of Waystation as separate maps) which offer strong Blink opportunities; and randomly proxying a Council on Alterzim probably still yields an absurd winrate. MC vs TY at the Dreamhack, PartinG vs SuperNova in Code S, Welmu vs YoDa in WCS Europe or Bbyong vs herO in the Season 1 Code S proved that Protoss players can still claim easy wins by building a Twilight Council on those maps.

Incidentally, getting rid of this even if the map pool was composed of 7 Yeonsu clones is absolutely trivial. When do Terran players start getting their second rax? If not going for a quick EB build, usually at 4'05 after a Reaper expand. This means the second rax is complete at 5'10, and the Tech Lab is done at 5'35; Stim takes 170 seconds and thus is complete at 8'25. If the Stim nerf was reverted, it would be done at 7'55. At what timing does 2-bases Blink complete, at the very earliest? Around 8'. Case closed.

Until that is done, the Blink wolf will keep prowling the night, always on the look-out each time a map offers a tempting cliff.

---

After the MSC sight range nerf, the Time Warp nerf and the introduction of the new map pool, Blink temporarily disappeared from the PvT scene; more due to the the psychological effect that inevitably follow nerfs than for actual gameplay reasons (as its recent return to prominence demonstrates). Nonetheless, Terran was for a while free from the Blink menace. Better yet, mech play was on the rise in TvZ, and the new Mines in TvP commanded respect from the previously out of control Templar openings. Every condition seemed to be there for Terran to rise from its ashes. When looking at the Terran list of Code A players, one could have expected 6-8 Terrans in Code S. SuperNova, Bbyong, Cure, Journey, FanTaSy, Flash, Bogus, TY, Reality: surely there was room for a Terran rebirth this season?…

The sequel of the story is known. There was no happy end. Compared with the previous season, only one extra Terran made it from the Code B pit: Bogus, i.e. the second or third Terran in the world (depending on where you place TaeJa). Terran was still at the bottom of the food chain, but this time it had rather fallen prey to a new predator—or perhaps I should say an old one, remembering the end of WoL: Zerg. The consequences of the Mine nerf had, at last, fully unfolded: Terran had diversity in strategies, as Blizzard had so kindly wished, but those strategies had also something uncalled-for in common: they were all equally insufficient against solid Zerg play.

Here again, the map pool excuse was drawn. The shift could be explained because the map pool was Zerg-favored: thirds were harder to secure and distances were longer. There is of course some truth to that. Yet was Terran struggling on Whirlwind at the beginning of HotS? Is it because of the new maps Terran no longer had any form of AoE properly working so their armies were repeatedly wiped out by mass banes regardless of how well splits were performed? Is it purely because of maps that the muta death cloud is unstoppable once it reaches a critical mass? Would Bogus, the undisputed master of macro games in TvZ, have felt forced to all-in Soulkey on cross Frost with the old Mines?

---

It is no wonder that Terran collapsed if you consider what changed between WoL and HotS. Out of the three main novelties bio play received in HotS (boost Medivacs, Mines and Hellbats), only the first one was left unscathed by the nerf hammer. The Hellbat nerf removed the unit from bio play for months before it came back in TvZ... and only because the Mine was no longer working as it had. Terran's bio play still largely operates on the WoL core, except it faces the HotS tools Zerg and Protoss received and kept. In a way, you could say Terran is one expansion behind. In jest... or not.

As for mech play—a long due promise of HotS—it is still completely extinct at the highest level of TvP. A few European players use it, but Goody or Strelok are neither top Korean players nor particularly successful among the European Terran ranks. Outside of old SuperNova or Mvp games, it was only played once recently in Korea (15); Maru shocked everyone by using it in a Proleague match against herO. Needless to say, success wasn't on the cards; despite an impressive defence from Maru considering how terrible mech is to handle Protoss' mobile threats, herO managed to pull him apart with Zealots/DTs harass while building a large fleet of Tempests.

In TvZ, mech (or rather mech into air) has been made stronger in the extreme lategame (16), but most timings have been killed by mutas, Swarm hosts or Vipers, and particular map layouts are needed so Terran can harvest out of the 10-12 necessary gas to build their extremely expensive fleet. Basically, macro mech is only viable on maps in which you can secure a fifth, and perhaps a sixth, without Zerg having the whole map against you; on the current map pool, this means King Sejong and Overgrowth in the best of cases (17). On other maps, Terran has too much troubles securing a fifth—or even a fourth—and is left slowly starving to mutas/SH harass (18). In short, mech is far from being the saving grace of HotS novelties for Terran. At the end of the day, among the Terran stars, Bruce Bio is still way more popular than Mike Mech.

In an eerie combination of factors, Terran's bio play faces common issues in TvP and TvZ:

1. Terran has less options, leading to a more predictable (and thus more easily "countered") play
   
2. Terran has a much higher vulnerability to all-ins (partially fueled by scouting issues)
   
3. Terran is way more unforgiving: mistakes and sloppiness are punished harder, and once the race falls behind there is almost no comeback potential (in particular due to the weaker reproducibility)
   
4. Terran has an inferior lategame.

The first point is the direct result of Blizzard's patches. The Balance Team indeed nerfed Terran while keeping untouched the corresponding defensive tools. For instance, dumb as the concept is, Photon Overcharge in its current form could justify itself in PvT when the powerful threats of pre-nerf Hellbat and Mine drops were still there. But now? Mutalisks regenerating one hit point per second was sound in the days of constant fights near Zerg's bases and lethal Mine hits. But now?

This observation applies to some previous Terran nerfs which are no longer justified by the stronger defensive tools at Zergs and Protoss' disposal (Queen/Overlord patch, the MSC). No doubt nerfing Stim was normal when SC2 was played on maps with wide open naturals and short rush distance, but now? What early game timing is to be feared in TvP when Protoss can have a 2000 hit points fortress as early as 5'30?

As said above, Blizzard could have used this to solve the Blink issue. Instead of adding yet another constraint on the maps, affecting the critical Reaper scouting in TvP in the process, simply revert the Stim nerf: Stalkers wouldn't be gods among the insects if they were immediately welcomed with stimmed bio the first time they Blink. Similarly, the earlier stim timing would make Protoss concede a few more units in the early game, automatically erasing all the grotesque builds that stay on one gate for 9 minutes. Why doesn't it happen despite this request being common? Only David Kim knows.

Overall, what leads to Terran issues is crystal clear. Bio play's strength is dependent on putting a "cascade of pressure" on your opponent; if you weaken pretty much every tool bio uses to achieve this while mech remains a niche strategy, the Terran ship is doomed to slowly sink. Hey, who could have guessed that? You're certainly Nobel material if you did.

---

Let us rewind the tape to examine the "4M was imbalanced" claims from the earliest days of HotS. The thesis can be summed up as follows: Marines and Mines being only or mostly minerals, Terran was very favourably trading minerals against gas (banes and mutas) all game long, blocking Zerg on Lair (and thus 2/2) while getting 3/3, resulting in such a cost-efficiency that Zerg would go bankrupt at some point and lose, asphyxiated by the never-ending stream of units rallying across the map (19).

This perspective was at best short-sighted, at worst flat-out wrong on several levels. Back to basics: before being gas, Banelings are minerals; an individual Baneling costs 25/25, but it comes from a 25/0 Zergling so its final minerals:gas ratio is 2:1. Unless my math is dramatically wrong, this still represents more minerals than gas. Besides, in a typical 4M vs lings/banes/mutas game, both players would start floating huge amounts of gas as the game went by.



A characteristic game from this era, Bogus vs Solar shows the Zerg dying with almost 5k gas in bank; something very odd indeed given the nigh mystical preciosity of the resource according to gas cheerleaders. Someone should have told this would-be Code S scrub to spend it! Bad luck, he could not; for the gas-heavy bulk of his army, mutas, was already complete, and he was mainly trading mineral units (Zerglings, and to a lesser extent banes). In the end, the very reason he lost is that he ran out… of minerals.



In this famous Scarlett vs Dream game on Red City, let us fast forward to 30 minutes in the game. Scarlett's resource report? Again, no minerals... and 4.5k gas. There are countless examples of this phenomenon (Bunny vs Jaedong on Habitation from the last Dreamhack, for a recent instance); past a certain point, Zerg too has no more practical use for his gas if he doesn't have a 4 mineral lines' worth of economy, because well controlled mutas seldom die, and assuming a dynamic game, a constant stream of minerals-heavy lings/banes is needed to shield them against Marines. (Nowadays, this imbalance in the resource ratio is much rarer since the days of constant trades are over; thus Zerg can get a better economy and spend their gas into Hive + 3/3 + ultras earlier.)

Another claim is that Terran was automatically favored in macro games because they would get 3/3 much faster, inevitably reaching superior cost-efficiency. If we keep the last game as an example, notice how Scarlett doesn't complete 3/3 before… 31 minutes, that is to say 12 minutes after Dream finishes his own 3/3. Can you imagine how much better Dream traded during all this time?

+ Show Spoiler +

Welp, no wonder there was so much anger about the Mine: what can you do against the unparalleled power of a 1.016:1 cost-efficiency, really? This might only be one example, but Zergs controlling well had no troubles matching Terran's 3/3 even when they were only 2/2 for a prolonged period of time (which could last up to 15 minutes at this time). After all, mutas are very cost-efficient since they don't die when Zerg microes them well and disengages at the right moment. Besides, the cost-efficiency of an engagement, for one side or another, depended on many, many more factors than a mere game of upgrades: to name a few, Mine detonations, Banelings connections, creep or not, the formation of both armies, Mutalisk count, Medivac count; just because a Terran was 3/3 didn't mean he was emerging triumphant from each engagement. And fortunately: if 3/3 vs 2/2 was a game-ending advantage, then indeed Zerg would never win macro games, even as of now. Is it the case? Of course not.

Shortly after the Overseer buff, games like Bogus vs Curious appeared: Zerg was stopping the parade push and actively taking back the initiative with muta harass, rewriting the "Terran attacks, Zerg defends" scenario of the past months. The following Code S season, DRG vs Bogus occurred; in a series which makes you wonder why Blizzard is so proud of this, the two players fought what is probably the most mechanically impressive battle of HotS TvZ. DRG emerged victorious with 2-1. The next month, Flash (who had insane TvZ, equal or only second to Bogus at this time) went 0-3 in bio macro games against DRG and Curious at the IEM New York. At the top, the Zerg adaptation was complete.

Yet after several months of a heroic resistance—can you imagine how much the arm holding the hammer nerf was itching during all this time?—Blizzard finally ceded and heavily nerfed the Mine in November 2013. The summoned pretext was "diversity," mirroring the cunning propaganda which, instead of claiming 4M was "OP," said it resulted in "stale" gameplay. Oddly enough, the Marines/Tanks composition played during 3 years of WoL TvZ had never been declared "stale." Go figure.

Naturally, since TvZ was no longer particularly favoring Terran, Blizzard gave from one hand what they took from the other: to compensate, an obsolete relic of the Terran arsenal (20) received a 7,1% increase in attack speed, an irrefutably fair trade considering the Mine probably lost half its potency with the nerf. With touching naïveté, Blizzard was apparently expecting that after those changes, Terrans would play bio/Mines/Tanks instead of bio/Mines without any reduction in efficacy (21) .

This was, of course, nothing more than a vain wish. In the reality of the game, such thing had absolutely no chance to occur as the Marines/Tanks dynamic, which doesn't work anymore in HotS (22) , is completely different from the Marines/Mines one. On one hand, Tanks trade awfully in low numbers: they're easily overwhelmed by Zerg's swarm of fast units, especially on creep, and are destroyed inflicting minimal casualties. Thus the Tank obeys to a logic of accumulation: to attack, you must first muster a healthy Tank count (which starts somewhere around 7-8); but the critical weakness of this modus operandi is that Zerg is given almost complete freedom before the big Tank push strikes, and we know what happens when Zerg is left alone minding his own business. On the other hand, the Mine is essentially a disposable single deadly detonation allowing constant pressure in rally mode should the opportunity arise: it hits generally only once or twice, but it hits hard; hard enough that Marines can fight on fair terms against the remnants of the Zerg horde. With Tanks, Zerg is given too much space in the air; with Mines, you can try to smother him.

If you compare both units, the superiority of the Mine is blinding. Want to build 8 Tanks? You need 1200 minerals, 1000 gas, 24 supply and 3 minutes from 2 Lab Factories. Want to get 9 Mines? 675 minerals, 225 gas, 18 supply and 2 minutes from 2 Factories (one Lab, one Reactor). Cheaper (23) , faster, hits air, dynamic, more quickly deployed and immediately more efficient: the Mine hands down beats the Tank in absolutely every aspect when it comes to fighting lings/banes/mutas (24) . Hoping that Terrans would build an ineffective unit because the main one had been nerfed was absurd; even post-nerf, Terrans are still much better sticking to the mediocre Mines than handicapping themselves investing in the dead weight that is the Tank.

Combined with Zerg's refined anti-4M play, the effects of the Mine nerf were flagrant. Terrans quickly lost their ability to pressure efficiently enough in midgame; keeping the creep in check and preventing Zerg from saturating too quickly his fourth was a critical point in ensuring that Terran doesn't get overwhelmed by the huge amount of units a 80+ Drone economy can produce. As a result of this slackened pressure, the original sin of the Queen patch, i.e. Zerg's development (creep, economy, tech) getting out of hand, resurfaced anew.

Games like this multiplied, in which Terran is left hopelessly witnessing an evasive pack of mutas ransacking his bases; barring a miracle of that kind, Terran has simply no answer.

O Zerg Micro, Where Art Thou?

Even more disheartening is the fact that with the Mine nerf, most of the finesse required from Zergs to engage is essentially gone. Just like the Baneling introduces the demanding Marine splitting technique, setting apart mortals from micro masters, the power of old Mines forced Zerg to actively defuse and split. Terrible a-moves were punished—as they should be.

It should be noted that the existence of strong AoE weapons on the battlefield is what introduces micro to units which don't necessarily have it built "inside them". See for instance Marines' behaviour when there is no AoE on the field, for instance against pure Roaches or Roaches/Hydras: in those cases, the optimal behaviour is to have a wide concave and a-move, something much easier to perform. Now take Hellbats, which by themselves are a simple a-move unit: in the presence of Banelings, they have to be splitted so they don't melt. Likewise, the old Mines forced lings/banes declumping, which complexified their in-battle optimal behaviour, considerably raising the necessary skill floor to win battles compared with nowadays.

Post-Mine nerf, simplistic charges too often overwhelm Terrans because Mines simply don't kill enough Zerg units so Marines can deal with the rest; as for Hellbats, they lack firepower against banes (the primary target of Terran's AoE weapons). What's more, with Hive being unlocked earlier, the Ultralisk now commonly appears on the field. Supreme winner in the contest of mindless beefy a-click units, the mammoth on steroids effortlessly chunks bio troops, further reducing the amount of micro needed to prevail against Terran armies.



soO a-moves a bunch of ultras offcreep, with Queens and out of mana Infestors lagging behind; several screens of hit & run later, 2 ultras fell for 70 supply of bio. Zerg's micro requirements in this sequence? Null: hands lifted from keyboard would have made literally no difference.

The disparity in micro required for each side in this situation is clear as day. You can see Maru microing his heart out of his units; meanwhile, Zerg is perfunctorily casting some useless ITs and a Fungal (25) ; only after 10 seconds does Jaedong awake to shift focus a few Medivacs with his mutas. Jaedong's awful engagement is left unpunished: cost-wise, Terran traded 1:2, but too few remains and supplies are even after the reproduction. Unless the fight takes place near one of Zerg's bases, Terran has no way to immediately capitalize on a won battle.

Is the reverse true? Alas, no: if Terran blunders and takes too bad of an engagement, the sanction is often immediate. Consider for instance the following sequence in Bogus vs Life, Frost, at the Dreamhack Bucharest: what happens after Bogus' forces get wiped (26) ? Zerg's army being well-educated, Bogus is kindly taken to the door of his natural, and there he is notified that he has been fired from the game, effective immediately. Against Jaedong, TaeJa is caught unprepared, his Mines are nullified and a disastrous engagement ensues. The outcome? Not even close: 30 seconds afterwards, Jaedong doubles TaeJa's supply. Bogus vs Life again? No difference: caught in a bad position, boom, dead. Trying to contain the purple tide in a spread position offcreep? Zerg unleashes the Swarm; goodbye, thank you, the end. Mistakes do not have equal impact in TvZ; one race has to avoid any imprecision at all costs to have a chance, the other forgives sloppiness and condones blunders.

Lategame is the phase in which this disparity is the most blatant. Something goes dramatically wrong for Zerg? Don't panic, for the great forgiveness of an immediate buyback is there: as long as resources are available, the legion of larvae awaiting near the hatches can be mobilized at once to raise a new army of mass lings and a few ultras; and voilà, you have your second chance. A large engagement results in a severe loss for Terran? Start packing your stuff. Even in midgame, when Zerg's bank (resources/larvae) is weaker, the Zerg powerhouse still produces enough current to mitigate small/medium failures; when Terran is ahead, converting the advantage is no easy task because engaging on creep is always delicate. When Zerg is ahead, Terran has no similar defender's advantage to hold the breaker and just gets overwhelmed.

They See Me Rolling…

To capitalize on Terran's newfound lack of a powerful AoE weapon, Zergs started morphing incredibly high amount of Banelings. The idea is simple: since Terrans can no longer get rid of enough banes via Mine detonations, all Marines are slaughtered and mutas reign supreme at the end of the engagement. This possibility was also aided by the fact Terran could no longer trade enough in midgame, so Zerg no longer needs to pour as much gas into Baneling replenishment, allowing the accumulation and eventual advantage.

Illustration with TaeJa vs Snute: reaching lategame, Snute builds the money composition of 30 mutas, a few ultras… and 66 banes, i.e. more banes than TaeJa has Marines. It goes without saying that TaeJa, despite no splitting mistake, fighting offcreep and 60 supply of mutas peacefully snoring nearby, gets absolutely annihilated. Maru suffered a similar fate when Symbol a-moved 77 Banelings to his 70 Marines. At this point, Zerg can almost alt-tab and browse Twitter if he fancies; even if Terran sweat streams trying to split every Marine down to the individual, he just gets mauled. The power of the mass is so ridiculous that Terran would still get rolled even if 20 banes had crashed on the Thor. Naturally, it barely matters if Terran has more Armored obstacles in the way: as evidenced by TRUE vs FanTaSy, all Zerg really has to do is right click the 70 banes past Thors or Marauders, and mutas will peck away at the rest while Marines are busy dying in different locations. But hey, those Terran scrubs know what to do if they want to win in that kind of situation:



They just need to take advantage of Terran's high skill ceiling: with the modest performance of 15k apm (translating to a mere 250 actions per second), this is absolutely doable.

Indefensible muta hords, uncontrollable masses of Banelings or unstoppable Hive armies: pick your poison. Actually, a delicate mixture is the most common option; with a healthy Mutalisk count still in the air, any kind of attempt from the Terran to stall with a never-ending initiative or to simply kill the Zerg through multi-pronged attacks/drops is made extremely difficult. Mutalisks are indeed the ultimate tool in TvZ: defending drops and harassment, rampaging Terran's economy and production, preventing the installation of extra bases, sniping Mines, killing Medivacs in engagements, cleaning up what ever might be left of Terran's army after each battle if Terran no longer has enough anti-air alive… Talk about versatility. WoL broods/infests at least had one clear weakness: mobility. HotS mutas + Hive units have none.

Perhaps the most depressing thing in this mess is that Zerg's skill ceiling is actually extremely high: in other things, DRG's monstruous mechanics (when at his peak), Leenock's cunning tactics or Curious' muta usage stand testimony to the depth of Zerg's lings/banes/mutas play in ZvT. There are so many aspects in which Zergs can shine, but it doesn't matter: half of them can remain dormant since mediocre a-moves of ultras and banes do the job just fine; Zerg probably has a lot of gas left in the tank, but it doesn't have to be used anymore. Why would you bother to multitask or to play in a dynamic way when you can sit and hoard gas juggernauts at home? If DRG and Bogus were to play the same series at the level they played back then, the outcome would not be a hard-fought, spectacular battle but two one-sided Zerg stomps.

My Kingdom For A Roach

At the end of WoL, seeping through the rotting shell of the Terran ship arose the infamous 1/1 Speedroach timing. The principle of the opening is very straightforward: Zerg saturates 2 bases, then produces mass upgraded Roaches and attacks. At this time, Zergs were taking advantage of the fact that Terrans were still building their infrastructure off the Hellions/Banshees opening. Nowadays, the same hole remains: most Terran builds still retained this vulnerability to the Roach onslaught. It is no coincidence if Roaches timings, a byproduct of Terran being forced to play on a razor's edge to keep pace with Zerg's macro builds, were the first haven Zerg found in the tough first months of HotS. The notorious hatch tech 3-bases Roach/Baneling bust was the grand Terran slayer back then, and it still makes its melodious voice heard from time to time.

Bogus went an insane 28-4 record in the current ATC season. Among his 4 losses? Only one non-Korean: Tefel, spamming the R key. What happened when Bogus was up 3-0 in the first HotS GSL finals against Soulkey? A simple bug made the machine jam. Symbol's recipe for success when he was advancing far in GSL? Bullying Terrans with Roaches cheeses. Flash lost 4 games against Zerg in Code A? 3 Roach cheeses in sequence. Maru's headache? Roaches. HyuN's grand ZvT plan? Overwhelm Terran with Roaches.

After all, it is no wonder why this kind of play is so popular when you consider the features of a timing like 1/1 Speedroaches:

1. easy to execute;
   
2. tricky to scout;
   
3. dangerous even when scouted (thin margin of error for the defender);
   
4. forces serious defensive commitment from most builds;
   
5. yet has smooth transitions to various forms of play (27) .

It is commonly said cockroaches would rule on Earth should humanity disappear after a nuclear war: Terrans can certainly vouch for that.

Hellbats: the savior?

Blizzard finally became aware of their own mess, and promptly tried to adjust things in an interesting way via the reintroduction of Hellbat pressure. The logic behind this move is sound: following the "principle of cascade" mentioned above, more early game options/threats could result in a stronger midgame, which in turn could appease the lategame woes.

Unfortunately, the patch thus far has been nothing but a massive disappointment: shortly after its release, Zergs have already adapted to the sparse new Hellbat openings. Maru's face below says it all, really.



EffOrt has no troubles defending the Hellbat timing with Roaches/Queens, and later counter-attacks with some acid of his own. Similarly, Life defended a hardcore variant of a Hellbat push in his Shoutcraft game against Flash. In order to have as many units on the field, Flash cut almost 15 SCVs for his attack (thus making it semi all-in), yet Life held despite immediately losing 3 Queens for nothing. Fortunately for Flash, Life unwisely opted to challenge SCVs/Medivacs, the best melee combo in the game, and failed his counter-bust; but otherwise he could have droned a bit and remained ahead (28) . Did Bbyong have more success against RorO in Proleague with his Hellbat drops? Not one bit. At the HSC IX, Scarlett fended off MMA's Hellbat timing as well.

The patch does introduce new threats, but not ones strong enough to fix the disparity of development as long as the usual scouting procedure and a Roach Warren (or an earlier Baneling nest) are incorporated into Zerg's build: after all those defences, Zerg is clearly ahead because in order to put that much pressure, Terran has to remain on 2 bases. Zerg can of course straight up die, but only if he didn't scout and kept playing pure lings/queens + dual evo; another possibility is that he mistimed his Roaches and suffered too much damage before the eggs hatched, but all of that falls under the category of youthful errors.

The core of Zerg's 3 hatch builds is versatile: you get an early third, but you can forgo its saturation until it has been made adequately safe; you can build a Roach Warren, but of course you don't have to make Roaches if you don't need them; you can make defensive Roaches, but only the minimal amount needed to survive until the Spire; and in order to confirm/materialize your lead, you can build 10-15 Drones at once if you glimpse a window of tranquillity after the timing is held. A contrario, Terran's builds don't have this flexibility because of the investments in infrastructure costs: you can't have a third and a deadly early timing; SCV production is constant and set, you don't have those brutal accelerations the larva mechanic provides (29) . Unless some kind of miraculous triple OC + Hellbat timing build is crafted in Korean forges, which of course is very unlikely given the inescapable infrastructure costs (30) , Terran will keep struggling in TvZ after Zerg's adaptation is complete.

It should also be noted that the current Terran disadvantage in TvZ cannot be blamed on Terran's lack of innovation. Terran players tried more or less everything: replacing Mines with Hellbats, incorporating more Marauders into their compositions, using Thors, going pure bio, taking the mech route from timings to Raven turtling and mass Battlecruisers; hey, Terran even came up with a never seen before mech into biomech play (31). But in the end, everything falls short in one way or another against solid Zerg game plans, so Terran is left with a good diversity… of bad compositions. Moral of the story: regardless of the strategies played, if Terran's tools to restrain Zerg are not efficient enough, the beast simply grows too fast and cannot be tamed.

---

The introduction of the MSC is by far the most groundbreaking change between WoL and HotS. Available at Core tech, i.e. as soon as 4 minutes, the MSC is gifted with the ability to equip any Nexus with the weapon of a Photon Cannon during a whole minute.

At the beginning of HotS, Protoss had of course not yet realized the endless potentialities contained within their new unit. Macro-wise, they were mostly playing WoL + MSC builds, content with simply swapping their robo and gates 2-3 timings. But as they would learn over time, Photon Overcharge single-handedly gave them way more than any WoL Protoss could have ever dreamed of regarding builds.

Consider for instance two classic, safe robo PvT openings heading towards dual forge play; one from WoL, and the other from HotS. What are the respective timings?



As you can see, in HotS, gas 2-4 are taken ~90 seconds earlier than in WoL. How much gas does a geyser yield per minute? 114. This means that the HotS standard build is ~500 gas ahead of his WoL counterpart.

But there's more. In WoL, Protoss had to build a few Stalkers and Sentries early, if only because Terran had various agressive options that could not all be scouted in time (32); hence they would generally spend 200-400 gas on 2-4 Sentries on top of the initial Stalkers. This is no longer the case in HotS: common builds tech out of 3 Stalkers, and get 1-2 Sentries later; some even entirely skip them.

To sum up: in the early game, Protoss' standard builds can be up to 800 gas ahead of their WoL counterparts. Meanwhile, with an old CC first or a 1 rax FE build, Terran is 100 minerals ahead because he earned the right… to skip his Bunker. How cool is that?

HotS Protoss bathes in gas compared to WoL. Ever wondered why Protoss was able to invest in a Stargate and an Oracle while remaining fine even if the Oracle instantly crashed on a Mine? Look no further. Your DT harass failed in HotS? You're 275-400 gas behind. Is that a serious problem? No; you've merely consumed a part of the generous advance on salary HotS gave you. Trying to catch up in macro after your DT harass failed in WoL? Goodbye. Back then, things mattered.

Consider the following build used by herO:


herO literally has 5 units (MSC, 3 Stalkers, one Colossus) when the Medivac push hits… and Cure cannot do anything.

In WoL, Protoss would have immediately questioned your sanity, even against a scouted fast expand: "You can't be serious. Completing Warpgate at 7'45? Are you aware that you're on 1 gate until 8'40? Looking for troubles against 4 rax? What would you do against a random Tank push or Banshee? You don't even have a single Forcefield available when the Medivac push hits!" But finally, the technology is there in HotS: press F on your Nexus, and all your early game problems are dispelled.

Note that the boulevard towards midgame opened by the MSC also had impact on chrono usage. In WoL, Protoss sometimes had to chrono their Core against some openings or in case they had doubts about Terran's intentions. On HotS, unless Terran is going gas first, your Nexus' energy can safely be spent on Probes. Earlier Nexus timings combined with more available chronos means, of course, a better economy.

Meanwhile, Terran's rhythm of development is largely unchanged. Expanding during the third minute and getting your third around 9' - 9'30 is still the norm. Even the HotS novelty of a Reaper expand into 4' EB only accelerates Terran's 1/1 (33) since WoL Terran could build a 9'30 Armory too for +2 attack. Having slightly faster Medivacs with 2 rax instead of 3 yielded little result.

Protoss will retort that they require faster development since Terran has better agression with boost Medivacs. That is perfectly true, but certainly not to the current extent. The removal of Terran's serious threats heavily skewed the initial dosage between Protoss' stronger defence and Terran's stronger offence. At the beginning of HotS, Hellbat and Mine drops indeed commanded respect. Now what remains? Marines/Hellions attacks are pretty much the only thing left able to deal substantial damage if Protoss stays on 3 Stalkers (34), but a prepared Protoss has absolutely no troubles blocking the attack (35). Protoss has so little to fear in the first 10 minutes that they can tech way too agressively in complete impunity: what they earned in superior builds far outclasses what remains of Terran's new offensive potential.

If the Balance Team is consistent, it should perfectly be aware of this issue. After all, they're the one who enforced the principle of "asymmetric design":


Yet what happens in HotS? Terran is progressively robbed of all aggressive early game options, which means Protoss enters midgame (and by extension lategame) in better shape than ever. Protoss were given a next generation motor on their shining racing car, and somehow Terrans are supposed to compete with their old tractor.

Take for instance the Flash vs Rain series in the last WoL Code A. In the Akilon Wastes game, Rain misplayed and allowed two full Medivacs to unload his main base while Flash rallied the rest of his Marines at natural; Rain immediately lost 10 Probes in the maneuver and was in difficulty for the rest of the game. In the Daybreak game, he tried an ambitious tech-heavy build on 2 bases and was likewise refuted by Flash's drop play. In HotS? Photon Overcharge combined with the massive tech advantage would make him impervious to such things; breaking a Protoss on 2 bases is pretty much unthinkable these days.

Due to this, and despite the fact Medivacs can often land troops by brute force with Boost, top Protoss are barely droppable: sOs vs Bomber, Akilon Wastes, Blizzcon; Maru vs herO, Outboxer, Proleague; Polt vs Rain, Frost, IEM Cologne; Bogus vs Zest, King Sejong, Code S RO16 or Maru vs Classic, King Sejong, Code S RO4 show Terran's risky attempts repeatedly fail to seriously dent the brick wall.

Even when Terran's aggression is somewhat successful like the Maru vs herO game above or Bbyong vs Zest, Frost, Shoutcraft, the outcome isn't always favorable for Terran: Protoss commonly takes his whole army, walks across the map and kills the Terran who, after his sacrifices, often simply lacks the units required to hold the counter. Against Protoss masters of defence there is little hope in going the drop route.

Three Pyramids For The Protoss-Kings Under The Sky



At the end of WoL, quick third builds were far from being unheard of in PvT. You may for instance remember PartinG's famous quick third into 8g on Daybreak. The idea behind the build is brilliant in its simplicity: at this time, Terran's standard build was 1 rax FE into 3 rax Medivacs, and with this opening Terran does not push on the map before Medivacs are out. PartinG's build cleverly capitalized on this to get a quick third, build a strong production and use the earlier economy kick to fuel a deadly timing attack. Bio builds have indeed very limited threats before Medivacs appear; unstimmed bio is fairly terrible, and basic Sentry/Stalker micro has no troubles fending off intermediary pushes. Thus, even in WoL, quick thirds were unpunishable before Medivacs with fast expand into bio builds, as for instance MarineKing experimented against HerO.

Nonetheless, the risks and vulnerabilities of such a build were obvious back then. Fast expands into bio were not the only weapon in Terran's toolkit. Besides, without the MSC, the ETA balance was still respected: you could have Economy, you could have Army, but you could not have Technology. PartinG's timing was no more than a classic mass of unupgraded Gateway units. The build had no detection. Even if you were theoretically safe from bio pokes, any mishandled Marine push could result in huge damage: Grubby experienced it against LucifroN. Facing 1-1-1 harass? Tough life. One Cloak Banshee comes forth and you don't have detection? Kiss your economy goodbye. Out of position against a simple dual drop? Dead. Deviating from the mass gates scenario and trying to tech too agressively? Better have perfect crisis management or the sanction is immediate.

Things are completely different in HotS. With the MSC safety belt + quad airbag pack, the absurd concept of fail-safe greed fully applies to that kind of quick third, greatly expanding its power and turning a fragile build with a coherent risk/reward to an unsolvable conundrum. Think about what Terran can do to punish. An intermediary bio push? Maru tried against Classic on Frost with a Concussive poke followed by an attempt at a stim timing; Classic had enough to hold both without even using Photon Overcharge. Engaging with the Medivac push? Protoss holds. Another attempt? Brick wall again. All-inning with SCVs then? Doesn't work either. A bit later perhaps? No more success there.

Naturally, most of those builds now include Blink in the transition so any drop attempt is a one-way ticket; perhaps you can land, but you certainly won't escape. The only option left is taking the macro game from behind. Unsurprisingly, this rarely yields very good results. Consult the following list: Polt vs Classic, Habitation, IEM Cologne; Flash vs PartinG, Star Station, Proleague; Bogus vs Zest, King Sejong, Code S RO16 (Zest "only" builds its third at 8'30); Maru vs Classic, Frost, Code S RO4; Bogus vs duckdeok, Akilon Wastes, Blizzcon; TaeJa vs sOs, Whirlwind, 2013 GSL Season 1; Polt vs First, Bel'shir Vestige and Whirlwind, WCS Global Finals. The common denominator? Terran loses.

The only ways to not end up behind are to instantly catch Protoss out of position, or Protoss botches the build (playing a weak transition is a third option); in short, unforced mistakes from the Protoss side. Another possibility, if Protoss plays Blink into third and thus has no detection (note that Oracles into third have it) is that Protoss runs into a WoL antiquity, the Cloak Banshee. This is, for instance, what happened in Maru vs herO. But wait—what unit did Photon Overcharge erase from standard TvP already?

One Does Not Simply Walk Into Lategame

It is no secret to anyone that Terrans are approximately as interested in entering lategame TvP as chickens are in visiting a wildlife park dedicated to foxes. Terrans were already particularly reluctant to play lategame in WoL, but things considerably worsened in HotS, with the MSC being able to support the main army with Time Warp (36), Protoss' stronger development, the arrival of the Tempest and, overall, the improvement of Protoss' harassment techniques and Templar control. As a result, the list of Terran's issues when the game stabilizes in 4v4 bases scenarii is quite long:

1. Terran has no massable supply-free static defence, and therefore little/no answer to Zealots/DTs raids while the Protoss main army dances in front of Terran's bases (Bogus vs Myungsik, Neo Planet S, IEM Shanghai Qualifiers; Maru vs Rain, Akilon Wastes, OSL finals; Flash vs herO, Frost, Shoutcraft). For Terran, this situation is extremely difficult to handle as your attention is required in two or even three screens at once. Looking away at the wrong time, even for a split second, means your army can die to terrible, terrible damage. On the other hand, Protoss can spend most if not 100% of his attention on his main army; Zealots are fine rampaging bases on autopilot. PFs are garbage and barely do their job to protect SCVs: for instance, simply parking 2 Stalkers behind the mineral line means Terran has to send troops to drive them out of the place. If Protoss is out of position against a drop, a Zealot warp-in will clean the threat with little attention required.
   
2. Superior supply-efficiency of Protoss' army: if Terran sends 20-30 supply away, instant loss can happen if Protoss attacks/counters as the main army is suddenly lacking the critical ranged mass necessary to keep up with Protoss' army: Flash vs Puzzle, IEM Shanghai KeSPA Cup (Flash had 2 full Medivacs in Puzzle's main). Meanwhile, Protoss can detach up to 20-30 supply in Zealots/DTs without danger; defensive Storms make it impossible for Terran to charge into Protoss' army. With Warpgate, Protoss can always outnumber Terran somewhere when given a moment's notice.
   
3. Protoss can keep securing expansions (one Templar and Cannons/Pylons are enough to deal with any mild bio pressure) while Terran has tons of troubles expanding because of Zealots/DTs raids; if you want to secure a remote base, for instance, you have to detach and park units during 50 seconds while the PF is morphing. But as seen with the above point, this is problematic. As a consequence, a common cause of death for Terran in lategame TvP is starvation on 4 bases.
   
4. Protoss has map control. This means he can build Cannons outposts at key locations and prepare Storm flanks to hit by surprise each time Terran ventures on the map. Zealots or DTs can always hold remote Towers.
   
5. Tempests stomp Ghosts/Vikings: Flash vs PartinG, Daybreak; Happy vs HasuObs, Daybreak, WCS Europe Season 1; TaeJa vs San, Newkirk, ASUS ROG. Compare with the Tempest-free WoL environement which actually allowed Terran to win lategame fights: ByuN vs HerO, Daybreak, Up & Down. The Terran range advantage from WoL is gone (15 > 9-10); one of Protoss' WoL weakness, the lack of vision, is gone thanks to Revelation from Oracles. Tempests tipped the scales into Protoss' favour because Vikings cannot handle both Colossi and Tempests while surviving Archon and Storm splash (not to mention that if Terran overproduces Vikings, he gets rolled on the ground). Besides, in a lategame scenario, the superior siege weapon tends to prevail. On top of that, Tempests make the already extremely rare Battlecruisers transition completely unviable; when fully upgraded, Tempests kill Battlecruisers… in 7 shots. Ravens are simply dead supply if Protoss gets rids of his few Tempests.
   
6. There is a massive disparity in the defender's advantage. Defensive Templar/Cannon positions are nigh unbreakable, especially with Tempest support, while Terran has nothing to hold when behind. This means that in order to win the game, Terran generally has to win several fights (basically until Protoss' bank is gone). On the other hand, Protoss only has to win once. Protoss can play from behind as Warp/Storms allow comebacks (e. g. KeeN vs First, Atlas, GSTL); Terran cannot overcome massive supply deficits. Protoss is not the Egyptian race for nothing: worshipping cats definitely grants you extra lives.
   
7. Protoss dictates Terran's unit composition. If Protoss has x Colossi, you need y Vikings. If you don't have them, you lose the fight. If you have too many of them remaining while Protoss remax on ground… you lose too.
   
8. One Prism entering Terran's main base is enough to inflict devastating damage. Turret rings are not 100% reliable as there are always weak points. Besides, Protoss can build a second Prism anyway to force the way.

Basically there are three ways Terran wins a real lategame scenario:

1. Terrans enters lategame ahead of Protoss (preferrably in significant fashion);
   
2. Protoss makes grave blunders: not having detection vs Cloaked Ghosts (e. g. KeeN vs Ruin, Star Station, Up & Down); move-commanding his army (HasuObs vs MMA, Bel'shir Vestige, WCS Europe); getting impatient and having most/all Templars disabled, in particular if they're clumped: Maru vs Rain, Star Station, OSL Finals; or simply complete nonsense such as in Maru vs Zest, Frost, Code S RO16: Zest has 52 (!) supply of Zealots sent in harass when Terran attacks, barely any protection for his Colossi and floats absurd amounts of resources with little production; and it's still rather close.
   
3. Terran somehow manages to weaken/destroy Protoss' economy before winning a major engagement: Protoss needs 60+ Probes to fuel continuous buybacks.

To have a chance at competing, Terran needs excellent Ghost/Viking control, which is hellishly hard to achieve in the accident-prone environment of lategame TvP; as seen above, Protoss has indeed numerous opportunities to actively cause mistakes by forcing defensive multitask. If the mistake happens, Terran is toast. Considering how many pitfalls must be avoided and how slippery the slope is, even the best Terrans end up stumbling over and falling:

+ Show Spoiler +

In order to give the illusory appearance that things are not that bad, games like TaeJa vs Rain, Whirlwind are sometimes taken as an example of lategame TvP being fine for Terran; as if starting lategame with 10 Orbitals to 3 Nexus after you caught 3 Colossi in midgame was the norm. As a matter of fact, the game precisely showcases why lategame TvP is not fine, as this eloquent comment from the LR thread summarizes (37):

+ Show Spoiler +

Note that TaeJa, hailed as the savior of lategame TvP, often wins in the "early lategame," i.e. before the game enters a fully entrentched 4v4 bases scenario. His wins against MC on King Sejong and Habitation at the HSC IX; Bogus vs San, Waystation, ATC or MMA vs Zest, Overgrowth, GSL Championship fall under that category: a Ghost/Viking timing before things get out of hand. When Protoss gets Tempests, not even Taeja's lategame prowess is enough:



There is also the preposterous idea that Terran auto-wins in the extreme lategame because they can sacrifice most of their SCVs to build an unstoppable 180 supply army. This myth ignores the practical requirements of such an undertaking: the absolutely exorbitant costs (resources and time) required means Terran systematically dies before even hoping to reach that stage. What's more, it is quite ironic people suggest Terran should get rid of his workers when the race primilarily dies in lategame… precisely because it cannot maintain his economy. One Orbital is "only" ~4,3 workers; if you lost 20 SCVs to harass and went down from 75 to 55, then your 5 extra OCs do nothing more than compensating. The idea also ignores what would happen if Protoss mirrors the idea and sacrifices part of his Probes to use his own bank; the sOs vs jjakji and Bunny vs Creator games above hint at the fun it would be for Terran to deal with 160+ supply of Archons/Carriers/Colossi/Storm.

The Derailed Train



Many Protoss were not content with the Mine buff. According to the pessimistic theory, Protoss was doomed in PvT:

1. Mines sounded the death knell of Templar openings;
2. Protoss were thus forced to open with Colossi, which "die to SCV pulls".

Does this cute theory survive the trial of reality? Let's take for instance Flash vs herO in the Shoutcraft finals: leading 2-1, Flash pulled SCVs three games in a row. The score in the end? 2-4. People in the LR thread were confused and did not understand why Flash kept pulling SCVs. Was he trying to illustrate the famous quote about insanity? Nope, he simply wanted to avoid 3/3 into lategame. Unfortunately for him, as the series demonstrated, herO's build was absolutely SCV-proof, and so are most blink colo dual forge builds: Protoss changed their builds and no longer play into Terran's hands with the vulnerable 3 Colossi into Storm which captured the image of SCV pulls in their best light.

From recent showings, it seems the 3-bases all-in is on the decline despite Colossus play being standard as of now. This is not good news for Terran: if midgame timings/all-ins get solved, then it means the race has to play even closer to lategame.

PokerCraft

List of Terran all-ins
+ Show Spoiler +

List of Protoss all-ins
+ Show Spoiler +

Jokes aside, there are 2 or 3 things for Terrans. See how solid and difficult to defend they are: 8 rax inbase? Dies to Probe scout. 11/11? Not entirely void of venom, but Mvp would certainly not have a fourth GSL title in HotS. Lifting to gold? A map-dependent gimmick relying on Protoss forgetting his WoL classics. From the Terran side, the macro game is forced (38). Is this a problem in itself? Not really. After all, in TvZ, there are no especially viable 1-base all-ins from either side, and no one is missing them. Unfortunately, there is no such equilibrium in TvP.

Some might retort that there is no problem: after all, cheeses and all-ins are a legitimate part of the game. That is perfectly true, but what makes Protoss particularly infuriating in that domain is the coinflippy aspect of those cheeses. If they were all easily scoutable and defendable, there would be no issue, but the reality is very different.

Consider for instance what could be called "the third Pylon problem". What buildings can be powered by a proxied third Pylon?

1. Robotics for Immortals: has to be proxied relatively close to Terran's base; generally scoutable;
   
2. Stargate for Oracles or Voids: can be proxied quite far since units fly and Oracles are extremely fast; hard to scout if the proxy is sufficiently far away from the Terran;
   
3. Twilight Council for Blink: can be proxied anywhere on the map, unscoutable;
   
4. Dark shrine for DTs: can be proxied anywhere on the map, unscoutable.

The problem is immediately flagrant: Terran requires different answers to those different threats, yet has no tool to tell them apart in time. Even if you have a Reaper still alive and active on the map, what are the odds you find that Twilight Council randomly proxied on Frost?

Maybe Terrans should just aim at a middle-of-the-road answer then? Alas, there is precisely no such thing. Specific adjustments need to be done with defensive structures (Bunkers, Turrets (39)) or with the composition and positioning of units (especially with 1-1-1 openings). If Terran cannot get the information, he's left playing a perverse game of heads or tails. If he happens to prepare for the right thing, he may survive. If he guesses wrong, he instantly loses.

So, Bogus: heads or tails?

+ Show Spoiler +

Bbyong is our next player. Heads or tails?

+ Show Spoiler +

As you can see, wrong defensive measures such as a quick EB and blind Turrets may lead to an instant loss against something else. But of course, terrible things happen too if you don't have detection when DTs show up:

+ Show Spoiler +

More shenanigans?



From dual forge to proxy 7g: Creator, the master of macro PvT in WoL, changed his tune in HotS. Ignore FanTaSy's opening (here his transition would die regardless of whether the all-in is scouted or not) and focus on what Creator is doing. What FanTaSy saw with his Reaper—an empty base—could even hint at… a quick third. See for instance how First tricked Polt, using his own scan for that: the fake Sentry moveout makes Terran fear an all-in, triggering useless Bunkers while Protoss is quietly taking the Giza route.

Scans are of course of no/little use to determine what is actually coming:

+ Show Spoiler +

Scans are pure gamble, but with an important opportunity cost. They can see everything, they can see something, they can see nothing. Needless to say, if Protoss bothers to hide his stuff, you will run into the third case. Mains are so large on some maps that building on the edges does the trick; and there is of course always the proxy case.

Metagaming or luck, in the Star Station game above, TaeJa hits the right answer: his Mines/Bunker defence is adequate to Sage's frontal Oracle bust. But had it been a Blink all-in, he would have been toast, just like jjakji vs elfi at the ASUS ROG:





It goes without saying that "counting gas" doesn't work at all: Stargate + an Oracle is exactly the same price as Council + Blink or Council + Dark shrine. Okay, there is 50 more gas in the first option. The difference is enormous.

Protoss can thus give birth to poltergeists like Has, odd creatures permanently living in the netherworld of cheeses and proxies, only occasionally stumbling over the Nexus key… to build it at their opponent's natural. Of course, Protoss' huge cheese repertoire is put to good use to coinflip superior players: even if the Terran is at the absolute top of the world, he's still in danger. On the other hand, can you imagine MorroW scalping Rain with a 1-1-1 all-in or some obscure GM knocking down PartinG because of a single Cloak Banshee?

Emergency Exit, Over There

To add insult to injury, Protoss is far from being dead when his "all-in" is scouted and defended. The following game shows, for instance, how Terran being forced to heavily commit in defence allows Protoss to take the macro route in impunity:



Polt loses the head or tails game and takes refuge in its main; Protoss simply contains and expands.

One of the most obnoxious traits of 2-bases Blink play, for instance, is that Protoss' options when building a Council range from extreme agression to what should theoretically be "greed," yet is safe thanks to the MSC. Regarding his own build/transition, Terran has to make a choice based on partial or nonexistent information.



A slight scouting mistake leads Flash to misread Billowy's play; the disaster is immediate. You can read Flash's frustration on his face.

Maru vs Classic, King Sejong, Code S RO4 is another example of Protoss' over-flexibility: Classic's attack packed enough power to instantly kill Maru should any mistake happen, yet Classic had a smooth dual forge colo transition and could even successfully build its third before Maru. Low-risk high-reward attempt at winning the game into being ahead anyway? I'd buy that stuff too.

The Power Of Protoss

Losing at the lottery is one thing; an utterly inferior lategame is another; but for Terran, the most depressing aspect of the match-up is probably the massive contrast in control needed to handle a Protoss and a Terran army in large engagements.



jjakji controls every part of his army: bio, Vikings, Ghosts; elfi doesn't even try to protect his Colossi from free shots and focuses nothing particular with his Stalkers. The superiority of Protoss' design particularly shines in that sequence: during the battle, elfi has enough time to stroke his cat should the noise of Ghosts yelling prove too stressful for the animal.

Let us watch again the decisive battle between MMA and PartinG. See how much Terran has to do in that situation: he would need to EMP all the Archons and focus the weak ones to kill them off while simultaneously retreating Ghosts in danger and hitting & running with the rest of his army; an insurmountable task if you're not called Automaton 2000. Meanwhile, PartinG can sip the drink of his choice pressing the T key. MMA will necessarily make a copious amount of mistakes solely due to the micro requirements; PartinG cannot be wrong a-moving Zealots/Archons.



Let us set the context. Before this, Bogus played against TaeJa, and their series produced one of the best games of 2013: on Newkirk, TaeJa's biomech fought Bogus' mech for nearly one hour in an amazing back-and-forth game. Despite Bogus actually being in a delicate position half of the game, it took a long time for TaeJa to finally break the Machine's exceptional resistance. Now head over the critical engagement. What was needed from a lesser opponent to knock out a top3 world TvP from the tournament? 1a + 2 Storms.



Bogus has to split in different areas while Welmu's troops scatter everywhere as he delivers the current. Notice how the fight is still dramatically close despite half of the Colossi being engaged in a hard-fought battle against a Refinery and Welmu thinking he works for the Bank of Finland.

What leads to this disparity and the unparalleled effort/effect ratio of Protoss' deathballs is not hard to conceptualize. In PvT, when nearing max supply, the core of Protoss' armies is composed of a-move style, beefy units (Zealots (40), Archons, Colossi (41)), which by definition are simultaneously much easier to use and much more forgiving than glass cannons whose efficiency purely relies on the quality of control. One Marine in the hands of Maru is much, much more efficient than in the hands of an average Terran; while one Zealot in the hands of PartinG is almost exactly the same a-moving berserker as the one in the hands of a random high master. Since other micro units such as Sentries or Blink Stalkers are progressively erased from Protoss' composition as the game goes on, the skill difference regarding in-battle micro ends up being mainly restricted to Templar control; there you can clearly see the difference between a micro master like PartinG and his lower skilled counterpart. For the rest…

How Is The Weather?



Last but not least in the list of TvP woes, Storms have the devastating potential to turn several minutes of Terran dominance into an immediate and unceremonious loss. Countless are the games in which Terran is ahead, and suddenly Protoss lands the money Storm and he's either winning or the game is reset. It happens even to the best; one slight attention mistake is all it takes. The brutality with which TaeJa loses the two following games speaks for itself:

---

Comparing games between top Korean Terrans and top non-Korean players from their respective races tends to reveal massive differences in the way games play out. Take for instance the series Bogus played against Bunny and Welmu at the Dreamhack Bucharest: the TvP was a laborous win for Bogus, who was within an inch from losing the series. The TvT? A formality. Another example would be the TvTs played by TY, in which he came out on top even when he had to negotiate delicate positions, compared with his series against Serral; if someone was to remove names and make you watch the two first games, would you ever guess those 30 minutes struggles in macro games feature a Code S level Terran against a top20 EU Zerg?

On top of being fail-safe "greed" features, the MSC and—to a lesser extent—the Queen patch scripted the vT match-ups in such a way that even outclassed players are almost impervious to various early game threats that would otherwise kill or cripple them. Blizzard paved the way so midgame can be reached without much trouble (42): you can't outplay a 2000 hit points Cannon.

Even if their level is lower globally, non-Korean Zerg and Protoss players can still put up a fight against Koreans. A contrario, it is a known fact that Terran out of Korea lies in the gutter for years. The weaker level of play can indeed only worsen Terran's issues:



+ Show Spoiler [Data used] +



+ Show Spoiler [Data used] +

Some food for thought: if Zerg and Protoss were as demanding as Terran, particularly against a superior opponent, there would be about four times less "upsets" in the Korea vs World rivalry. Terran respects the most the skill gap logic, while other races provide a higher chance to run off with the takings.

Because of this higher difficulty, Terran is the least played race at the top level. Make no mistake though: this under-representation is much older than the dreadful patch 2.0.12.



+ Show Spoiler [Data used] +



+ Show Spoiler [Data used] +



+ Show Spoiler [Data used] +

As you can see, even when the race was still fine, Terrans were the least represented.

You Be The Judge

Ultimately, the spectactors decide what's best. What do people generally prefer? Upon consulting the Best games thread, we find this:


You may dismiss this as purely subjective, and you would not be completely wrong. Besides, who knows? Maybe the judges were a bunch of Terran elitists running in circles around the Eagle totem. You hate Tank lines and you prefer PvP? That's your absolute right. As wise people once said, de gustibus et coloribus non est disputandum.

Still, it remains obvious that everyone—players and spectactors alike—would benefit from a healthier Terran. The race being weak out of Korea since years is unfortunately accepted, but the foundations are crumbling now that the disease spread even within the natural land of the species. Are we to burn a candle before the beginning of each Code A session?

Starcraft without Terran is like a Has game without proxies; it just does not have the same flavour.

The Eagle needs his wings back.

---

(1) Notice that Life, who won that tournament, played in the same way, often going for big Baneling busts at times in which Terrans were vulnerable with their new standard build.
(2) And indeed Terran won the most Premier tournaments during the first 7 months of HotS, more than the two other races combined (12 vs 5 for Zerg and 4 for Protoss).
(3) It is no hazard that the only Terran who lost in a non-mirror at the first round of this MLG was the only one not born in Korea.
(4) This principle (having to repeatedly switch your camera between the attack and your base) is also what makes an active Hellions/Banshees harass a very demanding opening in TvZ. The Marines/Mines attack in TvP has the same high mechanical requirements: notice how Bogus' bio upgrades are delayed compared with the completion of his add-ons because he's busy managing his two-pronged agression and cannot switch back his camera to his main base without having his Marines waste themselves doing nothing in the meantime.
(5) Besides, Probes would have survived the onslaught as soon as +1 armor would be complete, which can be quite early given how fast Protoss can build their first Forge in HotS PvT.
(6) You need extra resources, or a second fact, and you end up with way less Hellbats than before, which ironically leaves you vulnerable… to the composition you try to "counter". See for instance SuperNova vs Trap, Akilon Wastes, OSL: SuperNova tries to tech Blue Flame Hellbats and dies to a Zealots/Archons push anyway.
(7) Compare for instance with his DH Winter path to victory: Bogus ForGG sOs SjoW HerO MMA Life Life. The record (18-7) looks weaker, but the performance was way more impressive; excepting SjoW, all players he met were a serious threat.
(8) Flash's performance in Proleague could also be mentioned as an invariant.
(9) People keep talking about Blink "all-ins," but there was no such thing as a "Blink all-in" back then. Unless Protoss suffered disastrous Stalkers losses for nothing, a failed attack would virtually always have a smooth Templar transition.
(10) It took one year to turn the MSC from a flying scan to a myopic vessel.
(11) Even as of now, after the AoE bonus damage to Shields buff, Mines are but a shadow of what they once were against Probes.
(12) See for instance MC vs YoDa, Daybreak at the first IEM for one of the first games illustrating the death of detection-less all-ins in PvT in early HotS.
(13) SuperNova vs MC, Antiga, GSL 2012 Code S Season 1.
(14) Bogus vs Rain, Akilon Wastes, MLG. Notice how Rain is forced to delay the attack and build a robo to first deal with the Mine drop, leaving Bogus enough time to prepare his defence.
(15) It must be mentioned that ForGG uses it sometimes; e. g. ForGG vs San, Overgrowth, Fragbite Masters; ForGG vs Welmu, Alterzim, Fragbite Masters; ForGG vs First, Alterzim, DH Summer.
(16) Though ironically, WoL Battlecruisers/Vikings/Ravens fleet are still probably superior to their HotS counterparts because Vipers were not there to pull Battlecruisers into a swarm of Corruptors and Spores; see for instance Mvp vs sLivko, Metropolis, IEM Cologne or Brat_OK vs Goswser, Entombed Valley, Ritmix RSL IV. Of course, in practice those fleets were absolutely impossible to get, which is the main difference between the end of WoL and HotS.
(17) Even at this stage, Terran has to proceed carefully; at the Shoutcraft, Flash won his games versus Life on King Sejong, but at one point he was one Fungal away from disaster. Bbyong also narrowly escaped a deadly Fungal against Life in Proleague. Once Zergs refine their 7v5 bases creep contain into starvation techniques (a concept illustrated in Soulkey vs Reality, Star Station, Proleague), fighting over the critical 12th base, it is not said Terrans will have enough fuel left to run the machine.
Note that "turtle mech" is distinct from Hellbats/Thors/Banshees timings/all-ins played for instance by ForGG or SuperNova, and Mvp when the opportunity arises. Flash vs Life, Overgrowth, Shoutcraft also showcases an example of those "timings by opportunity".
(18) Bogus vs Soulkey, Frost, IEM World Championship Asia qualifier; Mvp vs RorO, Outboxer, Proleague.
(19) The Terran player is best pictured yawning as this happens.
(20) You may know it as the Siege Tank.
(21) Which is probably why scientific tests run in David Kim's secret lab determined that the original 3 → 2.7 proposition would lead to a massive upheaval, while 3 → 2.8 was fine, really; I assure you.
(22) And after the Queen patch, it wasn't properly working at the end of WoL either. Terrans kept playing Marines/Tanks because there was nothing else to do, but the old Marines/Tanks pushes were unable to stop Zerg's developement towards Hive: there was too much on too much creep, and the window was too short.
(23) The lesser gas requirement for the Mine is in particular critical to quickly build a dense Medivac cloud. It is not before ~15 minutes, shortly after 3/3 starts, that Terran starts banking gas.
(24) The much higher range and a more practical focus provided by the Tank don't result in a more efficient Zerg cleaner.
(25) Needless to say, the apm peak is due to the massive production of Zerglings.
(26) Yes, he did have an upgrade disadvantage for 4 seconds in the fight, but it is obvious he was already losing it when +2 Carapace completed.
(27) Roaches/Hydras (e. g. HyuN vs jjakji, Frost, SeatStory Cup), a normal lings/banes/mutas game (e. g. DRG vs MMA, Daybreak, ATC Playoffs Season 1), or more rarely lings/banes/roaches (e. g. Bogus vs HyuN, Newkirk, ATC Season 2). Pure Roach is a fourth option (e. g. Revival vs Polt, Habitation Station, WCS America; this is also what Shine played out of despair when he ended up behind vs Bogus on Waystation this Code S season). What makes the transition smooth is that Terran cannot land his third until the attack is dealt with, so during this time Zerg is free to saturate his own third and tech his transition. Note that even 2-bases Terran timings can get overwhelmed: Bogus vs Vortix, Derelict Watcher, ATC Season 2.
(28) Life had 41 Drones and 3 bases to Flash's 30 SCVs on 2 Orbitals with no third started until 10'. Life could not have turned into full Drones mode in case Flash committed again, but he could have progressively increased his Drone count.
(29) See also Flash vs Sheth, Ohana, MLG MvP Invitational. Both games featured a Roach/Baneling bust against a triple OC opening. Kas lost all his units, so Minato is free to build 20 Drones; Terran has nothing to immediately punish. The real price for this attack is delayed tech (upgrades + Lair timing) and creep (no extra Queens) rather than economy (still slightly weaker because Drones are delayed, but they're made). Sheth's bust was more of a failure, but Flash had no immediate counter-threat either. Scarlett vs Bomber, Whirlwind, WCS Global Finals is another example of this concept: Scarlett has not the usual defences in place when Bomber raids his Hellions in the natural. As a result, 15 Drones die, but larvae forgive: Zerg replenishes the fallen workers at once, still hits the 70 benchmark at 10' and lives to tell the tale.
(30) A hybrid build like the one used in Bogus vs Life, Frost, Dragon Invitational #2 results in a weaker timing; Hellbats are not that hot without Medivac support, especially when hitting that late. The following game, Life even held with pure lings/Queens despite this scenario being of course the most favourable for Terran.
(31) If you have hope in that strategy to redeem the TvZ match-up, don't hold your breath: Soulkey demonstrated the refutation.
(32) It should be noted that 1 gate expand into 3 gate robo could deal with everything, from 1-base all-ins to triple OC builds.
(33) Compare with former HotS builds which could begin their dual forge as early as… 5 minutes; e. g. PartinG vs Flash, Star Station, Code S RO16 or Avenge vs Flash, Whirlwind, Code A RO24.
(34) TaeJa vs Alicia, Akilon Wastes, ASUS ROG; Bbyong vs Zest, King Sejong, Proleague R3 Playoffs.
(35) Rain vs Bomber, Bel'shir Vestige, OSL RO4.
(36) YoDa vs Welmu, Frost, WCS Europe shows what happens when Terran's army cannot move because of a Time Warp.
(37) http://www.teamliquid.net/forum/sc2-tournaments/426686-wcs-day-2-season-2-grand-finals-2013?page=230#4593
(38) Even if you aim at all-inning with SCVs, you need 3 bases (2-bases variants are very rare because they're too obvious and lack the power of the third MULE).
(39) See for instance Dayshi vs Has, Alterzim, ATC to see how annoying DTs can be when you have Turrets in your mineral line, but not at the front of your base.
(40) Compare with SC1 Zealots who did not have Charge. In BW PvT, Terran was the faction with tons of AoE, but there was friendly-fire; Protoss would thus micro their Zealots to draw Terran's splash unto them.
(41) Compare once again with its spiritual ancestor. Colossi have less hit points per supply than Protoss' melee line, but as an individual unit they're still quite sturdy.
(42) But perhaps there is an underlying pattern: after the Queen patch in WoL and the MSC in HotS, maybe Terran will get some asinine Orbital Fortress in LotV to complete the trinity of mindless auto-defence buttons.

This article was written before MLG Anaheim.