WCS: Looking Back at the Round of 32

Before delving into the R16 and how Bunny and GunGFuBanDa must be raging at their bad luck, it's time to look back and take stock of the R32 for season two--from MaSa channelling his inner NXZ to GunGFuBanDa gifting Jaedong first place in his group.  Here's how the predictions turned out:

Aligulac 12-4 (9-7 last season, but very much in line with its season two Challenger predictions)
TL 11-5 (7-9 last season)
Me 9-7 (8-8 last season)

These are significantly better results from Aligulac and TL compared to season one.

Who were the biggest upsets and biggest surprises?  To my mind there are only three of note (one in the former category, two in the latter).  In Group C we had one of each, with Ret moving forward and viOLet becoming (again) the only Korean eliminated at this stage.  In Group G iaguz advancing was a surprise, as the Gimli Terran has never advanced this far before in WCS (he gives SEA it's second straight season with a player in the R16, as PiG accomplished the feat in season one).  Besides these three players, everyone in the R16 was selected by at least one of Aligulac or TL.

A few general notes: the stereotype about the weakness of NA and Latin America stood true this season (unlike last when Kane made it to the R8).  The king of Copa America, MajOr, failed to win even a series (again! although Kelazhur at least beat Snute on his way out), nor did Kane or MaSa, while Xenocider (the only American participant) fell out with only an unremarkable win over NXZ (the only player with an 0-4 score).  Just one player from China/Taiwan made it to the next stage (iAsonu), while the EU region stood strong again (9-4 overall).

As for the Koreans, they all lost multiple games; three besides viOLet actually lost a series (StarDust, Hydra, and Polt), with only ForGG and Jaedong making it through with two straight wins.  All of them looked beatable in the R32, except perhaps ForGG.  This isn't to say I expect a foreigner to win their own tournament, but rather we might see a Korean knocked out in the R16 unlike last season (Koreans didn't drop a map in the R16).

A quick look at balance (with 7 Zergs, 5 Terrans, and 4 Protoss moving forward, a change of one less Zerg, two more Terrans, and one less Protoss from season one):
ZvT 8-7
PvZ 6-5
TvP 4-1
The only note here is a slight edge for Terrans against Protoss (something barely perceptible in Challenger); the only Protoss player to beat a Terran was Lilbow over Kelazhur.

Map scores  (38 ZvT's, 24 PvZ's, 12 TvP's):
Coda (23) ZvT 5-7 PvZ 3-5 TvP 2-1
Vaani (16) ZvT 6-1 PvZ 3-2 TvP 3-1
Echo (13) ZvT 3-7 PvZ 1-1 TvP 1-0
Expedition Lost (9) PvZ 4-1 ZvT 3-1
Iron Fortress (8) PvZ 2-2 ZvT 3-0 TvP 1-1
Cactus Valley (4) TvP 1-1 PvZ 1-0 ZvT 0-1
Inferno Pools (1) ZvT 0-1

Most common ZvT maps: Coda, Echo, Vaani
Most common PvZ maps: Coda, Vaani/Expedition Lost
Most common TvP maps: Vaani, Coda, Iron Fortress

Coda is now the most popular map, beating out Vaani (which was slightly ahead of it in Challenger).  Expedition Lost was more common, although zero TvP's were played on it (there were only PvZ's in Challenger).  Inferno Pools remains deadweight (even more so than Alterzim from past seasons), with Cactus Valley nearly there itself.  The most stark advantages above are for Zergs in ZvT on Vaani, Terrans in the same matchup on Echo, Protoss' against Zerg on Expedition Lost, and possibly Zerg against Terran on Iron Fortress.  Here's a look at the map scores including Challenger (minus Hydra's stomp of Zeal):

Coda (35) ZvT 6-9 PvZ 6-6 TvP 6-2
Vaani (29) PvZ 4-8 ZvT 8-3 TvP 7-2
Echo (24) ZvT 5-8 PvZ 5-1 TvP 3-2
Iron Fortress (15) PvZ 3-5 ZvT 3-1 TvP 1-3
Expedition Lost (12) PvZ 4-4 ZvT 3-1
Cactus Valley (9) TvP 3-2 PvZ 1-1 ZvT 1-1
Inferno Pools (3) ZvT 0-1 PvZ 0-1 TvP 1-0

It will be interesting to see how these trends fair going forward.

This article is written by Peter Levi (@eyeonthesens)

WCS Premier Predictions: Groups G & H

Groups E & F have wrapped up with both Aligulac and TL at 3-1 and myself at 2-2 (at last I have taken to heart that MajOr just can't get it done in WCS once he's in studio).  Snute losing was the biggest upset (particularly being the first eliminated), but it was surprising that both of last season's finalists (Hydra and Polt) lost a match (the former to Lilbow and the latter to TLO).  Despite the upsets, the final results are not particularly surprising.   Speaking of predictions, TL did well with the SSL, although sadly provided no prognostication for Group D in Code S [subsequently found where they were 1-1].  Moving on: here are the predictions for the final groups in the R32 (with stats via Aligulac):

Group G
MaNa vs iaguz - I'm a fan of the Gimli Terran, who despite the odds (68%) has actually beaten the Polish player every time (3-0, but those games go back to WOL); while the Aussie's TvP numbers are good overall, against European Protoss' he's only 2-4 this year; on the other side MaNa is 14-1 in the match-up lately (somehow losing to frozz) and I expect him to win here
MaSa vs FireCake - can the Canadian do the scene a favour and beat his Zerg opponent?  The odds say yes (59%); MaSa is 22-1 in TvZ, albeit largely against weaker NA players (practicing against Hydra must help, surely); the French player hasn't had many series against Terrans recently, but he's 6-1, only losing to Happy (who he's also beaten, along with MarineLorD); I think it could go either way, but I'd like to think the talented MaSa is due for WCS success and I'll stick with the odds
Losers
FireCake vs iaguz - once again the French player is the underdog (53%); this is definitely iaguz's comfort zone (TvZ) and much like MaSa he's compiled an incredible record largely based on SEA/NA opponents (12-0); one of iaguz's flaws is predictability (so losing his second series against MacSed last season, or how he lost to NesTea in season two of Challenger in 2014); but this is a BO3 and he'll have the benefit of seeing how FireCake plays the match-up in the previous series, so I'll stick with the Aussie
Winners
MaNa vs MaSa - the Polish player is favoured (58%), even though MaSa's TvP record is absurdly good (54-11, built off the NA scene); I think MaNa's more impressive track record and better available practice partners solidify sticking with the odds
Finals
MaSa vs iaguz - the ROOT team-kill favours the Canadian (68%), although their score against each other is close (6-5 in games, 2-1 in series, all played in the last two months); MaSa is better in the match-up and I expect him to win here.  If this is MaNa instead the odds are above and I think we'll see a repeat victory (particularly if ROOT performance against Protoss can be gauged by Hydra and Kelazhur)

Aligulac and I have MaNa and MaSa moving forward (TL has the Polish player and FireCake).

Group H

GunGFuBanDa vs Has - the German is favoured (68%) against the wily Taiwanese player and he has a ridiculous PvP record of late (17-0); Has doesn't have the same kind of statistics to back him up (he hasn't played a PvP series since season one), but his unpredictable style makes him very dangerous; still, I'll stick with the odds here even though I'd love to see Has play Jaedong again (despite being a massive underdog; I think he has a good chance against the Tyrant)
Jaedong vs elfi - not surprisingly the Korean is favoured (67%) and over the years has had a lot of exposure to the Finn's play style; elfi is 1-8 vs Korean Zergs in HOTS (only beating Golden) and I see no hope for him here
Losers
Has vs elfi - the Finn is favoured (56%), but his recent PvP numbers are based off lesser European players and while both players are incredibly cheesy I think Has is the stronger player
Winners
GunGFuBanDa vs Jaedong - once again the Korean is favoured (63%); the German is 4-6 against Korean Zergs of late, granting his wins aren't against players of Jaedong's calibre; that said, he consistently takes games off Korean Zergs so an upset is possible, but with the Tyrant 11-0 against foreign Protoss of late I have to stick with him
Finals
Has vs GunGFuBanDa - the odds are above, but I think the re-match favours the loser, so I have Has moving forward; if this is elfi, the German is favoured and I think would comfortably win despite the Finn's record against him (3-1)

Aligulac has Jaedong and GunGFuBanDa going forward (TL agrees), while instead of the German I have Has moving on.  This is potentially a very entertaining group to watch and I hope that Has does get to play Jaedong (check out their match from last year).

This article is written by Peter Levi (@eyeonthesens)

WCS Premier Predictions: Groups E & F

Aligulac (and I) were 3-1 over the weekend (TL was 2-2; speaking of TL, they continue their odd struggles in predicting the R16 in the GSL).  Ret surprised the odds by making it to the R16 for the first time since season two in 2013.  The Dutch player had never won a series against a Korean Zerg in HOTs (0-5), and was just 1-13 against Koreans over the last year and a half.  The other surprise was the defeat of the first Korean in WCS, viOLet.  Along with StarDust, viOLet was the most beatable Korean in the system, but while the former MYI Protoss survived the previous weekend, viOLet failed for the second time in a row--this time he wasn't sick and he'd just returned from Korea, so there really are no excuses.  I suspect he's doing some serious soul searching at this point; he's still good enough to get to the R32 each season, but is that enough for him?  Putting last week aside, let's look into the Aligulac crystal ball for the first two groups of the week:

Group E

Lilbow vs Kelazhur - the French player is heavily favoured (71%), and if there's one kind of quality practice the Brazilian isn't getting in either the ROOT house or Copa America it's TvP; Lilbow is 9-2 in the match-up of late, while Kelazhur is 7-1 (granted, against weaker competition); I'll stick with the odds
Hydra vs Snute - potentially one of the best matches in the R32, the Korean is favoured (59%) and is 4-1 against the Norwegian; there is hope for an upset, as while Hydra used to be unbeatable against foreigners in this match-up, he's lost three times in the past month (to Petraeus, TargA, and PandaBearMe (!)); Snute's overall track record against Korean Zergs is quite good (24-29), but he's only 2-9 lately (including two losses to Hydra); given the latter fact, I'll stick with the odds
Losers
Snute vs Kelazhur - not surprisingly the Norwegian is favoured (61%) and beat the Brazilian in their only meeting (2-0 back in March); Kelazhur is 12-1 in the match-up of late, while Snute is 6-6, but three of those losses are to Bunny (the others are to MarineLord, Happy, and Beastqt)--so only the best beat him in ZvT--I don't think the Brazilian is quite in that category yet, so the odds seem on target
Winners
Hydra vs Lilbow - surprisingly the Korean is only a slight favourite (52%); the French player is a very respectable 9-11 in series against Korean Zergs, but that includes two losses to Hydra (both back when he wasn't losing to foreigners); speaking of not losing, Hydra still hasn't lost a series to a foreign Protoss (29-0!) and hasn't dropped a map since GungFuBanDa in early March--the odds make this look much closer than it should be
Finals
Snute vs Lilbow - the French player is favoured (54%), and is 10-0 in PvZ series of late; he also beat Snute in the last days of swarmhosts in early March (he's 9-8 against him historically); Snute is 9-3 in the match-up recently (losing to Harstem, PtitDrogo, and ShoWTimE)--who to pick?  The Norwegian has the bigger name and surely must be the default for most--I'll stick with the better known quantity

Aligulac has Hydra and Lilbow going forward, while I take Snute over the French Protoss player (TL agrees).

Group F

Elazer vs MajOr - the Mexican player is the favourite (72%); he hasn't played a lot lately (and losing to ZhuGeLiang is worrisome), but he's historically very good in TvZ and the Polish player's record against good Terrans is not great, so I'll stick with the odds
Polt vs TLO - the Korean is heavily favoured (74%) and other than losing to Kane back in January he hasn't lost to a foreign Zerg since June of 2014; it's been almost as long since TLO beat a Korean Terran (taking out Supernova last July), so there's nothing to suggest the odds are wrong--it's worth noting the German was also a huge underdog against Nerchio and beat him, but Polt is in a different category
Losers
Elazer vs TLO - the German is favoured (60%), although both have a similar, middling records in ZvZ; the difference is TLO loses to well-known players while Elazer can still fall to lesser known Europeans (Guru. Revolution, Revolver, etc), so I like the odds here
Winners
Polt vs MajOr - the Korean is favoured (60%) despite his opponent holding a slightly better head-to-head record (11-10); MajOr has actually beat Polt the last three times they've played (most recently at IEM in February), which puts doubt into the latter's excellent TvT record against foreigners (11-1); were it any Korean Terran other than Polt this would be an automatic decision for me, but I like the upset here
Finals
Polt vs TLO - see above, and I don't think the loser beating the winner (as is often the case in WCS in re-matches) will occur here; if this is MajOr instead of Polt he's favoured (67%), with a 4-1 record against the German

Aligulac and I both have Polt and MajOr moving forward (TL has TLO instead of MajOr).

This article is written by Peter Levi (@eyeonthesens)

WCS Premier Predictions: Groups C & D

Aligulac was a perfect 4-0 in the first two groups (3-1 for me and TL), as Kane bowed out in Group A (Bunny said "he's been a bit lazy lately"), and StarDust was able to overcome an initial loss to TooDming to make it through.  Looking forward, each of the following groups includes a largely unknown player (Mystery and NXZ), so let's look into the Aligulac crystal ball and see what's in store (I'll note TL's predictions whenever they get posted):

Group C
Mystery vs Ret - the Zerg is a massive favourite (81%); the Chinese player's PvZ is not impressive, while Ret has done well against mid-level Europeans (besides losing to Deathmask), so the odds seem on target
viOLet vs MarineLord - the French Terran is favoured (57%) and has a winning record against the Korean (3-1, granted that was over a year ago); viOLet's ZvT is all over the place, but he's hardly been playing of late so it's hard to know what to make of it; MarineLord's TvZ has been a mixed bag as well, so I see this as a coin toss--I'd like to think the Frenchman is trending upwards while the Korean is slowly spiralling down, so I'll stick with the odds
Losers
Mystery vs viOLet - the Korean is a huge favourite (84%) and I see no reason to argue; if this is MarineLord he's just as favoured (84%)
Winners
Ret vs MarineLord - the Terran is heavily favoured (73%), and while Ret has been quite good in ZvT of late, I don't think his opponents measure up to MarineLord; if this is viOLet instead see the odds below
Finals
Ret vs viOLet - once again the Korean is favoured (69%) and Ret's ZvZ is not impressive, so there's no reason to argue; if this is MarineLord, see the odds above

Both Aligulac and I see MarineLord and viOLet going forward (TL picks the Korean and Mystery).

Group D
iAsonu vs Xenocider - the Chinese player is a slight favourite (55%), despite not having played a series against a Terran since January; however, Xenocider's TvZ has been pretty bad the last couple of months so I'll stick with the odds
Harstem vs NXZ - not surprisingly the popular Dutch player is heavily favoured (73%); Harstem's PvZ has been excellent of late while NXZ has struggled with players like Blysk, so the odds seem spot on
Losers
Xenocider vs NXZ - there's a slight edge (51%) for the American and given how middling both are in the match-up it's tough to call; as the better known player it's hard not to pick Xenocider, however, he's been so bad in matches lately (2-5) that I'll take the upset
Winners
Harstem vs iAsonu - the Dutch player is favoured (66%) and his Chinese opponent hasn't had much luck in ZvP outside his region, so I'll stick with the odds
Finals
iAsonu vs NXZ - the Aussie is given the slightest of edges (50%); NXZ has been good against his fellow SEA Zergs of late (9-1), while iAsonu is 5-0 in the match-up of late--I think the latter scene is stronger so I'll favour the Chinese player; if this is Xenocider instead, re-matches generally go to the loser in WCS (we saw it again with StarDust this past weekend), but Aligulac would stick with the Chinese player

Aligulac and I have Harstem and iAsonu going forward (TL has the Dutch player and Xenocider).  It would be great for the SEA region for NXZ to pull off an upset, but it's not something I'm expecting.

Copa America has held its qualifiers for its second season (again, sadly, with no English stream covering it).  In the second qualifier Fenix came out of retirement to participate (do SC2 pros ever really retire?), but lost in a qualifying match to BLord.
Qualifiers from the first round were: Cham, XionS, Demonizer, Erik, ReasoN, ByRada, JimRising, and Smile.  All the players except Demonizer made it previously (he admittedly had an easy bracket).
Qualifiers from the second round were: BLord, OdiN, Spectrum, Shakti, siNder, Sektor, Akroz, and DominiC.  Only BLord and OdiN made it last season.
There are seven new players appearing in the main tournament, with virtually the same racial balance as last season (one less terran, one more protoss).  Notably Catz did not play in this season, meaning someone else will get the chance to be obliterated by MajOr or Kelazhur.

Just a quick note on predictions in Korea (which I don't do generally): I was curious what TL's accuracy was, so here's a look (I couldn't find the prediction posts for some of the SSL groups (A and D for season two, D in season one)).

GSL (S1): 20-11
SSL (S1): 8-5
GSL (S2, R32): 12-4
SSL (S2, R16/R8): 1-5

These are interesting numbers.  First of all, the GSL predictions are off the charts in places (24-8 in the R32, that's 75%), but how does that mesh with the SSL struggles (9-10), or similar ones within the GSL (3-5 in the R16, 2-2 in the R8)?  It's an interesting factoid, as this early round accuracy isn't reflected in either TL's SSL predictions (as noted) nor their WCS picks.  Perhaps it's simply a statistical oddity.  Regardless, food for thought.

This article is written by Peter Levi (@eyeonthesens)

StarCraft News & Notes

 

Lycan decided to tackle why foreign players don't measure up to Koreans on The Late Game.  This has been addressed numerous times, but it remains an issue.  The focus was on foreign habits: should players practice more? do they need coaching? should they play more custom games together? should they participate in Skype groups? Etc.  As interesting as all this is, none of it addresses the underlying problem.  Every region outside Korea lacks the infrastructure and money required to create a player base that can compete with Koreans.  There's very little money available in SC2--in terms of prize winnings last year only 10 foreign players made 20k or more; there's a shortage of teams that actually pay a salary (almost none of whom pay a living wage), so not many parents are going to indulge their children and allow them to focus all their time on becoming better.  Once a player is living on their own the idea of being a full-time pro is economically almost impossible.  Given how few tournaments a foreigner can reasonably expect to make a name for themselves in, they're left to streaming and doing that successfully involves skills that have nothing to do with pro play.  Until someone (a company, an individual) wants to spend the money to create the necessary infrastructure, Korea will remain forever ahead regardless of whatever tweaks players make to their practice regime.  As it stands that white knight can only be Blizzard and we can only hope they make the necessary changes for LOTV.

 

 

This leads into a lot of confusion about a full region lock--some fans wonder what the benefit is--won't players suffer by missing out on the stronger competition Koreans offer?  The answer to the latter is yes, but only to an extent and only in the short term.  Region-lock provides a space for players to make a name for themselves, to make some money, and targets to aspire too.  The expenditure needn't be extraordinary (Rotterdam thought 10k-15k would set up an appropriate EPS-style NA tournament, for example).  Whatever anyone thinks of how Copa America is run, it draws in a larger player base than WCS NA--clearly a region-locked league is attractive to the local talent.  Despite interest in the community to create something like this, at the moment it's just a pipe dream.

 

 

TL continues to work on their economy model and as well-intended as they are, comments like this do make me wonder just how much they've thought through things:

A nonlinear curve encourages various 1-base cheese and all-ins as well. Frankly, we didn't look at it much when designing Double Harvesting model

How did they not consider cheese?  Especially when people commenting on their model have specifically brought it up as a problem prior to this?  Regardless, putting incredulity aside, I'm curious what the end-game of the TL experiment is.  There's nothing wrong with perusing the point as an intellectual exercise, but unless they have a pipeline to the LOTV designers it's ultimately not going to impact anything.  On the positive side, if there's ever an SC3 it's the kind of data that could be used for design consideration.

Speaking of the economy, qxc offered his current thoughts on LOTV and they are well worth looking into.  He notes:

Overall, Legacy is filled with more action than HoTS. The 2 main reasons for this are the economic changes. The starting workers means that the first ~2 minutes are gone from the game. You jump right into making your first depot/rax almost immediately and have at least some units out shortly thereafter. This removes certain early game plays such as 6 pools, proxy 1 gate, and proxy 2 rax, but there still seems to be potential for larger econ busts. The timing window that a 6 pool took advantage of doesn't really exist for the equivalent 12 pool, but things like a fast 15 hatch into ling/bane aggression off minimal drones still seem reasonable. While some of the early builds are gone, the game benefits from their absence. Some of the most frustrating and coin-flippy builds were builds like 6 pool and so on. The early game of Starcraft becomes a little less luck based as the options for scouting open up and there's fewer threats before scouting is really an issue

I agree wholeheartedly about being happy to lose the early all-ins in SC2 (IEM Katowice last year between sOs and herO is a great example of how a series can flop because of cheese).  Among the other things in the post is about how powerful the viper's parasitic bomb is (which players are slowly starting to use in LOTV tournaments), along with how the buffed adept is helping Protoss.  He doesn't get into balance per se, but at least in match-ups with Terran he suggests TvZ is fine (the Lycan League, which gets better sign-ups than desRow's tournament, has had 3 Terrans and 1 Zerg as its top-four the last two weeks, with Bly winning both).  From what I've observed, Protoss still needs some tweaks--as good as adepts are, it's easy to counter them if that's all your opponent is going to do (as KiWiKaKi discovered in Take's Legacy of the Ultralisk tournament).

The latest LOTV update is out with a few nerfs to Terran, one to Zerg, along with the new Terran unit.  I'm surprised nothing was done to Protoss, which still has issues (the roach change will help PvZ a bit), but it will be interesting to see how the changes play out.

Maynarde gave a long and interesting interview over a week ago and those interested in the Aussie caster should check it out.

Viewership is one of those things SC2 fans like to fret about--more so than is rational in some cases--but as a game that was the vanguard of eSports when it came out it's no surprise that fans embrace the "ded game" meme and keep tabs on things.  Numbers aren't always easy to get hold of (apparently the Chinese scene can hit 100k views), but there's always Conti at TL doing his best to provide them.  A year or so ago (link below) I looked at how WCS was trending in terms of views and I thought it was worth revisiting the numbers to see how HOTS is trending in its final days (I picked WCS since that's Blizzard's flagship product for SC2).  First, numbers for season one and two (all via Conti: here, here, here, and here, with some rounding for simplicity):

Qualifiers
Season One
NA 17k/14k/12k/12k (average 14k)
EU 13k/13k/14k/13k/15k/13k (avg 13k)
Season Two
NA 6k/7k/8k/8k (7k)
EU 9k/18k/13k/12k (13k)

Challenger
Season One
NA 13k/14k (13k)
EU 19k/15k/20k/13k (17k)
Season Two
NA 9k/9k (9k)
EU 18k/19k/19k (19k)

Premier R32 (season one)
Group A 18k (Snute's group)
Group B 20k (Polt's group)
Group C 22k (Bunny's group)
Group D 25k (Kane's group)
Group E 32k (FireCake's group)
Group F 36k (NaNiwa's group)
Group G 17k (Hydra's group)
Group H 17k (HuK's group)
Overall: avg 23k; NA broadcast 18k, EU broadcast 29k
R16 35k/38k
R8-Finals 54k

In terms of comparison, using the R32 as a baseline, compared to season one in 2014 (granting that EU/NA regions are now combined), EU viewership for the R32 is slightly better on average (29k vs 25k), while NA viewership is largely unchanged.  The numbers for EU are not much lower than the HOTS high point of season one in 2013 (with a 31k average).  All of this is to say the common consensus that SC2 core viewership remains steady seems true, but clearly LOTV is required to push for growth (along with some tweaks to how WCS is done).  Other stray observations: NA viewership for the qualifiers and Challenger were down in season two, perhaps due to the involvement of more Koreans (the R32 matches this past weekend were unaffected)--certainly neither Hydra, StarDust, or Jaedong have done anything to boost numbers in NA.

A brief note about balance in the GSL/SSL, since we've moved far enough along in season two to compare (keep in mind the GSL is still using the old map pool--apparently they can't get enough of Overgrowth).

GSL Season One R32: 13 Terrans, 12 Protoss, 7 Zergs
GSL Season One R16: 6 Terrans, 5 Protoss, 5 Zergs
GSL Season Two R32: 12 Terrans, 10 Protoss, 10 Zergs
GSL Season Two R16: 7 Terrans, 5 Protoss, 4 Zergs

With the identical map pool similar results are to be expected and that's what we see here; the only notable change is that there are more Zergs in the R32, but once we hit the R16 everything is essentially as it was.

SSL Season One R16: 7 Protoss, 5 Terrans, 4 Zergs
SSL Season One R8: 3 Protoss, 3 Zergs, 2 Terrans
SSL Season Two R16: 8 Protoss, 6 Zergs, 2 Terrans
SSL Season Two R8: 4 Protoss, 2 Zergs, 2 Terrans

Given the general lack of change at the top level in either league I think the small variance is less a matter of balance and more a matter of who the better players are (8 players have been in the R16 of the GSL each season, and 9 have repeated in the SSL--herO, Life, Maru, and Dream have done it in both).

This article is written by Peter Levi (@eyeonthesens)

WCS Premier Predictions: Groups A & B

The draw for WCS Premier went off without a hitch and the groups are quite varied, from a soft Group D to a difficult Group E.  The first two groups take place in California, with a dynamic Group A and a less exciting Group B.  As always, I'll go through Aligulac's numbers and see what the odds say about who moves on to the R16 (I ran these numbers two weeks ago as well and I'll note changes were they occurred).

Group A
Jim vs Kane - the Canadian is favoured (64%) and his top-8 finish last season certainly emphasizes that (granted, pre-patch results); the two have never played before, but the Chinese player's PvZ has not been impressive of late so I'll stick with the odds
Bunny vs Petraeus - the Danish Terran is favoured (57%; it was 54%), losing to his opponent recently, but Bunny is the better player and Petraeus was not impressive at DreamHack, so I'll stick with the odds
Losers
Petraeus vs Jim - the Zerg player is an overwhelming favourite (72%; it was 74%), despite Jim holding a 5-0 record against him (and a lot of Petraeus' positive ZvP numbers being built off of B-grade Europeans); that said, Jim hasn't been impressive in the match-up lately, so I'll stick with the odds
Winners
Kane vs Bunny - the Canadian beat his opponent in the R16 last season, but the Dane is still favoured (63%; it was 62%); other than showmatches Kane hasn't lost to a Terran in 2015 (a match-up not dependent on the old swarm hosts), so I'll go with the upset
Finals
Bunny vs Petraeus - see above (although it's worth remembering in WCS re-matches the loser was 5-2 last season); if this is Kane instead of Bunny, then Petraeus is favoured (65%; it was 66%); given the paucity of the Canadian's ZvZ results the only argument to make is that the Canadian is the better player

For Aligulac it's Bunny and Petraeus, whereas I have Kane and Bunny (the latter echoed by Team Liquid).

Group B
TooDming vs StarDust - the Korean is heavily favoured (69%; it was 70%); the two have never met before and the sample size from StarDust recently is much too small to make anything of, but TooDming has never beaten a Korean of his calibre (Arthur is as close as it gets), so I won't argue with the odds
ForGG vs Sen - not surprisingly the Terran is massively favoured (73%), having crushed Sen last season in WCS; the Taiwanese player has one significant win over a Korean Terran (Bomber back in August), but it has been almost a year since ForGG last lost to a foreign Zerg, so I find the odds convincing
Losers
TooDming vs Sen - the Taiwanese player is the favourite (65%), having won the only previous meeting between them (2-0 back in 2013); Sen also has a much stronger track record in ZvZ, so all things being equal he should win here
Winners
ForGG vs StarDust - the Terran is once again heavily favoured (75%; it was 73%), with a good historical edge against the Protoss player (8-5); StarDust is only 1-8 in his last BO3 vs Koreans Terrans, while ForGG's only recent losses are against Rain and San; I like the odds
Final
StarDust vs Sen - the Protoss player is favoured (57%; it was 59%), the two never having met before; Sen hasn't had a significant result against a Korean Protoss since 2013, so I'll take the favourite

It's no surprise the two Koreans are favoured and despite StarDust being a beatable opponent I don't think either foreigner is well-matched against him, so I'll stick with he and ForGG (TL agrees).

This article is written by Peter Levi (@eyeonthesens)

StarCraft News & Notes

KiWiKaKi has re-appeared out of the aether to qualify for TakeTV's Legacy of the Ultras tournament (using an Adept-heavy style; admittedly not having to beat any heavyweights--Tarantius and ParanOid are the most notable); subsequently ROOT posted a short interview with him.  What stood out for me was this:

I never liked HOTS when it came out because of swarm hosts and it's a big reason I didn't come back.

As far as I know he's the only pro besides NaNiwa who specifically left the game because of that unit.  Whether he'll stick with LOTV remains to be seen, but it's good to have an NA foreigner return to SC2 after such a long break.  For those unfamiliar with the Canadian Protoss player, he was at his best in 2011 (here's an old Day9 episode that includes him).

DreamHack Tours has come and gone and on the whole was a great success (the finals were excellent and closer than the 3-1 score might indicate; for some analysis of the tournament you can check out The Late Game [I'll post the link once Lycan puts it up on Youtube]).  In terms of foreign performances it was a mixed bag.  Only MMA was knocked out by a foreign player (both Scarlett and Revolver beat him in the second group stage), although four Koreans lost in non-elimination matches (Lilbow beat HyuN, TargA beat Leenock, Verdi beat Heart, and Bly beat San).  Harstem made it the furthest (R8), largely on the strength of his PvZ against fellow foreigners (he lost to HyuN); Lilbow should have beaten MarineKing, but threw the deciding game.  Snute and MaNa also made it to the round of 16.

Speaking of tournaments, Lambo became the latest EPS champion (beating HeroMarine); while the German player hasn't qualified for Challenger this year, I'm surprised he's currently teamless.  Lambo seems to be virtually the only Zerg left in the EPS now that Hanfy doesn't participate (technically there's also KiToO, but he doesn't get very far).

The latest LOTV balance update has come out and includes a mix of strange and interesting changes.  The tweaks might help TvZ a bit, but there's nothing on the table to balance PvZ.  The Lurker nerf seems to be the only one everyone embraces.  Both Remax and The Late Game discussed the changes (in the latter I have no idea why Jakatak continues to complain about force fields and the warp-in mechanic--he's smart enough to know Blizzard isn't going to completely redesign Protoss, so he's better off with constructive thoughts within the realm of possibility--blanket balance whine about Protoss is pointlessly toxic).

I was surprised to see that some pros still believe Nerchio cheats in online cups (he discusses the issue in an old Team Acer interview); without proof its too speculative to take seriously.

Now that we're into season two of GSL, I thought it was worth taking a snapshot of how the Koreans who returned from WCS America/Europe have done against stronger competition back home (excluding ByuL who returned to Korea before the soft region lock, as well as players who are retired, inactive, or switched games: SelecT (NA), TheStC (NA), JYP (NA), NesTea (NA), Balloon (NA), Mvp (EU), Golden (EU), and Genius (EU)).

GSL
MMA (EU) - R4 (S1), Code S in S2
YoDa (EU) - R16 (S1), Code S in S2
Bomber (NA) - R32 (S1), Code S in S2
MC (EU) - R16 (S1)
San (EU) - R32 (S1), R32 SSL (S1)
Patience (EU) - Code S in S2, R32 SSL (S1)
TaeJa (NA) - R32 (S1), R32 SSL (S1)
HyuN (NA) - R16 SSL (S2)
First (EU) - R32 (S1)
Sacsri (EU) - Code S in S2
Seed (NA) - R32 (S1)
Heart (NA) - R32 (S1)
HerO (NA) - Code S in S2

Others: jjakji (EU), Pigbaby (NA), Alicia (NA), Arthur (NA), Revival (NA), Check (NA), CranK (NA), Ryung (NA), aLive (NA), and Oz (NA, failed to get into Code S in season one, now retired or inactive)

No surprises in the results, except perhaps for Patience (whose reputation in EU had sunk quite low).

Fenner posted an update on his earnings as a streamer, so for those interested in that check it out.  Speaking of Fenner, he recently played through the SC2 campaigns, which was pretty entertaining and worth checking out if he ever posts them.

This article is written by Peter Levi (@eyeonthesens)

StarCraft News & Notes

 

A lot has happened over the last month or so.  The WCS R32 groups were drawn, once again without the drama of season one.  On the whole we have interesting groups, granted that most of the marque names wound up in Europe.  In terms of competitive balance, there doesn't seem to be any group where it's obvious who will get both R16 spots, although I'll save specific analysis for my prediction posts.  This time only two groups (F and H) will feature just one non-mirrored match-up, which is excellent for the sake of variety.

The groups for Dreamhack have also been announced and hopefully the tournament will remain competitive all the way through (unlike Gfinity this past weekend).

Troy Pavlek addresses the perception of SC2 and see's plenty of signs of life.  I agree with him, although I do think more opportunities for growth need to exist and that the NA scene needs help.

Speaking of NA, on Remax Rotterdam talked about a desire to see an EPS-like system instituted for NA.  He has a plan for how it would work, but doesn't have the 10-15k required to fund it.  I think it's a good idea and hope that in the future something like it will appear, allowing space for NA players to truly develop and grow.

A few weeks ago Blizzard talked about the possibility of buffing the recently nerfed swarmhost (!), Snute offered some thoughts and the one thing I want to point to is this:

if Zerg starts underperforming

That's the real question, because we haven't seen that yet (on Remax the sentiment was that the current map pool is Zerg-favoured, which you can take with a grain of salt from five non-Zerg players).  One interesting note to all of this was Destiny saying he didn't think the swarmhost change would impact ZvP meaningfully and that he was much more concerned with ZvT versus mech--what's funny about this is that the swarmhost was most problematic in the former match-up, which is a good illustration in how a unit designed to do counter one thing can cause all sorts of problems when it does something else.

I've never seen so much ink spilled over the economy in SC2 before.  qxc made a post responding to the initial TL suggestion, then David Kim talked about it, and TL responded to the latter.  As interesting as it all is, I agree with Nathanias (via The Late Game) that until there's a good competitive sample we really won't know if there's a better way to do it than LOTV does now.  I also agree with Nate that the issues in Legacy are much more about balance than they are about the economy.

Speaking of LOTV, there's an interesting thread on TL about giving the zealot a small buff to make it more useful.  It's unlikely to happen, but it's an interesting idea.  The most recent balance patch still hasn't resolved the problems Protoss has in the beta, although ZvT might be more balanced now.

BaseTradetv's live event was posted on Kickstarter and has gone well past achieving its goal.  I think they did a good job with the various tiers they posted and I'm looking forward to the event.

Breaking Out wrapped up on the weekend with Canadian Bioice winning Feardragon's tournament (he hasn't posted the replays yet, but they are available on Twitch).  I enjoyed what I was able to watch and having player cams used while games were being played was a good innovation.

I caught some of a gold/platinum tournament casted by DBReaper and others a week or two ago and while I think it's great that there are tournaments for lower leagues, it would be nice if at least one of the casters brought some analytical knowledge to the game (watching chat having to explain the importance of combat shields in TvT was a little painful).

The top-50 streamers for April were posted on TL (with apparently a few errors judging from the comments), but for those interested, there it is.

There's also this:

And for some funny ABomB stories, check out the latest StarCraft Jeopardy.

A bit tangential, but The Daily Dot talked to Destiny about how exactly Twitch streamers make a living.

This article is written by Peter Levi (@eyeonthesens)

WCS Challenger Season Two: Looking Back

After a hectic week WCS Challenger has wrapped up across all regions.  Only half the players (16) are returning to Premier, although none of the Koreans fell.  In generally I think the turnover is good for the system (6 of the top-16 from season one failed to return, but only 1 from the top 8)--it's a more interesting dynamic to see new players join the mix.  In terms of the biggest upsets: Elazer beating ShoWTimE, NXZ knocking out PiG, and Mystery taking out MacSed stand out.  I wouldn't expect deep runs from any of these players, but it does illustrate some depth in the scene (particularly in regions like SEA and China).

In terms of balance:
PvZ 3-4 (includes Hydra vs Zeal)
ZvT 2-2
TvP 3-2
There were also 8 ZvZs, 4 PvPs, 3 TvTs.

Map balance (excluding Hydra's match):
Vaani (13): TvZ 2-2 ZvP 3-1 PvT 1-4
Coda (12): TvZ 2-1 ZvP 1-3 PvT 1-4
Echo (11): TvZ 1-2 ZvP 0-4 PvT 3-1
Iron Fortress (7): TvZ 1-0 ZvP 3-1 PvT 2-0
Cactus Valley (5): TvZ 0-1 ZvP 1-0 PvT 1-2
Expedition Lost (3): ZvP 3-0
Inferno Pools (2): ZvP 1-0 PvT 0-1

It's not a great map pool for Protoss (Expedition Lost, Cactus Valley, Iron Fortress, and Inferno Pools are all bad), while Coda and Echo seem bad for Zerg and possibly Echo for Terran.  It's a small sample size however, so it will be interesting to see how this shifts in Premier.

We are left with 14 Zergs, 9 Terrans, and 9 Protoss, which is similar to season one (one less Protoss and one more Zerg).  The swarmhost change has not (yet) shown itself in results, although MorroW showed mech players how not to play against late game Zerg.

In terms of predictions, here's how things went by region (Aligulac/me where different):
EU: 8-6/6-8
NA: 7-0
China: (didn't get the chance to make predictions beforehand)
Latin America: 2-0
Taiwan: 2-0
SEA: 1-1
Overall: 20-7/18-9
These percentages are much higher than last season's Premier numbers (74% vs 54% for Aligulac) and it will be interesting to see if that holds going forward.

A few specific observations
-part-time players got crushed (which is a good sign)
-ShoWTimE played poorly against Elazer, dying to a spine crawler rush and bungling a pair of cannon rushes
-I had high hopes for Bunny vs Beastqt, but instead it was endless doom drops by both
-GunGFuBanDa crushed NaNiwa, but the Swede has taken right approach in realising he needs to practice more and not rely simply on talent
-the NA branch of Challenger was incredibly lacklustre, mostly featuring 3-0 stomps by the expected winners; even the 3-2 series featured games that lacked drama

This article is written by Peter Levi (@eyeonthesens)