Saturday 29 August 2015

WCS: Looking Back at the Round of 32


The R32 is in the books and rather than delve into ForGG retiring after BlizzCon or the R16 matchups, I'd like to look back on the round that was.  I thought the most entertaining groups were D and E, while there was an abundance of one-sided series' in general.  The biggest upset was Hitman getting through, although taking out Elazer he's hardly knocking off one of the titan's of Europe.  The Chinese players all crashed and burned this season, with Shana and Jim the only 0-4's of the round.  Here's how predictions turned out:
Aligulac: 14-2 (!)
Me: 11-5
TL: 10-6

Both are personal bests this year for me and Aligulac (whose success approaches the ridiculous), while TL dipped ever so slightly.  For whatever reason the R32 has always been hardest for me, while the later rounds get easier--this is the opposite trend for both Aligulac and TL (thus far).

Racial balance
PvZ 5-6
TvP 3-5
ZvT 4-4

Map balance
Coda (22) PvZ 5-6 TvP 2-4 ZvT 3-2
Cactus Valley (15) ZvT 2-4 PvZ 3-2 TvP 1-3
Terraform (12) PvZ 3-3 ZvT 2-2 TvP 2-0
Iron Fortress (9) TvP 2-3 PvZ 1-3
Bridgehead (5) ZvT 3-0 PvZ 1-1
Dash and Terminal (3) PvZ 1-1 ZvT 0-1
Moonlight Madness (3) TvP 1-2

The Koreans make it hard to interpret these numbers, whose only notable is a seeming edge for Zerg in ZvT on Bridgehead.  Coda remains the most popular map, with Terraform replacing Iron Fortress as the third most after Challenger.

Viewership
NA broadcasts
Group A 23k
Group B 18k
Group C 24k
Group D 25k
EU broadcasts
Group E 17k
Group F 18k
Group G 16k
Group H 15k
Average: 20k; NA 23k, EU 17k

This is a drop off from previous seasons (viewership through seasons one and two was pretty similar), with about 3k less viewers overall and a significant drop for the EU streams.  I'm not sure if the drop can be blamed on LOTV coming soon, the inevitable march of the Koreans, or that the days were so short (incidentally, those who'd argue Koreans draw viewers should note the most watched stream, Group D, had no Koreans and only one top-foreigner).  The quality of games was not that great overall, featuring few memorable (or long) matches.  At any rate, statistically it's not the biggest of drops so there's no point in dissecting it too much.

This article is written by Peter Levi (@eyeonthesens)

Tuesday 25 August 2015

WCS Predictions: Groups E-H


The second half of the R32 of WCS starts today so it's back to predictions for me.  At the moment Aligulac is 7-1, while TL and I are 5-3.  There have been no major upsets or surprises, with elfi the only player to make it to the R16 who hasn't been there before this year (TL predictions):

Group E
Lilbow vs Neeb - the French player is an overwhelming favourite (77%); the American's limited history as a Protoss isn't that helpful, nor is his (relative) dominance in PvP in the NA scene; I see no reason to question the odds
Shana vs GunGFuBanDa - the German is also a heavy favourite (72%); the Chinese Terran's record is almost inscrutable--he's been good lately in his local scene, but as most of his opponents are players we never see it's hard to know what to think about it; the German's PvT is more explicably, but outside Germany it's more spotty; in the absence of information I'll stick with Aligulac
Losers
Neeb vs Shana - the American is slightly favoured (56%), but despite having been a Terran he's struggled in PvT (3-5 in series since the switch); the Chinese player has a much better record of late (11-1), but it's so hard to gauge his level without knowing the players he's beating (think about how the similarly obscure Mystery got crushed in season two), so I'll stick with the known quantity and go with the American
Winners
Lilbow vs GunGFuBanDa - the French player is only a slight favourite (56%); the German is 5-3 in PvP series against foreigners of late, while his opponent is 9-3 over the same period; I see no reason not to go with the odds here
Finals
GunGFuBanDa vs Neeb - the German is a heavy favourite (72%); no American has made it to the R16 this year (can't wait for the real region lock), and no North American since Kane in season one; I think Neeb isn't strong enough in PvP to see this one out so I'll stick with the odds; if this is Shana, see above

Aligulac, TL, and I have Lilbow and GunGFuBanDa returning to the R16.

Group F
Zanster vs Kane - not surprisingly the Swede is favoured (65%); looking shaky against HuK at HIAT doesn't seem relevant as his ZvZ is excellent; it's a surprise that the Canadian made it this far and while he has the talent to win here there's no reason to think he will (he hasn't played a ZvZ series since May, incidentally)
iAsonu vs ForGG - the Korean is a huge favourite (71%), and the Chinese player was unimpressive last season (0-4), so I see no reason to question the odds
Losers
Kane vs iAsonu - the Chinese player is favoured (55%); with the Canadian player semi-retired and not that great in ZvZ, I can hesitantly go with Aligulac (it's worth noting that when I've assumed Kane was going through the motions he's tended to win)
Winners
ForGG vs Zanster - the Korean is favoured (66%); the Swede struggles in ZvT and the Korean is a monster in the match-up, so I'll stick with the odds
Finals
Zanster vs iAsonu - the Swede is favoured (61%), and I'll take his dominant record in Europe over the Chinese player's solid record in his scene; if this is Kane see above

Aligulac, TL, and I have ForGG and Zanster moving forward.

Group G
Snute vs Jim - the Norwegian is heavily favoured (71%), and the Chinese player's record against well-known players in PvZ is awful, so I'll stick with the odds
Hitman vs Elazer - the Polish player is a slight favourite (54%); the American is a bit like Has where if people aren't prepared for him he can do very well, but if they are his chance to win almost disappears; this wasn't the Zerg player's best match-up, although he's done well of late (8-3, despite losing to kAnji)--I want the upset here, but there's no reason to predict one
Losers
Jim vs Hitman - the Chinese player is favoured (57%), although PvP is truly the American's element (he's beaten Arthur, Patience, and Sage, for example); we haven't seen Jim in a PvP outside his region in a long time--I think he's the better player, so I'll go with the odds, but an upset is possible here
Winners
Snute vs Elazer - the Norwegian is favoured (69%), and he's a monster in ZvZ (11-2 of late), while the Polish player is very much a mixed bag (8-5); with Snute's 10-2 lifetime record against his opponent it's easy to stick with the odds
Finals
Jim vs Elazer - the Chinese player is a slight favourite (54%), but his PvZ numbers get a huge boost from crushing no-name players from his region--he hasn't done well against Europeans; the Polish player's ZvP is hardly perfect, but I think there's an "upset" to be had here

Aligulac has Snute and Jim going forward, whereas TL and I have the Norwegian and Elazer.

Group H
viOLet vs TLO - the Korean player is favoured (59%) despite failing out of the R32 the last two seasons and looking awful in doing so; he has almost no recent ZvZ history vs foreigners recently, but he has been getting his head kicked in by fellow Koreans; the German player's record is hard to judge--in WCS this season he's 7-1 (only losing to Hydra) in the match-up, while piling up a mediocre record outside of that--clearly he puts much more effort into WCS and I think TLO will continue the trail of tears for the Korean
XiGua vs MaNa - I'm surprised both players are here, but the Polish player is heavily favoured (79%), and it's easy to understand why--his PvZ is excellent (11-2 lately), while his opponent has floundered against known players, so I like the odds here
Losers
XiGua vs viOLet - once again the Korean is favoured (76%), and this time I'll take it as the Chinese player's ZvZ record is all over the place (3-3 since going 0-3 in season one)
Winners
TLO vs MaNa - the Polish player is favoured (68%), although the German has the historical edge on his opponent (24-16); TLO's track record in ZvP has not been good this year so I'll stick with the odds
Finals
viOLet vs TLO - the odds are above and typically re-matches go to the loser, but I like the German player here

Aligulac has the Korean and MaNa moving forward, whereas I have the latter and TLO, and TL has the German and XiGua.

This article is written by Peter Levi (@eyeonthesens)

Friday 14 August 2015

WCS Predictions: Groups C & D


Group B is in the books with Aligulac again 2-0 (TL and I were 1-1 again).  It was hard watching Kelazhur's devastation when he lost to elfi, although it was fun seeing the normally cheesy Finn playing macro games against him.  Onto the odds and predictions for the weekend:

Group C

ShoWTimE vs Sen - the German is favoured (69%), and given how awful the Taiwanese player has been in WCS this year it's no surprise (0-4 in series, 1-8 in maps, although three of those losses were to Koreans); Sen has only played one ZvP since April making his current form difficult to discern; the German has struggled a bit lately in PvZ (losing to Elazer, Guru, and FireCake), but in general is very good in the matchup so I'll take the odds

Jaedong vs HeRoMaRinE - the Tyrant is only a slight favourite (55%), but it's difficult to see why that is given the German's middling TvZ record (5-5 of late); while Jaedong is susceptible to top foreign Terrans, I don't think that's what he's facing here (at least not yet), so I'll stick with the odds

Losers
Sen vs HeRoMaRinE - the German is favoured (56%), even though (as mentioned above) this is not his best matchup; once again there's virtually no good data on the Taiwanese player in the matchup (although historically in his local scene he's excellent), so I'll stick with the odds

Winners
Jaedong vs ShoWTimE - once again the Korean is only slightly favoured (55%), but the German has a terrible PvZ record against Koreans (1-10) and the Tyrant hasn't lost a meaningful ZvP against a foreigner since losing to HuK last August--so I'll take the odds

Finals
ShoWTimE vs HeRoMaRinE - truly an EPS final, the Terran is just slightly favoured (53%); he's won 3 of their last 4 meetings, including a BO5 just a month ago; I think the Protoss might be the better overall player, but the Terran is going to be in his comfort zone (the German scene is mostly TvP for him) and is clearly confident against his opponent, so I'll stick with the odds

Both Aligulac and I have Jaedong and HeRoMaRinE moving forward; TL likes the Tyrant and Sen.

Group D

MarineLorD vs MaSa - the Canadian, just like Kelazhur above, has been spending his time at the ROOT house and this is his last chance to make an impact in WCS with that as a springboard; Aligulac gives him the smallest of edges (51%), but that's based on him wrecking fellow NA players; while the French Terran does sometimes lose to the best Europeans, there's nothing in the Canadian's track record to suggest he'll win here

Has vs iaguz - players who created opposite impressions last season (the Aussie ascending and the Taiwanese player descending); the Gimli Terran knows it will be an all-in every game so he simply has to defend; the odds favour him (51%) ever so slightly; the Protoss player has only played one series against a Terran who wasn't Chinese or Taiwanese (losing to Bunny), but I think the Aussie has the chops to stop him as well so I'll stick with the odds

Losers
MaSa vs Has - the odds are with the Canadian again (60%), although much of that is built off crushing the NA scene (which is really no different than the Taiwanese player dominating his local scene); I don't have a strong opinion here so I'll rely on Aligulac; if this is MarineLorD instead he's heavily favoured (69%)

Winners
iaguz vs MarineLorD - the Aussie doesn't get to play TvT that often and I like the odds (65%) that the French Terran will move forward; if this is MaSa see below

Finals
iaguz vs MaSa - once again the Canadian is favoured (66%), despite having lost 3 of the 4 series the two have played; I think MaSa has the talent to win here, but I like the Gimli Terran's experience and I think he'll win out; if this is MarineLorD the French player is favoured (65%) again

Myself and TL have MarineLorD and iaguz moving forward, Aligulac likes MaSa and the French player.

This article is written by Peter Levi (@eyeonthesens)

Thursday 13 August 2015

WCS Predictions: R32 Group B


Aligulac won the day in Group A as Harstem seemed to panic in both base trade scenarios he faced getting eliminated from both series because of it (forgetting to bring probes with him against Polt and losing track of his mothership core against Petraeus), leaving TL and I at 1-1.  On to predictions for Group B:

elfi vs Starbuck - the Finnish player gets a slight edge (53%), but it's hard to get a feel for where the Protoss player is at in PvZ as he's only played one meaningful series in the last three months; the Slovenian's ZvP has been shaky since he returned to active playing, losing to DmC, fraer, and DnS recently; I don't feel confident in predicting this one, but without other information I'll stick with the odds

Hydra vs Kelazhur - the ROOT team kills have been unremarkable in WCS this year and I see no reason not to accept the odds (82%)

Losers
Starbuck vs Kelazhur - the Brazilian's numbers get skewed from him butchering players in Copa America and NA, so he's favoured (54%), but the Slovenian's ZvT track record (including beating Bunny to get here) is hard to overlook (the last foreign ZvT series he lost was over a year ago), so I'll take the upset

Winners
elfi vs Hydra - there's no reason to question the odds (85%), however much fun it would be to have an upset here

Finals
Starbuck vs elfi - a rematch and rematches typically go to the loser--given how difficult it was to pick previously I'll take the "upset" here; if Aligulac is right and this is Kelazhur see above

I have Hydra and Starbuck going forward, whereas Aligulac has the Korean and elfi (TL has the obvious and Kelazhur).

This article is written by Peter Levi (@eyeonthesens)

Wednesday 12 August 2015

WCS Predictions: R32 Group A


It's time for yet another round of WCS predictions begins--once again I'll go through Aligulac's numbers and see what the odds say about who moves on to the R16.  Just some general news beforehand: uThermal has had to withdraw from WCS due to poor health, with Elfi replacing him.  On to Group A [the TL preview arrived after I posted, so I've included their predictions in square brackets]:

Harstem vs MajOr - Aligulac generally favours the Mexican Terran (his rating buoyed by crushing players in Copa America regularly) and that remains the case here (64%); he does have a 6-1 record against the Dutch player (granted they last played in 2014); throughout this year I've put my faith in MajOr and he's utterly failed, so I've learned my lesson and I'm going with the upset [TL sticks with Aligulac]

Polt vs Petraeus - surprisingly the Korean gets only a slight edge (50%), but the Kiwi has never beaten a Terran of his caliber so I'll take the odds here [TL has the favourite]

Losers
Petraeus vs MajOr - the Zerg is favoured (56%); the players have split the games they've played (10-10), but the Kiwi has won their last two meetings; I like the odds here; if this is Harstem once again the Zerg is favoured (61%); the Dutch player has a slight edge in their matches against one another (7-5), but they haven't played since the swarmhost nerf [TL likes the Protoss over the Zerg]

Winners
Polt vs Harstem - as expected the Korean is a huge favourite (85%) and despite the Dutch player's win over GuMiho I see no hope for him here; if this is MajOr the odds are closer (66%) and the Mexican has a decent score against Polt (11-14), but lost the last two series against him [TL likes the Korean]

Finals
Harstem vs Petraeus - the odds are above (61% for the Zerg), but the Kiwi hasn't been as impressive in ZvP since the swarmhost patch (losing to players like HuK, Strange, and Verdi recently), so I like the upset here; if this is MajOr instead of the Dutch player, see above [TL has the Protoss beating MajOr]

Aligulac has Polt and Petraeus moving forward, whereas I have the Korean and Harstem [TL agrees].

This article is written by Peter Levi (@eyeonthesens)

Tuesday 11 August 2015

StarCraft News & Notes


According to Game Informer (reporting on Mike Morhaime's statement to Activision) Legacy of the Void will come out this year (2015).  This isn't a surprise--there was never going to be a lame duck WCS season of HOTS in 2016.  While no specific date was mentioned, at or after BlizzCon (November 6-7) makes the most sense.

Speaking of LOTV, Allied Commanders was officially announced (along with a video about it), and latter demoed at IEM Gamescom.  It's another interesting angle on the game and I'll be interested to see how it's received.


I had an idle thought while watching ASUS ROG: I wonder how successful HOTS would have been if the initial release was as it was now (no old school hellbats, no unnerfed swarmhosts, along with nerfed blink/msc/tempest)--not good enough to prevent the decline in SC2, but perhaps that decline would have been less severe.  As for the tourney itself, PtitDrogo had an excellent run to the final before running into Losira.  Because of the prize pool, the tournament is not a Premier event, so the French player doesn't become only the second foreign Premier finalist this year.

Speaking of tournaments, IEM Gamescom ended with an occasionally exciting finale between soO and INnoVation (won by the latter), while poor Gfinity was stuck with two B-grade Koreans in First and DeParture--I can't help but think how much better things would have gone for the latter if they'd made it a foreigner-only event (much love to Gfinity for running SC2 events, but I don't see the point in having Koreans involved unless you are going to bring the best).

Apparently the TeSL has cut SC2 without a formal announcement, but given they'd run nothing in 2015 it's not a surprise.


Feardragon's first NA Ladder Heroes tournament is in the books as JonSnow beat Drunkenboi to win (the latter inexplicably race-switching to Protoss one game and got crushed).  There were a couple of late no-shows for the tourney, but I think that was more a case of bad luck than a worrying sign.

Speaking of the ladder, SC2 unmasked "debuted" (Rotterdam and others remind people it's been around forever, but it's the first time I've ever heard of it), and among other things I was curious how many hackers/boosted accounts would be revealed (only NA ones are identified):
Hackers: 18 (5 in the top-50, 7 in the top-100, 13 in the top-200)
Boosted: 5 (3 in the top-50, all boosted by European pro Krr, who fans might remember was banned from season three of WCS in 2014 for account sharing; 3 in the top-100, 5 in the top-200)
BsK has also launched a similar site whose purpose is simply identifying smurfs/barcodes, but he seems focussed on EU from what I can tell.


JaKaTaK talked about the possible changes to Protoss suggested by the SC2 devs (including giving the nexus warp-in power), I thought he made the more radical suggestion that warpgate should no longer be required (the only requirement being building a pylon initially), but I apparently misheard him--that concept (just thinking about it) would require compensation for the other races as well, but it's an interesting idea.  He also talked about the proposed changes to the macro mechanics (something he's onboard with)--he was attacked by avilo about this on Twitter, but subsequently joined him on his channel to discuss the issue (not the most valuable discussion, but good on him for entering that space with his opinion).


There has been a ton written and said about the macro elements of SC2; with David Kim announcing their removal from the beta what has been theoretical discussion will be put into practice.  Let's briefly recap the two sides of the argument.  Stuchiu (TL) addresses the macro elements (bringing up my point that "fun" is not an argument), talking about how eliminating these elements would impact Zerg:
First, all units come from the hatchery. Second, you do not need to make a building to unlock supply; you make an overlord from a hatchery so you never have to do the extra actions of selecting a worker, pressing the hotkeys, selecting a location, and shift-queuing back. Finally, there is no end point to creep spread and inject. Those two mechanics can be done throughout the game ad infinitum with decreasing importance. To cut or put inject larvae on autocast is to simplify zerg macro beyond comparison. The inject larvae forces zerg to imitate the macro requirements of the other races (where they have to make production and pylons/depots). Additionally there is no end state for zerg. Eventually, a protoss/terran gets the maximum amount of production feasible. Zerg’s macro mechanics of inject and creep spread must be done for however long the game lasts, whether it’s 5 minutes or 2 hours. But in return zerg gets a larger strategical advantage the more larvae they bank and the more creep they spread.
It's an excellent point, although Stuchiu isn't arguing that macro changes shouldn't be made (you can read more about that in his article).  The most popular argument against this comes from a Reddit post by BigWiggly1 (which is perhaps even longer than the above), but boils down to this:
A fundamental to game design is to make the experience exciting for the player. Every interaction that the player has with a game should be exciting in some way. You'll notice that the term "exciting" will get thrown around a lot. I'm not talking christmas-morning-presents excitement or fireworks over the bay. I'm talking about small things that excite and encourage your brain and your decision making abilities. It doesn't take much, and it's one of the reasons that art styles are important to game design (I won't bring up art again).
He goes on to talk about "value added" (the popular business expression) and I think that's a better way to make his argument, although I don't find it particularly persuasive.  The idea of changing macro mechanics in the game is fine with me, albeit we have to see how it goes.  TotalBiscuit actually weighed in on this (an infrequent thing for him to do with SC2 of late); he thinks the experiment is worth trying and that overall eliminating the macro has more benefits than minuses to the scene.


I don't understand iNcontroL's hostility to tracking macro mechanics (ala injects) for broadcasts.  His argument was that it would make players look bad which would hurt selling them as good players.  I find this bizarre--it's simply a metric like any other, so even if a player isn't perfect (and none are) it's all relative and I don't see how someone looks "bad" unless they are far below the average.  If you look at other competitive enterprises, more information, more stats are always a good thing for the coverage of competitive play and the fanbase (look at the explosion of fantasy sports which thrives on this kind of data).  If you follow his argument to its logical extreme it becomes even more absurd.  Maybe I'm missing something, but on the surface it's an odd complaint and not very convincing.


I don't typically delve into the drama in the scene, but occasionally it reaches a fever pitch where it has to be mentioned at least--something former SC2 personalities always want to comment on as well(ala DJ Wheat this time around--any chance to talk about how negative the community is can't be missed).  I won't get into the specifics, but Destiny did a stream discussing it and desRow made a post about it, so you can indulge if you wish.  The best thing to come out of it was this from Day9.


Speaking of drama, some of you may recall the hubbub when it was claimed back in January that Kaitlyn and LiviBee (among others) got their accounts boosted into GM (if memory serves only the latter accusation was serious).  Interestingly, with SC2 unmasked we can see the former has three legit GM accounts in NA while the latter has one.

Also on the streaming front, another ex-SC2 player rose from the dead on Twitch as XlorD streamed for the first time since last October (he hasn't played competitively since 2013).  On the flipside, BlinG has decided to retire--if this had happened last year I would have been shocked, but in 2015 his performance has dramatically declined so it's an understandable decision.


Returning to DJ Wheat, he discussed eSports in the context of DOTA 2's International and not surprisingly believed that an online delivery system (ala Twitch) is the best one for it (as opposed to a more traditional medium like TV).  As self-serving as that is (he works for Twitch), for the most part I think he's correct that eSports doesn't translate that well to television (this doesn't mean it couldn't be rearranged to suit TV's format, but traditional media seems to be on the way out so there may not be a point to it).

This article is written by Peter Levi (@eyeonthesens)

Tuesday 4 August 2015

WCS Challenger Season Three: Looking Back


WCS Challenger is in the books and with the Premier matches already determined (thanks again to a giggling Apollo), so it's time to step back and look at the results (specifically NA/EU).  In terms of predictions, here's how things went for myself, Aligulac, and TL:
Me/TL: 14-7 (66%; NA 4-3/3-4, EU 10-4/11-3)
Aligulac: 13-8 (61%; NA 4-3, EU 9-5)
This is a slight improvement from me over season two (13-8), a slight decrease for Aligulac (15-6), and an improvement for TL (11-7; for some reason they didn't predict FireCake vs Zanster and I can't find their predictions for the first day of NA).

Balance:
PvZ 2-2
ZvT 3-2
TvP 3-2

Map scores:
Coda (12) TvP 4-1 ZvT 3-1 PvZ 2-1
Cactus Valley (10) PvZ 3-1 TvP 2-1 ZvT 1-2
Iron Fortress (9) TvP 2-2 PvZ 2-1 ZvT 1-1
Terraform (8) ZvT 3-1 TvP 2-1 PvZ 1-0
Dash and Terminal (6) TvP 2-1 ZvT 1-1 PvZ 0-1
Moonlight Madness (4) ZvT 2-1 TvP 1-0
Bridgehead (3) ZvT 1-1 PvZ 0-1

This is a pretty even distribution between the top-five maps (and it's no surprise older maps were favoured, with Terraform getting the most love out of the new maps).  It's difficult to get a balance feel from such a small sample, but Terrans enjoyed an edge on Coda over Protoss (continuing a trend from season two).

Other observations: the most entertaining matches in NA were the PvP between Neeb and StarDust (link) and the PvT between Hitman and Xenocider (link); most of the other series were one-sided stomps and the five-set match between Kane and HuK wasn't particularly dynamic.  On the EU side there were one-sided series as well (especially day two), but overall it was more competitive, including the most entertaining ZvZ I've seen in awhile between Zanster and FireCake (link) [a reader reminded me of TLO vs Bly was, so here's your link].  Quite a few players in both regions got caught using the same strategy over and over again and got crushed.

The R32 consists of 13 Zergs, 10 Terrans, and 9 Protoss, which compared to last season is one less Zerg and one more Terran.

This article is written by Peter Levi (@eyeonthesens)