ZvZ from WCS America Premier League

About a week and a half ago, WCS America season 1 Premier League had a group (stage 2, group C) with 3 Zerg players. With the patch out increasing Spore Crawler damage against Mutalisks, we got to see what ZvZ looks like when you can actually do different things.

Before this balance patch (which doesn’t have its own number, unfortunately), ZvZ was all Mutalisks. Early game was Zergling and Baneling trades, where the defender usually had the advantage. Once it got to Lair, Zergs would invest gas in Mutalisks, and getting the 1st volley or having 1 or 2 more Mutalisks was the difference.

Post-patch ZvZ looks more like WoL ZvZ. The early game is the same, the midgame is mostly Roaches, and then you mix in Hydras or Infestors (and eventually both) until the late game comes late. Most games would be decided around 3 bases in the midgame, however.

So the builds. Unfortunately, the matchups, cheeses, and technical difficulties mean that we have 3 games between Suppy and Revival and only 1 between Revival and Scarlett, but this is a good set of them. All of those are below in long form. Just a few notes ahead of time. I’ll be updating my Zerg strategy guide based on the recent changes to ZvZ, so read that for the bigger picture.

Continue reading

Spawning Tool upgrades: authentication and an improved schema

For all Americans, I hope you had a good Memorial Day weekend. If you’re unfamiliar with it, Memorial Day is a federal holiday for honoring departed members of the US armed forces. Given the time of year, many people will go out to grill and enjoy the weather. I did some of that, but I also got started on the new season of Arrested Development and built a number of new features for the Spawning Tool.

First, I added basic authentication so you can login and upload replays. This mechanism isn’t particularly interesting, but it’s necessary to restrict who can edit the tags on a replay. I once thought it best if anyone could edit any tags, but I think that’s inviting vandalism. If anyone feels so strongly to add tags to other replays, we’ll come up with a mechanism to enable that.

Second, I improved the Browse Replays interface. You can now do text search and toggle a few options. The actual list of replays is now much more readable thanks to built-in bootstrap styles.

Third, I actually incorporated more parsed data from the replay into the actual database schema for the site itself. The full explanation requires some context.

The Spawning Tool is primarily composed of 2 parts. First, there’s the spawningtool tool, which actually goes through the replay data and pulls out the build order. The result of this tool is basically just a bunch of structured text. Second, there’s the spawningtool site, which gets this text and turns it into a presentable website.

Previously, I was just taking the raw build order data as a whole and re-rendering it from scratch every time. It was easy and was sufficient for displaying builds. The downside was that the spawningtool site was pretty blind to what the data was. I re-implemented it so that the build order is stored in the database itself: instead of a raw build order in text, the build order itself is stored with times, units, and supply counts in the database.

The biggest benefit is coming soon: more advanced filters and statistics from replays. With this data, we can make database queries that translate naturally into questions like, “If you build a Robotics Facility before 7:00 in PvT, how often will you win?” It’s just a matter of finding all replays with builds that fit the criteria, then counting the wins and losses.

I’m very excited about the next steps for the Spawning Tool, and I think that advanced filters is a big part of that. Additionally, I’m working with ChanmanV to make the Spawning Tool a replay archive for practice games. Tune in at http://youtu.be/vstWWo0Gmao?t=29m30s for more about that.

One last thing: I’m in the process of moving the Spawning Tool off of my personal server and onto Amazon Web Services. As such, there may be some downtime in the near future. Deployment might be rough, but I think this will ultimately lead to a much more stable environment.

MVP’s 2 base Marauder/Hellbat/Marine All In

Game 3

Game 4

If you weren’t watching the WCS Europe Finals, you were missing out. I myself happened to only catch the finals after sleeping in, which were some fun games. Games 1 and 2 were pretty standard, but in games 3 and 4, MVP came out with a sick 2 base all in.

It starts with Hellbat drops shortly after 7 minutes, which is relatively normal and can do some damage. From 7 to 10 minutes, MVP gets a Hellbat Marauder composition, which can defend against Roach Zergling pushes (especially if the Hellbat drops force the Zerg to go all in). Then, he cranks out 4 Marines at a time with 2 Reactored Barracks to have Marines ready for 11 minute Mutalisks. The push then goes out slightly after that.

So here are the builds.

MVP v. Stephano, WCS Europe Finals Game 3, Newkirk District

Hellbat drops into a 2 base Marauder/Hellbat/Marine All In

  • 10 Supply Depot
  • 12 Barracks
  • 13 Refinery
  • 15 Marine, Orbital Command
  • 16 Supply Depot
  • 16 Marine
  • 18 Factory
  • 19 Reactor
  • 22 Command Center (low ground)
  • 4:45 Swap, Hellion x2, Supply Depot
  • 5:15 Armory, Hellion x2
  • 32 5:50 Starport
  • 6:00 Barracks Tech Lab
  • 6:10 Orbital Command, Refinery
  • 6:15 Hellbat x2 (continuous)
  • 40 6:40 Medivac (continuous)
  • 7:15 Stimpack, Marauder (continuous), Hellbat drop goes out
  • 8:00 Barracks x2
  • 9:15 Barracks Reactor x2
  • 10:00 Marine x4 (continuous)
  • 10:50 Concussive Shell
  • 110 11:15 Push, SCV pull

Some notes:

  1. MVP never gets a Reactor on the Starport. I think Terran players just instinctively swap, but he wants the Reactor for the Hellbats.
  2. There’s no 3rd Refinery, which may contribute to the lack of Medivacs
  3. There’s also no Engineering Bay
  4. He cuts SCVs from 7:30 to 9:00, then again after 10:00. He has the money, so maybe it’s a mistake, but this is MVP, so maybe he really need the money

And the second build. It starts with Reapers, but the rest of the build turns out similarly.

MVP v. Stephano, WCS Europe Finals Game 4, Star Station

Reaper Expand into Hellbat drops into a 2 base Marauder/Hellbat/Marine All In

  • 10 Supply Depot
  • 12 Barracks
  • 13 Refinery
  • 15 Orbital, Marine (canceled for Reaper)
  • 16 Supply Depot
  • 18 Reaper (both out to harass and scout)
  • 20 Command Center
  • 21 Factory
  • 23 Barracks Reactor
  • 24 Refinery
  • 26 5:30 Swap, Hellion x2, Armory, Marine
  • 6:00 Starport
  • 6:15 Hellion x2
  • 6:50 Barracks Tech Lab, Hellbat x2 (continuous), Medivac (continuous)
  • 54 7:30 Stimpack
  • 7:45 drop goes out
  • 8:00 Marauder (continuous)
  • 8:15 Barracks x2, Bunker
  • 9:25 Reactor x2
  • 9:45 Barracks x2
  • Marine x4 (continuous, x6 with
  • 10:55 Combat Shield
  • 120 11:15 Push, SCV pull

Some notes:

  1. The big difference is that he goes for a Reaper opening in this game. He doesn’t cut the 13th SCV to get out a faster Refinery, though, so perhaps it’s not too important
  2. Compared to the last game, he isn’t cutting as many SCVs. Again, not sure how to read that.
  3. At 9:45, he gets out 2 more Barracks for more Marines. Again, I’m not sure how to read this choice.

I like the build a lot, and because of that, Zerg players beware: you might be seeing a lot of this on the ladder.

Terran Strategy guide completed (and some Spawning Tool news)

I feel like I took far longer to get the Terran strategy guide up than either of the other guides. And this is despite actually playing Terran nowadays. In any case, I finally got the last of my edits in, so it is out of construction and worthy enough of your attention.

Like the other guides, I’m committed to maintaining it for the foreseeable future, so let me know if you see any mistakes or have any recommendations. Already I’m seeing that the Protoss and Zerg guides have gotten out of date. The Zerg guide is actually painfully underspecified, and Protoss play has changed dramatically. Some of that is Naniwa’s PvZ 1 Gate Expand, and some of that is better defense. So you may see some updates there, and I will of course post here about the big edits I’m considering.

In unrelated news, ChanmanV shared some exciting developments about Spawning Tool on Pro Corner earlier today. Tune in at http://youtu.be/vstWWo0Gmao?t=29m30s for about a 10 minute discussion of his thoughts (pretty well aligned with mine) on what the rough vision of the site is.

Development is going to be continuous, so look out for more updates there. In the meantime, please upload any replays you have handy. To do good analysis, we need both good tools and lots of data, so please send your replays our way.

Tagging Added to the Spawning Tool

It’s funny when StarCraft personalities mention their league promotions since they inevitably are lower than you would expect them to be. I believe the TotalBiscuit recently made Platinum, and Husky made Master, though for all I have heard them talk about StarCraft in the past, I always thought they were at least Masters. Well, you all should know by this point that I’m happily in Diamond, and between content for this blog, following Proleague and Fantasy StarCraft, and the Spawning Tool, I’m beginning to see why these personalities don’t play much. There just isn’t that much time.

The most recent task (and first major feature added to the Spawning Tool since launch) was adding tags to replays, which I completed Monday night. If you look at a replay, you can see a few gray boxes with red text, and those give context and metadata for the replay. Moreover, you can now browse replays by filtering by tags, so if you’re interested in finding all PvT games, you can easily narrow in on those replays.

Moving forward, I hope that tags become to primary method of organizing replays on the site. There is a lot of metadata present in the replays, which I could have (and still may) spit out into specific fields in the database, but I think tags encompass that feature and allow for more flexibility. Note that tags are filed under various categories, which I think should also improve organization. It’s easy to label a replay based on the map it was on, the event it was from, the players in the game, the build orders used, and so forth. I’m not exactly sure what tags are most useful for searching, but it should hopefully be self-organizing.

The strengths of flexibility and community input are also its biggest weaknesses. Because tags can be anything, we could see a lot of strange, esoteric, and unnormalized tags out there. For example, “My Favorites” isn’t helpful because it’s specific to a user, “1 Barracks Expand into Medivac Drops into Early 4th into Late Game Reaper” isn’t helpful because it’s just too specific, and “having both “1 Rax Expand” and “1 rax expand” leads to fragmentation*. Currently, the best solution I have is to include auto-complete in inputting tags so that you are guided to the right result.

Because I’m counting on community input, tags are also open to all users, registered and anonymous. I would like users to be able to upload replays anonymously, and they should be able to tag them as well. This does lead to an odd asymmetry in that I haven’t yet built deletion or editing of tags since that could lead to massive vandalism. I’m toying around with ways to split up permissions based on whether you’re registered or not, but I’m open to ideas.

Anyways, please take a look at the site, play around with tags, and let me know what you think. After the big hit on reddit, traffic has died down a lot, and I’m okay with that. The proof of concept came quickly, but I imagine it’ll be about a month of development before I make another big push for people to start using the site.

In the meantime, I was wondering whether I should put the word “Beta” on the site somewhere. Notice that the front page has a big disclaimer on it for basically the same purpose, but I think “Beta” has been overused and overextended heavily, and I kind of want to fight my little fight. On the other hand, it’s exactly the right description for the Spawning Tool while it’s still in heavy development. Vote below if you have an opinion.

Should I slap "Beta" on Spawning Tool?

  • Yes (75%, 3 Votes)
  • No (25%, 1 Votes)

Total Voters: 4

Loading ... Loading ...

* I considered normalizing all input by removing punctuation and reducing it to lowercase, but 2 counter-examples came up. First, “Roach/Hydra” is much easier to read than “roachhydra”, and second, “HerO” and “herO” are 2 different players

Apollo’s HotS Terran TvP

You all know how much I love Apollo’s tutorials, where he plays 4 hours with each race starting from bronze. He usually shows 1 or 2 builds for each race and focuses on safe, reactive, macro play. Well, he’s back with HotS tutorials, and he’s started with Terran.

The games are unfortunately wacky, so there’s not too much to learn from them. I’ll probably post again when he finishes the series, but I’m excited to share.

Terran versus Terran

In TvT, he starts with a Reaper-Expand opening. The early Reaper is mostly for scouting, but you can kill SCVs if the Terran player reacts slowly. Notably, there is no longer a SCV scout, so you should focus on scouting with that Reaper. In the early game, the big dangers are 1) Proxy 2 Rax (where they build 2 Barracks near your base and attack in the first few minutes) and 2) drops (Hellion, Hellbats, Widow Mines). If you can get through that, then you can focus on a normal Marine/Tank army.

He didn’t get through his build without disruption, but here’s the start:

  • 10 Supply Depot
  • 12 Barracks (at ramp)
  • 12 Refinery (delay next SCV slightly)
  • 15 Orbital Command, Supply Depot (at ramp), Reaper (scout)
  • 18 Reactor (Marine x2 when ready)
  • 19 Command Center (in-base)
  • 20 Factory
  • 21 Marine x2
  • 24 Marine x2
  • 27 Starport, Hellion x2, Swap, Orbital Command

Actually, he kind of got through the build the first time, but I disliked it so much, I assumed that he did something wrong.

Terran versus Zerg

In TvZ, he again uses a Reaper-Expand opening. Scouting is more critical since Zerg all-ins can come at any time. By walling at your ramp, you should be able to block any early Zergling aggression, especially since Reapers can outrun Zerglings. With a quick Factory and Reactor, you can get out 4 Hellions for harassing and scouting, then switch into Widow Mines for defense.

We have another abbreviated build for this matchup, though it should look identical to TvT

  • 10 Supply Depot
  • 12 Barracks (at ramp)
  • 12 Refinery (delay next SCV slightly)
  • 15 Orbital Command, Supply Depot (at ramp), Reaper (scout)
  • 18 Reactor (Marine x2 when ready)
  • 19 Command Center (in-base)
  • 20 Factory

Terran versus Protoss

TvP is also a Reaper-Expand, except he follows up with more Barracks instead of the Factory. I would explain more, but he actually through his build, and it should be pretty self-explanatory.

Apollo’s standard HotS TvP

  • 10 Supply Depot (at ramp)
  • 12 Barracks (at ramp)
  • 12 Refinery (delay next SCV slightly)
  • 15 Orbital Command, Supply Depot (at ramp), Reaper (scout)
  • 18 Reactor (Marine x2 when ready)
  • 19 Command Center (in-base)
  • (sees wall at ramp, cannon)
  • 24 Barracks x2
  • 27 Bunker (at natural)
  • 30 Orbital Command
  • 30 Supply Depot
  • 33 Tech Lab x2
  • 35 Stimpack, Marauder x2 (continuous)
  • 53 7:30 Refinery x2, Factory (reactor), Engineering Bay
  • 8:00 Combat Shields
  • 8:30 +1 attack
  • sees Twilight Council
  • 9:00 Turret at front, Starport
  • 9:45 Concussive Shells
  • 10:00 Swap Factory Reactor, Medivac x2 (continuous)
  • 10:15 Refinery
  • 10:30 Command Center (in-base)
  • 11:30 Barracks x3
  • 12:30 Engineering Bay, Armory, Attack

Conclusion

Since I actually play (and get crushed as) Terran nowadays, I have stronger opinions and totally plan on writing that guide later this week. To be honest, though, most of my StarCraft attention is currently focused on spawningtool, which continues to go well. Currently, it’s really focused on build orders, and a cursory glance at this blog should explain that obsession. Keep an eye out for development there!

Flash build orders from Proleague (TvZ Hellbat drops!)

At some point, my friend George told something like, “It just doesn’t feel like real StarCraft unless it’s Proleague.” Well, thanks to Fantasy Proleague, I have really gotten into Proleague, and it’s just about all I watch.

The timing is not terrible. Friday and Saturday nights work well, and I’m willing to stay up Sunday/Monday night until 1 watching games. The gaps between games give me time to take care of chores and get ready for bed. But of course, the real draw is that these are the best players playing high stake games and are willing to do all sorts of crazy all-ins, map-specific strategies, and builds tailored for the matchup.

This past weekend was the start of round 5, which is an all-kill round. Instead of each team picking a full lineup of 7 players, the winner of each match stays on while the losing team picks another player. There’s more variance for Fantasy StarCraft, and you don’t get as much exposure to new players, but the aces come out for more games. Innovation looked good in an all-kill of EG-TL, but I think the bigger bounce-back was Flash, who did poorly last round and got knocked out of his GSL Code S group of death (Life, Innovation, Parting, and Flash).

I have a few Flash builds written up. Let’s jump into them.

Flash’s Standard TvP

(v. Terminator on Neo Planet) (TwitchYoutube)

  • 10 Supply Depot
  • 12 Barracks, Refinery
  • 15 Reaper, Orbital
  • 17 Command Center, Reactor
  • 18 Supply Depot
  • 19 Bunker
  • 20 Engineering Bay
  • 20 Marine x2 (continuous)
  • 27 5:20 +1 attack
  • 30 5:45 Barracks
  • 6:20 Barracks
  • 6:35 Refinery
  • 6:50 Tech Lab
  • 7:10 Factory
  • 7:15 Refinery
  • 7:20 Stimpack
  • 7:30 Tech Lab
  • 7:55 +1 armor
  • 8:15 Reactor (on Factory), Starport
  • 8:35 Combat Shield
  • 9:15 Medivac x2
  • 9:40 Armory
  • 9:45 pushing out, delaying Protoss 3rd
  • 10:00 Command Center

This build should look pretty normal: Reaper opening into Bio. The funny part of htis build is the very early Engineering Bay. This lets him get out +1/+1 upgrades for the usual 10-11 minute push. Otherwise, this should be reassuring since it looks like Apollo’s progression.

Flash’s TvZ Hellbat Drops into Mech
(v. Roro on Akilon Wastes) (Twitch, Youtube)

  • 10 Supply Depot
  • 14 Command Center
  • 15 Barracks
  • 16 Refinery
  • 19 Marine, Orbital Command
  • 22 Refinery, Factory
  • 23 Reactor
  • 26 Supply Depot
  • 27 Bunker
  • 29 Starport
  • 6:00 Swap, Hellion x2 (+ a few more), Marine
  • 6:10 Armory
  • 6:35 Medivac (continuous)
  • 7:15 Hellbat x2 (continuous)
  • 7:30 Factory, Tech Lab (will swap onto Factory)
  • 8:00 First drop out, Refinery
  • 8:45 Siege Tanks start
  • 9:00 Command Center
  • 9:20 2nd drop arrives, Refinery
  • 9:35 +1 Attack
  • 10:05 Factory

I myself have been looking for a Hellbat drop build, and it looks roughly like what I should have expected. CC first if you like, or maybe don’t: I certainly won’t. Otherwise, it looks like Polt’s drops, but you add in an Armory for the Hellbats. Disclaimer: none of us can micro like Flash, but Flash was continuously dropping at 2 locations. The Starport and Reactored Factory are an endless stream of drops, and even Roro was losing tons of workers. This is a great way to punish quick 3 bases from Zerg.

Flash’s TvT Mech
(v. Reality on Whirlwind) (Twitch, Youtube)
10 Supply Depot
12 Barracks
15 Refinery
16 Marine, Orbital Command
17 Supply Depot
17 Marine
19 Marine
20 Command Center, Reactor
22 4:20 Factory, Supply Depot
23 4:50 Marine x2
5:20 Starport, Refinery, Swap Reactor, Hellion x2 (continuous)
Orbital Command
6:10 Viking (defend drops)
6:50 Tech Lab (Barracks)
7:20 swap onto Starport, Raven
7:50 fends off a drop and run-by (takes some damage)
? 9:00 Hellion pressure
8:30 Command Center
9:00 Armory x2
9:45 Refinery x2
9:50 Factory x2
10:30 +1/+1 vehicle
11:00 Siege Tanks x2, Hellbats x2 start
11:20 Engineering Bay
12:00 Medivacs, Missile Turrets

I don’t have much to say about this build: it kind of speaks for itself. I have just 2 things to draw attention to. First, there’s the early Raven. I have been looking at ways to integrate a single Raven into my own builds (great in TvZ for clearing creep, right?), so I guess the answer is just to go for it. Second, Siege Tanks start really late for mech play. Barring any signs that you need the defense, I guess it’s okay to go up to 3 bases before getting Tanks. I guess it worked in this game because Reality revealed his hand in the 7:50 Marine/Hellion pressure, but even so, that’s really late.

I hope you like the builds. To be honest, all Terran play is starting to look very similar to me, but I guess that makes it easier to put together a guide. On that note, I heard that Apollo will be doing another set of videos, so you should look forward to those. On a related note, I’m feeling good enough to write up a Terran guide at this point. Maybe you can look forward to that.

Finally, Blizzard is getting serious about replay analysis, and I’m very excited. They apparently have enhanced the data provided and released an open source library for parsing that data. I have already put together a tool for extracting build orders that I’m calling the Spawning Tool, so check that out when you get a chance. Note that this is primarily just a proof of concept, so there are a lot of bugs. I would go into more detail, but that’s not really what this post is about, so expect more news about that soon!

Play Fantasy Proleague!

The 5th round of Proleague is starting tomorrow night, and in the 36 hours before that, sick nerd ballers everywhere are filling out their Fantasy Proleague teams. I highly recommend that you all play.

If you’re unfamiliar with Fantasy sports, the idea is that you put together a team of players, and based on their performance in real games, you get points based on their statistics. For Fantasy Proleague, each player has a value. At the beginning of the round, you put together a team with a value up to 30 for your main team, and at least 13 points on your anti-team. When players on your main team wins, you get points. When players on your anti-team win, you lose points. Each week, you can make a few trades, and player values fluctuate based on their performance. That’s it.

Is this nerdy? Tremendously. But if you watch Proleague, it will definitely enhance your knowledge of the games, and if you don’t watch Proleague, you should. In the Americas, it’s conveniently on in the evenings over the weekend. You will pay a lot more attention to individual players and follow their performance much more closely, which hopefully enriches your StarCraft knowledge. And fantasy in itself is fun.

To get started, you just need to sign up at http://www.teamliquid.net/fantasy/proleague/Entry.php. If you want to do additional research, here are a few more resources:

I hope you play this round and enjoy!

P.S. Here’s my team right now. I might change my mind multiple times in the near future:

Main Team

  • 8 – Innovation – might be the best SC2 player. Also can win a lot of games against EG-TL this week
  • 6 – SoS – people seem to think he’s undervalued, and I would agree. He’s good, and he’s on Woongjin Stars
  • 6 – CJ herO – he didn’t do so well last round, but he’s still the CJ ace and should bounce back
  • 3 – Brown – he starts in week 1, is on SKT1, and did well in his first Proleague round in round 4
  • 2 – Mekia – he starts in week 1 and is on Woongjin Stars. Good enough for me
  • 1 – Bear – I didn’t see any 2 point players I wanted, and he’s on STX Soul, so that’s worth something
  • 4 – SKT1 – They have Rain, Fantasy, and Parting. And then they have the rest of their lineup

Anti-Team

  • 5 – Crazy – I wanted to go 5-4-4, and had a really hard time picking a 5 point player. I went with Zero, but WJS are just too deep, so Crazy and KT seem safer
  • 4 – Hitman – KT hasn’t looked good recently, and Hitman is pretty deep on their lineup (5th best player, behind Hydra who I already don’t like)
  • 4 – Mini – again, he’s deep on their lineup (maybe 5th best player, 3rd best Protoss)