Starcraft II

[MountainGo]

My latest crushing defeat is against protoss. They went massive cannons (I mean, I ran into a line of 10 cannons with my early zealot rush) and then transitioned into upgraded void rays. How do people find the money for this sort of thing?

The standard wisdom is that if you see that many cannons, just stay the heck away from them and expand immediately (if I saw THAT many cannons, I would take an early third as well). You should be able to maintain a large advantage in both economy and army until eventually not even their cannons can save them. And make sure you don't let them expand anywhere not guarded by those cannons. As for how they can find the money for Cannons and Void Rays, I'm guessing they just didn't make any other units. Hence why it's so easy to expand like crazy against an over-turtler.

I've been focusing on a 3 gate zealot push, with delayed warp gates and a balanced economy (9p 12 gate, 15 gate, 16 z, 16p, 18 core, 20 gate, get warp gates, and pump out units, while constantly producing workers)

This build seems questionable to me (but I could be wrong). First of all, three production buildings isn't enough to spend all the money from one base. Second, I don't see the point in going heavy Zealots and delaying the Core. Normally you only see a delayed Core if you're going for the very early rush with two Chrono boosted Gateways, generally proxied. This works because it maximizes the strength of the attack in the early game. But if you're not attacking until after you've made the Core anyway, then I don't see the point.

and this build works fairly well, unless I run into early void rays/mutalisks.

And there you go. Suddenly your opponent is not punished for cutesy fast teching because you can't hit air. If I were you I would just change to a more standard 4-gate or 3-gate robo.

In general it sounds like your problem is that your macro is slow, but I couldn't say without seeing a replay. Obviously that will improve with time, but for now I'd recommend taking a more standard approach with your builds.

[MountainGo]

Now it's my turn to ask for advice. I'm finding it difficult to maintain a clear mental state when I play. I tend to get hyper-focused on whatever I'm looking at, rather than being able to take a step back and evaluate the overall game situation, what build I think my opponent is using, etc. Today I played 30 games, trying to focus on improving my awareness of the opponent. I planned to make sure to ask myself each game exactly what production buildings I thought my opponent was getting at the beginning of the game. I forgot to do this every single time (realizing at the end of each game and thinking "I can't believe I forgot again!") until the very last game I played. I just get sucked into the game and forget everything I'm trying to improve upon. I'm sure it's at least partly due to the mild panic that comes from trying to keep up with my macro, but I think it's also due to general stress about trying not to screw up. Anyone have any experience with finding ways to keep calm and lucid?

p.s. Grand finals of MLG Dallas, anyone? Thanks, Blizzard. I'm going to kill myself.

[CSamurai]

And there you go. Suddenly your opponent is not punished for cutesy fast teching because you can't hit air. If I were you I would just change to a more standard 4-gate or 3-gate robo.

My opponent is never punished for going early air. Ever.

The one sure way to beat me is to push void rays out like a punk. No matter what I do, fast expand, 3 gate robo expand, 2 gate 2 robo expand, 3 gate 3 star, expand phoenixes, I lose to air. Air kills me. Air destroys me. Early air wins. I have almost taken to resigning when I see the first void ray because nothing, and I mean nothing, I do is going to kill those void rays. I could build an equal number of void rays, and they'll beat me on upgrades. I could build a massive amount of stalkers, and they'll harass my production until I'm unable to build anything.

I can't beat air, ever.

As long as I'm not going to beat air, I might as well beat everything else.

[fwiffo]

That was embarrassing. I was irritated that the casters had to make excuses for Blizzard about how it wasn't a problem with Battle.net, but a problem with some Internet connection. The thing is, it doesn't matter. They could build perfect, awesome Battle.net servers with five nines of uptime. You would still need a LAN option because the Internet itself is intermittently unreliable or slow.

[MountainGo]

And Facebook integration is just an annoyance for me. So I guess I should just plug my ethernet cable into a rock.

[MountainGo]

I can't beat air, ever.

As long as I'm not going to beat air, I might as well beat everything else.

I know how you feel. Sometimes I feel like there's nothing I can do against Banshees. But you just have to stick it out and keep trying and retooling your approach until you get better and better.

Also, realize that if a game goes longer than, say, 18 minutes, then a big part of the apparent impossibility might come from mistakes made earlier on, such as falling behind in economy by staying on one base. When I evaluate one of my replays, I don't bother looking at anything after I fall significantly behind.

[schilds]

Now it's my turn to ask for advice. I'm finding it difficult to maintain a clear mental state when I play. I tend to get hyper-focused on whatever I'm looking at, rather than being able to take a step back and evaluate the overall game situation, what build I think my opponent is using, etc.

Sit on your hands .

Ok, so that's go advice ....

[Solomon]

Now it's my turn to ask for advice. I'm finding it difficult to maintain a clear mental state when I play. I tend to get hyper-focused on whatever I'm looking at, rather than being able to take a step back and evaluate the overall game situation, what build I think my opponent is using, etc. Today I played 30 games, trying to focus on improving my awareness of the opponent. I planned to make sure to ask myself each game exactly what production buildings I thought my opponent was getting at the beginning of the game. I forgot to do this every single time (realizing at the end of each game and thinking "I can't believe I forgot again!") until the very last game I played. I just get sucked into the game and forget everything I'm trying to improve upon. I'm sure it's at least partly due to the mild panic that comes from trying to keep up with my macro, but I think it's also due to general stress about trying not to screw up. Anyone have any experience with finding ways to keep calm and lucid?

Don't take this the wrong way, but I think an attachment to win/loss ratio and ladder points play a part. At least, this is the case for me - when I play on my anon account with an empty friend list and pretty much no one knows about, I find myself calm and level-headed even in situations where I'm getting proxied. The win/loss ratio isn't very good and the ladder points aren't as high, but I don't care because it's a punching bag account anyways. Instead of caring about winning, I always try to execute a particular BO I watched in a high-level game / work on some strategy that I normally don't feel comfortable doing (e.g: blink stalker rush), and basically just focusing on ways I can improve rather than win.

Now on my other account that most people know me by, I always play the BO I'm most comfortable with (pretty much after playing it over and over again on my other account) and focus 100% on winning, leading to the mild panic symptoms you mention and a lot of adrenaline-filled games that tend to be more stressful than enjoyable (but still enjoyable ).

Actually now that I think about it, the feeling I get is remarkably similar to when I play on an account on KGS that everyone knows me by versus an anon acount. IIRC, I even wrote an extensive blog post about this a while ago...hmm.

I don't think getting a second account is exactly the ideal solution, but anything you can do to have yourself detached from the numbers is probably a good thing.

[CSamurai]

Oddly, guessing my opponent's builds is one of my few strengths. Comes from looking at a lot of replays (I always want to know how they won). I don't even have to see all the buildings go down to know what's what. If I see airforce, of course they've got 2-3 stargates. Zerg, it's easy to spot a ling/muta build because the early portion of the game is very ling heavy.

Early scouting can tell me whether someone has vespene early (Heavy marauder/tech builds for the terrans, heavy airforce/4 gate for the protoss, roaches/mutas for the zerg) or whether they're focusing on 2/3 baracks/gates/lings. I tend to be able to guess when and where they've put down a hidden expo because of the number of units I'm seeing, and am able to guess how much they've built of what based on what units I'm seeing.

I do a lot of scouting by attacking in the midgame, throwing small units up ramps that I know I can't get through, just to see what I hit. If I hit a massive ball of MMM, I know what I'm up against. If I hit a small MMM and a seige tank and a couple vikings, I know how to deal with that too.

Zerg are actually easiest to attackscout, since they don't have easy wall offs the way most races do. Also, I like playing against zerg since the possibility of proxy gate/barracks/other cheap inside my base rushes is pretty limited.

The important thing to remember is that units come from somewhere. If you see an airforce swooping in from an unusual direction to harass, it's a good idea to check that direction out for hidden air production/bases, even if it's just tossing out a zealot to go die if he finds something.

Size/quantity of units tells me as much as attacking does though. If I see a trully massive number of marines marauders, I know someone went 4-6 rack 2 base on me. Watch replays. Attack with stuff just to see what kills it. Scout early and often, and always assume that terrans are about to build baracks beside your base and fly them up the cliff.

What's important to me is figuring out what to do with this knowledge once I have it. I have real trouble adapting what I know is a bad build to a situation, or even capitalizing on a good build. For instance, I had an observer parked in enemy base. I saw them dropping void rays (they had 2) so I dropped 3 stargates, and started throwing out phoenixes and stalkers, with my 2nd expansion. They kept rallying void rays, and when I attacked, I had about as many phoenixes as they had Vrs, and as many stalkers.

I guess they had dropped all other production and gone only void ray because when I decided to attack, they had more void rays than my entire phoenix/stalker army, hidden up a ways. Had 4 stargates. I didn't think anyone would commit that much to a single unit. How wrong I was. And, for all that phoenixes are the 'counter' to void rays, I can't ever seem to make a phoenix airforce that will win against VRs.

Phoenixes are great against terran armies, but seem to fail like punks against void rays, I don't understand it.

How do you guys beat massive numbers of VRs?

[MountainGo]

How do you guys beat massive numbers of VRs?

Above all, try to put constant pressure on your opponent by attacking them. Then they won't be able to sit back and mass specialized units like VRs.

Respond proportionately. (Day[9]'s phrase is "Don't overreact.") If they're only building a few VRs, don't tech to anything just to counter, except maybe researching Blink so long as you already have a Twilight Council. If they're getting ridiculous numbers of them, feel free to build a Stargate and get Phoenixes, but don't expect to fly over and kill them with just the Phoenixes. Think of them more as a boost to your existing anti-air as well as a tool to chase down fleeing VRs.

EDIT: Also, I should really mention that the root of your troubles is probably your macro. As you get faster at making units and taking expansions, you'll see the nature of your games transform. So just make sure you're always putting the majority of your efforts into improving that.

[Solomon]

How do you guys beat massive numbers of VRs?

All I'll say is that SC2 isn't so simple as "if X then Y". I've won games where my army of mass stalkers beat out an army of mass marauders and lost games where I had double Stargates pumping phoenixes lose to mutas. This is an exaggerated example, but if you have an opponent who's turtling off one base to build a 200/200 army of VRs while you're on 5 bases with map control, I'm pretty sure you could macro up anything that can shoot air and still win. It's a bit narrow-minded to try and think "well what 200/200 army should I ensure on getting to win against his 200/200 army of VRs?".

[flOvermind]

How do you guys beat massive numbers of VRs?

All I'll say is that SC2 isn't so simple as "if X then Y". I've won games where my army of mass stalkers beat out an army of mass marauders and lost games where I had double Stargates pumping phoenixes lose to mutas. This is an exaggerated example, but if you have an opponent who's turtling off one base to build a 200/200 army of VRs while you're on 5 bases with map control, I'm pretty sure you could macro up anything that can shoot air and still win. It's a bit narrow-minded to try and think "well what 200/200 army should I ensure on getting to win against his 200/200 army of VRs?".

That reminds me of one of my first SC2 multiplayer games, on one of these tutorial maps. My initial attack found that the opponent turtled with siege tanks behind the rocks. My first thought was "great, now I can get another expansion for free"

I then went on to take map control and tech to VRs. Later in the game, I got careless while harassing with my VRs and lost my ground force. Then I panicked and lost my VRs, too. At this point, I had no units left, so I lost one base right away, and all workers of a second base.
I still won the game rather easily, because by that time I had about 10k resources, 7 or 8 stargates and a few warpgates just waiting for my army to drop below 200/200

The point here is: Once you have a big enough economic advantage, it doesn't really matter what you do, you'll still win. Losing the entire army just delays the win a bit. And a turtling opponent is the best opportunity to gain an economic advantage.

It's a bit like in go. When your opponent is only playing ultra-solid territory building moves, the way to win certainly isn't to try and invade that territory. Just build your moyo, sooner or later the opponent has to come out and invade, otherwise you'll just win by territory (or, back to SC, when the resources run out ).

[MountainGo]

It's interesting you bring up tank defenses. I struggle with that, because it seems like putting resources into Siege Tanks doesn't prohibit a Terran from choosing at any point to go on the offensive with them. For being so superb on defense, I don't understand what the downside to Tanks is.

[Tooveli]

Any Zerg players here willing to help me out?

I've recently been playing a lot of Zerg and been getting pummelled on the ladder. I was planning on transitioning to Random, but every time the game is loading I'm thinking 'please not Terran!' Terran just feels so .. ugly. I don't know why.

Anyway, with Protoss I feel like I understand what is going on in the game. I try to constantly make probes and constantly build units out of all my unit producing structures and I attack/expand at sensible timings. But I realised (after playing Zerg) that I don't really understand the game at all. It just happens that simply building units and attacking with good macro is very powerful with protoss. It gives you the impression that you're on top of things becuase everything feels so simple.

With Zerg however, I feel totally lost. Here are my win ratios since I started using Sc2gears.

Code: Select all

Race      Games    W-L    Ratio    APM 
----------------------------------------
Protoss      54    36-18    66%     87
Zerg         71    26-45    36%    115

I think it's interesting that my APM is higher as Zerg. (This is not just because I've been playing Zerg more recently as my Protoss APM is still only around 87 in recent games). I don't know if this is down to heightened panic or tension or because I'm simply doing more stuff. I wonder if I managed to reduce it for Zerg I'd actually be playing better.

A lot of the losses are just due to inexperience with Zerg. The kind of mistakes that you (hopefully) don't make twice. eg. getting too aggressive with mutalisks or thinking your infestors will come save the day from the banshees but not knowing the can't cast fungal growth immediately (unless the extra strating energy has been researched).

With Protoss I felt like I could hold everything in my head at once. With Zerg I get overwhelmed. Having the scouting and awareness to know when to build drones and when to build army is a lot more demanding than constantly building probes.

Hmmm, it's pretty obvious after writing this that I just need to practice more and work on my multitasking but it'd be a shame not to post this anyway.

EDIT: Is there a better way to do a table?

[Tooveli]

PS. GSL finals tomorrow!!!1!!overlyexcitedpunctuation!!11!

Spoiler for those 0% of you that both care who's in the final and don't already know.

It's such a shame it's not (fake)boxer vs boxer. I was really looking forward to that. It'll be interesting to see if (fake)boxer sticks to his mass marine style. NesTea looked so solid against boxer I'm hoping we get some entertaining matches.