Make sure you read the OP before asking a question. Asking a question already addressed in the OP will result in moderation action. Also, please put some effort into your questions.
Thank you so much for this post. I'll be working on all of these things. : ) It also explains why I've been having such serious issues in TvP in general -- too much marine investment which has less impact than in my TvT/TvZ.
Gotta work on getting my mechanics up to par and fixing this stuff, thanks again.
On April 19 2015 21:29 00Zarathustra wrote: Guys I need some serious help, I thought I was improving a lot. I was playing with macrobuilds like polt 3CC +2/+2 and 14 CC based builds. But I recently played vs a master friend and he manhandled me so much.
I just don't know how he outplayed me so much. I won't even know how to start trainning to beat him. Also he isTHe one Ill play on the National Finals of my country's tourney. I though I just needed better micro but just watch my games and you willl see when the battle comes he just haves more banelings than I have marines!!!
So what should I start improving first? Any advice would be appreciatted.
I looked at each game. I'll note general things that I noticed.
G1 you say you are doing a macro build but you are doing much more of a semi-allin build--you execute it decently, but against better scouting you simply can't win with it. If this keeps working as well as it did for you in this game you are probably ok, but will have to change it eventually once you face good enough players you can't do the damage you need to do (kill the third and plenty of drones.) Your followup isn't very good though, you delay your third forever but get double ebay anyway. Followup should be 3rd cc then double ebays then starport and rax 4 and 5. So just switch around the order you build things and try not to lose your first army with medivacs.
There are some general errors with execution, like supply blocks and late buildings. In G2 you could have exeuted the build a lot better. What you did should be a straight counter to what he did, but since you don't macro anywhere near as well, you can't beat it. Here's a guide for that build. G4 your build was kinda random; you did a very strange build that put you very far behind. The standard followup to CC first into 3OC is gas 2x rax 2x ebay starport 2x rax, with the factory making reactors for a rax and the starport, or just the starport if you prefer. This would have given you the medivacs at the same time with 90 second faster upgrades and more army as well. In addition, you also miss a lot of production time because you etiher get supply blocked or forget to build units/workers.
Try practicing one or two 3OC builds vs zerg, if you like that, until they are perfect. On paper, if your friend keeps doing his life-style 3base ling-baneling, then you can beat that very easily with a well-executed 3OC build, providing you retain your hellions, or with a well-executed version of your G2 build. If you had executed the build up to 12 minutes perfectly in every one of these games you probably would have won. As it was, you nearly won in G2 despite losing tons of scvs, just because your build lined up so well against his. So just practice the builds on ladder or against practice partners until you can do them in your sleep. If you want more information on how exactly CC first into 3OC should be structured, respond and I can post replays or detail it more. Look to the guide I gave you for the gasless rax expand into 3OC build.
Yeah I know I didn't Execute my bomber 200/200 build well. I just read that guide like 3 days ago and have been practicing that build. BUt my question about that game is even when I was 200/200 He just had more banelings than I had marines. How can I win a fight like that? NEver played against ling bling only. I really didnt know what to do against that many units.
The game you say it is semi all in. It actually isnt. Its the build Polt used against Hydra. He never gets to kill a hatch in those games but he still ends in a very good spot. I just followed up poorly because I didnt multitask well.
The other games Yeah my reaper first build is fully improvised I really don't know how to play 3oc with reaper opening. Either I do bomber 13' 200/200 or I open CC first.
I would really appreciate to see some reps with shows some Macro oriented builds. And you now a bit of theory about how to face that many units. Or general TvZ mentality . I think Im missing something that everyone else just seems to understand.
Widow mines and try to tank the blings with marauders (I havn't checked your replays) And if your opponent played pure ling/bling you should've dropped more and try to pull him apart that way.
On April 16 2015 06:04 NotTrulyEvil wrote: What variation of the 2-3 base scv pull is the best in TvP? Seems like there is a lot of versions, with 3rd, without, with ghosts etc. I'd appreciate vods
3 rax third ~55 SCVs 5 rax 2 port no Ghost
Feels like how flash plays out that build it will die to 50 % of what protoss can do. Very little scouting etc
On April 16 2015 06:04 NotTrulyEvil wrote: What variation of the 2-3 base scv pull is the best in TvP? Seems like there is a lot of versions, with 3rd, without, with ghosts etc. I'd appreciate vods
Feels like how flash plays out that build it will die to 50 % of what protoss can do. Very little scouting etc
Maybe it feels like that, but also keep in mind that players like flash rely on non-obvious tells (what units you see near your front at certain times, lack of opponent scouting, etc.) to make educated guesses about what is coming. Plus, there are often built in reactions into the build in order to deflect various options which you don't necessarily see unless the attack actually comes.
Died to roach/queen nydus in front of my natural 2 times in a row today, trying to go mech with hellbat/banshee. Too many queens for banshees to handle, any tips how to hold this? I guess it needs to be scouted early and then rush out tanks or marodeurs with mass repair the wall.
I've gotten back into playing SC2 and as usual my TvZ is pretty terrible.
It seems now the most common way to play TvZ on the Zerg side is to get a Warren off 2 bases and make maybe 5-7 Roaches mostly defensively while taking a 3rd and going into Lair/Mutas as the follow up. This confused me heavily the first few games because I saw 2base Roach and assumed I was going to get Roach/Bane all-in'd but it never happened then mutas came flying in. This happened literally 4 games in a row against 4 different Zergs so it must be a pretty common build.
So my question is if I see this coming as the basic setup is there a good way of punishing this? I generally go 12/12 Reaper into Hellions and a Cloak-less Banshee off 1gas then drop my 3rd CC. But after that is where my issues come. Since I see the Roach Warren I generally don't float my 3rd CC off immediately and I get 2 more raxes instead of ebays. But this appears to be the wrong call.
So I guess my basic question is this: Is the above outlined build strategy by Zerg at all common or did I just run into a streak of luck getting off builds from the Zerg and if it is common what is the most appropriate response?
On April 22 2015 17:22 Meavis wrote: what kind of positioning? defensive? offensive? whilst pushing?
To be more exact:
When im ready to move out how do i get to enemy base without falling to an ambush w/o seiged tanks. Do i scan infront of me all the time? Or units getting in the way of eachother. Right now i have 1 cg for ground and 1 for air, but units tend to clump up, And choke points are horrible for me,
how do i identify a good spot to unseige my tank(s) Defensive and offensive
how do i push forward , and what units in what order, if any (tanks, mines, ground, air) how do you leapfrog to a new spot?
As a new Terran player, I've found that tracking certain key benchmarks in all my games (whether or not a game is a win or a loss) has been very helpful. At the moment, I track three stats: (i) spending quotient (SQ), (ii) % of time supply capped and (iii) # of SCVs at the 10 minute mark (the ideal number of SCVs ofc varies depending on build - e.g., CC first would have more SCVs than a 1-1-1).
One additional set of stats which I think would be quite helpful is: idle production time by building. For example, if I build five barracks, a factory and a starport over the course of the game, I'd like to see what % of time each one was idle. For example, let's say I play a 15 minute game and my starport finishes at the 8 minute mark - also, let's say that the starport was idle for 45 seconds (after it was built, ofc), then the % idle time would be = 45 seconds / (7 minutes * 60 seconds) = 11% idle time.
TL;DR: -Are there any programs/replay handlers that show idle production time by building (description above)? Based on some google searches, nothing of the sort exists today (atleast, publicly available) -Do replays contain the data for us to be able to extract this information? And, if so, could you please point me to any resources that would help me understand the API for analyzing replay data?
P.S. I understand that the idle production time metric is not perfect. For example, if we attempt a tech switch of some kind, we may well expect certain types of production buildings to remain idle.
P.P.S. Sorry if this isn't the right thread, but, given the various races' mechanisms of production, this type of analysis seems most relevant for Terran (which is my race of choice)
On April 23 2015 04:39 dchaudh wrote: As a new Terran player, I've found that tracking certain key benchmarks in all my games (whether or not a game is a win or a loss) has been very helpful. At the moment, I track three stats: (i) spending quotient (SQ), (ii) % of time supply capped and (iii) # of SCVs at the 10 minute mark (the ideal number of SCVs ofc varies depending on build - e.g., CC first would have more SCVs than a 1-1-1).
One additional set of stats which I think would be quite helpful is: idle production time by building. For example, if I build five barracks, a factory and a starport over the course of the game, I'd like to see what % of time each one was idle. For example, let's say I play a 15 minute game and my starport finishes at the 8 minute mark - also, let's say that the starport was idle for 45 seconds (after it was built, ofc), then the % idle time would be = 45 seconds / (7 minutes * 60 seconds) = 11% idle time.
TL;DR: -Are there any programs/replay handlers that show idle production time by building (description above)? Based on some google searches, nothing of the sort exists today (atleast, publicly available) -Do replays contain the data for us to be able to extract this information? And, if so, could you please point me to any resources that would help me understand the API for analyzing replay data?
P.S. I understand that the idle production time metric is not perfect. For example, if we attempt a tech switch of some kind, we may well expect certain types of production buildings to remain idle.
P.P.S. Sorry if this isn't the right thread, but, given the various races' mechanisms of production, this type of analysis seems most relevant for Terran (which is my race of choice)
I don't know of any programs that can record idle production time by building, it would actually be quite difficult and not all that useful IMO (for the reasons you mentioned + add-on swapping, ect). AFAIK it would be possible to access this information though. I assume you're already using scelight or sc2gears if you're tracking SQ? If not, check out their respective threads on TL.
edit: Watching the replay and seeing when your production is idle would be more useful, since you'd be able to see when it was on purpose and when it was a mistake.
On April 22 2015 17:22 Meavis wrote: what kind of positioning? defensive? offensive? whilst pushing?
To be more exact:
When im ready to move out how do i get to enemy base without falling to an ambush w/o seiged tanks. Do i scan infront of me all the time? Or units getting in the way of eachother. Right now i have 1 cg for ground and 1 for air, but units tend to clump up, And choke points are horrible for me,
how do i identify a good spot to unseige my tank(s) Defensive and offensive
how do i push forward , and what units in what order, if any (tanks, mines, ground, air) how do you leapfrog to a new spot?
any links/threads would be appreciated
ok, when you play mech, you're playing a mindgame of what when and where units are built/positioned, so you want to be aware of when your opponent is where and with what.
assuming you have this info available you should be able to tell when to push where.
to get this information, you can use hellions scouting around the map, hellions/hellbat drops, or banshees. scans are also an option, but you usually save those to see what tech zerg is getting/has, this can also tell you if his army is at home/out on the map, and how careful you need to be.
other options include forcing a zerg out of position, more often than not you can force a zerg to fully retreat by sending out 2 packs of about 4 hellions, when you notice the zerg taking the bait you can push out in safety.
as for positioning, you either have to know exactly where the zerg is, or have at least some tanks sieged at all times. simply a-moving forward while occasionaly sieging and unsieging some tanks will provide a small line of tanks sieged up, that you can retreat in to if you overstepped. a good way of setting up your tanks is to have them spread rather well, with about 3-5 range between tanks, this will make it so you can have quite some reach, but the zerg has to move deeper and deeper into your tanklines to kill them all, resulting in far more losses for the zerg compared to when all tanks are in a single spot, it also prevents the zerg from abusing the minimum range of tanks. the room between the tanks can also be used for hellions/hellbats/thors to maneuver through, allowing them to assist the tanks without taking to much friendly shelling.
I am looking for Reaper into 3CC into cloacked Banshee / hellion Build VODs / Replays of pros. (most preferably 2 reaper into 6 hellions into 2 cloacked banshees into bio transition with doulbe rax double ebay etc. just as many as you can find =) I want to study them). Thx!
2nd Question:
If you scout your opponent to go for gas first how can pros reliably differenciate between Banshee or Marine Hellion Drop? I almost never see pros scout or scan the opponent after the initial scv scout, thus not being able to see their 2nd gas but still making the correct call of going for Ebay or units. Is this just preperation? Even if the opponent builds a hellion he can still go for banshee so that does not tell you much. I am oppening 15 Gas in TvT.
On April 24 2015 00:15 forsakeNXE wrote: Hey guys,
I am looking for Reaper into 3CC into cloacked Banshee / hellion Build VODs / Replays of pros. (most preferably 2 reaper into 6 hellions into 2 cloacked banshees into bio transition with doulbe rax double ebay etc. just as many as you can find =) I want to study them). Thx!
2nd Question:
If you scout your opponent to go for gas first how can pros reliably differenciate between Banshee or Marine Hellion Drop? I almost never see pros scout or scan the opponent after the initial scv scout, thus not being able to see their 2nd gas but still making the correct call of going for Ebay or units. Is this just preperation? Even if the opponent builds a hellion he can still go for banshee so that does not tell you much. I am oppening 15 Gas in TvT.
Ty.
to the 2nd question, you can't tell if it's banshee or whatever if you go 15 gas and only see one hellion / nothing. Pros usually commit 3 turrets if unsure, alternatively you can go viking > viking and save scans.
If you commit turrets it's very important to get the 5:30 ebay and cut scv's completely to get up 3 turrets, start viking and have an as good hellion production as possible, favor hellions / viking over SCV's.
What are good heuristics/rules of thumb for figuring out how many production facilities AND research facilities can be supported on 2 bases? 3 bases? This would be for a marine-heavy bioball strategy so relevant production facilities are barracks (some reactored, some w/ techlab) and reactored starport(s); relevant upgrades are stim/shield/concussive/attack/armor.
Unfortunately, SC2Calc doesn't take research into account. Moreover, I'd prefer to rely on something that more experienced players use in-game as there are always factors that the simple income/spending model doesn't consider.
For example: these are (most likely) wrong but these would be the types of heuristics I was hoping someone might know of: e.g., 1 e-bay on 2 bases, 2 e-bays + armory on 3-bases. The ideal rules would ofc need to consider production facilities also but that might be overcomplicating it.
I know there are builds/videos that go up until the 10 minute mark but it would be helpful to have a few simple rules of thumb to follow in case *@?! hits the fan at some point in the early/early-mid game and I need to adjust on the fly.
On April 26 2015 13:23 dchaudh wrote: What are good heuristics/rules of thumb for figuring out how many production facilities AND research facilities can be supported on 2 bases? 3 bases? This would be for a marine-heavy bioball strategy so relevant production facilities are barracks (some reactored, some w/ techlab) and reactored starport(s); relevant upgrades are stim/shield/concussive/attack/armor.
Unfortunately, SC2Calc doesn't take research into account. Moreover, I'd prefer to rely on something that more experienced players use in-game as there are always factors that the simple income/spending model doesn't consider.
For example: these are (most likely) wrong but these would be the types of heuristics I was hoping someone might know of: e.g., 1 e-bay on 2 bases, 2 e-bays + armory on 3-bases. The ideal rules would ofc need to consider production facilities also but that might be overcomplicating it.
I know there are builds/videos that go up until the 10 minute mark but it would be helpful to have a few simple rules of thumb to follow in case *@?! hits the fan at some point in the early/early-mid game and I need to adjust on the fly.
It all depends on playstyle ect. Here are a few general rules to help you figure out what works for you.
1 Base Greed= None 1 Base Eco= 1 Rax 1 Base Aggressive= 1 Rax 1 Fact 1 Port 2 Base Greed= 1 Rax, 1 Rax 1 Fact 2 Base Eco= 3 Rax 1 Port 2 Base Aggressive = 5 Rax 1 Fact 1-2 Port 3 Base Early (not fully saturated) = 5 Rax 1-2 Fact 1-2 Port 3 Base Later (fully saturated, potentially taking 4th) = 8 Rax 1-2 Fact 1-2 Port
edit2: Generally speaking, 1 Tech Lab Rax then the rest Reactors in TvT, 2 Reactor Rax then the rest Tech Labs in TvP, 1 Tech lab Rax then 4 Reactor Rax then the rest Tech Lab Rax in TvZ. Factory is 1 Reactor 1 Tech Lab in TvZ, 1 Reactor in TvP (if you want to play widow mines, otherwise none), and all tech lab in TvT. Starport = Always Reactor (unless adding more than 1-2 for an air transition).
edit: You need an additional Starport vs Collosi earlier, usually on 2 base, and in super late game I"ll add on even more rax (sometimes as many as 11->13) depending on how often I'm trading my army (general rule of thumb, less trading = more production).
About Upgrades; Upgrades have a high up-front cost but a very low cost / time. For example, +1 attack costs 100/100 but has a research time of 160. This means that you're only spending 37.5/37.5 per minute on the upgrade. Comparatively, you spend 120/0 and 200/50 per minute on marines and mauraders one at a time respectively, and 133.33/133.33 per minute on medivacs. The cost of upgrades over time is almost negligible, only the upfront cost is important. Typically, starting upgrades on 2 bases is reccommended, and then researching them nonstop until they're complete.
Here's one example of a game going horribly wrong, i feel like i'm quite ahead and decide to scv pull on three bases.
Any help is appreciated, i know the theory of how to engage a Protoss but i feel most of the time i'm really cost ineffective and end up losing solely because of awful fights.