| Forum Index > StarCraft 2 Tournaments |
|
TheDwf
France19747 Posts
On December 29 2013 23:02 opterown wrote: i think you should try getting ravens, PDD is good against stalkers You don't get it, you need to "somehow get 4 Thors" (quoted from a Dreamhack Winter thread in which someone did seriously suggest this). | ||
|
vthree
Hong Kong8039 Posts
On December 29 2013 23:15 Qikz wrote: just getting a single tank and defending it defends blink attacks early on. I don't play anymore but going mech the amount of times I won by just having a tank or two were pretty impressive. Of course if you're not going to ever build anything but MMM and there's a build that can exploit the current meta builds vunerability please be sure to blame balance and not stubborness of the players. Maybe MMM can't hold everything off early and that's a good thing, build something else. So, pro Terran players play bio, bio/tank, mech in TvT, yet they just refuse to play it in TvP, I wonder why? I am not sure why Protoss always use this stubborn argument. It is like when WMs were claimed to be OP'ed, yet all Zergs did was ling bling muta, how is that different from MMM? Why didn't zergs go roach hydra to counter WMs? That is the same reason Terrans don't go mech. And sorry, were your 1 tank builds holding off Rain on Yeonsu? You simply cannot cover all the possible blink in areas with 1 or even 2 tanks unless you sac your natural. And having 2 tank means you have no mauraders and you can't even push out when medivacs are out. So when Protoss sees you have 2 tanks turtleneck on one base. They just laugh, do a soft contain and tech behind it. The problem is the blink build is that it is not all in, if the Protoss doesn't like the setup when they get there with their MSC, they have multiple options. Aggressive builds that have a good chance of doing game ending damage should have have transitions if the opponent reacts correctly. | ||
|
gingerfluffmuff
Austria4570 Posts
On December 29 2013 23:41 TheDwf wrote: You don't get it, you need to "somehow get 4 Thors" (quoted from a Dreamhack Winter thread in which someone did seriously suggest this). Hm, good idea. Thank god we can reactor them out of the fact in case of an emergency. Just cancel CS, should be enough money for 4 of them. OT: gogogo SKT T1! | ||
|
vthree
Hong Kong8039 Posts
On December 29 2013 23:24 Hider wrote: So I think alot of people are missing what actually happend this game. Rain went for an economical blink timing. Normally, against 1-base blink you lift your expo into main, get 3-4 bunkers around your main and delay your starport tech (because you can't afford 3 rax'es and bunkers + quick factory/starport). While Dream went for relatively late factory/starport tech, he didn't lift his nat into main. Why? Cus he saw Rain getting an expo behind the blink-pressure, hence he figured that lifting his nat into main would put him behind. But he simply couldn't survive against this timing which hits before medivacs are out and defend locations (nat and main) at the same time. Against a normal 2-base blink pressure build, Dream could turtle on 2 bases with a couple of bunkers and 2 siege Tanks and be fine. However, this wasn't a normal 2base- blink pressure build, but an inbetweenber which caught Dream totally off guard. In my opinion, Dream should have played against this like it was a 1 base blink build and lifted nat into main. To optimize his situations, he could proceed to make these 3 adjustments; 1) Cut bio production a bit 2) Get slightly faster Medis out instead 3) Retake expo when 2 Medis are out He would be able to do this since Rain had less Stalkers than in a normal 1-base blink all in. But overall, we can't blame Dream for not adjusting proplery to this build since he probably doesn't have that much experience against it. Ok, so what you are saying is that Ps have an Eco expand build that can force the Terran to give up his natural even if he invests in multiple bunkers. This makes sense how? | ||
|
Hider
Denmark9433 Posts
On December 30 2013 00:27 vthree wrote: So, pro Terran players play bio, bio/tank, mech in TvT, yet they just refuse to play it in TvP, I wonder why? I am not sure why Protoss always use this stubborn argument. It is like when WMs were claimed to be OP'ed, yet all Zergs did was ling bling muta, how is that different from MMM? Why didn't zergs go roach hydra to counter WMs? That is the same reason Terrans don't go mech. And sorry, were your 1 tank builds holding off Rain on Yeonsu? You simply cannot cover all the possible blink in areas with 1 or even 2 tanks unless you sac your natural. And having 2 tank means you have no mauraders and you can't even push out when medivacs are out. So when Protoss sees you have 2 tanks turtleneck on one base. They just laugh, do a soft contain and tech behind it. The problem is the blink build is that it is not all in, if the Protoss doesn't like the setup when they get there with their MSC, they have multiple options. Aggressive builds that have a good chance of doing game ending damage should have have transitions if the opponent reacts correctly. Eh... the real reason he didn't get a tank out in this game is because Rain went for a quicker blink allin, and thus Dream didn't have factory tech ready when the timing hits. Ok, so what you are saying is that Ps have an Eco expand build that can force the Terran to give up his natural even if he invests in multiple bunkers. This makes sense how? So what I was discussing wasn't related to balance-whining, but simply optimal strategy. If you study the timings with my suggested adjustments, you will notice that Dream only had to be on one-base for a very short-whil before Medis came out. Relative to a "normal 1-base" blink-allin Dream could reclaim his natural faster in this build if he made the suggested adjustments. Taking into account that Dream began mining from his expo faster and has Mules + Rain probably cut probes a bit, Dream would (probably) be a bit ahead once he reclaimed his natural. | ||
|
vthree
Hong Kong8039 Posts
On December 30 2013 00:33 Hider wrote: Eh... the real reason he didn't get a tank out in this game is because Rain went for a quicker blink allin, and thus Dream didn't have factory tech ready when the timing hits. It was a delayed 1 base with late expand and Rain messed up his warp gate timing build... | ||
|
Hider
Denmark9433 Posts
On December 30 2013 00:36 vthree wrote: It was a delayed 1 base with late expand and Rain messed up his warp gate timing build... Ehh? Yes I know that, and that's my entire point. In the actual game he would almost have time to get a Siege Tank out, but if Rain hadn't forgot it, it would have been absolutely impossible. Thus, he simply had to lift his natural into his main and put the 2 bunkers that were placed in the natural into the main. | ||
|
cutler
Germany609 Posts
| ||
|
vthree
Hong Kong8039 Posts
On December 30 2013 00:33 Hider wrote: Eh... the real reason he didn't get a tank out in this game is because Rain went for a quicker blink allin, and thus Dream didn't have factory tech ready when the timing hits. So what I was discussing wasn't related to balance-whining, but simply optimal strategy. If you study the timings with my suggested adjustments, you will notice that Dream only had to be on one-base for a very short-whil before Medis came out. Relative to a "normal 1-base" blink-allin Dream could reclaim his natural faster in this build if he made the suggested adjustments. Taking into account that Dream began mining from his expo faster and has Mules + Rain probably cut probes a bit, Dream would (probably) be a bit ahead once he reclaimed his natural. I think you underestimate the time it would take to retake the nature. Dream would have to guess if Rain went back to saturating his natural or kept making stalkers, if Rain kept warping in stalkers, 2 medivacs won't hold it off since he no longer has bunker in the natural. | ||
|
Hider
Denmark9433 Posts
On December 30 2013 00:41 vthree wrote: I think you underestimate the time it would take to retake the nature. Dream would have to guess if Rain went back to saturating his natural or kept making stalkers, if Rain kept warping in stalkers, 2 medivacs won't hold it off since he no longer has bunker in the natural. No. In that type of situation you can always retake your expansion once you have 2 medi's out (since he knows Rain invested ressources into a Nexus). As a general rule; when bio units can afford to stim against blink stalkers --> your safe. This is ofccourse assuming Rain hasn't engaged him (which he can't/wouldn't if Dream lifted his expansion in the main) --> his medis would have full HP. In terms of balance, the problem in TvP isn't that builds like this can't be beat. The problem is that protoss has 100 different builds with 100 different variations. And against each build terran has so many ways to lose and always needs to rely on "lucky scouts". If terran reacts perfectly, then the downside as a protoss player is very limited. For instance I believe if Dream had reacted as I suggested, Rain would probably just be 45-55 behind. Given all the ways for Dream to lose that game, that's an absurd risk/reward. | ||
|
Hider
Denmark9433 Posts
Even if he saw a Robo coming down he still couldn't salvage bunkers as the blink stalkers still can do too much damage by poking at bio units without medivacs. Had he not wasted that scan, he could have afforded an additional bunker or faster medivacs. So bascially the loss in that game came down to Dream being very unfamilar with the blink stalker variation shown by Rain. | ||
|
vthree
Hong Kong8039 Posts
On December 30 2013 00:46 Hider wrote: No. In that type of situation you can always retake your expansion once you have 2 medi's out (since he knows Rain invested ressources into a Nexus). This is ofccourse assuming Rain hasn't engaged him (which he can't/wouldn't if Dream lifted his expansion in the main) --> his medis would have full HP. In terms of balance, the problem in TvP isn't that builds like this can't be beat. The problem is that protoss has 100 different builds with 100 different variations. And against each build terran has so many ways to lose and always needs to rely on "lucky scouts". If terran reacts perfectly, then the downside as a protoss player is very limited. For instance I believe if Dream had reacted as I suggested he would probably just be 45-55 behind. Given all the ways for Dream to lose that game, that's an absurd risk/reward. I would argue that if Rain did not start making probes with the second nexus, he could have enough stalker a to take on 2 medivacs with no bunkers and time warp on the ramp. Of course, if Dream had perfect information or get 'lucky', he can hold it off. But like you said, the risk/reward is just absurd given that it was a delay expand build. | ||
|
vthree
Hong Kong8039 Posts
On December 30 2013 00:49 Hider wrote: Another "mistake" by Dream last game was that he invested into a scan shortly before Rain attacked him, yet he didn't utilisie the information in anyway. He saw no addiitonal infastructures were being build by Rain, and still didn't lift natural into main. Even if he saw a Robo coming down he still couldn't salvage bunkers as the blink stalkers still can do too much damage by poking at bio units without medivacs. Had he not wasted that scan, he could have afforded an additional bunker or faster medivacs. So bascially the loss in that game came down to Dream being very unfamilar with the blink stalker variation shown by Rain. He scanned because Rain missed his timing and he was wondering why he wasn't. It wasn't a variation, Rain just messed up. I think it would have been too late to put up bunkers after he scanned. | ||
|
Hider
Denmark9433 Posts
I would argue that if Rain did not start making probes with the second nexus, he could have enough stalker a to take on 2 medivacs with no bunkers and time warp on the ramp. What are you basing this off? Against a 1 base blink-all in I can (from my experience) always straight up beats blink stalkers with bio + 4 Medis. In the case where the protoss player has 2-3 less stalkers, it seems logical to assume that you only need 2 medviacs to break out. This is ofc. assuming that your maurauder/marine ratio is pretty high (e.g. you have 2 tech labs and 1 barrack without any addons) He scanned because Rain missed his timing and he was wondering why he wasn't. It wasn't a variation, Rain just messed up. I think it would have been too late to put up bunkers after he scanned. Yeh I expected this response, because it is a typical reason for scanning "I am confused what is happening now?" But in reality, its not logical to think like that. You should only scan, if your planning to change the way you play depending on your information. At the time Dream scanned, he wasn't gonna change anything really, and thus it was just a poor decision of him to scan. All this simply comes down to him being in an unfamiliar situation, where he didn't have time to figure out the optimal response. So I feel that Rain deserves some more credit for making developing this type of build. Wouldn't surprise me if he perhaps had studied Dream and knew he likely was gonna react the wrong way. | ||
|
vthree
Hong Kong8039 Posts
On December 30 2013 00:57 Hider wrote: What are you basing this off? Against a 1 base blink-all in I can (from my experience) always straight up beats blink stalkers with bio + 4 Medis. In the case where the protoss player has 2-3 less stalkers, it seems logical to assume that you only need 2 medviacs to break out. This is ofc. assuming that your maurauder/marine ratio is pretty high (e.g. you have 2 tech labs and 1 barrack without any addons) Yeh I expected this response, because it is a typical reason for scanning "I am confused what is happening now?" But in reality, its not logical to think like that. You should only scan, if your planning to change the way you play depending on your information. At the time Dream scanned, he wasn't gonna change anything really, and thus it was just a poor decision of him to scan. All this simply comes down to him being in an unfamiliar situation, where he didn't have time to figure out the optimal response. So I feel that Rain deserves some more credit for making developing this type of build. Wouldn't surprise me if he perhaps had studied Dream and knew he likely was gonna react the wrong way. I am basing it on how blink stalkers work vs medivac bio. Because of the way blink works, the damage/hp is shared and stalkers that don't die is still giving full DPS. So against blink stalkers, it is about keeping up your DPS (ie. marines and marauders not dying). Yes, 4 medivacs will work, but with 2, you have 1/2 the healing and your marines are still dying too quick vs 15% DPS lost from the 2-3 stalkers. Once you lose your first 4-5 marines without forcing most of the stalkers low, it just snowballs. The scan was done so Dream could confirm if Rain was continuing his all in or going into other tech. When he didn't see additional tech, he kept making bio and started another bunker on the low ground. Maybe you do not agree with his response, but it was really a confusing action, he just did see the attack he expected at his timing and he had to see if he needed to keep building army or go more eco to match Rain. | ||
|
TheDwf
France19747 Posts
On December 30 2013 01:23 vthree wrote: I am basing it on how blink stalkers work vs medivac bio. Because of the way blink works, the damage/hp is shared and stalkers that don't die is still giving full DPS. So against blink stalkers, it is about keeping up your DPS (ie. marines and marauders not dying). Yes, 4 medivacs will work, but with 2, you have 1/2 the healing and your marines are still dying too quick vs 15% DPS lost from the 2-3 stalkers. Once you lose your first 4-5 marines without forcing most of the stalkers low, it just snowballs. The scan was done so Dream could confirm if Rain was continuing his all in or going into other tech. When he didn't see additional tech, he kept making bio and started another bunker on the low ground. Maybe you do not agree with his response, but it was really a confusing action, he just did see the attack he expected at his timing and he had to see if he needed to keep building army or go more eco to match Rain. Yeah, the scan was not illogical at all considering Rain didn't hit at the standard timing; funnily enough, Rain was somewhat rewarded for his blunder (forgetting Warpgate) here. What lost the game for Dream was not pulling SCVs to shield his bio in his main. | ||
|
Hider
Denmark9433 Posts
The scan was done so Dream could confirm if Rain was continuing his all in or going into other tech. When he didn't see additional tech, he kept making bio and started another bunker on the low ground. But scanning and making one additional bunker doesn't make sense (if that's your only adjustment). Why? Because its cheaper to not scan (and mule instead) and make one additional bunker anyway. What do you mean going into other tech? What exactly is he changing? He was going factory tech anyway at the same timing. All the scan did was that he had to cut bio production marginally + he could afford fewer bunkers after having wasted the scan. he kept making bio As a outlined above, you are going to do the actual opposite after having wasted a scan. In order to afford factory tech and bunkers you have to cut bio production a bit after having wasted that scan. So note that he is going to have less army stuff, not more after that scan. to keep building army or go more eco to match Rain Going more econ? The only real thing he could have done here was to add a 3rd, but there is no way he was going to do that before adding factories anyway. For instance 3 rax --> 3rd --> before factory is just a terrible build regardless of what the protoss is doing. So if your going to reply - Be specific. What exactly would he have changed depending on what he would see. (btw we know he would never ever see a Dark Shrine with that scan since Rain wouldn't place it in that obvious of a position). Yeah, the scan was not illogical at all considering Rain didn't hit at the standard timing; funnily enough, I already noted it in my previous posts, so you should be aware that I have taking this into account. It only helps to explain the reason why he scanned (he was a human being that was confused, and humans makes errors in unfamiliar situations). But that doesn't make it a logical decision. The point still stands - What is he actually gonna do differently after scanning and is that worth the opportunity cost? | ||
|
TheDwf
France19747 Posts
On December 30 2013 02:16 Hider wrote: This is, however, irrelevant (and I already noted it in my previous posts, so you should be aware that I have taking this into account. The point still stands - What is he actually gonna do differently after scanning and is that worth the opportunity cost? Had he spotted some tech transition (forge/archives or forge/robobay, or even dual forge + another tech building), he could have salvaged part of or all his Bunkers, resumed full SCV production instead of producing units with short hiccups regarding workers, favor Marine production over Marauders to add extra rax or a third CC, etc. Inversely, had he scouted extra gates (Protoss can go up to 7), he could have remained on the defensive. | ||
|
Hider
Denmark9433 Posts
On December 30 2013 02:28 TheDwf wrote: Had he spotted some tech transition (forge/archives or forge/robobay, or even dual forge + another tech building), he could have salvaged part of or all his Bunkers, resumed full SCV production instead of producing units with short hiccups regarding workers, favor Marine production over Marauders to add extra rax or a third CC, etc. Inversely, had he scouted extra gates (Protoss can go up to 7), he could have remained on the defensive. - He can't afford 4+ raxes at that point anyway. - If he scanned some type of tech, he would indeed be able to salvage a bunker. But only the one at his natural. If there were no bunkersm blink stalkers would then can poke at the bio units too efficiently. - How valuable is constant scv production (compared to semi-constant) relative to muling? In this situation he wasn't getting a quick 3rd anyway, and if you go 3-rax + starport tech you are always gonna be a bit oversaturated on 2 bases anyway which makes the value of constant scv production less significant. - How valuable is a slightly higher Marine/Maurauder ratio in this situation. Even if Rain isn't commiting, he still has a decent amount of Blink Stalkers. Compare this to the situation where he A) doesn't scan and B) Doesn't salvage the Bunker in the natural, C) Uses the additional ressources on extra bio units. - He was always going to be in the defense, regardless of the scan untill Medis are out (possible exception is extreme cases where he scouts 3 tech buildings or more, but that's quite unlikely). So I assume your talking about building additional bunkers. Yet, the problem is that at that point in time you don't really have excess ressources (after scanning) to afford mass bunkers and medivas with constant bio production. I think the trade-off your thinking about is something that comes 2 minutes later (or so) at the point where he has to consider whether to get a 3rd, 4+raxes (before 3rd) or get additional bunkers (against a 7-gate). But at the time he scanned, he wasn't gona change his infastructure setup in any significant way really (besides salvaging one bunker). | ||
|
Hider
Denmark9433 Posts
Nevertheless this shows the problem of relying on this type of scan. Like let's go through the options; * Rain adds like 3+ extra gates in main and Dream scouts it --> Dream gets additional bunkers. In this scenario Rain may chose to cancel them, and take a 3rd --> Rain gets a lot ahead. * Rain adds 3+ gates in the natural, Dream doesn't scout this since he scans natural --> Dream gets ebay and just dies to the 6-7 gate timing. * He scouts Robo tech and/or forge. He can choose to salvage one bunker, and produce scv's constantly. But he still can't afford a 3rd for the time being anyway and has less stuff out compared to the case where he doens't scout. * Rain does a delayed timing --> Perhaps Dream could have held it off better if he pulled scv's asap, but he would almost always end up behind. Regardless, your always better off by not scanning. The only situation where he is better off (relative to not scanning) is when Rain gets additional gateways in the Main and chooses to commit to the attack, rather than cancel them. | ||
| ||
StarCraft 2 StarCraft: Brood War Dota 2 Counter-Strike Super Smash Bros Other Games |
|
GSL
WardiTV Team League
The PondCast
WardiTV Team League
Replay Cast
Replay Cast
CranKy Ducklings
RSL Revival
WardiTV Team League
uThermal 2v2 Circuit
[ Show More ] |
|
|
EPT
