|
On June 19 2015 01:51 puppykiller wrote: What about the carrier vs a really smart Reaver that is shooting something standing right next to it? well the reaver shoot every 5 seconds or so, a carrier with all its interceptor shoot like 10 times a second ^^
|
On June 19 2015 01:45 oGoZenob wrote: spells don't count for the dps, we're talking about the base damage of units, and I'm pretty sure it's the carrier (I'm 100% sure it is in sc2 aswell) actually it seems the thor and battlecruiser have higher dps than carrier in sc2
|
On June 18 2015 17:28 puppykiller wrote: I guess in BW if a spell can target something other than the caster (unlike stim, cloak or burrow), then it can't target itself. I don't think it is a balance issue I think it is just how the game is designed. Makes sense. Thanks.
|
On June 18 2015 23:41 Assault_1 wrote: What fully upgraded unit has the highest dps? I'm not sure the DPS of suicidal units but I guess technically it's zero so they don't count. Technically, or practically? Technically, you can arrive at a number that's the highest, but practically, seems like you have to take into account things like splash and how many times armor counts against the attack. IIRC Carriers have armor count against their attack eight times, one for each 'ceptor.
Not taking that into account, in terms of fully-upgraded dps, seems like it's very close between archons, ultralisks, and carriers, all of which are just shy of 2 dmg per cooldown, not taking armor into account.
But, reavers with their +25 dmg upgrade seem to edge all of those out, doing about 2.1 dmg per cooldown. Winnah.
A dark horse that's surprisingly in the mix, though? Stimmed firebats. They apparently have a 'third attack' that hits sometimes, so they do 24 dmg base, +9 with full upgrades, for 33 dmg per 11 cooldown, or 3 per cooldown. AND they do splash.
However, said third attack only seems to land vs large units at certain angles. And large units only take 25% from concussive, so that'd seem to disqualify that.
Even DQ'ing the third attack, stimmed 'bats still do 22 dmg fully-upgraded per 11 cooldown, or 2 per cooldown, which puts them in second place. Not bad for a unit that costs all of 50/25.
|
On June 19 2015 04:35 Assault_1 wrote:Show nested quote +On June 19 2015 01:45 oGoZenob wrote: spells don't count for the dps, we're talking about the base damage of units, and I'm pretty sure it's the carrier (I'm 100% sure it is in sc2 aswell) actually it seems the thor and battlecruiser have higher dps than carrier in sc2 The liquipedia page on that is really outdated. But it would make sense for the thor, on ground at least
|
Japan11285 Posts
On a related note, because carrier 'attack' is split between 8 interceptors at max, armor has a larger effect on the actual damage than on other attacks.
|
On June 15 2015 23:51 Soulforged wrote: What is an 'even' unit ratio in midgame PvT? Speaking about 3-4 fact pushes.
As in, protoss has no zealot speed, mostly goon army, but can have a shuttle with zealots; terran has no +1. Talking about defending terran pushes, not breaking a defense line with scvs pulled. Assuming both sides micro well.
If no reply comes to mind, maybe you can link a VOD with such an engagement that traded fairly evenly, without big mistakes on either side?
Check out this game http://www.teamliquid.net/tlpd/korean/games/127412_Reality_vs_Shuttle/vod
For a simple answer, 22 goons + shuttle with zealots, then switch to getting rounds of speed zeals. Obviously its more complicated than that.
Terran is the initiator in that if he commits the attack, Protoss must reactively also commit. Equally Terran is the initiator in that if he spreads too thinly, Protoss must reactively seize the opportunity to do damage. This could be from around 8 goons onwards. Depending on how many goons are lost will affect what units you decide to make in each round of production. You want to have atleast 16 goons bare minimum though at any time for a core army. Think of it more in terms of rounds of production though rather than unit composition. Im making goons, then more goons, I identify the attack with observer, I make zealot and shuttle, then more goons, we fight, more goons, we dance, more goons if needed, else begin rounds of speed zealots. Thats a basic kind of flow.
It ties to how many gates you have too. For the most generic plan, I would say 3 gates then 3rd nexus (vs fac opener), (identify the 2 base factory attack with observer) then 5 gates, then up to 8 gates (3 base). Massing is the priority vs 2 base factory. Goons to 20/22 then waves of speed zeals and keep attacking the army. Clearing out the army is all that matters. If you lose all your goons in the big fight but it goes well, you can possibly do 1 round of speed zeals but then replace the goons. Ideally retreat to save 10 goons minimum though. Sometimes P will be going 2 base tech on 4-6 gates. I dont recommend it anyway and I dont recommend it as a response to a 2 base Terran factory attack. Protoss can stay on 2 base but still mass is better than tech. Speed zeal is the only tech you need.
Watch that game I linked, its perfect food for thought. Its an example of the most powerful a 2 base terran factory attack can be. The big fight at 13 minute goes quite well. It could have gone better but Shuttle is still in the game at 13.30. At 14 minute he could have attacked the contain again. But instead he back stabs and this doesnt work. If Shuttle had attacked the contain again rather than backstabbed, he would have cleared out most of the contain. He would probably have still lost his 3rd nexus though. However losing the 3rd nexus in that situation would have not been fatal. He could have cleared out the contain properly with the next round of units and probably got a decent map control from that point on, allowing to retake the 3rd not too long after. Would have been hard for sure, but thats the nature of a powerful factory attack. The crunch time comes earlier. Sniper Ridge is incredibly hard for this type of play aswell. The little things add up and Shuttle would have probably won on a different map (sounds a bit ridiculous but whatever).
Simple questions simple answers, I hate this thread lol. Let me know if you have any questions about the game I linked.
|
Thanks for the link, CardinalAllin. That game was great in a sense that Terran got off from an economic opening, delayed his e-bay and academy after 4th factory, so the attack was really strong(but not going into the crazy scv-cut and completely skipping ebay territory, though).
PvT, judging army strength doesn't come naturally to me, as I manage to [win by landslide] / [lose by landslide] / [trade evenly] in seemingly same situations. So it is hard for me to understand if it was a micro mistake on my side and something I need to work on, or if I actually really needed more units, in the first place. It isn't helped by the fact that my micro is poor, and often so is my opponent's.
The initial forces in the game you linked showcase 9 goons vs 5 tanks + 2 vultures + 4 rines on initial moveout; and in this situation, Terran felt safe enough to move out at that point, but had to at least partially siege his force to progress onwards, so he could be delayed - and by the time extra gates kicked in, he still wasn't sieging up anything meaningful. This ratio is something I'll consider to be 'enough to have to win more time'.
That was pretty much what I was looking for. That amount of goons was enough to slow down terran to get 4 zealots in a shuttle and trade; and then, 4 goons and for the 3rd wave of reinforcements the extra gates would already kick in, with terran sieging up the front door, but not quite there yet.
Now, in this game, if P had less initial goons, Terran would be able to move forward much faster(without having to siege at all), and most likely would siege up the natural exit point at least a wave of production faster; making it much harder to break out, and soon breaking the pylon wall at the 3rd.
Though the highground areas in that game that T had to cross made things a bit different, but well, that's a map specific thing. P did not have access to center to cut off reinforcements, and so on.
So I'll try to aim for 2:1 goon to tank count for initial engagement(speaking of timing attack moveouts of course, with vultures reinforcing from just finished factories), and see how that goes. Maybe a bit less that 2:1, if there are no marines and less of a buffer in general. With the goal of 'having enough units to get a 3rd as fast as possible against most aggressive terran follow-up to his opening, prior to getting scouting information, and adjusting accordingly once I do have it'.
P.S.> Yeah, this discussion isn't exactly for this thread; but I don't think it warrants its own thread, so...
|
the goal of 'having enough units to get a 3rd as fast as possible against most the aggressive terran follow-ups to his opening, prior to getting scouting information, and adjusting accordingly once I do have it'.
I love this sentence.
Ah it seems you mainly wondering when to get the round of zealots for shuttle bomb... Thankfully that has been worked out. Instead of the ratio mindset, I much prefer to think in terms of rounds of production. 3 goon expo identify its a factory expand 1 more goon and 2 more gates at the same time Then a round of 3 goon Another round of 3 goon Round of zealots and shuttle
So thats 10 goons, 3 zealots and a shuttle. Horang2 and Jangbi used this opener to great effect. Its very rigid with the only requirement being 'was it a 1 fac expo?'. 10 goons really is the magic number in this case.
You mentioned
Maybe a bit less that 2:1, if there are no marines and less of a buffer in general. This level of detail doesnt affect decision making in terms of rounds of production. It only affects micro when to pull back. The only thing you need to identify; is it rax expo or fac expo. It doesnt matter what his composition is. The build above deals with everything that can come after a fac expo. (Note that technically this is again oversimplifying)
Another scary 2 base fac attack seen here: Bisu vs Bogus http://www.teamliquid.net/tlpd/korean/games/96968_Bisu_vs_INnoVation/vod
Rax expo, 3rd nexus off 1 gate, 15 goon i think before round of zealots. I probably shouldnt keep posting games of Protoss losing lol.
|
^ *wall of text incoming*
Nah, I was running into an issue with certain scenarios. Thinking in production waves doesn't help much with that. I wasn't thinking much about when to get the shuttle/zealots, actually, that certainly can be reactionary while slowing down the terran moveout. If Terran never moves out, I'd probably want it approaching max. Instead, what I wanted to know, was how many goons I need as compared to terran army, to make him unable to proceed 100% unsieged. Because if he can do that, chances are my extra production will not kick in before he's either shelling my 3rd, or containing my nat choke.
Here's an example of such issue.
An example would be, a reaverless 2base obs->arb build, against a non-rax CC/non-14CC terran opening. Around 07:40, I'd have 2nd obs(that's off a gate/nex/gate/robo, not gate/nex/robo/gate), with 1st observer reaching the Terran for information, the 2nd observer reaching my 3rd nexus location to clean up mines. At the same time, around 08:00 I'd be approaching 50 probe count, and my 2 bases - 2 gases, 9 and 7 mineral patches, would be fully saturated. I'd need to get a 3rd around that time.
But, if I keep up probe production non-stop, don't slow down the tech(and getting forge/+1 starting soon is important too), and keep up unit production off 3 gateways, I just won't have money for the nexus before around ~09:10. And during that minute I'd have some 14 probes that aren't improving my income at all.
What do I do, then? I need to cut something. Delaying tech is the worst option overall, as with that style getting arbiter faster is more important than anything else(except not dying this instant, but yeah it is more important than economy). Cutting 3rd gateway earlier, for constant 2 gate goon production, is not an option because it will get me killed against things terran can do before I scout him.
I have an option to cut worker production. If I cut 6-8 probes to get a 3rd nexus a minute faster - and that's enough time to make 6 probes off of it, and earlier transfer for more efficiency more than makes up for the probe deficit - that's an option.
Another option, would be to cut something like 66% of a single wave of goons, which allows to both get the nexus early, and keep up the probe production. Also can get you killed.
That's a decision I can make after my observer reaches the terran, so I just need to identify when it is and when it isn't an option. Faster probes in this case mean that max-oriented probe goal count will reach faster, and cutting them(after having enough) faster allows for money for 4th/5th faster, and so on - it can speed up the eventual gateway explosion by an entire minute, so that after hitting the max, and after that ~12:50 recall, I'd already have 16 gates ready to remax at will, ...blah blah blah.
Other opening scenarios run into similiar issues, and I wanted to figure out adaptations. But that kind of depends on what I can get away with, and that depends on experience in fights where both sides control properly while having good execution before that point, and that's a thing I don't have.
Watching a number of protoss fpvods, they sometimes cut probes for just a bit to get that 3rd nexus, sometimes they cut units for just a bit, and sometimes they cut neither and delay the nexus - and it is a hard to figure out what exactly they observe to makes their decision.
It is easy to note when they cut both nexus and the probes to add gates to defend an allin, but more subtle things, where terran either takes a delayed 3rd while moving out, or actually took a 1fac 3cc 4fac ebay greed, and has both an army and a 3rd; or took a 3rd off 2fac ; or took a 3rd, got 2nd fac/ebay/acad/turrets, and only then added facts; other variations - those are harder to know how to react to. Same thing with siege CC openings, how they have surrender map control but have more tanks for initial timing attacks, as they never stopped to make vultures off their initial addon factory, and so on. Their initial move out would be stronger off the same factory count/timings, for example.
I'd love to be able to just count the tanks, factories, and decide what I can get away with, from there.
|
Hmm I found it really difficult to answer your post. I wrote a long reply but I wasnt happy with it. I just didnt think it would have been helpful in the end.
Perhaps if you watch a game in the future (that is on youtube not twitch, lags for me normally)(not fpvod too) that has some things where you wanna know more post it. Or a replay of your own game or pro game.
I think analysing specific games would be most helpful. Heres another game by the way. Stats vs TY on Sniper Ridge http://www.teamliquid.net/tlpd/korean/games/98903_Stats_vs_TY/vod 3 goons expo vs siege expo with no marines. Stats takes 3rd nexus on 2 gates (5 goon I think). TY goes into a 4 fac and later 5 fac attack. Stats observer sees and he adds gates to 6, and citadel. You can see him using the 9 initial goons, similar to the Shuttle vs Reality game (even though that is rax expo). I think Stats gets 3 zealots for his first bomb run.
|
That was another awesome game, thanks! Loved the 3rd nex off 2gates, and also loved the 3rd/4th fact positionings. Even though the obs did check for them. The reaction was heavily pronounced in this game; pretty much the second that observer saw the extra factories, all the goons abandoned they dropship/vulture runby blocking positions and went to the terran natural.
I think it was still map specific, though. That sniper ridge highground allows to delay pushes easier. I kind of feel like TY mismicroed, though. Not only the push was delayed, but also quite a few tanks were sniped. It was 9 goons against 9 tanks, he didn't really have to siege up the front tanks and let goons trade evenly by sniping them....that, and it happened when he had 5 tanks in front and 4 somewhat stuck in the main. Hell, he traded 3 tanks for 2 goons and some shields, with those numbers.
I think if not for that, he could just roll Stats there, to be honest. Well, maybe the 'sniper ridges' would make the difference anyway, though But with those tank losses, it resulted in roll by Stats, instead. I don't think it was very holdable if both played perfectly, though. If it was, that'd probably be because of the map.
Pity I can't understand a single word commentators are saying
|
there is no number of goons that is required to stop a unsieged advance. it doesnt work that way obviously you need a decent number of goons, but what im trying to say is that number isnt defined. theres no rule terrans dont walk up to you unsieged when army sizes are relatively equal for a reason; they need mines. unless the opponents vultures somehow planted mines way ahead of where his army is currently and you didnt manage to clear them, terran cant arrogantly a click your army protoss style. all the dragoons have to do is hit and move vultures while the terran advances and then when you see a shortage of vultures you a click your zealots into the tanks. if terran attempts to advances with his tanks first and vultures behind so that the vultures dont get shot down by dragoons, you just engage with your whole army and watch terran die. unsieged armies dont work without mines, which is why a typical terran push will always have protoss faking an engage to force siege and then backing to buy time for zealots or arbiter as for all the other stuff, you seem to already have the answer in the wall of text. react based on what your observer sees i personally dont keep probe production up non stop unless i see terran getting a fast 3rd. non stop probe production is useful if you know you are gonna get a quick 4th and 5th as well, but otherwise you usually need to cut probes for a while to get your arbiter out or add a couple more gateways
|
@Soulforged You are correct that TY micro as he moved out was very poor and was critical mistake. Infact 4 tanks died for only 2 goons. Then a zealot bomb mine drag kills 2 more tanks on the centre ridge. However, you say:
I think if not for that, he could just roll Stats there, to be honest.
I don't think it was very holdable if both played perfectly,
This is not the case. Stats was in a great position to defend the attack. If it had been on any other standard map, he would have still been in a great position to defend the attack aswell. Taking the 3rd nexus when he did is safe and correct vs siege expo as there is no pressure coming at you, and you cant apply pressure equally.
Next he is planning to go up to 4 gates minimum, and maybe more if he sees factories. He does see factories, so instead of going up to just 4 gates he goes straight up to 6, then citadel, then 9 gates. Its the correct response and it comes down to execution. You are right that this game ends up being a complete stomping, and that is a shame for study purposes. However, if we pretend that Stats had managed to snipe only a single tank and had to retreat all the way back to his own ridge, and then Terran had set up turrets etc.. this would have still been an acceptable position for Protoss to win. In other words, Protoss would have still been even, and not in full panick mode or anything. He would know that he has to commit to a contain break, and that this would be the critical fight. He would know that this critical fight would require full concentration and would be on a knife edge (as we see in the other two example games) but that is the nature of starcraft. At the end of the day both players are trying to kill the other, 1 has to lose.
If he had scouted armoury and 3rd cc being made, he would have gone to 4 gates, then forge and citadel, then archives and star and 6 gates.
The decision making flowchart really is that simple. Its not how many tanks he has or if he made academy or ebay before 3rd fac or something. (I am worried that Im going to get crucified for saying this). Its all about 3rd cc and armoury vs factories, its all about the order in which these core buildings are made. Ebay and academy are support buildings.
Is this post actually helpful? Im not sure. If it is, I will do a similar quick analysis for the Bisu vs Bogus game if you want. But post how you interpret that game first.
|
@evilfastsh1t "if terran attempts to advances with his tanks first and vultures behind so that the vultures dont get shot down by dragoons, you just engage with your whole army and watch terran die"
That's my point exactly. This does not happen if you don't have enough goons, and terran microes.
@CardinalAllin
I have really strong doubts that 9 goons can even attempt to snipe a tank off 9 tanks. And those were all his goons...
Tanks have more range, and deal more damage, I don't think terran needs to siege ever on that moveout. First production wave is enough to keep the ratio, not change it, since extra gateways don't kick in yet(and vulture reinforcements kinda teleport). So, on a ridgeless map, if a protoss can't force a terran to siege, I figure the first siege would be on the protoss natural? And having 9 gates doesn't help when terran is already containing the choke point.
That's something I see a lot in Flash's games. When he timing attacks before lategame, it is very rare for him to siege before reaching mid map. He does those attacks when he has that unit advantage rolling out...and mostly he siege crawls only last screen and a half.
|
fine to put it generally, if you have 2 dragoons for every tank he has, its pretty safe to say he cant a click you. this is assuming you have zealot support for any vultures he has. you send in your zealots head on and even if they die youre trading zealots for tanks
|
I've been curious....how do you guys 1a2a3a(etc.)? Thumb on 'a' or pinky on 'a'? It's awkward with the thumb when it gets to 4/5a onwards and it's awkward to start 1a2a3a with the pinky. I'm contemplating just hotkeying important units on 1,2 and just start from 3a4a5a6a(etc.) instead.
Thoughts?
|
'A' on pinky is essential if you are right handed. My first year of Starcraft I used to use the thumb and sort of flip my hand upside down as I went up in control groups but shifting to pinky was significantly worth it.
I hotkey units that don't get 'a' moved as often on 1 and 2, (vessels, tanks in bio tvz, dropships, lurkers, defilers, scourge, sairs, shuttles, arbs etc.) and I also highly recommend this even if it is less common.
|
Thumb/pinky? Weird, I always use my ring finger. I don't even know how using your thumb would be possible.
|
On June 21 2015 12:26 Sero wrote: Thumb/pinky? Weird, I always use my ring finger. I don't even know how using your thumb would be possible. I use my thumb, mostly. It feels much more ergonomic for me than using the pinky. Place your ring finger on the 1, and then tuck your thumb under your hand so that it is roughly under the index joint and laying sideways on the a-key. I usually use the ring finger for 1,2, and 3, and middle for 4,5,6. Sometimes the index finger gets involved as well. It's whatever feels more comfortable at the time I guess.
|
|
|
|