| Forum Index > SC2 General |
|
Faust852
Luxembourg4004 Posts
On March 16 2014 23:28 bo1b wrote: I think it would be pretty terrible, mech vs sh with narrow corridors like that. You mean like Heavy Rain where Jjakji did 2 of the most epic TvZ Mech of the year against Hyun and Revival ? | ||
|
Big J
Austria16289 Posts
On March 16 2014 23:06 Salient wrote: @Dwarf: you have no idea what you are talking about. Ohana, Akilon, and Belshir had less blink in points ( compared to Polar Night, Heavy Rain, etc.) Ohana was pretty decent for blink allins, Akilon too. The main point though is the following: Because of blink allins, a quarter of the maps historically used would currently be considered problematic. Another quarter or more would still give that allin a reasonable winrate. Independed wether you put Ohana in the first or second category, it is too much. Even if the game is balanceable by maps, I don't see why we have to take that option if it limits the game to narrow map usage where you cannot do any strategy or tactical move requiring a cliff, just because one build breaks such setups. If there is a way to nerf down blink rushes, I don't see why we wouldn't take that option instead, it's really not like Protoss needs to be able to perform those in any matchup to be competitive. If it breaks macro games, we are talking. But if we find ways that do not break macro games, then nerf the rush not the maps. Imo an easy way to fix it would be to give the blink research an extra requirment, say a forge. Tinkering with the research time has probably been done enough, and tinkering with the blink itself is too much of a macro nerf. | ||
|
bo1b
Australia12814 Posts
On March 16 2014 23:31 Faust852 wrote: You mean like Heavy Rain where Jjakji did 2 of the most epic TvZ Mech of the year against Hyun and Revival ? There have been amazing games on daedalus point tvp, amazing games zvp on yeonsu, I'd argue that the potential for amazing games doesn't remove the very easy fall back of a terrible raven/viking/tank vs sh/muta style ala reality vs soulkey. | ||
|
Faust852
Luxembourg4004 Posts
On March 16 2014 23:34 bo1b wrote: There have been amazing games on daedalus point tvp, amazing games zvp on yeonsu, I'd argue that the potential for amazing games doesn't remove the very easy fall back of a terrible raven/viking/tank vs sh/muta style ala reality vs soulkey. It won't change depend the map imo, hell it's even the contrary, because The terran might be much more aggressive on map with narrow path, because it's much easier to lock down some paths with fewer tanks, leaving a big chunk of the army free to move. +On Bel'shir there are a shit ton of Cliff, giving even more incentive into attacking for the terran. And you still can play bio on these map, Bel'shir is known to the heavy style on the 4th of the zerg for exemple. IIRC, the worst game mech vs SH was on one of the most open map ever, Star Station. | ||
|
Qwerty85
Croatia5536 Posts
On March 16 2014 23:27 Salient wrote: The issue with Deadalus is the size of the blink in space in the natural AND main. It has a huge cliff face. Bigger by far than the good old maps I mentioned. And those old maps could be revised or just used for inspiration (e.g. make Ohana a bit longer with a bit wider chokes). I'm trying to be constructive here. But I feel like many people in this thread just QQ, blame Blizzard, and aren't open to other ideas. It is not QQ. I just don't think that balancing protoss through maps is a good approach at all. You either remove all maps that give protoss any bigger advantages (and IEM stats, although small show 86% win rate for strongest version of 2 base blink) which would probably make for very similar games all the time (granted, seeing 2 base blink 6/9 TvPs is equally predictable), or make a map pool that equally favors other races to "balance things out". For example leaving Deadalus as it was before the ramp reduction to make it great for zerg against protoss. But we all saw how bad that looked like, even casters who usually remain professional, constantly kept bringing up how bad map is. I am not even sure how a map that gives terran a potential 86% win rate strategy against protoss would look like. Or how would such map affect TvZ. I just feel like it would make a mess out of things. That approach could maybe work long after Lotv release and when the game gets figured out for the most part, but right now I see it causing more problems than benefits. We all know that MSC is the main issue behind the strength of the blink. The unit was originally designed to be attached or tethered to the nexus but Blizzard changed the design to give protoss "cool new abilities" to use. It did help make PvP better than in WoL but it created a mess in other matchups, especially TvP. Not to mention that MSC being only a defensive unit would probably help PvP even more because even now we see way more 1 base play than in other 2 mirror matchups so it is still hard to safely expand, just not as hard as in WoL. And now you have people acting like protoss was unable to survive mid game TvP at all in WoL and that MSC is crucial for the matchup which isn't true at all. | ||
|
Djangoobie
13 Posts
I don't think the issue is Blink. It's the coolest ability protoss has and it's enjoying to watch when crazy micro happens. It's Timewarp in combination with the MSC that is the problem. Just some random ideas: Timewarp - Description: Make Timewarp a sentry skill, remove it from the Mothership Core. When doing blink all-ins the stalker count gets lowered because they have to build sentries. Stalkers also can't destroy critical high ground tech without getting punished. - Cast range: I don't want it to be a spell that can trap whole armies from a distant. So it should have a short casting range, but not too short. - Radius: As big as a Force Field. - Duration: 15s. Same as FF. - Cost: 40-45 energy. Force Fields - Description: Divide Force Fields in 2 spells. One Force Field which is destructible(~50 energy) and one expensive Force Field which can't be destroyed(~60 energy) like the current. Destructible Force Fields should have some reasonable amount of health to hold of zerglings and banelings, but should be weaker against ranged units, like Hydras or Roaches. This would create more fun to watch games. Players also have to use sentry energy more efficient. For example they can cast several Force Fields + one or two small Timewarps to intentionally trap them in choke points. They can also slow down locusts more efficient. Of course just an idea, I don't play SC2 so don't know if changes like this will break everything. | ||
|
bo1b
Australia12814 Posts
On March 16 2014 23:38 Faust852 wrote: It won't change depend the map imo, hell it's even the contrary, because The terran might be much more aggressive on map with narrow path, because it's much easier to lock down some paths with fewer tanks, leaving a big chunk of the army free to move. +On Bel'shir there are a shit ton of Cliff, giving even more incentive into attacking for the terran. And you still can play bio on these map, Bel'shir is known to the heavy style on the 4th of the zerg for exemple. IIRC, the worst game mech vs SH was on one of the most open map ever, Star Station. It's not so much the openness of the map as how easy it is to get a 4th and defend it all with a ring of turrets and planetaries, something heavy rain simply can't do as easily. But you're right, the map isn't to blame as much as the units in this situation. Having said that forcefield + collosus is horrendous on such a map, and I really don't want to see it in action again. | ||
|
Foxxan
Sweden3427 Posts
On March 17 2014 00:00 Djangoobie wrote: As a spectator I can't stand watching Blink "all-ins", it's really ruining the game. The patch didn't seem to have much effect in this regard. On some maps it changed, but one some it's nearly impossible to defend. After watching CJ Hero play lately, I can see he always goes Blink all-in when he doesn't have much faith in fighting a normal match. It's really boring to watch. I don't think the issue is Blink. It's the coolest ability protoss has and it's enjoying to watch when crazy micro happens. It's Timewarp in combination with the MSC that is the problem. Imo blink as it is now is terrible overall. Balance aside, its usually micro for the stalkers but not for the enemy. He fight when he can be costeffective or he run away | ||
|
mozoku
United States708 Posts
On March 16 2014 23:01 Salient wrote: Unfortunately, we would need hundreds of games for the statistics to be meaningful. That's not necessarily true. The sample size needed to draw conclusions is heavily dependent on the true blink winrate, what you consider an "imba" winrate, and your desired level of confidence. Here are some example numbers: (I can go through my calculations if someone really wants, but I'm not going to take the time to unless asked. Basically, I'm calculating the sample size required for a one-sided confidence interval for a proportion with a (one-sided) margin of error equal to the difference between the point estimate winrate and the specified imba level. This gives us a confidence level for the lower bound of the pvt blink winrate.) To be 95% confident blink pvt is has a winrate of at least 55% (specified "imba-level") using the 67% number as our sample winrate: < 45 blink pvt games --- To be 80% confident the true blink pvt winrate is at least 60% using the 67% sample winrate: < 45 games To be 90% confident the true blink pvt winrate is at least 60% using the 67% sample winrate: 76 games To be 95% confident the true blink pvt winrate is at least 60% using the 67% sample winrate: 122 games To be 99% confident the true blink pvt winrate is at least 60% using the 67% sample winrate: 245 games ---- To be 95% confident the true blink pvt winrate is at least 60% using a 70% sample winrate: 56 games --- To be 95% confident the true blink pvt winrate is at least 65% using a 75% sample winrate: < 60 games -- To be 95% confident the true blink pvt winrate is at least 70% using an 80% sample winrate: < 75 games The " < a " are those where I calculated a lower number, but we need "a" games in order for my distributional assumption to hold (i.e. to get a good estimate). You could probably calculate these required sample sizes with a more advanced approach. It is possible that the required sample sizes for some of these "< a" cases would be much smaller (i.e. 10-30) if proper analysis was applied, especially with the higher sample winrates. --- As a disclaimer, this assumes games are independent of each other and that each protoss has the same chance of winning against each terran with a 2-base blink build. Whether those assumptions approximately hold or not, for any sample, is debatable. Source: I'm a statistics PhD student (though I didn't use anything that isn't taught in a basic undergraduate statistics class) -------------------------------- So while we don't have enough games for the statistics to be meaningful from IEM Katowice alone, ~40-60 is an achievable number if someone wanted to look through the past couple months of top level games. It's certainly possible, unlike 100s of games. Of course, you'd have to be pretty selective with the games you pick in order to satisfy the assumptions though. | ||
|
DinoMight
United States3725 Posts
The one game that stands out to me where someone made more rax than usual is Illusion vs. Minigun from WCS AM, where Illusion went up to a silly number of barracks before Starport vs. a scouted Blink build and won handily making mass Marauders. A friend brought this up as we were watching IEM Katowice and I didn't know.. Blink is really strong, surely you need more rax to defend it? EDIT - please give me a real answer not a "oh, there's dinomight again... he is so biased" answer. Something like "it delays XYZ too much" is an acceptable answer. It's an honest question!  | ||
|
TheDwf
France19747 Posts
On March 17 2014 05:39 DinoMight wrote: Honest question: why do Terrans try to hold Blink allins with so few barracks? I've seen games where Protoss is up to 7 gates but the Terran is only going off 2-3 rax. The one game that stands out to me is Illusion vs. Minigun from WCS AM where Illusion went up to a silly number of barracks before Starport vs. a scouted Blink build and won handily making mass Marauders. A friend brought this up as we were watching IEM Katowice and I didn't know.. Blink is really strong, surely you need more rax to defend it? The amount of rax is not at all the problem. You can smash a 7g blink with 2 rax Medivacs; better than with 3 rax Medivacs (one lab + 2 naked) actually. Rax beyond 3 are useless. | ||
|
sitromit
7051 Posts
![[image loading]](http://i.imgur.com/9Dvkfyr.jpg) It's not only Terran that can't win tournaments in the reign of Protoss. Protoss is also dominating all 3 WCS regions now too, and Zerg is doing even worse than Terran in 2 out of 3 regions. | ||
|
Terence Chill
Germany112 Posts
| ||
|
TheDwf
France19747 Posts
| ||
|
sitromit
7051 Posts
On March 17 2014 06:15 TheDwf wrote: Zerg cannot really have outstanding performances in WCS Europe because there is no Korean Zerg there. Just saying. Zerg is not having an outstanding performance anywhere, unless you count 2 Zergs in ro8 in Korea vs 1 Terran as an outstanding performance. | ||
|
Loccstana
United States833 Posts
| ||
|
TheDwf
France19747 Posts
On March 17 2014 06:18 sitromit wrote: Zerg is not having an outstanding performance anywhere, unless you count 2 Zergs in ro8 in Korea vs 1 Terran as an outstanding performance. Couldn't be clearer than my previous post. Also, had Soulkey advanced to RO8 you would have probably said nothing about Code S. You're quoting "2 Zergs vs 1 Terran in RO8," but you forget there were 13 Zergs vs 3 Terrans in RO32. | ||
|
Loccstana
United States833 Posts
| ||
|
TheDwf
France19747 Posts
On March 17 2014 06:31 Loccstana wrote: Maybe to counter protoss blink builds, Terran should get some sort of command center cannon ability. For 100 energy orbitals can become a stationary siege tank for 60 seconds. We have enough of one MSC in the game! | ||
|
Loccstana
United States833 Posts
Whats wrong with that? I am just trying to use David Kim's thought process to balance the game. | ||
| ||
StarCraft 2 StarCraft: Brood War League of Legends Super Smash Bros Other Games |
|
WardiTV Weekly
IntoTheTV X SOOP
PiGosaur Cup
The PondCast
Kung Fu Cup
IntoTheTV X SOOP
FEL
FEL
Sparkling Tuna Cup
Replay Cast
|
|
|
EPT
