|
On April 27 2012 10:31 Andreas wrote:
David Kim just oozes of incompetence. All he talks about are statistics, and all he can show to is Blizzard's "justified winrates" which we don't know if are reliable stats or not. The stats we do know are real (map winrates) are skewed as hell. TvP is my main concern right now, and I'd love how David Kim thinks it's balanced that Terran has a slightly better chance to win in the first 13 minutes, and then a really low chance to win once it gets past that point. Insead, all I get to know is that their justified TvP winrate is 50% in Korea, so I guess it's fine?
Much more reliable than any one person's opinion.
|
On April 28 2012 04:00 KiLLJoy216 wrote:Show nested quote +On April 28 2012 03:53 SupLilSon wrote:On April 28 2012 03:49 SuperYo1000 wrote:On April 28 2012 03:36 SupLilSon wrote:On April 28 2012 03:23 crocodile wrote:On April 28 2012 02:55 SuperYo1000 wrote:On April 28 2012 02:54 architecture wrote:
Man with those imba scans, it's clear that T plays with full game knowledge! Therefore, T should never lose to stuff like 2 base bane/roach timings or 1 base trickery, or getting fooled that 2base8gate is not 3nexus.
No one that has any objective view on the game should be able to say that any race has good, or even perfect scouting. All races play extremely in the dark, and have the make decisions as such. Go watch any top T replay, most of the time, there's practically no extra information other than the implicit stuff that they read from unit counts/gasses etc.
you do realize that terran IS strongest early mid game BECAUSE they can react so well with there ease of scouting right? He said, without providing any evidence. trololol yea, itd be nice to have some substantial evidence ;\ ..... time frame win rates do provide evidence but Im certain that you will never be satisfied really? Have you never heard of confounding variables? Or the phrase, "correlation is not causation". Just because Terran has good early game win rates doesn't mean it's because of scouting. There are a number of reasons that could be. That's why I said "substantial". Rofl. You are right, its not only because of good scouting. It is also because marine marauder ball is stupidly good early/mid game. Just because correlation is not causation doesn't mean that this is one of those cases. In this instance the correlation can indicate the causation.
It is you who bears the burden of providing proof of this. You can't just say "This is a special case" and leave it at that. If "It is also because marine marauder ball is stupidly good early/mid game." is your "substantial" evidence/proof than I am guessing your' peers will reject it as it as a very lazy and narrow-sighted attempt.
|
4713 Posts
On April 28 2012 04:43 caradoc wrote:Show nested quote +On April 27 2012 10:31 Andreas wrote:
David Kim just oozes of incompetence. All he talks about are statistics, and all he can show to is Blizzard's "justified winrates" which we don't know if are reliable stats or not. The stats we do know are real (map winrates) are skewed as hell. TvP is my main concern right now, and I'd love how David Kim thinks it's balanced that Terran has a slightly better chance to win in the first 13 minutes, and then a really low chance to win once it gets past that point. Insead, all I get to know is that their justified TvP winrate is 50% in Korea, so I guess it's fine? Seriously? Did you even read the article? It contains composition analysis, feedback from pros, tournament results, internal metrics and analytics, regional breakdowns, meta analysis... How would you propose he analyze balance, by listening to one person? By figuring it out with a pen and paper on his own? When you're trying to address something as complex as this, you need a variety of information to paint a detailed and nuanced picture. He does this and you say it's incompetence because the results aren't either what you expected or what you want to see?
Are you serious?
The article has some random numbers thrown in there and just one stat.
And Blizzard's stats are worthless since they obviously include ladder games as well, otherwise they couldn't have arrived at their almost perfect 50% win rates for Korean and their "within the margin" 54% and 53% TvP and TvZ.
The TLPDs reports from Korean and International tournaments paint a far different picture.
Also Blizzard didn't even mention what pro's they talked too, on SoTG Thorzain was even asking "Who are these pro players?".
Meta analysis? Laughable, he didn't address the issues of late game TvP or TvZ, he only talked a bit about the low representation of Zergs in tournaments.
A good job from Blizzard I'd consider if they would only analyze GM and tournament stats, if they could come up with a program or method to quickly analyze win rate distribution per minute played of games, and if they formulated a list of 33, Terrans, 33 Protoss and 33 Zergs, 11 from NA, Kor and Europe, that they contact every month or every two months to talk to and ask opinions related to the state of game balance.
That would be much, much better way to gather data. From what I've seen of Blizzard's stats they are worthless because they take too much ladder into consideration, they aren't aware of a lot of meta-game issues, and for all we know they could be talking to some random NA GM and asking them for balance opinions.
|
[quoteMehukannu ][
Well there is winrate and game length statistics that shows some kind of evidence about the weaknesses of races during a certain time period, but it is still far from perfect./quote]
Wow this is very telling, All the graphs should be similar to that of pvz with an almost parallell linear representation.
|
Honestly you all are tripping too hard. The only real issue with the game balance right now is late game Protoss vs Terran. Everything else is so minor and much more dependent on personal skill than balance.
It makes little sense to implement drastic balance changes when the new expansion is on the way and close, they will only have to start over in the end. And as it's actually pretty close to 50% in all areas I kind of agree with his decision.
Who knows, maybe in a few months we will start to see some good shifts in the metagame and people will figure out how to deal with the lategame doom army.
|
On April 27 2012 11:27 Jaegeru wrote:Show nested quote +On April 27 2012 11:17 windsupernova wrote:On April 27 2012 11:12 SarcasmMonster wrote:
Every time David Kim does one of these, the community takes it as an opportunity to QQ. For the love of god, stop doing these Blizz. Haha this. Its especially baffling considering that now most of the QQ comes from T who.... are half and half with P in Code S, won Dreamhack, and last MLG? P lategame may be powerful but the MU is obviously winnable. That may be true but as demuslim said on stream it seems unfair that terran has to do so many more things than protoss being dodging storm's, getting emp's off, sniping templar, microing vikings and microing bio. Whereas protoss has to split templars, try hit a storm and focus vikings when they can. Don't misinterpret this as balance whine I'm just pointing out that you have to be so fast and have crazy apm to win. it's winnable if you play really well.
The difficulty and aggravation of Protoss having to keep Templar split and moving towards a kiting Terran army, while having to figure out if there are cloaked ghosts there ready to snipe, is why it is the way it is. You can't actually ever catch Terran if they decide to run away, stim or not, because HT are so slow. Look at the Parting storms... he gets them because he positions them for incoming Terran armies and spots them with observers. Without the observers he could never get those storms off. And late game without observers all HT potentially die to cloaked ghosts. Terrans just need to go out of their way to kill the robo and snipe observers. Game over.
|
On April 27 2012 11:29 usethis2 wrote:
Slightly improved ovie speed might be a good fix for Z's early scouting. I always thought it's a bit silly that T's buildings float faster than (unupgraded) overlords. But to be frank, the current trend is that T doesn't really much care about being scouted. Simply aim for the 150+ food with 3/3 marines and a handful of seige tanks and from there Z has an uphill battle for the rest of the game.
Since changing unit stats will likely break the game, I think the easier route for better balance is map changes. Maps like Cloud Kindom, Entomed Valley, Antiga Shipyard are all too turtle-friendly. (main/2nd/3rd are triangle-shaped, pocketed in the corner)
Personally as a Protoss I'm sick of overlords and overseers seeing every single thing I do unless I go Stargate tech. But really it's late game overseers that are out of line. I was happy for the overseer to 50 gas change way back when with the mindset that more scouting for all races lead to better games, but now extreme late game Zerg has 10+ overseers and they mass changeling block you every 20 seconds. All the late game changelings has become just stupidly annoying. You can't kill them without 15 clicks or without wasting a storm... I mean really.
|
On April 28 2012 05:08 Plansix wrote:Show nested quote +On April 28 2012 04:57 Thrombozyt wrote:On April 28 2012 04:39 Plansix wrote:On April 28 2012 04:29 aksfjh wrote:On April 28 2012 04:22 Plansix wrote:On April 28 2012 04:19 aksfjh wrote:On April 28 2012 03:21 Mystgun wrote:
I think you guys have to realize that changes are not made in isolation and you can't address something like "TvP late game" without creating issues in other match-ups at other points of the game. SC2 is Blizzard's flagship eSports franchise and you can bet that they have people working hard to address balance issues, but they are not going to make a bunch of changes based solely on qualitative community feedback. I also think ZvP is a bit broken right now thanks to Stephano, but I can't conceive a solution that wouldn't break other match ups ("nerfing roach" will create more problems than it will solve).
I agree that these statistics are a little bit superficial and it's really to show that they have some data to back up their decisions, but I don't really expect them to upload a spreadsheet of all the match-up data from battle.net either. It would be nice to show some trends over time to see how the metagame is evolving and what are the variances from week to week.
The map statistics are a bit surprising. I didn't think of Cloud Kingdom as a PvT favored map but I guess drops in the main are difficult which basically shuts down Terran in the late game. 70% is a huge discrepancy and his casual tone is a bit alarming. Really, I'm not mad that nothing is on the table for TvP, I'm just concerned that they didn't even make a nod towards it. There was no mention of struggles between Terran lategame at all, just that the early game might favor Terran in TvZ, which really hasn't come up in months before now. It feels like they're really focusing on Korean numbers and US/EU feedback, and Terran is still "fine" in Korea. Maybe it is not as big of a problem as people think it is. Maybe there are silent terrans out there having no problem at all. Or Blizzard sees that the problem is correcting itself as the metagame moves on. But that seems to be the opposite approach they've taken so far. The last time I remember them taking a really hands-off approach to balance was the initial lategame TvZ stance, where they said something about infestor/BL combo maybe being too strong, and we heard nothing for months until the situation "flipped" and ghosts got more screen time in major tournaments. I mean, looking back at the Thor energy change and Ghost nerfs, they were all fairly fast paced without waiting for strategies to catch up. It's not like Blizzard to wait things out like this. Also, Terran pros have made comments about TvP, especially non-Koreans. I will agree they jumped the gun with the snipe, but they have always been reluctant to mess with late game issues otherwise. The other changes were more for all ins of specific types. People forget that when the thor had no energy, players like Nony were trying things like using phoenixes to pick up stunned immortals(a terrible plan) in an attempt to deal with thors and the strike cannons. It was a silly time. But the current “issue” in PvT is not some all in, but late game with 200/200 armies. When an issue is so far in the late game, it is far more likely to work itself out. How much time passed between a serious thor build being played in tournament and the thor nerf? IIRC it was about 2-3 weeks and pretty much only one prominent game. It was so fast that Nony never really tried things but was thinking about things. It was like 5 months from what liquidpedia shows and that was a while. I don't think that thor needs mana to be balanced, but strike cannons with a cool down it makes it so it can auto kill immortals, which is no good With marine support, there isn't a lot else that can really touch the thor is one of those 2 base pushes. Really, if anything, strike cannons should be removed.
Nice try... the Throzain-MC-series that sparked the thor discussion was broadcasted april 23. May 10th was the patch...
|
On April 28 2012 05:30 Blacklizard wrote:Show nested quote +On April 27 2012 11:29 usethis2 wrote:
Slightly improved ovie speed might be a good fix for Z's early scouting. I always thought it's a bit silly that T's buildings float faster than (unupgraded) overlords. But to be frank, the current trend is that T doesn't really much care about being scouted. Simply aim for the 150+ food with 3/3 marines and a handful of seige tanks and from there Z has an uphill battle for the rest of the game.
Since changing unit stats will likely break the game, I think the easier route for better balance is map changes. Maps like Cloud Kindom, Entomed Valley, Antiga Shipyard are all too turtle-friendly. (main/2nd/3rd are triangle-shaped, pocketed in the corner) Personally as a Protoss I'm sick of overlords and overseers seeing every single thing I do unless I go Stargate tech. But really it's late game overseers that are out of line. I was happy for the overseer to 50 gas change way back when with the mindset that more scouting for all races lead to better games, but now extreme late game Zerg has 10+ overseers and they mass changeling block you every 20 seconds. All the late game changelings has become just stupidly annoying. You can't kill them without 15 clicks or without wasting a storm... I mean really.
IMBALANCE!
|
On April 28 2012 05:01 mrjpark wrote:
Reading through about 20 pages of this, really no conversation about Zerg. Just TvP and PvP. Pretty much what the current pro scene is TROLOLOL.
Or because all Zergs outside of KR GM are doing more than fine. Honestly, IDK what there is to complain about Zerg if you're not playing against MKP/MVP/etc. You're Hive tech is pretty much the trump card against T unless you've lost more than 35 dones to harass.
|
On April 28 2012 05:19 Destructicon wrote:Show nested quote +On April 28 2012 04:43 caradoc wrote:On April 27 2012 10:31 Andreas wrote:
David Kim just oozes of incompetence. All he talks about are statistics, and all he can show to is Blizzard's "justified winrates" which we don't know if are reliable stats or not. The stats we do know are real (map winrates) are skewed as hell. TvP is my main concern right now, and I'd love how David Kim thinks it's balanced that Terran has a slightly better chance to win in the first 13 minutes, and then a really low chance to win once it gets past that point. Insead, all I get to know is that their justified TvP winrate is 50% in Korea, so I guess it's fine? Seriously? Did you even read the article? It contains composition analysis, feedback from pros, tournament results, internal metrics and analytics, regional breakdowns, meta analysis... How would you propose he analyze balance, by listening to one person? By figuring it out with a pen and paper on his own? When you're trying to address something as complex as this, you need a variety of information to paint a detailed and nuanced picture. He does this and you say it's incompetence because the results aren't either what you expected or what you want to see? Are you serious? The article has some random numbers thrown in there and just one stat. And Blizzard's stats are worthless since they obviously include ladder games as well, otherwise they couldn't have arrived at their almost perfect 50% win rates for Korean and their "within the margin" 54% and 53% TvP and TvZ. The TLPDs reports from Korean and International tournaments paint a far different picture. Also Blizzard didn't even mention what pro's they talked too, on SoTG Thorzain was even asking "Who are these pro players?". Meta analysis? Laughable, he didn't address the issues of late game TvP or TvZ, he only talked a bit about the low representation of Zergs in tournaments. A good job from Blizzard I'd consider if they would only analyze GM and tournament stats, if they could come up with a program or method to quickly analyze win rate distribution per minute played of games, and if they formulated a list of 33, Terrans, 33 Protoss and 33 Zergs, 11 from NA, Kor and Europe, that they contact every month or every two months to talk to and ask opinions related to the state of game balance. That would be much, much better way to gather data. From what I've seen of Blizzard's stats they are worthless because they take too much ladder into consideration, they aren't aware of a lot of meta-game issues, and for all we know they could be talking to some random NA GM and asking them for balance opinions.
Yea, same things were going through my head. He must have read a different thread and posted in the wrong place. Because I sure as hell didn't see any pro player analysis in there, or meta analysis (w/e that is) or composition analysis, or pretty much anything he listed lol..
|
On April 28 2012 03:46 Rokoz wrote:
I don't understand this at all.
Is Blizzard's balance team so simple-minded that they take these figures as the definition of balance? Read the post and all I saw is that the game is so balanced since in tournaments none of the match-ups don't favor any races.
I see no talk about early, mid- and late-game. No word about Terran forced to stick with T1 units while both others races are frequently using their T3.
I can also keep changing maps to which favor the weak race to make those figures be 50/50 too. But that's not balance.
Actually, no, it's the general TL.net member that is single minded. Starcraft 2 is an RTS played by hundreds of thousands of people (that in itself is quite a feat), 80-90% of which can be considered casual. Casual players want to have fun in a balanced game as well. If you balance around pro gamers only, you risk alienating the larger part of your playing fanbase. There's no way Blizzard can afford to do that from a business point of view.
Besides, people are incredibly quick to scream "imbalance". The game is still very young and the metagame is constantly evolving. Being able to overcome difficulties against all odds is what seperates the great from the merely talented. The game is far from figured out and with two expansions in the making, won't be figured out for another six years or so.
|
Our community is so angry all the time.
Keep working on it Kim! Sc2 is a wonderful game and I hope we reach something we can call balanced soon, or something close to it at least. Dosen't really matter at my level, or for most of the angry people whinning in this thread. The pros should team up and send their collective thoughts, least make em heard. Instead of this angry mob culture we got here lol.
|
if zerg had better early game scouting they can hard counter any aggression that terran does and also if the terran goes greedy, either do guranteed win all in or Super greedy macro with queens and a few roaches early. This game is balanced around terran messing with the other races early on and if the zerg scouts it and responds properly then the zerg should get ahead. If the zerg could scout exactly what is coming everytime early enough, then they in theory should never lose. Also any mech builds the terran would do would be pretty much bad since the zerg would know about it super early and since mech sort of relies on being found out after damage has been done or they have enough to establish a third base, earlier scouting for zerg would sort of make zerg invincible with their ability to mass eco or army when they scout certain things. Im not even talking about 1base all ins, i mean after u hellion expo, u follow up with either banshee, more hellions (mech), third base (greedy), bio medivac hellion, 2base marine tank push, ect ect. and in theory with zergs production capabilities, they can deal with each strategy in an optimal manner if they have perfect scouting early on.
|
On April 27 2012 11:45 Msr wrote:
finally they address zerg scouting.. zerg will actually be a good race now if they don't mess this buff up
yeah whats with protoss scouting? protoss has worse scouting options and less flexible unit production mechanics.
the issue with zerg scouting is, that larva mechanic is actually a broken ability.
if zerg had better scouting, they could prepare perfectly always in time, and therefore win everytime, because certain builds both of terran and protoss are designed to at least trade somewhat cost effectly or force units.
If this is removed, zerg will have to be nerfed in other areas, because a zerg doing a perfect read earlygame almost always gives a very good shot at winning the game.
|
i feel cheated. blizzard isn't nerfing my race anymore it sounds like. i gotta start picking up my game
|
It would be nice if Blizzard released the data sets along with the results so people could look for themselves. I think it's just whining to insist the data must be wrong without having any data of your own, but I'm curious about how or where they got their numbers.
|
On April 28 2012 06:09 freetgy wrote:Show nested quote +On April 27 2012 11:45 Msr wrote:
finally they address zerg scouting.. zerg will actually be a good race now if they don't mess this buff up yeah whats with protoss scouting? protoss has worse scouting options. the issue with zerg scouting is, that larva mechanic is actually a broken ability. if zerg had better scouting, they could prepare perfectly in time, and therefore win everytime.
I do agree with the point that toss has serious early game scouting problems. Once the first ling or marine is made, you have to retreat or loose your probe. Until you get hallucination or obs, you are completely blind if they are competent. If protoss is fast expanding, the time until they get an obs out is even longer. You can't really risk moving out, because if your opponent did make speedlings or mm with concussive, your army can't retreat against them, and you will be on defense for a very long time after loosing early units due to blind aggression.
However, zerg or terran can scout protoss during their blind period with reaper/scans/overlords. It's really easy to tell exactly what a protoss is doing just by seeing the gateway and sentry count. Even with a zealot stalker poke, you usually just see a bunker, and then you just make your best guess as to what they are doing.
Meanwhile they know exactly what you have and can hold the towers for early warning. This is why protoss is in it's current state of blindly all in from 2 base or turtle until deathball. You can't punish greed without an all in and you can't be greedy unless you turtle. It's one extreme or the other, putting pressure while expanding is a bad idea in most cases because they see it coming and the units you send can't fall back. If the opponent is prepared they will chase your army down and kill all of your units. This allows zerg and terrans to cut corners in the early game, or prepare aggression that won't be scouted early. Protoss units are like spartans, no retreat no surrender, you either overwhelm them and kill them or you loose all of your units.
Also, they should release relevant stats like early game/mid game/late game win percentages, and stats by league. The numbers they released here are meaningless. I'm really curious why they would release numbers that are not relevant to actual racial balance instead of numbers that are. For example what are the terran win rates vs protoss in the early/mid/late game in grandmasters on all 3 servers? These numbers are much more relevant than what has been presented here.
|
On April 28 2012 06:20 coverpunch wrote:
It would be nice if Blizzard released the data sets along with the results so people could look for themselves. I think it's just whining to insist the data must be wrong without having any data of your own, but I'm curious about how or where they got their numbers.
It wouldnt help. They informed people on Bnet that terrans were, in fact, not disappearing from the ladder. The response was people freaking out and challenging them, posting links of SCranks. Blizzard can't win no matter what they say or what data they release.
|
On April 27 2012 22:04 jdsowa wrote:
It's just remarkable that Terran's tier 3 units can't win lategame versus Protoss's tier 1 and tier 2. David Kim must be blind or else he would have fixed this by now.
Fixed
|
|
|
|
|
|
EPT
