As was foretold, we've added advertisements to the forums! If you have questions, or if you encounter any bugs, please visit this thread: https://forums.penny-arcade.com/discussion/240191/forum-advertisement-faq-and-reports-thread/

[StarCraft2] POST IN THE NEW THREAD. DIABLO KILLED THIS ONE.

1515254565799

Posts

  • fatalspoonsfatalspoons Registered User regular
    edited May 2012
    MMMig wrote: »
    Careful though Spoons.
    Ever since that Cycles betrayal, I've had a baneling with your name on it.

    We won didn't we? Sometimes the strong must sacrifice the weak for the greater good.
    That's what my dad told me when he made me sleep in a tent in the alley so he could turn my room into a gym/entertainment room.
    That never happened

    fatalspoons on
  • 815165815165 Registered User regular
    whitera's building placement is the sexiest of things

  • mEEksamEEksa Registered User regular
    http://www.teamliquid.net/video/streams/TaKeTV

    TakeTV Invitational is on right now. It has a super stacked line-up.

  • AnzekayAnzekay Registered User regular
    3clipse wrote: »
    Joe K wrote: »
    Biosys wrote: »
    wasn't sirlin the guy who did sf2:hdr?

    i think so, he's be the balance dude for a lot of fighters. Personally, given Browder's history with the C&C franchise, I'd put money on Sirlin doing a better job. I can't think of a C&C that didn't have outright abusable strats that never got addressed.

    Browder's never done balance though. He's always game design. Balance is a totally different group of people.

    Also at the risk of sounding like a broken record, SC2 is a really, really balanced game. Yes, there are issues, but all of the races can and do win tournaments, which in and of itself is a feat for a competitive RTS (lol any of the Dawn of War games?).

    Well, technically speaking balance is still under the umbrella of Game Designer's work, but in larger companies like Blizzard they tend to use a separate team that works closely with QA for proper balance testing and development. The sort of people you find working on balance stuff is still likely to be guys who can work on other aspects of Game Design, though. That said, I don't believe Browder has been directly involved with balance at Blizzard.

  • BiosysBiosys Registered User regular
    hero v supernova and taeja v squirtle tommorow in code s

    god i hope those are better matches than today's T.T

  • 815165815165 Registered User regular
    my german so rusty

  • shwaipshwaip Registered User regular
    815165 wrote: »
    my german so rusty

    Que?

  • fatalspoonsfatalspoons Registered User regular
    donde esta el burrito

  • tuxkamentuxkamen really took this picture. Registered User regular
    Sprechen sie Photon Discs?


    Games: Ad Astra Per Phalla | Choose Your Own Phalla
    Thus, the others all die before tuxkamen dies to the vote. Hence, tuxkamen survives, village victory.
    3DS: 2406-5451-5770
  • kedinikkedinik Captain of Industry Registered User regular
    edited May 2012
    Anzekay wrote: »
    3clipse wrote: »
    Joe K wrote: »
    Biosys wrote: »
    wasn't sirlin the guy who did sf2:hdr?

    i think so, he's be the balance dude for a lot of fighters. Personally, given Browder's history with the C&C franchise, I'd put money on Sirlin doing a better job. I can't think of a C&C that didn't have outright abusable strats that never got addressed.

    Browder's never done balance though. He's always game design. Balance is a totally different group of people.

    Also at the risk of sounding like a broken record, SC2 is a really, really balanced game. Yes, there are issues, but all of the races can and do win tournaments, which in and of itself is a feat for a competitive RTS (lol any of the Dawn of War games?).

    Well, technically speaking balance is still under the umbrella of Game Designer's work, but in larger companies like Blizzard they tend to use a separate team that works closely with QA for proper balance testing and development. The sort of people you find working on balance stuff is still likely to be guys who can work on other aspects of Game Design, though. That said, I don't believe Browder has been directly involved with balance at Blizzard.

    Browder seemingly played a really big role in balance early in the life of SC2. Analyzing data, running unit smashes, coming up with ideas for surveys to send out, providing general help when Kim needed another set of eyes for a test. At least that's the impression I got from that old balance panel they had.

    Whereas it seems like for maybe the last 6 months or whatever it boils down to David Kim reading ladder stats, watching tournaments and reading the pro surveys by his lonesome. Probably appropriate for how well balanced the game seems to be.

    kedinik on
    I made a game! Hotline Maui. Requires mouse and keyboard.
  • Dropping LoadsDropping Loads Registered User regular
    3clipse wrote: »
    If you push roundhouse kick faster than I push roundhouse kick, there is an upper limit to how much the speed difference matters. At some point, we will be producing the same number of kicks per second, even if you are hitting the button twice as fast. Reaction time and anticipation are retained as individual player skills. Sirlin would argue that those are more "cerebral" skills than seeing whose finger can move the fastest. As a contrast, we know that a robot in SCII would never lose because of perfect micro while never dropping macro.

    It's just a point of comparison. It depends on what you want the game to be about. Within human limits, I think it's fine that some people have more effective APM than others. Sirlin would argue that's like letting your opponent take 2 turns in a row in chess. For SCII, I'm fine with that.

    See, I hate his position. I see it as analogous to saying that certain pitchers in baseball can only throw so fast. Or in american football a running back can only run so fast. Etc. It's a dumb arbitrary limit. Really impressive APM and multi-tasking are indicative of a very high level of skill/talent at the game, and capping them artificially just strikes me as completely ridiculous.

    I understand his basic point, it's just that then it's a different game. It's like saying "Who is a better athlete, Joe Montana or Nolan Ryan?" What does that even mean? Any way of measuring them against each other is going to contain arbitrary metrics, and fans of one or the other will pick a set of metrics that favors the person they like. Starcraft 2 fans seem pretty happy without an APM cap. Someone else can make a RTS that does have an APM cap for people that prefer that system. It doesn't mean Starcraft 2 is broken, just that the arbitrary metrics of "being better" requires different skills to pull off.

    Sceptre: Penny Arcade, where you get starcraft AND marriage advice.
    3clipse: The key to any successful marriage is a good mid-game transition.
  • mastmanmastman Registered User regular
    hells yeah go grubby

    ByalIX8.png
    B.net: Kusanku
  • jaziekjaziek Bad at everything And mad about it.Registered User regular
    815165 wrote: »
    whitera's building placement is the sexiest of things

    Everything white-ra does is the sexiest of things.

    Steam ||| SC2 - Jaziek.377 on EU & NA. ||| Twitch Stream
  • GraviijaGraviija Registered User regular
    edited May 2012
    jaziek wrote: »
    815165 wrote: »
    whitera's building placement is the sexiest of things

    Everything white-ra does is the sexiest of things.
    I don't know. I really like White-ra (as should everyone), but he can be pretty iffy sometimes. Like, he has interesting and effective builds most of the time, but his second-to-second decision making is sometimes pretty bad. Especially vs terran, if his stream is anything to go by. Like taking out a third, killing all the SCVs there, then attacking the natural in the most awkward possible position and losing everything, when he could have just macro'd for a few minutes and steamrolled.

    Graviija on
  • redraptorredraptor Registered User regular
    edited May 2012
    3clipse wrote: »
    If you push roundhouse kick faster than I push roundhouse kick, there is an upper limit to how much the speed difference matters. At some point, we will be producing the same number of kicks per second, even if you are hitting the button twice as fast. Reaction time and anticipation are retained as individual player skills. Sirlin would argue that those are more "cerebral" skills than seeing whose finger can move the fastest. As a contrast, we know that a robot in SCII would never lose because of perfect micro while never dropping macro.

    It's just a point of comparison. It depends on what you want the game to be about. Within human limits, I think it's fine that some people have more effective APM than others. Sirlin would argue that's like letting your opponent take 2 turns in a row in chess. For SCII, I'm fine with that.

    See, I hate his position. I see it as analogous to saying that certain pitchers in baseball can only throw so fast. Or in american football a running back can only run so fast. Etc. It's a dumb arbitrary limit. Really impressive APM and multi-tasking are indicative of a very high level of skill/talent at the game, and capping them artificially just strikes me as completely ridiculous.

    I understand his basic point, it's just that then it's a different game. It's like saying "Who is a better athlete, Joe Montana or Nolan Ryan?" What does that even mean? Any way of measuring them against each other is going to contain arbitrary metrics, and fans of one or the other will pick a set of metrics that favors the person they like. Starcraft 2 fans seem pretty happy without an APM cap. Someone else can make a RTS that does have an APM cap for people that prefer that system. It doesn't mean Starcraft 2 is broken, just that the arbitrary metrics of "being better" requires different skills to pull off.

    I feel like people who say that a robot would never lose at SC2 have a very poor understanding of AI and RTS as a whole or at least its current face. The reason you can produce such impressive micro simulations as what is out there is that they have very clear boundaries to program within. Expanding this framework to the whole game is devilishly complex to program within. Sure an AI would never miss an inject, but macro is as much decision making as it is mechanics.

    Say even if the program is based off of a database of 100k games of its opponent that is uses to calibrate its best guess at when to start making units versus economy based on available information, a human can just cheese the shit out of that robot from somewhere it is not scouting in a manner it had no data on.

    This is a game of imperfect information, much of it is divergent. If the AI was say so suspect of being cheesed that it sent out 3 workers every game to comb every corner of the map, well then it has opened up a vulnerability in itself that running a micro script with its smaller army won't do anything to fix.

    Also the idea that Sirlin turned down being lead designer or what not of Starcraft 2 to start a board game company is wholly unbelievable and sounds more like some PR bullhog.

    redraptor on
  • DhalphirDhalphir don't you open that trapdoor you're a fool if you dareRegistered User regular
    @redraptor - I think that the point was that a robot, theoretically, has a higher maximum skill cap than a human, no matter what. It may not be a point where its ever actually relevant, of course.

  • Dropping LoadsDropping Loads Registered User regular
    edited May 2012
    Raptor, there were SC1 robot-only tournaments that were unbelievably broken because of infinite micro. It had nothing to do with decision making and everything to do with APM, which is why I brought it up. I didn't say that speculatively, it's been demonstrated:
    http://arstechnica.com/gaming/news/2011/01/skynet-meets-the-swarm-how-the-berkeley-overmind-won-the-2010-starcraft-ai-competition.ars
    Dropship micro had to be banned from these tournaments because it became degenerate.

    Re-edit: The link above does not contain the dropship micro comment I thought it did, I'll keep looking. It's still an interesting look at how these AIs actually work to account for hidden information.

    I agree with you about the Sirlin actually working on SC2 thing was speculation. His opinion on the game, based on his class at UC Berkeley, has some references though:
    The actual lectures on youtube:
    edit: killing the ref b/c I didn't want to embed the video. It's in youtube.
    watch?v=L7XiE_V0PZ8&feature=relmfu
    Sirlin's take on the lectures are on his website:http://www.sirlin.net/blog/2009/1/30/uc-berkeley-starcraft-class-week-1.html

    Dropping Loads on
    Sceptre: Penny Arcade, where you get starcraft AND marriage advice.
    3clipse: The key to any successful marriage is a good mid-game transition.
  • 815165815165 Registered User regular
    Graviija wrote: »
    jaziek wrote: »
    815165 wrote: »
    whitera's building placement is the sexiest of things

    Everything white-ra does is the sexiest of things.
    I don't know. I really like White-ra (as should everyone), but he can be pretty iffy sometimes. Like, he has interesting and effective builds most of the time, but his second-to-second decision making is sometimes pretty bad. Especially vs terran, if his stream is anything to go by. Like taking out a third, killing all the SCVs there, then attacking the natural in the most awkward possible position and losing everything, when he could have just macro'd for a few minutes and steamrolled.

    This is all true.

    But look at that line of pylons and gateways, man. Holy fuck so pretty.

  • Joe KJoe K Registered User regular
    Raptor, there were SC1 robot-only tournaments that were unbelievably broken because of infinite micro. It had nothing to do with decision making and everything to do with APM, which is why I brought it up. I didn't say that speculatively, it's been demonstrated:
    http://arstechnica.com/gaming/news/2011/01/skynet-meets-the-swarm-how-the-berkeley-overmind-won-the-2010-starcraft-ai-competition.ars

    I agree with you about the Sirlin actually working on SC2 thing was speculation. His opinion on the game, based on his class at UC Berkeley, has some references though:
    The actual lectures on youtube:
    edit: killing the ref b/c I didn't want to embed the video. It's in youtube.
    watch?v=L7XiE_V0PZ8&feature=relmfu
    Sirlin's take on the lectures are on his website:http://www.sirlin.net/blog/2009/1/30/uc-berkeley-starcraft-class-week-1.html

    well, he was at least in contact with Pardo about ranking systems. http://www.sirlin.net/blog/2010/7/24/analyzing-starcraft-2s-ranking-system.html

    i see both sides of the "limiting APM" argument. Yes, being able to complete your keyboard gymnastics faster and more efficiently does separate player levels in both BW and SC2. However, due to the rudimentary nature of the controls (yes, sc2 is better than bw), a lot of the APM could be viewed as "busy work" or something to do just to have you do it, such as having to babysit larvae injects, sways the game more in favor of the "real time" aspect of the game, and less in the "strategy" level.

  • DhalphirDhalphir don't you open that trapdoor you're a fool if you dareRegistered User regular
    Anyone see what White-Ra just did on stream?

  • redraptorredraptor Registered User regular
    I'm well aware of Starcraft AI development. I'm just telling you that without heaps of data on a human player they are so ineffective but they don't react well to unpredictability. And if you are going to bother debating an APM limit you mine as well attack the bigger issue which is if I stick a robot vs a human and get it shit tons of data on its human opponents tendencies, well does the human player also get info on the types of decisions the AI is programmed to make? And if so, how does these things ever come to equilibrium allowing them to compete. There is no discreteness like getting Big Blue to play versus Kasparov.

    And for all the analysis people do of the SC2 matchmaker they forget its pretty meaningless other than lumping players into decent sized skill groups which it does a fine job of. Beyond that the focus of any game searching should be on enabling players to get the types of games they want, say practicing a certain matchup or finding new UMS games within a genre, which it sucks ass at.

  • Dropping LoadsDropping Loads Registered User regular
    edited May 2012
    Dhalphir wrote: »
    @redraptor - I think that the point was that a robot, theoretically, has a higher maximum skill cap than a human, no matter what. It may not be a point where its ever actually relevant, of course.

    This is what I meant. Think of IBM's Watson, the computer programmed to play Jeopardy. They didn't allow Watson to plug directly into the board to ring in for an answer. They built a light sensor (a light goes on when the players are allowed to answer) and an actuator for Watson to physically push a button. Even with these limitations, Watson can still almost always ring in before a human contestant. If it was plugged in directly, it would always answer first. Whether or not it knows the answer is a big AI hurdle, but it's different from APM.

    Edit: Answering your question, sure let the human have as much data on the AI as they want. The difference is the AI can play hundreds of games simultaneously and analyze every game ever played in far less time than the human. It's a processing power argument.

    The APM argument is independent of the data argument. I fully agree that computers are better at abstract strategy games then they are at limited information games. The point was to say that there are many ways to design a game where a computer does not have an advantage (typically because of complexity, i.e. http://en.wikipedia.org/wiki/Arimaa) but limiting APM is another way to do that.

    Dropping Loads on
    Sceptre: Penny Arcade, where you get starcraft AND marriage advice.
    3clipse: The key to any successful marriage is a good mid-game transition.
  • redraptorredraptor Registered User regular
    edited May 2012
    It's such a herculean and shaky design project I think I'd rather attempt to get a man on Mars before I'd sit down to program a Starcraft AI that would take a series off a Code S player, even one who just drank his weight in Korean beer.

    redraptor on
  • DhalphirDhalphir don't you open that trapdoor you're a fool if you dareRegistered User regular
    So I decided to take some pictures of a present the wife made for me a long time ago, back in 2011.

    I don't know why I didn't post it then, but I think I forgot or something.

    Anyway here they are
    20120503074124.jpg
    20120503074112.jpg

  • KambingKambing Registered User regular
    http://us.battle.net/sc2/en/blog/5308170/Call_to_Action_Balance_Testing-5_2_2012#blog

    This is why I'm impressed with the bnet/sc2 architecture. Nice to be able to pull stuff off like this.

    @TwitchTV, @Youtube: master-level zerg ladder/customs, commentary, and random miscellany.
  • mEEksamEEksa Registered User regular
    Oh, Blizzard is actually addressing some of the issues I was concerned about. I feel like upping queen energy by that much will be too strong though, creep spread will be ridiculous.

  • interrobanginterrobang kawaii as  hellRegistered User regular
    ooo cool queen change

  • KambingKambing Registered User regular
    mEEksa wrote: »
    Oh, Blizzard is actually addressing some of the issues I was concerned about. I feel like upping queen energy by that much will be too strong though, creep spread will be ridiculous.

    Yeah. My openings look very different when I get an early tumor down (or not). Should be interesting.

    @TwitchTV, @Youtube: master-level zerg ladder/customs, commentary, and random miscellany.
  • SaarutoSaaruto Registered User regular
    Dhalphir wrote: »
    So I decided to take some pictures of a present the wife made for me a long time ago, back in 2011.

    I don't know why I didn't post it then, but I think I forgot or something.

    Anyway here they are
    20120503074124.jpg
    20120503074112.jpg

    Time to get a matching Terran mug for the Mrs.? Then the inevitable Zerg one down the road?

    If you can chill, chill.
    Steam ID
  • marty_0001marty_0001 I am a file and you put documents in meRegistered User regular
    Your heart will forever belong to the void, Dhal.

    In fact I believe there's a void ray quote where your name is mentioned.

  • redraptorredraptor Registered User regular
    edited May 2012
    Good call that the test realm is completely unneeded and they should just put forth balance iterations through custom maps.

    Pro players might even ... try it, rather than complaining 24/7 of being no partner in design. When they always had the option.

    redraptor on
  • DhalphirDhalphir don't you open that trapdoor you're a fool if you dareRegistered User regular
    marty_0001 wrote: »
    Your heart will forever belong to the void, Dhal.

    In fact I believe there's a void ray quote where your name is mentioned.

    thats a dark templar quote

  • DhalphirDhalphir don't you open that trapdoor you're a fool if you dareRegistered User regular
    edited May 2012
    Just loaded up the custom map and played through a couple games vs the AI with Zerg and then Protoss

    Two immediate thoughts playing Zerg

    1) Wow, overlords move FAST

    2) The queen energy change is only significant for players who have the skill to use it. For most players, its just going to end up meaning they end up at 100 energy faster

    Playing Protoss

    1) lol, observers build so fast

    Dhalphir on
  • tuxkamentuxkamen really took this picture. Registered User regular
    Well, what I'm thinking is that not using the extra energy on a tumor but rather saving it for an early extra inject on a macro hatch is going to make it too strong early on. I'm okay with the OL speed change (but maybe split the difference between the current and test speed). Also, note that the observer build time decreased by 10s--yay!


    Games: Ad Astra Per Phalla | Choose Your Own Phalla
    Thus, the others all die before tuxkamen dies to the vote. Hence, tuxkamen survives, village victory.
    3DS: 2406-5451-5770
  • KambingKambing Registered User regular
    Yeah my overlord zipped around. Now it isn't a liability to try to get him into good positions in the early game before marines get out.

    The queen change is pretty damn nice. It'll be hard to see the ramifications of the change because of what Dhal said, but I think it'll go a long way towards making zerg safe in the early game.

    @TwitchTV, @Youtube: master-level zerg ladder/customs, commentary, and random miscellany.
  • DhalphirDhalphir don't you open that trapdoor you're a fool if you dareRegistered User regular
    A non-chrono'd observer now comes out in the same time as a currently chronoboosted one.

    A chronod one comes out within the duration of chrono.

  • narwhalnarwhal monodon monoceros Registered User regular
    Dhalphir wrote: »
    So I decided to take some pictures of a present the wife made for me a long time ago, back in 2011.

    I don't know why I didn't post it then, but I think I forgot or something.

    Anyway here they are
    20120503074124.jpg
    20120503074112.jpg

    You're going to have to get her to make you a new one now that you're terran. Way to ruin her gift.

    Narwhal#1834
  • shwaipshwaip Registered User regular
    queen change is crazy. if you do a 4queen build, you get so much creep so fast.

  • redraptorredraptor Registered User regular
    edited May 2012
    Conversely you can actually skip tumors with the new queen energy, and get a couple more drones earlier, pushing your next queen back a round of larvae. Then you catch up on the first two tumors with the second queen, and don't see a net increase in tumors until your third queen and maybe fourth iteration of tumors or so.

    redraptor on
  • DhalphirDhalphir don't you open that trapdoor you're a fool if you dareRegistered User regular
    It's certainly going to shake up zerg openings if it goes live

This discussion has been closed.