Here is the transcript of Heartbourne’s liveblog coverage of the BlizzCon 2010 StarCraft II Multiplayer Panel.
1:21 heartbourne here. Starcraft 2 multiplayer panel coverage coming at 1:30 pst
1:29 just a couple more minutes
1:31 panel starting!
1:32 “all of our tools fail us at some point – community feedback, internal testing, pro feedback, etc”
1:35 the creativity of the community really surprises them for strategies that covers what internal testing cannot
1:35 they are referencing the epic morrow v idra mass reaper strat for the “pro feedback” effect
1:36 downside to pro feedback? they usually only play 1 race, making it tough for them to judge race imbalance and it doesn’t represent the majority of player experience
1:37 “if there is a tiny crack in our game balance, these guys are going to drive a truck through it”
1:39 “everyone who works on balance is a diamond random player”
1:39 switching over to concrete tools for testing, like spreadsheets
1:41 its great for number-crunching, but doesn’t take into account variables like difficulty of builds, terrain, pathing, etc
1:41 change something small like unit size can have huge implications
1:42 showing a tool called “make combat” that they can quickly use to adjust numbers and quickly see the effect by running battle simulations
1:43 weaknesses: usually terrain is varied, and unit compositions are hard to match up with real builds
1:44 other tools: battle.net stats
1:45 they can look at per player stats: posters complaining about say ZvP can actually be fine, just a subjective problem
1:46 battle.net race win stats as officially reported by blizzard! screenshot coming
1:46 MMS doesn’t know at what levels races do better, etc, so how do you make it 50/50?
1:47 MMS tries to make race win percentages 50/50 as well to help with balance by matching weaker races with weaker players
1:48 MMR system uses not only relative win percentages of players and their ratings, but larger relative “race ratings” as well
1:48 for example, if zerg is weak in ZvP but zerg wins, they get bonus points in aggregate
1:49 adjusted for skill levels, the adjusted percentage for TvP is 40/60, PROBLEM
1:50 for diamond, its not nearly as bad
1:51 strategies and win percentages often widely vary across regions – korea leads in strategy innovation and stats
1:52 though TvP is broken in most places, its flipped in diamond league in korea: terran is stronger in that league
1:53 this is why they need to be careful: they consider korea diamond to be the highest level on battle.net
1:54 combat tool suggests marauder stim is not OP. they don’t want to change it because its so borderline
1:55 Pro opinions on TvP: of 5, 2 said terran is way OP, 2 said protoss is way OP, last person said balance is good. mixed opinions are rare
1:56 some early opinions: terran is strong early, protoss is strong late. it could be a larger design problem
1:57 the thing they are looking at now: stim balance and psi storm balance
1:58 here is the equation they use for race balance stats
1:59 doing achievements offline? they are considering solutions to online/offline play
1:59 buff mutas against light? mutas are fast and attack air and ground. they think its very versatile and don’t want to fiddle with it right now
2:01 custom hotkeys coming? yes, we wanted to push the game for release, but we want to do it. the UI is complicated and we want to make it simpler so new players don’t screw it up
2:02 there is a guy asking about map marketplace, same guy as the map panel :|. they say earliest release is HotS
2:04 “what do you think about rushes and proxy stuff?” those are easier to fix than greater balance problems. we think its pretty fine now, low priority
2:06 “hyrdas vs zerg?” in small numbers, try zealots and stalkers. bigger, templar and colossus
2:06 err, protoss
2:07 “score screen updates with more stats coming?” we really want to, we have lots of things planned for game analysis and feedback
2:08 look for it in HotS. could be in a WoL patch, but probably HotS
2:09 “score screen from replays and batch analysis?” we want to do more, but UI is hard to figure out
2:11 “was zerg designed for macro and the others for choke point play?” they designed zerg to make it hard for them to exploit chokes, so lots of maps have terrain, etc. question answers itself to a degree
2:13 “patch balance issues flip flop relative balance” this is why we are careful. we did it fast during beta since it was for testing. we want a balanced, stable game, and careful changes are important
2:14 we intentionally kept to 3 races and 12-14 units per race to make balance more possible
2:15 hey, its excaliburZ, the guy that made the matchmaking thread on TL
2:16 blizzard: “your analysis isn’t perfect, but its ‘close enough’. we do take league and division skill level into effect a little, etc”
2:18 “how do devs experience tournaments? as fans or as developers?” both
2:19 “are warp gates OP? they make units sooo fast.” they’ve been nerfed over the past year. we think every race has a ‘broken aspect’, but we think warp gates are fine right now
2:21 “hunter seeker missile getting better?” they used to be too OP, trading mana for enemy resources and degenerated games. right now its situational, as spells should be
2:22 we might make separate servers to play older patches for tournament play so patches don’t catch pros by surprise
2:24 “do you think roach and reaper changes were too hasty? community could have found solutions?” TZ was absurdly powerful in 2v2 and it was pretty strong in high and other levels of play. wanted to reduce terran openings
2:26 “do you think thor and ultras are too different?” they have different purposes, not as THE end game solution. ultras are for slash ground, thors are more tactical
2:27 zerg macro is really good, so ultras are fine for now
2:28 “do you take spawn locations into account for stats, matchmaking, etc?” don’t think so. we cut a couple maps and map balance is a real serious concern
2:29 map stats are gathered and watched, but the balance guys think that right now its drilling down too far
2:29 multiplayer panel closed! heartbourne AWAY