Back at end of July, at ChinaJoy and IEM Shanghai, I was able to sit down with David Kim, Senior Balance Designer of the Starcraft II development team, for half an hour. With him I discussed various aspects of the game including balance and features that Heart of the Swarm brought to the game.
![[image loading]](http://i.imgur.com/1rpHpt9.jpg)
Reminder: This interview was conducted before 2.0.10 patch, the recent ladder map pool changes and balance test maps, because of scheduling and other work/real-life commitments I was just able to fully transcript it yesterday, Some information might not reflect the current situation or metagame.
Thanks monk and the TL Strategy team for providing several questions and help formatting this thread!
What's your feeling about the overall balance after the recent Terran changes? Are there any other aspects of the game the team is currently monitoring?
DK: I think it's slightly safe to say the changes are good; initially I think it's decent. The main thing we wanted to address in the Hellbat change was to reduce Hellbat drops in TvT, mostly because while we like to have more harassment options in the game, we don't want to have harassment options like Hellbat drops that are a little too easy to execute. So for example if you see Marine drops executed by the highest level of Terran players versus a average level pro player, you can clearly see the difference, while Hellbat drops doesn't look like that because it's really easy to execute. That's why we want to see less of it. TvT has become a lot better since the change. As for the Banshee we are sort of hoping to see a bit more of Banshee play in HotS since we almost stop seeing it altogether. The reason why we want to see more of it is because of the same reason: Banshee micro is really awesome to watch and I think a lot of Terran fans and viewers will agree so we want to bring that back. Initially, players are trying new builds, so we want to keep a close watch on what they are doing. On top of that we don't really have specific changes we are trying to do any time in the near future. But our general stance is that we want to make sure if there is a race that is over-performing or under-performing, we can quickly identify the problem and deal with the problem in a way that makes the game a lot of fun to watch. So for example say Protoss is a bit weak right now, if that's the case then we want to buff units like the Oracle, or other units that promote you to have more skirmishes through out the game.
We currently have two different map pools for WCS, one for WCS AM and EU and one for KR, what's the reason behind this?
DK: So what we have now is a bit complex: in American and Europe we are using 7 of 9 ladder maps, so we have 9 ladder maps on the pool. Korea is a exception, we tried to include the two maps this season as well, but because of timing issues and we need to fix a lot of bugs on the map, stuff like that, so we couldn't make it into the pool. For Season 3 we are currently working with all the partners to make sure and check if it's viable to have a complete alignment of all the maps on ladder and the three regions. So the current plan is try to work out a list of 7 maps so some maps will include TLMC winners and some maps will include the new GSL maps they are creating right now. Ultimately, the goal is try to have a completely aligned map pool between the regions next season.
Can we expect more changes to the team ladder maps? Players have been asking for it since forever.
DK: We are definitely aware of that feedback. We are planning to include two new team play maps every new season. Next season we're working on a couple of new 2v2 maps, so the season after that we might move on to new 3v3 maps and so on. Every new season we will try to include some new team maps as well as 1v1 maps. But with that said, I don't know how feasible that is because our season is quite short now but we will try our best to make sure to rotate out map pool of every single bracket, not just 1v1.
Will the new team maps come from TLMC #2?
DK: Yeah sure some of them will be coming from TLMC as well as Blizzard maps, but right now I think the only maps that will come to the map pool from the contest are 1v1 maps, but the season after that you can definitely expect some team maps to enter the rotation as well.
In custom games we don't have the game list feature from BW and War3. Can we actually expect a return of that? Not to replace the custom games system we have now, just an alternative.
DK: So we have the open game list feature for both custom games and arcade. Compare to say, Warcraft III, the current open game system will eventually have all games filled and much faster, but you can't make your own customized names in open games. I think this is an area we are definitely looking at, as the ways to improve not only to the open game system but also the custom game experience in general. We don't have anything final right now but we definitely are talking about improvements.
Despite the HotS changes to promote more harassment and multiple engagements, Protoss is still the race that encourages deathball play, especially in PvZ. While the warp prism buff is a step in the right direction, is Blizzard considering any other options to change this often-complained about aspect of Protoss?
DK: Not just Protoss in specific, just across the board every single race should have more harassment options, the more skirmishes there are in a given matchup, I think the better. And the reason is because like back in WoL right before HotS came out, if you take any matchup for example PvT, we saw many very similar games because Protoss players are defending drops so well that both players just macro up and don't fight for like 15 minutes and the game ends in 2 or 3 big engagements, if you watch 10 PvT games even on different maps, you see exact same games over and over again. We don't think that's a good thing, across the board we try to do our best to make sure we have more skirmishes. For Protoss specifically we currently don't have plans to change anything right now, but if Protoss or any race appear to be a little weak, the way we want to buff those races is by adding more options to be more on the aggressive.
Oracles were added to encourage harassment and aggressive plays but still it's a do game-ending damage or nothing at all unit. Though professional players are beginning to employ Revelation more usefully, they still seem somewhat as a dead weight unit if they don't do significant damage in the early game. Compare this to the widow mine, which can act as an early game harass unit, but maintains its usefulness throughout the game. All in all, the concept seems to reinforce the idea that Protoss has either only gimmicky or deathball units. Is this a major concern both for this specific unit and for the overall design of Protoss?
DK: So I think the Oracle right now is the Go-For-It unit if it isn't there already. It's a unit we want to see throughout the whole game, every time you go in with it to harass a worker line, maybe you don't get 20 or 30 worker kills like we saw in the beginning of HotS, the defenses has got a lot of better for all races, you don't see huge Oracle wins but at the same time, we want the unit to utilize its movement speed, just stay as a threat throughout the course of a game. A few months ago we tried a movement speed and acceleration buff for the Oracle, I think that was a good test to see if the Oracle is really difficult to kill. If we have a unit that stays in the game for a long time, just doing small things like scouting, using Revelation to check the army composition but not doing crazy significant game ending damage, then we can create a lot more small skirmishes so that each game is more exciting to watch since there is a lot more action throughout the game.
The Stephano style Swarm host based compositions in ZvP are beginning to feel like infestor broodlord compositions in WoL. The Zerg would often hunker up on 4-5 bases with free units and starve out the Protoss in a split map situation. How big of an issue does the development team see this as?
DK: It's definitely the problem if that's the main way to use Swarm Hosts, right now we seeing two different ways to use Swarm Hosts, one is what you mention, the super boring mass spine, you defend and slowly push using Swarm Hosts. The other I think it's more fun to watch, the Zerg uses Swarm Hosts to send locusts then unborrow them right away, move them to a different location, constantly play this hide and seek game. Depending how it works out, if it becomes the case that the Stephano style is the most useful way to use Swarm Hosts, we will take measures to fix that problem. Right now we are in the process of still evaluating which one is the most prominent and effective way to use Swarm Hosts, but we do agree that stale and boring strategies are not good for the game, our two main reasons to patch HotS right now would be: first, if a unit is too overpowered, second is like you say a unit or strategy is causing the game boring to watch.
In WoL, Protoss could react to mutalisks with either blink/templar or phoenixes. But with the mutalisks and phoenix buffs in HotS, only phoenixes are now viable as a counter, causing PvZ to become very centralized around stargate play. It also complete got rid of the storm vs mutalisk dynamic we commonly found in WoL, which seems counter-intuitive since a stated goal of the design team was to encourage different styles to be viable. Was this focus on phoenix/stargate play the intention of the design team and what do you think about this new mutalisk dynamic in PvZ?
DK: In WoL we really only have the Robo tech and the Templar tech viable. The Stargate was just for making a small number of Phoenixes or Void Rays early on and then you completely forget about it. So in HotS we wanted to make the Stargate as powerful as Robo tech or Templar tech. In this specific case, I agree only the Phoenixes is the best counter to Mutas, but I think you can still defend with Blink Stalkers until you get your counters out. We don't believe when you fight against this specific strategy every tech tree must be viable. We want every single tech tree to be viable in specific scenarios. Against Mutas, Stargate will be the strongest but if someone is going mass Hydras to counter that, going Robo or going Storm will be a lot better. I think as a whole we are at a pretty good space, what's interesting is we buff the tech tree that was never being utilized in WoL, because of that a lot of people feel like this tech path is a lot more OP than it actually is. Say for example if no one uses Robo, no one uses Colossi in WoL, we buff them to be what they are now, a lot of people will be complaining that Colossi are overpowered, when in fact that's not the case right now. So I think it's important to see the big picture and see how viable each tech tree is in that sense I think we are quite successful not only with the Phoenix changes but also with the Void Ray changes as well.
During the Beta I had a interview with Dustin Brower, and he stated that making mech more viable was an important focus of the design process. How successful do you think you were with this goal? How have your goals changes since then?
DK: So I don't think we completely hit that goal but for example in TvT mech is definitely viable, we see a lot bio vs mech and mech vs mech games already. In the other two matchups it's not as viable as bio but I think if we have to choose, mech being viable in TvT is the most exciting because it makes the mirror match-up a lot more diverse than players go to Marine-Tank all the time. As far as we try to explore how to make mech better, we have to be more careful because the balance of the three races are pretty good right now, each match-up is quite fun to watch right now even including ZvZ and PvP. We don't want to make drastic changes just to switch up the current metagame. Like I said before, I think our two main focuses for when we patch HotS is if there is problem with a unit like it's too overpowered or if a strategy is boring to watch. But if you look at bio play, especially in TvZ, it's very exciting. There are so many skirmishes and sometimes you even have game where players fighting over a location for like 20 minutes. So since the game is exciting to watch, I think it's dangerous to touch other strategy or units.
The recent nerfs to hellbat, with the stated goal of nerfing hellbat drops, had the unfortunate side effect of nerfing hellbat-bio compositions in TvP. While the blue flame upgrade is easily accessible in mech compositions, a factor with an attached tech lab is hard to come by in bio-based compositions. This is reminiscent of the snipe nerf to ghosts, which while were intended as a nerf in TvP, caused the unfortunate side effect of nerfing snipes vs zealots and ghost openings in TvT. How heavily did Blizzard consider this side-effect when choosing this particular hellbat nerf?
DK: Sniping on Zealots isn't that amazing even before the Snipe nerfs (blame monk for asking that lol). But that aside, for the Hellbat it's all about player hasn't been used to play that way yet, I think the perfect example is in TvZ: the most popular style is to play Bio-mine, and Terran players don't have problem going Reactor Widow Mines while getting Drilling Claws upgrade, so kinda similar to that, I think when players figure out the best timing to get the blue flame upgrade for the Hellbats, I think Hellbats can still be utilized as before because they are still very powerful units. Their stats hasn't been nerfed and they still dominate mass Zealots and so on. It's just a matter of timing and learning the best time to get the best upgrade and work that into your build order. eventually you'll get Hellbats back to your late game composition.
With all of the effort to make mech TvX a viable strategy in all match-ups, can you explain why the siege tank has been left relatively untouched besides free siege upgrade?
DK: I think that was a huge change because it allows Terran to play a lot diverse openers because the defense is a lot stronger. On Tanks in general we receive a lot of feedback that Tanks needs a buff, most specifically have more damage and lower rate of fire. Doing that we don't think it will do that much in like PvT, because the counters to Siege Tanks are Chargelots and Immortals. If you increase the damage a bit, it won't change much because Immortals and Zealots have a really harsh counter relationship with Tanks; you might be able to kill a bit more of them, but I don't think it will change too much. Our worry is that Tanks are good units in TvT, so what that buff will mean in that matchup? So I think it's difficult to touch units that are useful otherwise in other matchups. And as I said, our goal is not making every unit be as powerful as, say the Widow Mine or Void Ray. The goal for us is to create exciting games to watch and as long as strategies are good, as long as every matchup is diverse, I don't think it's a specific problem to specific units like Siege Tanks. But at the same time, if the metagame evolves into certain spaces that games are boring and dull, then I think Siege Tanks is a good area to improve on.
Are there any plans to try to further integrate creative content from outside sources other than Blizzard, directly into the client? Is this something you've looked at with the beta desktop client?
DK: Recently we released the WCS Portal. It's a place to check standings, schedules, and stuff like that. But like you said, there could be improvement like the in-game client, but for now our focuses is getting WCS right, getting the portal as best as possible, and then after that looking at other resources as well. For the desktop client, I have no idea actually haha. We are still in beta, so it's really hard to say but we wouldn't rule anything out in the future.
In Valve's Dota 2, they instantiated a system that places reported individuals in a low priority pool as a sort of penance for their maligned actions. Can you comment on what you think the value of this kind of system is in deterring what would be flagged as unwanted behavior? Would a similar system have value in SC2's environment?
DK: I don't know if the reporting system is the best for Starcraft II, but we do believe that improvement can be made for average players' playing experience, especially in team games. It's a thing we are looking into right now, and in the future hopefully we'll have something that really fits our game and make improvement to Battle.net that way.
Previous Blizzard titles have in the past had a simple but very effective hotkey setup to help with semi mouse-less navigation within their respective clients (battle.net). As your player base gets more and more accustomed to the Battle.net UI, do you have plans to further add to this existing functionality beyond what's already been implemented (such as adding [Tab] or [Enter] into the design)?
DK: We currently are more focusing on features rather than polishing issues like that. Because our focus is heavily on creating so many different features like training mode, matchmaking vs AI, leveling system and watch replay with others. The polish part is a bit neglected I guess. I think going forward if all these systems turn out to be pretty good, then the next step can be looking at how to make our systems easier to navigate.
Before we end, can you say something to our Chinese players?
DK: I think Chinese players are doing good in WCS right now, in particular I'm cheering for Jim and MacSed, I really hope one of them wins the whole thing! Hope you guys not only play the game, but also watch esports and cheer for your favorite Starcraft II players.
해석은 제 능력이 안된답니다
