Talking Balance with David Kim
Overall
Based on our internal data, tournament results, as well as the feedback we receive from pro-players and the community at large we believe that the balance between the StarCraft II races is in a good place. Still, we’re going keep a close eye on things to make sure that gameplay remains fun and balanced even as we prepare for the exciting changes that will come with Heart of the Swarm and beyond. In today’s report, we’re going to look at some recent shifts that might merit further investigation, where we are in terms of balance over all, as well as how things look in the pro scene.
We’re currently looking into a couple of potential issues. Zerg has recently shifted from making a very strong showing in tournaments, to having slightly weaker representation only at the highest levels of professional play. We’re also continuing to see a slight advantage for terran in terms of opening flexibility and scouting denial. In response, we’re considering offering zerg better scouting options in the early game.
I think that it’s worth pointing out that these are actually relatively minor concerns in the face of the statistics we’re seeing. Assuming nothing drastic changes in the meta game that would shift the win rates further out of balance, there’s a strong likelihood that it won’t be necessary to make balance changes until Heart of the Swarm arrives. Still, if a significant issue arises that is beyond the reach of metagame changes to resolve, then we will react as quickly as possible to make the adjustments necessary to restore balance.
Adjusted ratings
Here’s the latest global adjusted ratings data for all regions:
|
All leagues |
North America |
Europe |
Korea |
|
PvT |
54% |
56% |
50% |
|
PvZ |
54% |
53% |
50% |
|
TvZ |
50% |
50% |
52% |
Please note that the way we do this calculation factors out player skill. These numbers are also constantly in flux. For the purposes of interpreting this chart, a 45-55% win rate suggests that there is no sign of imbalance. Also, because these numbers change from week to week and day to day, it’s not uncommon for a race that shows a 54% win rate this week to show 46% next week. Overall, we have been seeing win rates in the 45-55% range in every matchup and in every region for a while now.
We do have a new method of calculating player skill more precisely. The different races tend to be slightly stronger or weaker at different league levels , and since most players only ever play one race all the time, racial strength was an influence on their hidden rating. For example, if terran is slightly weaker at the gold level, the player would have a lower hidden rating than if he were to play a different race at that level. When using this method of calculation, we are seeing good balance at the highest levels of the ladder - race balance is good enough at those levels to have a negligible influence on player performance.
We’re checking all of these numbers on a frequent basis in order to react quickly if something big happens.
Tournaments
My earlier comments about zerg notwithstanding, overall, we are seeing good win ratios and stats from tournaments.
One interesting thing to note is that there still is a skill gap even at the pro level. This is what we’re seeing:
- Protoss has good representation in terms of the number of protoss tournament qualifiers at the general professional level, but has the lowest representation at the very highest professional level.
- Zerg has the lowest representation at the general pro level, but has been showing the best highest level tournament standing until very recently.
- Terran has good representation over all.
We are concerned that protoss players seem to be having a slower start compared to zerg or terran players. For quite some time after StarCraft II was released, especially at the highest levels of professional play, protoss players were primarily only focusing on 1 base or 2 base warp gate all-ins. It was only after we toned down these builds that protoss players really started exploring their various tech options in the same way that the other two races have been doing for a long time.
Terran representation remains strong even after recent nerfs. We think that the adjustments we’ve made over time have kept pace well with meta game shifts, and that overall terran is in a good place.
Again, we’re closely watching every major tournament, and working with various tournament organizations around the world in order to keep close tabs on both specific games and overall tournament data.
Maps
In our current map pool, we’re seeing that not every map we’ve added to our pool is balanced for every matchup. For example, our data shows a 70% PvT win ratio on Cloud Kingdom, a 62% win ratio in PvZ on Korhal Compound, and a 37% win ratio on Metalopolis for TvZ. It’s worth noting that these represent the heaviest imbalances in the maps on our ladder, but that this also isn’t necessarily a problem.
Our current stance on this is we believe slight imbalances in maps actually make the game more interesting, as long as the imbalances aren’t too great. Various new meta games and strategies develop depending on the map, and it’s interesting to see the game play out in different ways as players exploit their own strengths and weaknesses from map to map. We believe a lot of the fun comes from this as well. On top of that, we have our veto system on the ladder and tournaments have their own player map veto systems that also deal with this issue.
That said, if a map pushes the limit too much and a matchup becomes significantly imbalanced, we will definitely take steps to adjust the map pool accordingly.
Community Feedback
The pro player and caster feedback regarding which races are overpowered or underpowered is very evenly distributed among the three races, which is a very good sign.
The feedback from you all in the player community has been very interesting too. A lot of the feedback we used to see on the forums was often related to perceived imbalances in units or strategies at the pro level. These days, it seems that we’re seeing more conversations that focus on what’s happening at the league level directly - which we think could be related to a more visibly balanced pro scene. That’s good for us because, while we like seeing very solid results at the pro level, we are also always working towards a balanced experience across all skill levels. We will continue to pay close attention to both high level players and discussions within the community in order to continue pursuing this goal.
At this moment, it feels like we’re in a good spot when we look at the current state of the game from various angles. However, we are always on the look-out for new issues that could arise, checking our data, and gathering new feedback on a regular basis, and we’ll make adjustments to the game as needed.
Thank you for your continued support and valuable feedback!
Muex has already covered why protoss early scouting is fine as it is.
When terran walls off and we have 0 info, they can either be expanding or teching or multiraxing - all three of which requires very different responses. At this early stage its a coin-toss which response to take.
And Zerg doesn't really work around 'generally defensive'. Getting a Roach Warren, Evo, Spines, Lair, Spire, Baneling Nest, appropriate researches, etc., to be safe against everything puts Zerg too far behind.
I never liked how terran can get the most info and deny the most info too. This advantage that they have was amplified when the maps were much smaller.