Hi everyone! I had a great time last week answering your questions on Emongg’s stream and hearing your feedback about Stadium. Today we’re pulling back the curtain on some important (and interesting) Stadium statistics!
Game balance is something that’s always at the forefront of our minds. Heroes, Items, and Powers have to regularly be tuned up or dialed back to create a more even playing field for everyone. Prior to launch, we’re always testing and ideating internally, but there is one very important aspect to the mix that we can’t test for: you!
Player data is an invaluable tool we use to see what’s working, what isn’t, and what could use a little more attention. To better show you how that process happens, our data analyst Ben Trautman is here to deep dive into how the team uses data on Builds to influence balance changes.
Hello hello! Ben the analyst here to chat about Builds, one of the trickier (but very fun) things we get to think about with Stadium. In our Stadium Reddit AMA I went deeper about how we discover Builds in the first place, but in short, it’s useful to think about Builds as a combination of Powers and Items that are distinct enough from one another to alter how you play a specific Hero. Thinking of Builds allows us to consider whether an individual Power or Item—or a combination of many of them—is the engine behind a player’s success in a match of Stadium.
Let's take Juno as an example of how Build data feeds into balance changes. Our first balance patch for Juno nerfed one of her Powers, Cosmic Coolant, which gives a shorter cooldown for Pulsar Torpedoes.
Our data shows that Juno's most successful Power by win rate is Pulsar Plus—giving Juno an extra charge of her Torpedoes—and Cosmic Coolant follows by a small margin. The synergy between these two Powers means that many of Juno's Builds, by our model's estimation, have both Powers as a must-pick.
The dominant Build variant (let's call this Variant 1) from a usage perspective takes a trio of Powers: Pulsar Plus, Cosmic Coolant, and Medicinal Missiles. A second variant (Variant 2) notably takes Pulsar Plus, Cosmic Coolant, and Torpedo Glide. A final variant (Variant 3) takes the usual duo of Pulsar Plus and Cosmic Coolant but hybridizes more by choosing Weapon Powers like Stinger and MediMaster.
The win rates of these three variants diverge strongly: Variant 2 leads the way with 55%, followed by Variant 1 with 52%, and Variant 3 with 48%.
The difference in the success rates of these Builds gives us more specific information about exact balance changes to make. Variant 2, with Torpedo Glide as its standout difference, suggests that the cooldown reduction of Pulsar Torpedoes (enabled by Cosmic Coolant) was the main driver of success. By tuning down Cosmic Coolant, other Builds have a chance to succeed.
I also pulled some additional Stadium stats on May 5 to better illustrate how things changed—or didn’t change—after the May 1 hotfix patch.
Across all platforms and ranks, Soldier: 76 has been the most-played damage Hero and Reinhardt was the most-played tank by a decent margin. Juno is still the most-picked support, but Mercy and Moira are nipping at her heals (get it?) thanks to the hotfix patch.
For those Heroes, certain Powers are your clear favorites: Juno’s Pulsar Plus has a whopping 78.1% pick rate while Soldier’s Super Visor has a 75.4% pick rate, even after minor nerfs in the latest balance patch. Reinhardt players are making slightly more diverse Builds, choosing either Magma Strike or Smashing! 32.8% of the time.
Being popular isn’t everything, though! Lúcio continues to have the highest Hero win rate on both PC (58%) and console (57%); however, Mei is matching that win rate in higher skill divisions on console. Until her recent adjustments in Stadium balance patches, D.Va had the highest tank win rate on console by a wide margin. Nerfs allowed other tanks to taste success on console and D.Va is now second place in high skill divisions.
It’s me, Aaron again. Thanks, Ben!
We also have an update to share on leaver rates in Stadium. Initially we reported a 7% leaver rate, but after some tweaks in how we collected data, we are seeing that leaver rate is closer to 12% (on PC). This seems to be largely coming from first-time players, with leaver rates dropping the more matches one has under their belt. Since Stadium is still so new, we expect the leaver rate to drop and settle at a much lower rate in the coming weeks.
The philosophy around Stadium balance is a bit different from the rest of Overwatch 2. We want you to feel as though you can have MVP moments in a match—riding the cusp of feeling overpowered—while also seeing the mode as something that rewards skill in both combat and Armory strategy. While we want players to feel OP at times, we’ve already started to see some builds that have high burst potential with a relatively low skill floor. We believe that some of these have gone too far and will be looking into adjustments to them.
Variety is still deeply important for Stadium, too. We want you to feel like your Hero’s strength grows and morphs in different ways depending on the match. Sometimes you’ll be the carry and sometimes you’ll take a supporting role. Stadium won’t be as delicately balanced as our Core competitive modes, but our goal is to keep it “casually competitive.”
Thanks for all the excellent feedback about Stadium and we hope you keep on climbing the ranks! As always, let’s make a great game.