★Indie games by Greg Lobanov & friends★
This post is a sequel to an article I wrote last year on the same subject!
I find all balance patches fascinating... they represent the living connection between game and audience, and they also make manifest the oft-invisible magic of game design. When I wrote our first post, we were in our first few months post-release and were finding our footing in how we identified and addressed balance issues. And now that it's been a year and 4 major balance patches since then, I thought it would be fun to return to that data, look at our results after a year of updates, and talk about how my perspective has shifted from experience.
In the last article, there was a graph of Beasties mapped by their popularity and winrate. Here's that same chart again, but with the names of the most significant outliers revealed this time!
At some point after my previous article, I developed a fairly simple formula that can sort the above data into an ordered list of balance suspects. (I'd love to talk to an actual data scientist or big-studio game balancer someday... I'm so curious whether their process is anything like mine...) For each character, I calculate their "Meta Influence" as:
1000 x (W / P) - ([W - Wc] / [P - Pc])
- W = the sum of all wins by all characters
- P = the sum of all games played by all characters
- Wc = the sum of all wins of an individual character
- Pc = the sum of all games played by an individual character
To put it into English, we basically measure the impact of subtracting each given Beastie from the global results. If a Beastie is used very often, and wins or loses very often, then you'll measure a big impact either positive or negative. But if the Beastie's winrate is very close to 50%, or they're used too little to be significant, then the impact will measure closer to 0. Having a formulaic sort gives us a helpful starting point when sifting through 107 unique Beastie+Trait combinations. With a meta that's spread so wide, there are always some Beastie+Trait combinations that get used by a only a few active players over the course of one balance patch... so those Beasties' winrates will tend to reflect those players' skill level more than the Beastie's actual parameters. This formula is very helpful for identifying what is actually outside the expected curve, having the greatest impact on player experience with the most data supporting it.
If I apply this formula to the dataset pictured above and sort by "Influence", you wind up with this ranking of the top 10 balance suspects:
And if you want to fully enjoy this anthropological study of Beastie balance in March 2025, you can look at the patch notes we released based on the above data, with changes corresponding to just about every suspect. When Beasties have too much dominance over the game's meta, we apply careful nerfs. And when we see a Beastie that many players are failing to find success with, we give them a thoughtful buff. Our end goal with these sorts of changes has always been to create a space where every player's favorite has a reasonable opportunity to succeed.
What's also nice about a formula is that we can measure results across patches. Here's a (redacted) ranking of the top 10 balance suspects as calculated in December 2025, shortly before our 3rd milestone balance patch, based on a data set about the same size as the previous one. You'll notice that, after going through a couple balance patches, we'd successfully reduced the extreme ends of "influence" so that the meta is less centralized to certain Beasties:
These days, it takes more and more time and data to see extremes emerge in the meta. But... there ARE still extremes. You may naturally now wonder, at this rate, how many changes will it take before everything is actually, finally balanced?!
As players raise in skill level, they increasingly prefer systems that are more predictable and "fair" so that their skill is consistently expressed in their results. Over time, all of the bumps and spikes in the game system will eventually appear to be flaws that demand flattening. And the more peaks we flatten, the more tiny hills will suddenly appear to grow into tall peaks. If we follow this trajectory of "more balanced" forever, we would erase and erase until there was no game left to play.
The truth is that balance simply cannot exist in a game with any asymmetry. There will never be a time when every Beastie has the same influence on the meta. But that variation in influence, and the way it changes over time, isn't just an inevitability... it's an essential texture to the competitive experience.
In fact, even if we were to entirely stop making changes, the game would continue to change shape and develop new peaks on its own. You can look at fighting games like Street Fighter or Melee whose metas have changed constantly without a single balance patch in decades. At any given time, some characters are seen as dominant and others as unplayable. But with time and practice, players will find new ways to leverage old tools and upset the status quo. In fact, sometimes I wonder if the promise of balance patches has a negative impact on the meta's innate ability to change. I'll sometimes see players get frustrated by a strong element in the meta, and then give up on it, expecting a balance patch to address it for them, rather than looking deeper through existing tools to find an answer.
All this almost makes you wonder what the point even is of making balance changes in the first place. After all, I do want to step away from Beastieball someday and leave it set in stone. How can I do that when balance is impossible and ever-changing? When do we stop?
Understanding that it's impossible for the game to ever be truly balanced at any given time, we're shifting our perspective towards how we can design a game that has longevity. Instead of trying to make all things equal, we're trying to make a game hardy and diverse enough to produce answers for any of its own problems. We want players to be able to continue exploring the game in 1, 2, 5, or 10 years and still see new metas develop, with no convergence on a single dominant strategy.
Our process up until recently has focused on quashing issues by flattening the outlier Beasties until they came closer to equal in terms of potential. But as we approach a version of the game that has a growing diversity of winning team compositions, we're finding greater value in making more things more crazy, even when it flies in the face of reason. In particular, we've put more attention towards re-examining Beasties who aren't used as much, or whose gameplans aren't as interesting, so see how we can spice them up. After all, Beastieball is a game where every single creature is laboriously hand-animated, so we can't afford to let even one Beastie become irrelevant in game design. In terms of character design, every Beastie is designed to be somebody's main character... so in gameplay, we want them all to feel overpowered.
We aren't making Beasties more "spiky" like this just because it's fun to design. It also helps us build a possible future where, no matter what strange shape the meta takes, an existing Beastie can reemerge as something that excels and wins against the current dominant one, thus keeping the meta in constant motion. I think that game has never been lacking in this quality per se, but we definitely are considering it more now.
A learning experience for me came recently when we were examining Mistic, a ghost-like Beastie whose Intangible trait makes them immune to Body-type attacks. Taking zero damage from 1/3rd of all attacks in the game is certainly a ridiculous trait to exist. And as greater meta threats had been mitigated, some players began to fixate on Mistic as the next Big Problem. This brought on attention from myself and co-designer Damian, taking it upon ourselves to try to finally "find a fix" for this.
Over the course of several experimental balance patches, we tried a couple different re-workings of Intangible, including one where Intangible would become Tangible after neutralizing its first Body attack. That change felt fun and creative and fresh, and it was certainly more balanced, and made it possible for Body-focused players to defeat Mistic without having to give up entirely on Body plays. But Mistic was no longer the terrifying presence it used to be. Body-attacking players didn't need to pay attention to Mistic anymore. And I very quickly realized that even though we'd made the game more mathematically sound, we'd done it by flattening away an essential corner, and turning something unique into just another Beastie. So we walked this change back and never let it into a proper milestone release (to the chagrin of a loud minority of players, I should add...).
Of course, the best design will always be something a little bit in-between. If everything is too spiky, then every matchup feels one-sided, the result always determined as soon as the teams are selected. The ideal is for every match to approach that exciting near-but-not-exactly-50% sweet spot where one side always has a slight disadvantage, but can often overcome it. That's why we decided to address players' unease with Mistic partially by adding a new play, True Strike, a powerful Body-type attack which can ignore traits, including Mistic's Intangibility. The catch was that we only gave this play to a selection of under-used Beasties with lower attack power. Now Body-favoring players have a clear answer to a common threat, but it requires some compromise and adaptation. Do you go all-in on hyper body offense, foregoing low-power attackers and hope you don't see Mistic? Or do you field a slower, more stable composition that can defeat Intangibility, but might lose to faster-paced teams? This introduces one more fold to the fabric of the game, a question whose optimal answer can ebb and flow based on the current metagame.
As a designer, I really can't help but love the art of building balance patches. They're an endless source of problems which become opportunities for design cleverness and creativity solutions. But it can also feel like building sandcastles, beautiful and doomed to be flattened by the inevitable tide of player innovation. As we pivot to thinking about "1.0" and I think about an end goal for balance, I've felt a renewed sense of hope and joy in this work. It's apparent from our stats and anecdotally that the game has made tremendous progress as a competitive platform for those players who want it. I'll keep striving to maintain that trajectory and leave it as something singularly great in the end.