Designing UniVersus: Atlanta Regionals Reflections
Bill Stark
Our recent Atlanta Regionals was the largest Regionals in nearly two years, and a thrilling event from beginning to end. As game designers we look at events like Regionals as opportunities to check our work and see how our designs are landing with players. We examine a format’s health to evaluate if we need to take any Ban and Errata actions, and to see if any play trends provide guidance on ways we should adjust game designs in future products. Today we’re going to talk through some of how game design assessed the Atlanta Regionals as well as providing some of the hard data from that event starting with looking through the lens of old design paradigms versus new.
Defining Eras
With last year's release of Jet Burn, UniVersus officially entered its UVS Games era. The change was met with a new card back, new card front, the beginning of a shift in how we approach game design within the game, and a lot of Attack Orange color. These changes came with many, many goals, not least of which were appealing to more players, reducing mental load and complexity for new players, keeping players in the game longer, and improving collectability. We’ve now delivered just shy of a half-dozen booster sets in this new era, as well as introduced new product lines like the Challenger Series. When we examine the results from a large tournament like Regionals, we’re examining closely how this new era of design is landing with players and how it compares to the sets from the older design mentality.
Are new cards being played? Is there a diverse representation of characters, including those from the UVS Games era of design? Do products make up their approximate “fair share” of a metagame, or do they over- or under-perform? The success of what we do in design is dependent on our tactics working; if the data shows us changes we’ve made aren’t working, we want to pivot our approach to find something that does accomplish our goal. So, are players playing the UVS Games era cards?
Card Breakdown
In a word, yes. This is the breakdown by booster set of the cards played in the Atlanta Regionals tournament:
We calculate the “percent of cardpool” by comparing the cards from a given set played in the event to the entire cardpool of all cards played at that event. (I’ll note: we analyze the entire format across all products, but for brevity’s sake, I’m presenting just the booster sets here for ease of viewing). The sets from the UVS Games design era make up the Top 4 slots, or put another way: approximately 60% of the cards played in Atlanta were from the UVS Games era.
What do we conclude from this? Certainly that the changes we’ve made are connecting with players and seeing lots of play. The newer sets were not only the most popular for competitors in Atlanta, but our internal data also shows us that Attack on Titan: Battle for Humanity is the best performing set of all time in attracting new-to-UniVersus players. While not everyone loves every single change or every single card we’ve made in this new world, as a whole the cards are popular and seeing lots of play. That’s very exciting for any game design team!
So the new sets are seeing play and doing well for players at events. How are the new character designs in this world landing?
Characters
Our overall Character inclusion in Atlanta was pretty diverse. In total over 60 unique characters were played, with Smiling Titan seeing the most play at 15 total copies. In fact, here’s what the top 10 most popular characters looked like (note: there is a 4-way tie for tenth place in popularity):
When we think of “format health” relating to characters in Game Design we want to see a wide variety of characters being played, as well as no particular character seeing significant play over others. For example, in a 100-person event, if you have 51 characters represented by 50 people playing the same character, and 50 other people each playing a unique character, your format is probably unhealthy. Our standard benchmark for a character that may potentially be problematic is if it makes up 10% or more of the field. We didn’t see that in Atlanta, though Smiling Titan and Ryukyu are getting close, and are characters we will continue keeping an eye on as the format develops.
Another consideration we make in examining tournament data related to characters: are many different sets or products being represented? We were excited to see that nearly every single product designed for competitive players had some character representation at the event from Challenger Series to Booster Set. In our Top 10, the majority of cards came from the UVS Games design era, though we still see some earlier highlights with Toga 3 and Recovery Girl. This is a strong sign that modern character designs are landing with players, more welcome news just like the popularity of UVS Games set designs.
Continued Learnings
Atlanta Regionals brought a lot of good news, but we still have some work to do in adjusting designs to accomplish our goals. I was excited to hear that round times--something we were very concerned with in the pre-UVS Games era--have improved significantly. That was something we took conscious effort to design around, particularly in Attack on Titan, and our hard work seems to be paying off. There was even time for a lunch break in Georgia! Our stated goal of building Standard formats where a best-of-three match ends in 50(ish) minutes is going well.
One area we aren’t seeing the results we want: the prevalence of turn-2 kills. For healthy game play, our desired “key turn” for winning is in the turn 4-5 range. Both players have had a chance to establish their game plan and “do stuff,” so they can feel they’ve had some autonomy over the outcome of the game. A turn-2 kill fails to accomplish this as one player barely got their foot in the door and felt no agency over their win or loss. That burns players out and will turn new players off of the game entirely.
In Atlanta, we saw games ending on the second turn too frequently. For example, of our 30 streamed games, 7 of them ended on the second turn. For a healthy game balance we’d want that number to be closer to 0-1 for that sample size. We knew this would be a challenge to land because a lot of variables can be tweaked to impact this outcome, and we’ve been cautious about tweaking them to avoid going too far and risking breaking things. For now we’re looking at the cards contributing to this outcome the most, and may take action on those with bans or errata, while also looking at making more adjustments to long-term game designs to better deliver this goal.
Up Next
At the time I’m writing this, our OCE Regionals has just concluded and we’re beginning to examine the results of that event as well. In the meantime, I’m excited to see the work we’ve done over the past year is resonating with players and we’re seeing some great successes in the UVS Games era. However, we’re keeping our eyes open on adjustments we need to make to reach our next set of goals while continuing to grow UniVersus to ever greater heights.