Table of Contents
Hello all! Welcome back to this week’s breakdown of what’s happening over on the Marvel Snap official Discord! This week, we get answers to when to expect some news about card acquisition changes, did Arishem really get nerfed just for being popular, why did devs add a higher price season pass right now, and more! If you want to stay up to date with what’s coming and what answers developers have for the community, make sure to check back here each week!
This week’s edition comes after the launch of new Series 5 card Peni Parker! If you don’t know if this card is worth Spotlight Caches, make sure to check out our weekly Spotlight Cache guide, as well as our Bonus Challenge guide! Also be sure to read up on the latest OTA changes as many questions refer to the update.
Answers and questions may be slightly rephrased for more clarity and ease of reading. This week’s topics will be divided into Card Specific Questions, Other Questions, and Questions From You!
Remember, you can check out our new Marvel Snap Developer Tracker to see all questions and answers in real time! Topics can also be searched, filtered, and sorted.
Card Specific Questions
Q: This season you’ve created cards that fall into territory with card design a lot of people assumed you never would. Doom 2099 interacting directly with Doctor Doom, Rocket and Groot not only being a team up card, but 2 characters already in the game, and Bruce Banner not only being Hulk, but also having percentiles on the card. What lead you to these designs and character choices?
A: Glenn “I don’t know that I agree about some of this space being so unexpected—we have Jane Foster interacting with Thor, Symbiote Spider-Man as another Spidey, and Danger Room using percentages. But certainly Rivals was a unique partnership and challenge.
The most obvious challenge was that their character lists overlapped a lot with characters we already had, so we would have to pick spots to innovate upon. We also wanted to make choices that highlighted Rivals’ own strengths, such as their team-up mechanic and Hulk’s transformation, but in ways that made sense for our engine.
Some of it was also experimentation—will Rocket & Groot open the door for future partnerships? Will Doom 2099 demonstrate an appetite for a larger number of namesake two-card combos in our system? Time to find out!”
Q: Could we see Nocturn return to a 3/5?
A: Glenn “Maybe, but probably not. Nocturne was the default “solid 3-Cost card” for a while, and since her change we’ve seen more diversity in that slot which is a healthy thing. Our goal for R&G is similar, to create another option without overwhelming the rest.”
Q: What is the status of Kang? After all of this time, it seems strange not 1 idea has been at least tried out (like Zabu).
A: Glenn “We’ve tried several, but not found something we like for that update. However, redesigning Kang isn’t a top priority relative to new seasonal content, game modes, balance analysis, and other critical work. Zabu was in a similar spot—he was just easier to solve for us.”
Q: The OTA mentions Arishem meta share is higher than would be ideal, but it’s somewhat self evident that Arishem and card acquisition are intimately linked. Anyone can build an Arishem deck with Arishem and their best 11 cards. Does the team agree this might be a factor to his high playrate?
A: Glenn “Certainly. And as the notes mention, Arishem got a longer leash than many decks do because the deck doesn’t homogenize gameplay as much. We also recognize it’s popular because it is fun.
But Arishem also has held firm at a play rate that is unprecedented relative to other decks—some weeks it has been double the next highest deck in games played on our primary balance slice. Ensuring a diverse metagame has always been an aspect of our balance philosophy.”
Q: Hela and Arishem were both killed because their playrate was considered too high. How do you balance meta diversity with killing a deck due to popularity?
A: Glenn “We don’t believe either of those decks have been killed. And as we’ve said many times, our balance changes aren’t aimed at killing any cards, but rather ensuring they live in a healthy range. If we nerf something too hard for that, we’ll often revisit the card and ensure it finds that spot.
I don’t think there’s such a thing as a card being too fun, but it can definitely be too popular. We’re all playing this game together–if half of a player’s opponents are playing the same deck, that’s going to negatively affect the fun for that player. That’s why diversity is an important component of our balance responsibility, and also why we never aim to “kill” anything.
(To be clear, Arishem wasn’t remotely close to a 50% average anywhere. But I’m sure some players played days where that frequency was true for them, given how high it was and how numbers distribute.)”
Q: (Follow Up) If, in theory, a card has horrible stats but has a massive playrate, do you still nerf the card for playrate? And is there a point where you say “we can’t nerf this card anymore?”
A: Glenn “I don’t think the hypothetical point really exists in practice, but to answer the question: play rate is just as important a metric as win rate and cube rate. We look at all 3 constantly, as I established in the first balance changes I wrote for Zabu and Silver Surfer in 2023. It’s not a tolerable state for a fun metagame to have a deck be 25% of the games played any more than it would be to win 65% of the time or average a cube per game, and we balance accordingly. (These are all hypothetical numbers, no deck has hit any of them.)”
Q: Would you host a competition where players submit their best Kang rework ideas, and the best one gets pushed to the game?
A: Glenn “While developers vary, I have an ethical issue with using unpaid design work in-game. I’m stoked that people love to make their own SNAP cards, but the design team doesn’t review those.”
Q: Are the nerfs to cards like Typhoid or Hela effective when you can just negate the effect with Luke Cage?
A: Glenn “Yes, these nerfs are effective. Increasing the intensity of any downside places more significance on enduring it or mitigating it. For example:
The best Surtur deck according to the data wasn’t playing Luke Cage. Are they likely to start? Probably not, given how important Surtur and Cosmo are. So they’ll either have play through the -2, change Typhoid Mary for a weaker option they weren’t playing before, or add Luke and weaken the deck in another way to make room.
The Hela deck is already quite reliant on drawing Hela. Adding another card they need to find increases the vulnerability of their combo. We often call this kind of dynamic a parlay, because you need multiple things to go right.”
Q: The OTA mentioned Surtur decks were “the clear best deck”, but most high ranked players would disagree. How do you view your data to get this conclusion?
A: Glenn “We use a variety of different slices for balance reviews. The ones we use most are all players, the top ~0.5% on MMR, and Series 3 complete players. We have different weights for these to create a complete picture, but the one we weigh heaviest is Series 3 complete.”
Q: With the change to Arishem, Loki will likely disappear. Would Loki be getting a possible buff due to this?
A: Glenn “We’ll be reviewing the data. It’s not our goal to make any card disappear, and if Loki merits an adjustment we’ll make one. I lean towards it being likely.”
Q: With Arishem’s new update, there is the new “Game start” keyword. Will other cards like Thanos and High Evolutionary be getting that keyword too?
A: Glenn “We’re planning for these cards to receive their update with the next patch.
For clarity, these aren’t really keywords—it’s a text change that saves space and helps them read better, we think. We’re likely to add mechanical weight to it later, though.”
Other Questions
Q: How does the team handle the disparity between what cards people are complaining about vs. what cards are actually overperforming? If everyone is complaining about Card A because they think the meta is “solved”, but Card B is actually turbo busted and no one is talking about it – what do you do?
A: Glenn “Players’ complaints provide a lot of useful context that the data can’t. When we’re trying to decide how to change something, understanding the nuance helps us identify more satisfying changes. For example, we might find two nerfs that accomplish our goal—but which part of a card’s strength is aggravating people most? Which parts might the people using the card want to keep?
We won’t always get that right, but by monitoring sentiment we can better measure our effort and produce better results in the future.
As for your examples, we would usually let Card B sit until players also understood what was happening. Nerfing a card players haven’t found strong feels weird, and discovering that strength would feel good. If players are complaining about Card A but our data indicates it’s not problematic, we’d consider the context more closely. What is the bad experience? Where is the feedback coming from? Is it localized to specific kinds of players? That kind of thing.”
Q: Does the team feel the monthly updates are working in a sustainable way?
A: Glenn “Small caveat: we update every 4 weeks, not monthly.
Regular patches are a part of any live service game. Their purpose is to address changes that require a client update. Many of those changes will be bug fixes, VFX additions, etc. Features, as large releases, always require a client update at some point, which means we will use patches to deliver them. But the primary purpose of a patch isn’t to deliver features, it’s to deliver ongoing game improvements—features are a subset of that.
Our cadence isn’t particularly unusual—I’ve seen 4, 6, and 8 weeks often used. I’m sure our engineering team has selected the cadence they most prefer—as I’m not an engineer, I couldn’t comment much beyond that.”
Q: With community sentiment right now, did you think it was a good idea to introduce a more expensive season pass with this specific patch?
A: Glenn “We lock patches a month or more before release due to the necessary testing and approvals stages with various app stores required to deploy them. So they generally will never be influenced by the weeks immediately prior to their release, and will largely contain the elements we are functionally ready to add to the game at that time.
I understand there are pain points the community would like addressed, and we are working actively on many of them every day. There are different prioritizations the team could have considered in 2024 for some of these cases—hindsight’s always 20/20. But we have also released several significant features and game improvements during that time, worked on many that aren’t quite ready for release, and even worked on some that ultimately had to be scrapped (and in the case of Character Mastery, resurrected by community response).
But trying to constantly develop in direct and speedy reaction to community sentiment would ultimately mean less features made, at lower quality. That’s not to say we don’t value the feedback and want to make the community happy (again, see Mastery)—but it is a practical reality. Imagine you’re building a house, and the client comes in every few weeks to change the layout, which side of the bathroom has the sinks, where the lights go, etc. These changes may seem simple to the client, but they’re not responsible for wiring the home, assigning the best tasks for each worker, accounting for code changes, getting other houses built, etc.”
Q: What do you think about niche decks disappearing from the meta?
A: Glenn “Underperforming cards and strategies are often considered an opportunity for future reworks or new designs. Attuma and Scream are both good examples of that in practice.”
Q: Are the next Series Drops in March of 2025 showing a future pattern of Drops every 3 months?
A: Griffin “The planned Q1 Series Drops are a one-time thing with the goal being to have an update to the entire Card Acquisition system after. This could mean the need for Series Drops as they are now would change entirely.
As we get closer, we’ll update everyone on our plans well in advance.”
Q: There are often large gaps in time between feature announcements and actually going live. Are there ways you can gather feedback in these time gaps about the upcoming features?
A: Glenn “We do solicit feedback on many new features from external groups under NDA during these interim periods. Generally, features tend to lock about 6 weeks before they go live.”
Q: If new features are often locked 6 weeks in advance, why are they announced sometimes just days before launching?
A: Glenn “It can vary, and ultimately is the discretion of how our marketing group wants to stagger things. In my own experience, announcing details too early distracts from what’s currently live, gives people time to forget what you announced (so you need to do it twice), and creates confusion due to a lack of in-game context for the information.
Delaying also avoids some other inevitable risks, like patch delays, unexpected bugs we find late, localization complications, etc.”
Q: What would you define as a new “feature”?
A: Glenn “For my purposes, anything that’s additive to the game itself—not just additional content for existing structures—is a feature. The lock points are basically all the same since they’re a function of QA and approvals.”
Q: With the recent announcement of card acquisition updates that the team is working hard on, do you feel that the unique card collection philosophy has worked out overall?
A: Glenn “I do think it remains meaningful to SNAP. We just don’t have as many cards as most games, and deck sizes are much smaller. That means the path to optimization is really short at various points in the player journey without taking steps to encourage and reward more diverse play choices.
Q: Have you gotten feedback that players enjoy unique collections?
A: Glenn “Yes, we do have positive feedback around crafting different player journeys via gameplay content. There are also negative gameplay outcomes we’re avoiding—it’s harder to recognize those in their absence, but they exist.”
As we’ve grown, I believe we’ve seen that element needs different tuning for longtime players and in the context of significantly more endgame content. Excited to share more about that as soon as we can.
I don’t think this need is especially related to live service, but releasing new content on a weekly frequency definitely makes it matter. That aspect remains a core goal of SNAP’s play philosophy.”
Q: Are casual players a big consideration when adding new content to the game? If so, how do you keep a balance with the hardcore fans so that everyone can have fun?
A: Glenn “Yes, they are. We believe SNAP’s accessibility is core to its appeal. The balance is delicate—we design cards across a range, with some offering simpler quests (Galacta, Doom 2099) and others more complex to evaluate (Luna Snow, Peni Parker). Ditto game modes, though we’re early in that process. Each one has taught us a lot about what players want to complement the core gameplay.”
Q: How does the team define or think about transparency as it pertains to community management?
A: Griffin “Transparency, in regards to Community Management, is a goal in everything that we do and share. However, there’s a balance that needs to be maintained as TOO MUCH transparency can set both devs and players up for disappointment.
A great example is the Dev Roadmap. Our goal is to share everything we’re working on. The realities of game development mean that it’s inevitable that projects can change, be delayed, get shelved, or entirely scrapped.
We want to reach a certain level of confidence in shipping a feature, product, or update before we introduce or announce it to players. However, there’s also a lot of value in early feedback. This is where, as Glenn has outlined before, we introduce early designs and concepts to groups of players and creators under NDAs so we can get trustworthy and honest feedback while we iterate.
TL;DR: We think about this balance a lot and there’s no perfectly right answer that applies to each and every scenario.”
Q: Does the team track player sentiment on hot locations?
A: Glenn “We do track them. Barring a few small exceptions, hot locations have neutral-to-positive impacts on our metrics, which is to say that players play the same or more when they’re active. When we identify one that goes poorly, we stop using it. We also see a mix of feedback on social media.
I imagine we will experiment further in the space, though.”
Q: With new game modes being introduced, Conquest remains untouched since launch. Does this mean the mode has been abandoned in favor of new modes?
A: Glenn “I imagine the rate and significance of updates will vary based on the specifics of each mode. Right now, our focus is on learning which game modes players enjoy most and why, which will in turn inform their future.”
Q: After the series drop announcement you have been incredible coy about when we can get news about improvements. “We’ll share more news when we are ready.” Being holiday season, can we at least get confirmation that we won’t get any news at all during December?
A: Griffin “I’m pushing to have it ready by next week at the latest. That’s not a guarantee but it’s my goal.
I want to be clear: this will NOT be some sort of instant ‘miracle’ plan or fix.”
Q: Peni Parker looks like she is severely underperforming, but she didn’t get an OTA buff. What are your criteria for buffing or nerfing a card quickly?
A: Glenn “Well, the card has to be out to get a buff—we lock OTAs a week before they go live, so Peni Parker didn’t have a winrate to reference at that time.
Generally speaking, we want to see about 2 weeks of data before making a balance adjustment. That’s because we want to give players an opportunity to solve their metagames—it’s a fun challenge—and if they can, we may not even need to make a change. If we make a change faster, it’s likely because the data is exceedingly clear.
We’ve stated in the past that we’re committed to new cards finding healthy play rates. So if a card releases too weak, we’re very likely to adjust it. We want everyone to be happy with their new content, which means it needs to be viable.
As for nerfs, we mostly just want confidence. The buffer helps there too, because it gives players time to optimize the deck and helps us ensure we aim at the right targets.”
Q: How do you aim to balance newly released cards pre-release versus post-release?
A: Glenn “I think erring strong in testing is better, certainly. It is easier to pull back than add on strength. However, the primary reason to err strong when in doubt for releases is that we’ll miss high and low in those cases. If we miss too low, then the content fails to impact the metagame, and it’s not really content.”
Questions From You!
Each week, I ask readers of the weekly dev update to leave their questions down in the comment section. That way we can ask those questions on your behalf, or let you know the answers if those questions have been asked before! I read each comment you leave, so I will be adding this section at the end of each week’s edition to highlight your questions that you asked last week!
We didn’t have an update last week due to the Thanksgiving holiday, but almost all of you voiced your thoughts on state of the game in the last edition. Your questions and concerns are flooding the discord, so they are definitely reading your thoughts.
For any questions you may have regarding card acquisition changes, please know they very likely won’t be answered by devs at this time. I know that is most player’s main question they want answered, but keep your voices active even without answers. If you have any other questions, be sure to also leave those in the comments.
Q: It is possible that there will be more cards with no ability, and would they work with High Evo decks?
A: Glenn “Not at this time. [If we did make a new no abilty card] we’d think about it then, but lean no with the possible exception of a 7-Cost”
Keep Your Questions and Feedback Coming!
That’s all for this week’s update! Be sure to check back here at Marvel Snap Zone for next week’s update! If you enjoyed the amount of content in this edition, make sure to keep asking your questions to the developers by submitting it in their official Discord in the “#ask-the-team” channel. If you have questions and don’t use Discord, leave your questions for the developers in the comment section here, and I will make sure your question has been answered by the devs!
If you have feedback or changes you would like to see with these weekly updates, also let us know in the comment section!
More Content