Table of Contents
Welcome back to this week’s Developer Update! This week’s edition comes after the launch of the new Series 5 cards, new season, and the developer roadmap! If you haven’t seen the latest update on the future of Snap from Second Dinner, be sure to check that out!
If you don’t know if
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!
See all individual questions and answers updated in real time!
Card Specific Questions
Q: With so much energy ramp in the game now, why does Arishem still wait until turn 3 to get 1 energy?
A: Glenn “Rather than require setup, Arishem’s balancing element is a penalty that affects the kinds of decks that can use him. It’s a different constraint, ergo it’s a different effect.”
Q: Does Snap have it’s own lore? Last I checked, Isca was not Apocalypse’s wife, but rather his sister in law.
A: Glenn “Just a misunderstanding in the marketing copy. Sorry!”
Q: Did Star-Lord’s overperformance lead you to releasing En Sabah Nur weak?
A: Glenn “Nothing about this season would’ve changed based on Star-Lord. It was already done and in the process of approvals by the time he was released.”
Q: Why haven’t you created any Anti-Destroy cards? For example: an End of Turn card that banishes enemy cards that enter the destroy pool to better counter Knull and Death.
A: Glenn “There are several meaningful counters to both of those cards, nor are they particularly strong performers at present. However, they can make for some splashy and memorable endgames.
We may explore the design space you’re describing in the future, but it seems pretty unlikely to me that it will be because these cards are too strong.”
Other Questions
Q: How often does the tech dev team check the game for any deck shuffling bias?
A: Glenn “We don’t actively check the algorithm, but we have a comprehensive testing suite that ensures (among other things) randomization remains within acceptable parameters. These tests are automatically run whenever any change to any part of the game is made, so failures get tracked immediately.
To any CS students out there: Write your tests! Tests are great!”
Q: Why isn’t is it common for studios to self publish like you have? Has there been changes that players could notice since you took over?
A: Glenn “It’s uncommon because it requires an additional layer of investment–financial, personnel, technology, etc. There are services that exist to do this, and the help of an experienced publisher can be a huge boon.
It’s not dissimilar from the way movie studios often handle the marketing and distribution of a movie, while the director and crew are responsible for making it.
There aren’t a ton of direct and clear player-facing elements that publishing handles. The big ones are things like the webshop and advertising. Much of the rest is behind-the-scenes work. It had a larger-ranging impact for us because we had to develop the capacity to do it, growing the company and adding core expertise that we didn’t have before.”
Q: What are your thoughts on the public data that shows a downward trend in players reaching infinite each season?
A: Glenn “To my understanding, the data asserts to show the percentage of a small set of players (~1400) achieving Infinite over time.
The data asserts that these specific players are hitting infinite less often over time, which is to be expected. Among any sample, that would be true for any threshold achievement in a game over time; players take breaks, quit, miss, find new goals, etc. and the other players can’t break the threshold more than once. So a downward slope is typical.
I can’t speak to the precise reasons for any player; I can only look at different sets of data and correlate outcomes. I look at this chart and think it reflects the influx of months with popular/multiple LTGMs is providing these players with goals they’d rather pursue than Infinite, and they’re not increasing their play time to do both. Generally when a game provides choices and players choose the new ones vs. doing the old one by default, that indicates a more satisfying option is now available to them. There’s more to do, and they’re picking what they value.
If the data was mapped against the sessions/time played for these players and also showed changes there, that could indicate otherwise. But it isn’t, and ours doesn’t indicate concern.”
Q: Is there is a “core” group of vanillla locations that always have a fixed % while newer ones share a % that dilutes with every new location release?
A: Glenn “We do have some distinct groups, but that really only affects the Recruit pass and lower ranks. Everyone else is playing within the same frequencies.
Locations can have different weights, which are essentially rarities. So some of them will be more common than others. And newer locations are more often high complexity or high impact, so they more often get assigned to that higher rarity weighting. But we don’t have specific buckets for ‘old’ and ‘new’ locations.
The structure of the weighting system we use does necessarily mean that every new location slightly dilutes the pool for all other locations.”
Q: It seems like the new Objective keyword has a unique icon next to it to highlight the keyword. Could this icon system be implemented in other keywords, or is it unique to Objective mechanics?
A: Glenn “It’s unique for now, but could be expanded.”
Q: Will we get an update to the card shop to better target cards we want?
A: Griffin “We don’t have any short-term plans for updates to the Card Shop right now. We’ve heard the feedback that players would like a better way better target individual cards beyond waiting for them to pop in the rotating spotlights. Discussions on options to improve card acquisition are always on going here, whether that’s to let players catch up or trying to get that one card they really want.”
Q: Is it possible we ever get any of these features: Friend list, replays, card/variants trading, or being able to customize friendly matches?
A: Griffin “While these features aren’t at the top of our priorities right now, they absolutely have been discussed, with the exception of Card/Variant trading.
They’re all really fun ideas and we’ll keep talking about them as we continue to work on Snap going forward. Keep your ideas coming for future content coming!”
Q: Do you track which variants are used most frequently?
A: Griffin “We definitely track the popularity of cards and variants. That’s a key piece of info we use when it comes to planning future content. The art team here at SD does a great job and trying to get all kinds of different art types for Variants and when one really ‘pops’ they do their best to continue creating more.”
Q: Do you ever have competitions (internal or between artists) to see who’s variant is more popular?
A: Griffin “I’m not aware of any sort of competitions like that. Either internal here or with artists. It’s all viewed as a great big artistic feast for the eyes and everyone is a supremely talented chef cooking up their specialties.”
Q: Why does the “fill all locations” bounty trigger if the match was terminated from a disconnected player?
A: Glenn “Sounds like a bug. My guess would be the game is ending before the location slots are instantiated, but the endgame achievement check still runs and sees no empty slots.”
Q: Does the team keep a list of card abilities that are being saved for the right character match?
A: Glenn “Kind of. We keep an archive of interesting designs that get developed over the course of any season, and keep an eye out for future opportunities to use a design we liked but didn’t ship for whatever reason.”
Q: Is reopening the Japanese webshop among your roadmap plans, but not included update news?
A: Griffin “Yes! We’ve made significant progress. Hopefully we can share firm dates soon.”
Q: There are locations that Ongoing and On Reveal decks have to play around. It’s a bit unfair that these newer (and often very powerful keywords) don’t have to deal with that, no?
A: Glenn “We want to be sure there’s a meaningful enough concentration of these effects in the game before introducing such a location. Otherwise the location is just Ruins too often, which isn’t fun and doesn’t serve the primary goal of locations, to create dynamic gameplay.
I don’t think it’s particularly unfair, no. Locations like Knowhere and Isle of Silence more often add texture to games than have a particularly large impact on the outcome. Their range of impact and appearance frequency aren’t meaningfully nerfing any archetypes.”
Q: What exactly does reporting a bug in-game do? And how many bug reports do you get in a day?
A: Glenn “The in-game Bug Report codes provide a snap shot of your account along with key data of what happened to you recently in your games. It allows our teams to easily investigate individual issues by narrowing down the parameters.
It can take a single report to identify an issue but a single bug may have multiple vectors or effects which could require more than one report to see the entire impact. Also, identifying a bug is just the first step. Actually fixing it can be quick or it could be a larger undertaking.
I don’t have the exact numbers, but it’s under 1000 a day.”
A: Addison “To build on what Griffin said, when you submit an in-game bug report we get lots of ‘stuff’ with it.
Sometimes the value in the report is in the custom message players fill out (ex: X Card keeps Aw Snap’ing). Often times though the value is in the client logs that get sent with the report and some of the player telemetry we receive as well.
An example of something we do pretty frequently when we’re trying to reproduce an issue locally that Griffin referenced is we’ll pull a player’s accountID from the bug report they sent and then download what we call the player’s ‘Snapshot’ which is a copy of that player’s state and data. We can then apply that snapshot to a test account on a test environment to see if the issue reported simply reproduces by virtue of the account data itself (independent of any behaviors or device/OS stuff).
Additionally with just the accountID attached to the bug report, we can check all automatically reported exceptions sent from that account within a given period of time as well. On the client side, one of the things I like to do is check an accountID’s exceptions on a visual timeline and see what exceptions might have fired immediately before their manual report because sometimes that can be a clue as to what went wrong that prompted their report to begin with (though other times it can be a wild goose chase that leads to nowhere!).
As to how many reports it takes to bubble something up, sometimes it’s very few because the issue reported and the indicators that we see are incredibly obvious. Other times it can be more tricky and so we’ll look for trends in the reports over a given time. If we see a hundred reports of a certain card behavior acting funky, we tend to look at that and go ‘huh, there’s probably something here!’This can get a bit tough because often times people will use the manual bug reports as a way to send us…colorful messages, so we have to filter through those lol.
And to echo Griffin on the numbers, it can range from a couple hundred a day to a couple thousand generally. We tend to see big spikes with the release of a new feature or the start of a new season, but most days are on the lower end.”
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!
Q: Do bots use the newest cards?
A: Glenn “Bot decks are generated automatically from the metagame, and they learn to play their decks over time. They don’t use actual games being played on live or any matchup data to do this–they teach themselves.
…We [do] warp their decks. They don’t build decks themselves, so without guidance they’d just pick the same version of each deck every time. We force them to diversify builds to provide a variety and different strengths/weaknesses in the pool.”
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