Q&A w/ Oktagon - March Edition (5/1/19)

Brigby

Hi Everyone,

Thank you for submitting your questions from both March and April. We unfortunately weren't able to get the April questions compiled and answered in time for this combined Q&A, however we will be sure to include them in May's session.

We want to thank the following team members over at Oktagon for taking the time to review and answer the questions provided: Alain Valchera and Guilherme Evangelista!

Without further ado, here is the March Q&A Session w/ Oktagon!
----------
How does the team prioritize bug fixes?
We have a priority guide that divides issues into "priority tiers". The community reports are always classified as Critical and are moved to the front of the line. Crashes and freezes are classified as Critical too, but we depend on many factors to reproduce and fix those, because in a lot of cases the problem revolves around both hardware and software. Bugs in the events, leaderboards, or rewards are Critical too, but usually easier to fix than the rest.

Issues regarding cards and mechanics that do not trigger infinite loops or crashes have Major priority. TO be more specific, we start to fix cards starting from Masterpieces and ending with Commons because, in most cases, cards in higher rarities generate more reports (often due to complexity)

What kind of schedule can we expect for when new sets arrive?
The focus is currently on the development of War of the Spark, but we had some space for a new interesting event we hope makes the cut; if not for the War update, then for a future one. We'll have the new updated schedule up soon. We'll be able to present the major changes related to scheduling and events when the Seasons system is closer in development.

Is improving Duel Decks still the priority over Planechase and Draft?
Yes. The updated Duel Decks event is ready at the moment, and if its snags have been solved and everything runs smoothly now, then new duels should be available sooner and easier.

Can game settings be tied to Facebook login?
Sadly that isn't possible in the current conditions that player data is stored. It is something we would like to try, but we have yet to estimate how complicated this shift would be.

Does the team feel like event experience is in a good spot right now, or will there potentially be adjustments in the future?
Experience gaining and curve will be adjusted once the Season format is implemented, give that it provides other means to gain experience. For the time being, the current configuration is in place so that leveling provides somewhat palpable rewards.

Are there any additional plans to optimize and improve the speed at which the game functions?
There are many plans to optimize the speed of the game, although they often conflict with what we're able to render functional within its original structure. The loading times are our biggest concern, which pull data from different places and these often take longer than necessary to communicate and confirm these requests. It's being looked into, but will take a while. Regarding content such as animations and game speed, improvements in that area should fall in line with the Options updates.

What is the benefit of ending a player's turn after a certain amount of time, as opposed to other methods?
It's a little more about what we can and can't track in relation to turn data in the game. Our main focus was to return certain cards to their original (and written) effects, allowing them to be as powerful as they should while still making them possible to play, especially played against. But some things we can do, others just don't work and break the game.

For instance: counting the number of times a card has been cast before ending a process generated a problem with the library and drawing cards. It's not possible to disable specific cards after four copies had been cast, and disabling all cards after that would have made even stranger interactions. Given we can't have a judge to identify and rule over all MtG situations - especially in Legacy - allowing an amount of time for each phase is the only way of covering all current and future cases where interactions may take too long, and still conserve those cards' real effects and powers.

In regards to that though, we've been pushing for a different system; a player-controlled way of stopping loops and long turns; although other deadline constraints take priority.

Will Mana Runes ever be used for something other than leveling up Planeswalkers?
Probably not. While there were some concepts related to alternative uses for Mana Runes, they unfortunately didn't make the cut.
----------
Thanks for checking out the March Edition of the Q&A w/ Oktagon. If you have questions you'd like to ask them for the April + May Edition, then please submit them HERE

Kinesia

Thank you, some of those answers (while they will still get grumbling) are very informative and useful.

James13

Interesting.  I'd like load times to be a bigger priority than game engine speed, personally.  Encouraging to hear it's being looked into.

Also a minor personal gripe is that the game is unplayable unless an active connection exists at all times.  My train commute going underground makes MTGPQ fully unplayable as you get that connection popup preventing inputs during ongoing matches.  It would be nice if I could continue to play and just require a connection in order to "turn in" the match results at the end.

Mburn7

James13 said:

Interesting.  I'd like load times to be a bigger priority than game engine speed, personally.  Encouraging to hear it's being looked into.

Also a minor personal gripe is that the game is unplayable unless an active connection exists at all times.  My train commute going underground makes MTGPQ fully unplayable as you get that connection popup preventing inputs during ongoing matches.  It would be nice if I could continue to play and just require a connection in order to "turn in" the match results at the end.

Seriously.  It used to be that the connection I get in each station was enough to end a match and then start a new one.  Once my opponent loaded I could start a match even without a connection.

Nowadays with how long loading takes I can't do anything while I'm underground.  And there is an extra connection check after your opponent is loaded but before the match starts that has crashed my app a few times.

wereotter

Same thing with the subways except the stations here go under a mountain, so there’s no connection at all once it goes underground till I come back up.

Not sure if it’s the app or me phone, but a frustration I have is that if I’m mid-match, sometimes the app will restart itself when I come out of the subway, forcing a forfeit of whatever match I was in the middle of when the train went underground. 

Mburn7

Brigby said:

Hi Everyone,

What is the benefit of ending a player's turn after a certain amount of time, as opposed to other methods?
It's a little more about what we can and can't track in relation to turn data in the game. Our main focus was to return certain cards to their original (and written) effects, allowing them to be as powerful as they should while still making them possible to play, especially played against. But some things we can do, others just don't work and break the game.

For instance: counting the number of times a card has been cast before ending a process generated a problem with the library and drawing cards. It's not possible to disable specific cards after four copies had been cast, and disabling all cards after that would have made even stranger interactions. Given we can't have a judge to identify and rule over all MtG situations - especially in Legacy - allowing an amount of time for each phase is the only way of covering all current and future cases where interactions may take too long, and still conserve those cards' real effects and powers.

In regards to that though, we've been pushing for a different system; a player-controlled way of stopping loops and long turns; although other deadline constraints take priority.

This answer has been bugging me, and I think I finally figured out why.

It is definitely possible to track how many cards have been cast in a turn, since objectives have been tracking that since Events were introduced.

And we now know that it is possible to disable cards' effects and stop them from being cast, as the LPS system shows us.

So, why couldn't the devs just implement our proposed plan to put a limit on the number of cards that can be cast in a turn?  Setting a sufficiently high number (probably around 30-50) would make it impossible for "normal" decks to be effected and not effect strategies that involve a large number of buffs and triggered effects while stopping strategies like Naruduplicate and Expansion-BSZ from being instant-wins.  Note this has nothing to do with counting or disabling specific cards being cast, just count every card and then disable them all.

Looks to me like either the devs completely ignored our ideas or completely misunderstood what we were saying.

@Brigby I think a blog post on exactly how the LPS was designed is in order.  Show us the ideas that were proposed, how they were tested, and how the team determined that setting a timer on animation speed (which is what this LPS basically is) was the best way to go about bringing in loop decks.

Brigby

@Mburn7 Take this with a grain of salt, as I'm not a designer, but I speculate the difference between tracking objectives and counting cards is dependent on what exactly is being tracked. For an objective, the system just has to track if, for example, a card with the Support classification is being Cast onto the battlefield or a card that cost 10 mana or less is being Cast. Things like mana cost and card type are all static and generic variables.

In order to count the number of times a specific card in a deck is being cast though, the system would need to be able to track dynamic variables, since everyone's deck is different. That would probably explain why they mentioned the library and drawing cards, because it'd also be dependent on how that system operates as well. That's not even counting what might happen if you steal a creature from an opponent's deck and "bounced" it multiple times. 

My speculation is that in order to make the LPS count specific cards, they'd essentially have to rework the entire system for how fights are calculated and processed on the back-end.

As for a blog post explaining exactly how the LPS is designed and how it interacts with the system, while I like the vision and intent of it, my personal worry is that it'd be too much of "showing how the sausage is made," and could potentially lead to someone reverse engineering it for malicious purposes.

Mburn7

Brigby said:

@Mburn7 Take this with a grain of salt, as I'm not a designer, but I speculate the difference between tracking objectives and counting cards is dependent on what exactly is being tracked. For an objective, the system just has to track if, for example, a card with the Support classification is being Cast onto the battlefield or a card that cost 10 mana or less is being Cast. Things like mana cost and card type are all static and generic variables.

In order to count the number of times a specific card in a deck is being cast though, the system would need to be able to track dynamic variables, since everyone's deck is different. That would probably explain why they mentioned the library and drawing cards, because it'd also be dependent on how that system operates as well. That's not even counting what might happen if you steal a creature from an opponent's deck and "bounced" it multiple times. 

My speculation is that in order to make the LPS count specific cards, they'd essentially have to rework the entire system for how fights are calculated and processed on the back-end.

As for a blog post explaining exactly how the LPS is designed and how it interacts with the system, while I like the vision and intent of it, my personal worry is that it'd be too much of "showing how the sausage is made," and could potentially lead to someone reverse engineering it for malicious purposes.

I agree that tracking specific cards being cast may be tricky, but I see no reason given why the LPS couldn't have counted all cards being cast and just set the cutoff number sufficiently high to account for that.  I am concerned that there was some sort of miscommunication with ideas that led to the team exploring an unfeasible option and discarding it outright instead of a much more doable and fair option.

As for the "how the sausage is made" conundrum I don't see how its possible for someone to use the knowledge of the failed tests and balancing concerns involved for anything malicious.  And players have already figured out how to occasionally bypass the system anyway, so what further harm could be done even if there was?
(also, thanks for responding and indulging me in this discussion, I like a good back-and-forth)

Brigby

Mburn7 said:

I agree that tracking specific cards being cast may be tricky, but I see no reason given why the LPS couldn't have counted all cards being cast and just set the cutoff number sufficiently high to account for that.  I am concerned that there was some sort of miscommunication with ideas that led to the team exploring an unfeasible option and discarding it outright instead of a much more doable and fair option.

As for the "how the sausage is made" conundrum I don't see how its possible for someone to use the knowledge of the failed tests and balancing concerns involved for anything malicious.  And players have already figured out how to occasionally bypass the system anyway, so what further harm could be done even if there was?
(also, thanks for responding and indulging me in this discussion, I like a good back-and-forth)

I think on a conceptual level though, wouldn't setting a cutoff for the number of counted cards be essentially the same as setting a time limit? Some players will want it to count less. Some will want it to count more. 

A community suggestion that our team liked was putting control back into players' hands. That's why we plan on adjusting it into a Loop Control System, so players can decide when they want their loops to end, while still have a system that stops the opponents' loops from going infinite.

As for the blog post, if it's simply "this didn't work" or "that didn't work" then that's certainly a more feasible perspective to provide. I think at that point it'd just be a matter of finding the bandwidth for our team to potentially draft something like that up.

Mburn7

Brigby said:

Mburn7 said:

I agree that tracking specific cards being cast may be tricky, but I see no reason given why the LPS couldn't have counted all cards being cast and just set the cutoff number sufficiently high to account for that.  I am concerned that there was some sort of miscommunication with ideas that led to the team exploring an unfeasible option and discarding it outright instead of a much more doable and fair option.

As for the "how the sausage is made" conundrum I don't see how its possible for someone to use the knowledge of the failed tests and balancing concerns involved for anything malicious.  And players have already figured out how to occasionally bypass the system anyway, so what further harm could be done even if there was?
(also, thanks for responding and indulging me in this discussion, I like a good back-and-forth)

I think on a conceptual level though, wouldn't setting a cutoff for the number of counted cards be essentially the same as setting a time limit? Some players will want it to count less. Some will want it to count more. 

A community suggestion that our team liked was putting control back into players' hands. That's why we plan on adjusting it into a Loop Control System, so players can decide when they want their loops to end, while still have a system that stops the opponents' loops from going infinite.

As for the blog post, if it's simply "this didn't work" or "that didn't work" then that's certainly a more feasible perspective to provide. I think at that point it'd just be a matter of finding the bandwidth for our team to potentially draft something like that up.

The difference is it is an objective measure rather than a subjective one.  Since a timer counts for animation times different phones can get different length loops.  I have stopped using cards that give and abuse repeatable buffs, for example, because they will trigger the LPS extremely quickly due to how slow the buff animations are on my phone.
For example, Jace Unraveler of Secret's ultimate ability caused the LPS to kick in after I cast just 2 BSZ in one turn, whereas in a Brokhan Sunbird loop deck I can frequently cast 5-10 BSZ before the timer kicks in.  I haven't tried March of the Multitudes and Path of Discovery together because of how long and boring it is, but that is another example.  A different phone will have different limits, making it an inherently unfair and frustrating system.

If you simply count the number of cards cast than the animation time is irrelevant to when the LPS ends the turn.  You simply know that if I play 7 BSZ it will account for 49 cards that turn and thus end my turn.  Its easy to account for and easy to track.

As for a manual "end the turn" button that sounds like something that can be abused.  I can let my turn go on forever but stop my opponent's turn as soon as physically possible every time.  In addition to making it so that it is easier to build more and more degenerate loop decks, that will also make it almost impossible to lose to said loop decks, since the AI cannot end your turn the way you can end theirs.  Be very careful implementing asymmetrical effects to competitive gameplay.

Tremayne

@Brigby - fair enough that you are not a developer but your answer to @Mburn7 does not make any sense.

if for some reason the system can’t cope with counting all cards, then count cards that have a cost smaller than 200 or some other arbitrary value the system can handle which is also higher than any current cost of existing cards (right now 41 is the smallest value higher than any card)

It seems more and more like the developers are not in control of the source code for some strange reason. 

GrizzoMtGPQ

You guys aren't being serious are you? Most games use this pattern:

https://gameprogrammingpatterns.com/command.html

This is probably what they're doing and it's likely how Battle Log is implemented. They absolutely could count every card played. You can do it too! All you need to do is open Battle Log and count every time there is an underlined word. 

Seriously, computers are good at counting things and programming them to count things is as easy as it gets.

khurram

@Brigby I'd like for you to take into account that what mburn is suggesting was just his idea. The community did not have a big vote off where everyone agreed that limiting the number of cards being cast in a turn would solve all our problems.

Just because one player posts frequently or makes lengthy posts about something does not make that thing a depiction of what the mtgpq community wants. 

Mburn7

khurram said:

@Brigby I'd like for you to take into account that what mburn is suggesting was just his idea. The community did not have a big vote off where everyone agreed that limiting the number of cards being cast in a turn would solve all our problems.

Just because one player posts frequently or makes lengthy posts about something does not make that thing a depiction of what the mtgpq community wants. 

I mean, it was one of several ideas that were proposed here:

https://forums.d3go.com/discussion/75385/infinite-loops/p1

And it seemed to be met with a good bit of positive feedback, and was linked to several times in the bug forum in response to loop reports and dev responses.

But yeah, there definitely wasn't some massive community support for any loop ending issue (partially because it was in the suggestions forum, partially because a sizeable chunk of the player base doesn't seem to want to rein in loops, and partially because it is definitely not a perfect solution by any means)

But, I mean, if they were going to put in something that (as far as I remember) literally nobody suggested that would cut a turn short, why not just use the solution that at least some people suggested (that is objectively better) that would cut a turn short?

Unknown

This content has been removed.