Hello and welcome to
another episode of Radio Royale. – I’m Cip.
– And I’m Bangs. And we’re here today with the man,
the myth, the legen… dary? – Stefu!
– Hi-o! Stefu, who are you?
Have we seen each other before? – We work in the same team.
– Oh, right! You’re that guy
that sits next to me. Oh, yeah! – I sit next to Bangs.
– Three desks over. – I never look there.
– He was familiar. Yeah. So, who are you again? I’m Stefu. I work as a game designer
in the Royale team. Right. What does that mean? I mostly work on the balancing, and the system design and game economy. But we’re all game designers in the team.
That’s how we roll. We all discuss all the design questions
very openly. Stefu, I know you had
reservations coming on the show, because this is a very touchy topic
for the community. Yeah, it’s a tricky topic for me.
It’s, of course, very close to my heart. This is what I work on.
I do take it very seriously. – You seem serious.
– Uptight, almost. Uptight, yeah. I just wanted to let you guys know
who’s listening he’s not that way. He’s a pretty fun guy. Sometimes. – Definitely.
– Definitely. He walks around the office
like a Royal Giant. Yeah, yeah. I’m a Finn.
I’m mostly naked in some way. – Playing Clash Royale.
– Playing Clash Royale. – That’s going to get the burn somewhere.
– Oh my God! We’ve got lots of questions
on the forums and reddit to ask you, Stefu. But I think before we get into that, it’d be awesome if you could give us
a short overview of your GDC talk. Yeah. The highlight reel, as it were. It’s a 60-minute talk and even that
had to have content cut from it, because it’s so much to talk about
when discussing the metagame myth and balancing of Clash Royale. But if you could summarize
that content in 3 minutes… – Oh, wow.
– Go! Oh my! Okay, so… The title was
“Quest for the Healthy Metagame: Balancing Cards in Clash Royale.” So, first, what’s a metagame?
I define it like this: “The card’s combinations and strategies
that the community at large have found most effective.” What kind of control do we have
over the meta? Balancing is the only tool we have
to try and affect that but it’s the community and the players
that create the metagame. What they choose, what they play,
that’s what creates it. Sometimes it’s not balance changes
that cause changes in the meta. Sometimes you can see that… The Shanghai tournament
was a good example. Yeah, that’s true.
We’ve seen, for example… In the Shanghai tournament last summer, the top 8 on stage, almost all of them
had a Giant Poison deck, and that really started
the infamous Giant Poison meta. What’s an example of a healthy metagame? So healthy metagame, which is the goal, that’s where the game overall feels fun. There’s a variety of different strategies, there’s many different deck archetypes
that are popular. This variety keeps it fresh. If you face the same deck every time,
it can be… That’s the main thing,
where you really see it. When you see the same thing all the time. – That’s the most boring thing.
– Variety is the spice of life. Yeah. So how do we choose what to affect when we think about balance changes? It breaks down to three things.
We listen to the players, we look at data, we look at the metrics,
but at the end of the day it’s really about intuition, and trying to figure out what works,
what doesn’t, what should be changed, what are the effects
and ramifications of different changes. What’s the key metrics you look at
to judge the health of the metagame? For metrics, we mostly use
use rates and win rates. We check that from different arenas, we check from top players,
we check from top tournament players… So these use rates… Does every card have to have
the exact same use rate? That’s a good question. Good point.
Because we have 71 cards right now, after the Bandit that has been released. That means that…
Maybe I should do the math for this, but it means 11 billion
possible deck combinations in the game. It’s not really realistic that everything
would be used perfectly evenly. Obviously, a very versatile card,
like the Zap or The Log, will have a higher use rate just
because it fits almost in any deck. One thing I got from your talk was that cards like the X-Bow,
which is, as you said, a “spicy card”, should have a lower use rate.
Can you explain that a little further? Yeah. The idea is that there are cards,
like the X-Bow, that we consciously didn’t want to have
an insanely high use rate. But we still wanted it to be
a viable card and option, so the X-Bow is actually quite powerful
if used correctly. But it requires a lot of skill
to actually use it well. A good way to see if a spicy card
is in the right place, is that it has a low use rate,
but it has a pretty high win rate. It means that not that many are using it,
but those who specialize in it, and those who have high skill with it
can find success with it. If you want to find out more information
about Stefu’s talk, be sure to check that out.
It may be a little more technical, so feel free to view it if you like.
If not… – We’re moving on to questions!
– Yes! My favourite time. Question time!
Did we receive a lot of questions? – We received all of the questions.
– Five thousand questions! So for this question,
I’m going to just summarize, because about 4 thousand of
our 5 thousand questions have been on this subject. And there are two cards that
are the most controversial in our game. Certainly controversial in the community. – Can you guess what they are?
– My guess is the Knight and the Valkyrie? I don’t know.
Stefu, do you know? I’m trying to come up with
something funny… It’s the Royal Giant
and Elite Barbarians, right? So, why are they so controversial? What is it about the Royal Giant
and Elite Barbarians that people get so enraged about? Or passionate, perhaps? They are a really tricky pair of cards. First of all,
they are quite strong offense cards. And we are, in general, quite reluctant
to nerf offense options, because we want the game to have
several different cool offense options. They really put the skill test
to the defender. They are easy to use as an attacker,
but the defender has to defend them well. Yeah, and certainly for Elite Barbs… A game developed where you put Elite Barbs
and then someone else puts Skeletons, and you predict they will put Skeletons
so you get your Log… It’s this kind of back and forth… The kind of counterplay you explained
is quite interesting, but for Elite Barbarians,
it’s a bit tricky when… If you defend very well,
you get no damage at all, but if you make a small mistake
then your tower is down. Yeah, and that perhaps
leads to frustration. But another thing is
they’re are actually highly used only at a certain trophy range,
roughly 3,500 to 5,000. But it’s a fact that at that level
they are used quite a lot to the point that it’s actually
quite unhealthy for the game. I guess we kind of measure the health
by the use rates specifically. If you face something all the time,
I call that an unhealthy metagame. You want variation,
and different strategies to be prominent. The Royal Giant and the Elite Barbarians
are used way below average in the top 100
and also in the top challenge play. So this makes it a really tricky topic.
Should they be nerfed or not? It also highlights that these cards
rely on over-levelling for their success. Because if you are at level cap,
like in a challenge or if you are in the top 100 on a ladder,
it doesn’t appear to be a problem. Yeah, it’s easier to upgrade these cards
because they are commons, so that makes them more appealing
for many players to focus on them, because it’s easier to get them
to the maximum level. It becomes a tricky problem.
Is the problem with the game economy? Or with the cards themselves? We’ve explored things with the economy. We were thinking about changing
the rarities of the two cards, but that becomes quite complex. What if we changed them to rares? What would be the conversion? If we change a max level card to
a max level card, nothing really changes. If we would make the conversion relative we’d actually downgrade
Royal Giants and Elite Barbarians, and that’d be quite nasty for players
who have actually upgraded them. The solution that is fair for our players
doesn’t also fix the problem, because they are already over-levelled, and it means that new players,
or players who are rising, don’t have this opportunity
like others players do. Another thing that we’ve explored
is capping cards at King level. The more we were exploring this,
the messier it started to feel. The idea was that you can’t use cards
that are a level higher than your king. It may have made sense if it was done
from the beginning of the game, but to this established live game,
that’s pretty tricky to do. I feel that maybe that sort of change
might mean the players get stuck, because, at the moment,
you can progress because of your skill, or because of your card levels.
It’s a combination of those two things. But if we were to cap
the levels of the cards, you can no longer progress
with those cards in the game, so you have to rely on your skill
to push to the next arena. Many players have relied on these cards
to get their trophy levels high, so, if we suddenly
cut that away from them, they’d plummet in the trophy levels, because they wouldn’t have
any alternatives to use. It’s a lot bigger thing to do
than just simply nerfing the cards. The important lesson that we’ve learnt
for future cards releases is about when we release a card
and it’s a common card, what’s the danger with over-levelling? We definitely need to be more aware
of this than we were formerly. This wasn’t an issue
early on in the game. Now that the game is one year old, it has become an issue, because there are
more players who are at that state. What we are currently looking at
is simply doing some direct nerfs. What are those nerfs? Right, for Royal Giant, we’ve decided to increase the deploy time
from one second to two seconds. Why is that a solution?
What things did we try? We tried many different things:
the damage, reducing the tile range too, the hitpoints,
but they all felt a bit wrong. And there’s also the fact that
the Royal Giant isn’t good in challenges and in top play
when you have an equally levelled card. The thing that makes the Royal Giant
nasty to play against, is when he’s placed at the bridge
and he immediately shoots at your tower. Or the situation where one tower is down
and you place it in that little pocket and it immediately shoots at your tower. By increasing the deploy time, the defender gets a bit more time
to react to the Royal Giant, but this might actually
slightly buff the playstyle where you placed the Royal Giant
behind your tower… It feels there might be
more skill in making it useful, so if you are placing it behind…
like from the back, it’s not about the Royal Giant, it’s about the support troops
you can place with that push versus placing it in the bridge
and get chip damage. But you can still use it quite well
at the bridge or close to the bridge to quickly tank for example
against the Mortar or an X-Bow, and then counter it… And especially when you catch
your opponent with less Elixir as well. It still becomes viable… Now that we’ve taken care of the Royal G,
let’s talk about the Elite Barbs. What are we doing with them? They are getting
a little bit of a haircut as well, so hitpoints are reduced by 4% and the time for the first hit
is 0.1 seconds slower. This makes them slightly easier
to deal with when defended properly, so they don’t remain alive when you
have put like everything against them and then they still come with this level
of health and destroy your tower. They are still a real threat
when they get through but just slightly easier to deal with. It’s kind of cool to do this,
then we can see the behavioral changes, and see the card use is changed. Perhaps that leads us
to the next set of balance changes. Something I want to add
is that when we nerf cards, we never want to kill
those cards completely. The idea here is not to kill
Royal Giants and Elite Barbarians, and remove them from the meta
or from the game, but ideally they are
a bit more tolerable to deal with and maybe their use rate
also could go down a little bit. PekkaridingDragon asks “If you were a card which one would you be?”
He said, he would be a wizard. I’d be an Elixir Pump. Just chilling… at the back. Just collecting. – Just laying on your couch.
– Just pumping away in the back. Luckythunder asks “If there’s one card
you wish you could make better, “what would it be
and how would you rework it?” But actually I feel,
Sparky is in a funny place right now. I think
when I was progressing through the ladder, Sparky was always a difficult one
to deal with when I was… It’s always scary. It’s like “Oh no!”. Sometimes I’d be like “Good game!”. Then I started learning ways counter it. I remember Yarn did a video
on how to counter Sparky. And after I said “I get it. It’s cool”. I think Sparky is just an example of
if you know how to counter Sparky, then it’s really quite easy
to defend against. – But if you don’t know, it’s impossible.
– It’s the worst thing, yeah. Sparky is so swingy. Either you get
insane value or you get no value. – Exactly, yeah.
– Absolutely! I don’t know what we could do
to rework it at this point. Maybe we could in the future
play around with a couple of things. We don’t have any plans to rework Sparky
but if we did start going back to mechanically changing some older cards
that would be the first one to consider. We actually did one rework
that it comes to mind now. It was the Lumberjack.
When we released it… It was way slower than it is now.
It was a medium-speed unit and it hit roughly
the same as the normal barbarians. But… It was hardly used and it felt like
it didn’t have its own role. We made a big rework by making it
very fast and hitting really fast. That was a rework we did,
but it was quite… fast after it was released
so people didn’t really… weren’t really used to it at that point. Another example might be also Poison. With the Giant Poison meta, we worked
the Poison to remove… Was it the slowing? Yeah, we did. I think even the Log had a few changes
that felt quite underwhelming at first, but then we added that push back
and extended its reach… And now the Log used to be the most… The worse of the legendary… And now people are like “I got a Log!” People used to post pictures saying
“I got a legendary! Oh! It’s a Log.” And now it’s the opposite:
“Yeah, we got a Log”. The Log when it was launched
it did have the push back, but a bit later we introduced
that it pushes back everything. I wouldn’t call it a complete rework,
it was really just… – An evolution!
– …giving it like more role, giving it that unique legendary status. It has its own soul in a way! Yeah! When it was released
it was The Log, but now it’s “The” Log. It was actually introduced
as a counter or a nerf for the Princess. That was
one of the design goals with the Log. – Can I ask you guys a question?
– No! – Okay.
– Next question. PiratemanKale asks
“Why won’t you buff tournaments? “They’re pretty much dead”, question mark. Yeah, I think we can see
from the use rates of tournaments that they have been on a decline. Yeah, I think we want
to revisit them at some point and take a different approach with them.
Rework them entirely. It’s still in the discussion phase,
we’re trying to see what we want to do. How we can make it work. We’re talking about it. No concrete plans
on how we’ll move forward with this. But I’d definitely like to see
in the next year that we rework the tournaments
into a structure that we’re happy with. T3000P asks “Can Cannons fart?”. No, and I think
it would be seriously OP if they could. That would be totally broken. Maxquash55 is asking “What game engine was used
to make Clash Royale?”. Bangs, can you answer that? Yeah, it’s a good question. We actually used an in-house engine. It’s one that we’ve developed,
it’s a Supercell engine. It’s written in C++,
and all of the Supercell games share this core tech
and are built of this core tech. I think the technology comes
originally from Clash of Clans, at least from the early days of Supercell
and it has matured over time. We find that sometimes
a lot of the game teams will create… create code and create features
that are reusable and at that point we moved
that code into the engine so that other games can benefit from it. An example of this would be where Clash of Clans
have recently added Friends feature where you can add friends
outside of the social networks, and maybe that’s something
that’s coming to Clash Royale soon. Spoiler! TAW96 asks “What does P.E.K.K.A
officially mean to you guys?, “and/or How did you create
or come up with the name?”. It’s a secret. No one knows. But Pekka
is a male name, kind of like “John”. – A traditional Finnish name.
– Yeah. How do you say it in the Finnish way? Great! Skull310 asks “Have you guys ever tried to make
a Legendary building?”. What would it do, Stefu? – We tried once.
– Really? – Cards we killed.
– We once had the Ice Wizard Tower. – Ah, yes.
– Cool! But it was too defensive. It just stopped everything. Yeah, it froze everyone in place. The thing with the buildings
is that they are often quite defensive and we don’t want to like… Aggressive gameplays…
We want to focus on aggressive gameplay. Furnace and Goblin Hut for example
are quite offensive buildings. Barbarians Huts too.
We need an offensive building. What’s the most offensive building? An offensive building?
Probably one that shouts out curse words. That’s Mega X-Bow.
The Elite X-Bow. Oh, yeah. The Elite X-Bow that shoots… Shoots curse words every 0.2 seconds. Yes. That would definitely be Legendary. – Can I mute it?
– No. Unmutable, because it’s Legendary. Zaramon95 asks, “Was the
‘Opponent left the match’ text removed “because I have made people rage-quit “and know that at least some of them
have left the match?”. The text was removed so that
if people accidentally disconnected, people couldn’t take advantage of that
right away. Yeah. I guess it’s nice to see
when that happens that you’re winning, they just go “I’m not playing anymore”. But at the same time.
people could take advantage of that, it’s like,
if you suddenly lose connection you don’t want your opponent knowing it – …because they go “Quick, rush, rush!”.
– Right. Tinmanhole asks “Is the developer team
going to fix True Red and True Blue?”. Yeah. True Red and True Blue is about slight asymmetry between the players, and, yes, we have
fixed that in the last update. There might still be some edge cases.
But it should be fixed. Yay! Cloudx writes, “How does a Doot
take care of his calcium supplement “if they can’t drink a glass of milk?”. This is something we looked into. We had to study their behaviour
outside of the game. – What are they doing on the downtime?
– Right. One of the conclusions was that we actually found them
chilling out in a milk bath. They’re pretty small as well,
so you don’t need a lot of milk, but they just chill out there
and absorb the calcium and then dry themselves in the sunshine. Which is good. They get the calcium
and the vitamin D as well.
That’s how the skeletons
maintain their bright white appearance. Yeah. All right!
Cip, here’s a question for you. – Specifically for me?
– Yes. ThinWG is asking, “Why did you change
the Executioner’s face?”. – Oh, here we go.
– Why did you do it? We thought like the Executioner’s face wasn’t really reflecting his personality
as well as we thought it would. Also, it was sort of poorly executed,
so we sort of reworked it, made him… I guess,
more reflective of his personality. What is the Executioner’s personality? He’s kind of goofy,
obsessed with his axe, and we just wanted to get that across. Bennymon9 is asking,
“Will you add in-game friends “instead of through social media?”.
Like Clash of Clans. Absolutely. We will be adding this. We can’t guarantee
it will be in the next update, but it’s something that will come. Mashhad112 asks
“How do you come up with the 2v2 idea?”. If you listen to our previous episode
you will find out. This question has come up a few times. For us to explain
what happened in the last update where a few cards whose attack speed was affected in peculiar ways particularly the X-Bow,
Dart Goblin, and other units that had a very high attack speed. There were a few unexpected changes. So what we have changed
was the targeting AI. And then, there was
a logic fix in the game. I’ll start with the logic fix. There was a bug
with the timing in the game. It was a really subtle thing,
and for the update we fixed it. What we fixed was the scenario
where two Ice Golems if you rocket them the rocket ends up dealing double damage.
That was the bug we fixed. It’s amazing that that bug has absolutely
nothing to do with the attack speed. The thing is that’s what we fixed, but what happened was
every range unit suddenly shot slower. The change was most visible
for X-Bow and Dart Goblin because they are really fast units,
so what actually happened was that every range unit had their hit speed
increased by 0.05 seconds. It doesn’t sound like too much,
but for the X-Bow that was 17%. We did the fix and, unfortunately,
we didn’t notice it until it was live. And then it was staring at us. It was embarrassing
that slipped through. We wanted to quickly do a maintenance
and address the issues. We did a few changes
in that maintenance break to address
some key interactions that happened. So we increased the projectile speed
of the Princess Towers, the interactions for example,
against under-levelled Fire Spirits remained as they were before. That was not an interaction
that we wanted as a permanent change. We also changed the Blowdart Goblin
and the X-Bow’s hit speed so they acted like they acted before. The thing is the X-Bow had been
shooting faster than what the stats said. The stats say 0.3 seconds,
but actually they had been firing at roughly 0.25 seconds. We didn’t want to
revert that bug fix, because it was a bug
and we wanted to fix it. So we did these tweaks
to address the changes. But, any other changes that happened,
any other hit speed changes, they are now considered to be
permanent minor tweaks to the game. Of course further down the line,
we may do balance changes to these cards. But that’s how it is now. One thing that was pointed out was that
the Archers fire at the same target. How do we fix that? That relates to a change
we did to the targeting AI.
How do we fix that? That relates to a change
we did to the targeting AI. What we tried to fix there
was this situation where, for example, Sparky
is just about to shoot and the Princess Tower shoots an arrow
and the Musketeer decides that “The arrow is going to kill
Sparky. I don’t care.” Then, Sparky actually ends up shooting
before the arrow reaches it. And, as a result,
Sparky gets one last shot. That could have been prevented
by the Musketeer. We wanted to improve that situation. We did some parameter tweaks, but we didn’t revert it completely
to the old system, so it should be better than
it has ever been. Gilmentacore asks “How do you keep
the game files, client, and the whole game in a very low memory requirement?”. He says he’s very impressed with this.
Thank you. I think, by memory,
what they’re asking for is disk usage, the actual size on disk. We’re always trying to keep
our games under 100 megabytes, because that’s the limit for what
you can download without having Wi-Fi. Which is very important, because it allows
more players to have access to the game. Yeah, if we release an update
and you can’t download it, you can’t play. So it’s important that we keep that limit
so people can always get access to that. If they don’t have Wi-Fi on their mobile,
they could still download the update. Right. And many people use
prepaid cards for Internet usage, so keeping it smaller is much better. Absolutely. There’s many
technical solutions we’ve got. One of the things we’ve done recently is as we’ve been adding more content
and more units to the game, we started getting it really difficult
to keep on to that 100 megabytes limit, and the game now actually
doesn’t have all of the content as part of the initial download. And some of the high quality textures
actually get downloaded in the background whilst you play the game, so that we are able to still add
new units to the game but not have
this restriction of 100 megabytes. Politehawk27 asks “I was wondering if Supercell would ever
make the emotes interchangeable?”. It’s definitely something we thought about
and hopefully it’ll come soon. Yeah, that’d be cool.
Well played. Good luck. Baronboss is asking “Why is there no public test realm
like Overwatch? “Would you ever consider public-testing
balance changes and updates “to guarantee the right changes
and are bug-free?”. With the Overwatch example,
with that public test realm, they have their own app on PC which you use to download
and play the game. That obviously gives them
a lot of advantages and makes it a lot easier for them
to release a test version of their game, because they sort of control
that environment. For us, we would need to put out
a second app on the App Store that is like the main game of Clash Royale
but, perhaps, we have a “Beta” at the end of it. The two games would be
in completely different worlds, with different servers, and you wouldn’t be able to migrate. And the beta may become the main app,
and we would be back into the problem. On mobile it’s quite restrictive
and difficult to do this, currently. Trainertim asks
“What decks do you guys use?”. Let me pull mine up.
Here’s what I got: I got a Mega Minion, Minions,
Lava Hound, Guards, Miner, Arrows, Archers, and Lightning. Nice! That’s cool! It’s a slow push.
I’m really good at double overtime, but in the beginning
it’s all about the defense. – What’s your trophy high?
– I just got to 4,000. – Oh, congratulations.
– 4,010. – Did you get into the last league?
– I did. So you’ve got your Draft Chest. Nice. I’m personally a Mortar for life. – Mortar for life.
– Mortar for life. I’ve got my Mortar deck with my Bowler,
I got an Inferno Dragon there, just to deal with those Royal Giants, and a Skeleton army
obviously for the Elite Barbs, and a Furnace for some chip, and my secret weapon
that I always keep hidden: the Rocket. – Oh, dirty.
– Just to finish them off. So many cry faces
come from that. I love it. Stefu? Lately I have been experimenting
with this funny deck with Battle Ram and the Bandit, the Miner, the Goblin Gang,
the Electro Wizard, Inferno Tower, Poison, and the Log. That’s quite an off-meta deck as well.
That’s quite cool. Yeah, it’s pretty good in the tournaments,
in challenge mode. – What’s your trophy high?
– My trophy high is 4,145. Am I the lowest out of us? I think so. I got quite high.
I almost got to Masters. – I got to 4,775 last season.
– Crazy. My play style is usually that
in Clash Royale and other similar games. I like to collect a bit of everything. I upgrade everything very evenly and like
to be able to try everything all the time. I like to experiment
with decks and cards. If you just play a Mortar deck
your whole life you go slowly insane… go slowly insane… go slowly insane… Right. Yeah. All right! So ACO3, “What can we expect on the next update?
Maybe give us a hint to look out for.” Something we have experimented with
in the last week in the development build
is a new 2v2 game mode. Off the back of the popularity of the 2v2
and how the development team love it, we have played it loads, and seeing
the reaction from the community, we were like
“How can we take this further?”. I can’t give you any concrete promises
or details about this yet because it’s still in the works,
but it will be awesome. It’s 2v2 Draft. Please like, comment, subscribe,
Youtube 5 star, Pinterest… Done! Let’s go! Thanks for joining us.
Thank you for listening. I’ve been Cip. – And I’ve been Bangs.
– And I’m Stefu. Yes. And we’ll… See you in the…! Ah…