Hearthpwn

Dean Ayala (IksarHS) Discusses Arena and Rogues

Dean Ayala (IksarHS) took some time today to talk with the users on reddit about Arena as well as the direction of Rogues in MSG and the future, "fun" cards and more! Grab a cup of coffee for this one because he had plenty to say!

Quote from Dean Ayala (IksarHS)

"Yong Woo: I think there's sometimes a little bit of perception from the community that we don't think about Arena, that we don't care about it. That's just not true."

You know actions speak louder than words...
I agree. It will be a multi-stage approach. The first stage is MSG. We took arena into heavy consideration when choosing rarities so I hope that contributes to a more balanced format. The second stage is to make some changes to how often certain cards in certain classes are available, and how often certain types of cards are available to all classes. Right now we're leaning towards a more spell heavy environment where we make things like Yeti/Spider Tank less prevalent than they are now in the hopes that arena becomes less about getting ahead and snowballing to victory. The third stage and beyond is a bit more hazy, but is about looking at the whole of arena format. Drafting by rarity? Wild vs Standard? Entrance Fee? Matchmaking by record vs Matchmaking by skill (huge implications there)? Lots of questions to be answered, but all things in discussion currently. Looking forward to any feedback you all have :).

Matchmaking by skill would basically kill the format for f2p infinite players as doing well would PUNISH you for future runs.
That's why I think the entry/rewards scheme would have to be updated alongside a change like this. Matchmaking by record is always going to make the format relatively low population because it is so unfriendly to new or less skilled players. It's hard for me to justify that being correct, but as you said there are some complications with just changing the way we matchmake. It sounds like you would rather matchmake based on record like we are now?

I believe they really need to add in an algorithm to increase the number of spells offered. Its kinda ridiculous how I often have decks with 1 or 2 spells in them because I was only ever offered minions or extremely shitty spells.
Agree. The distribution between quality minions vs quality spells is a bit higher than we would like. Something we are looking to adjust.

To start, they need to make spells appear more often and put arena on standard. There are too many cards to play around as it is now.
Spells appearing more often is the direction we are leaning. Cards like Spider Tank (powerful vanillas) showing up less and cards like Shadow Madness, Lightbomb, Hellfire (power board swings) showing up more seems like a better experience to have. Too many games come to getting ahead early and snowballing a victory rather than evaluating the risk of expending all your resources into a board clear effect.

Have you ever considered removing rarity as a factor of what cards show up for the non-legendaries? That instead of getting 3 commons, you could end up getting an epic and 2 commons to choose from for example? I think if all non-legendary cards had equal odds of showing up, it would also make it easier to pinpoint which cards are actually too strong and which ones are total junk too.
Yes, this is something we've talked about at length and will continue to debate. In some ways I like that rares and epics show up less often because it makes them feel more special when they do show up. Also when commons show up at a very high rate, it's easier for experienced players to predict what kind of cards they are up against. On the flip side, there are a lot of rares and epics we would like to show up in arena more often because they are fun or give people opportunities to draft a particular style of deck. Right now it's hard to do that when rares and epics show up so infrequently. We're undecided on this, but I don't think it will change in the near future.

Give players a choice like Raven Idol, you either get to pick between 3 spells or 3 creatures. Or maybe only 10 picks give you this choice, so we don't have to face freeze mage in arena.
Things like this are on the table, but probably for a much later date. The general idea of 'special arena cards' is pretty cool. Things that only show up or make sense in the arena. Cards you can draft that affect your future drafts, cards you can draft that allow you to have 4 losses instead of 3, cards that allow you to re-draft your deck, cards that are face-down? Lots of cool ideas, but our focus is more directed at balance and overall arena gameplay for the moment.

Mage isn't the best Arena class...lmfao, good one!
It probably is, but I have some doubts. I think Shaman is probably 1st-3rd in terms of ranking right now but shows up statistically as the 5th most played class and the 3rd highest in terms of win rate. In general, the highest played classes are also the highest in win rate because word spreads quickly.

Might be worth keeping in mind that Mage being perceived as the number one class so widely is a two-way street. You're considering the bias it builds against other classes, but the pro-Mage selection bias also means that lower skill-level Arena players are more likely to insta-pick Mage and bring it's winrate below the optimal level of the cardpool.
You might be right. I think one of the best ways to check is to look at the records of all individual players and seeing what their personal best class is. I would guess that a smaller number of players 'best class' is Mage than you would think. We currently don't track these statistics, but I've been curious. I think there is a spreadsheet for streamer stats somewhere I'd like to check out.

Quote from Dean Ayala (IksarHS)

It sounds like a lot of the negativity here is centered around Ferryman, so I'll address that. Ferryman is meant to be an option for players that build a deck completely centered around bouncing multiple targets or a single target back to your hand as many times as possible. As some of you have pointed out, this will most likely not be a meta-defining 10/10 power level card, but it still exists for some portion of the audience.

There is some value in creating cards that give you a more realistic opportunity to do the core fun piece of your deck (bounce X minion to your hand to replay) regardless of whether or not that results in the next tournament worthy performer. There are some similarities to Purify here, but as a general purpose card ferryman isn't nearly as weak as that card. Purify was released in a set where Priest was the lowest win rate class and we only had a few class cards to work with. Outside of the timing though, the design is something we definitely stand behind. Building a deck around silencing your ancient watchers and eerie statues is a super fun concept to a ton of people. Giving you more ways to consistently pull that off I think is a good thing. Ferryman is an example of a card that more consistently gives players an opportunity to do the fun thing they built their deck to do.

So what is there to do for the mega-competitive focused Rogue group? Well, we think Miracle is as strong as ever and coin will make a meaningful impact there. Rogue Jade is also interesting and probably the Jade class that scares me the most as someone who works on balance. Cards like Prep and Shadowstep allow Rogues to start the Jade train earlier than most of the other classes and snowball it very quickly. The amount of times Aya Blackpaw was discovered off of Journey Below was pretty high in playtesting, I'm sure someone will do that math there. As you can imagine, playing two of that card in a control matchup swings pretty heavily in the Rogues favor.

I hope this gives some amount of insight to what is going on with Rogue currently, we're happy to continue the conversation.

So the focus is on Ferryman right now, but you brought up Purify, and since the parallels are there (over-costed card for a niche deck archetype) -- why do you add these cards then stop further deck support in later releases?
Jade is an example of an archetype that wants to work with Bounce mechanics, so you could argue we are doing that with the set we're releasing soon(tm). As with Purify, there is a small subset of players enjoying that deck, the goal isn't always to make every single deck 'meta' but to create a variety of options for people to play. If every deck was a 1% deck then that would be a pretty interesting environment to play in. That said, I don't think it's unreasonable to say Silence priest get a tool or two in the near or slightly less near future.

But why make it nearly strictly worse than brewmaster?
I think strictly worse than brewmaster is a bit of an overstatement. Better and worse in a variety of situations is probably more accurate. I think Johnny in general is less concerned about a stat point here or there.

How does Blizzard determine that "people like playing" an archetype, card, etc.?
We have plenty of internal statistics to help with that. At any given time we can look at any skill range of players across any date and see what deck types are played the most/least along with the win rates of said deck types.

How many players are playing bounce rogue or silence priest?
A small pocket of players that I hope are having an awesome time. If we can introduce 2 or less cards to help 1/2% or 1% of players enjoy a weird niche deck then I would consider that a great success. 1% a card adds up pretty quickly.

You guys don't do enough with the statistics you have. There are cards that are overwhelmingly unbalanced that you need to balance. E.g. why does "Thing from below" have taunt?
Shaman has had Hex since the beginning of (hearthstone) time without it being a huge problem. I think revisiting what Shaman is all about is a real concern though. It's easy as card designers to look at all the things a class doesn't have and shore up all of their weaknesses because that's 'new and fun!' but we have to be pretty careful about doing that. What should Shaman be good at? What should they be not so great at? I think single target removal should probably be a weakness, but it worries me to remove a high power level card from a relatively weak base set of cards when we just changed rockbiter. It isn't unreasonable to me to say Hex shouldn't be as powerful as it is, we just want to understand where that leaves Shaman a year from now first.

So is Blade Furry going to stay a 4 mana card? Because it seems to me if you're going to cater to niche decks with these types of cards (Ferryman, Purify), it seems reasonable to print a board clear that caters to control rogue decks.
We don't have any plans to change Blade Flurry, I would be quite surprised if it ever changed. The future of Rogue is likely to include very weak AoE or life gain and very strong single target removal and card draw.

Class identity should not include the class's access to necessary tools to compete on ladder. With that mentality the same classes will always dominate ladder and the same classes will always languish in obscurity.
I don't think a class needs top tier AoE, top tier healing, or top-tier taunt in order to be able to compete on ladder. Many players are successful with Rogue currently and have been for the past year (years?). If you absolutely need to have healing, taunt, or AoE to be competitive then that is a problem with Hearthstone and not Rogue specifically.

It needs at least one of those things to play a non-combo deck.
I mean, miracle very clearly has none of those tools and does quite well (has for a long time). It's possible for sure if they just continue to do a better job at what they do now. Single target removal, drawing cards, utilizing things like prep, etc.

I literally just said it needs one of those things for a NON-COMBO deck. Miracle is infamously combo.
Oh, sorry I misunderstood. In general I think Rogue is a class that will be more heavily based around combo-centric styles. Jade can sort of do either, but Rogue in particular should probably be taking advantage of things like shadowstep/shadowcaster. Gimmick is one way of saying it, but there have been a lot of statements from players saying that is the style they would rather be pushed for Rogue in particular.

So do you want rogue to permanently lose to Hunter?
Yes and no. I imagine a meta in which hunter burn is popular is a meta where traditional rogue strategies will struggle. Hopefully there are enough strange neutral cards that allow rogues to modify a strategy to figure out those metas too, though. But as a general statement to answer that question... when Hunter is true to its identity, has a strong strategy for that identity, and is very popular in that current metatgame.... Rogue will probably struggle. We think that is okay.

My biggest problem with the Timmy argument of hearthstone is that there is absolutely no area of the game where these decks have any influence.
While this is also a reasonable argument, I think you might be surprised how many people are playing 'janky decks' vs how many people are playing 'meta decks'. It varies at each skill range, but I wouldn't say the 'janky deck' population is small. Some people just like making weird decks and playing them for fun it turns out (yay). Sometimes I wonder what we can do to promote just having a good time in a stress free environment. The majority of people that play Hearthstone are playing in ranked mode, which I think limits just how hardcore of casual we can make it when we have to cater to such a wide range of player types. What should playing Hearthstone be about for casual players? What should Hearthstone be about for experienced players? Right now I think the focus for both is improving your deck and your deck performance, which could be incorrect. I could see a world where casual players are focused on building their collection and playing weird decks and hardcore players are focused on playing a single or multiple decks at the highest level and are totally separated from the casual group. There is always a cost to separating the population into a bunch of different groups, but maybe it is correct. Ideally there is a way to satisfy both types of players within a single system.

I think the real issue you guys need to address internally is the ridiculous ratio of neutral:class cards. If you had more room to print class cards in your sets (a constraint you impose on yourself), you'd be able to throw something new to more deck archetypes at once.
I think this is a reasonable argument. We have had multiple conversations about whether or not there should be more/less class cards. The upside is that we can make more powerful cards that are more exciting to the general populous of people that just want powerful cards (reddit I think falls into that category sometimes). The downside I think is that for people that like 1-2 classes, less cards they open in packs feel like they are for you. In a world where we have no neutral cards and only play 1-2 classes, the chances you open a pack and 0 cards are for you is pretty high. I think neutral cards help that ratio of 'cards that i can play' a lot. Anyway, it's a reasonable argument thanks for bringing it up.

In comparison to Purify though, you are getting something that was not traditionally available to silence priest, which was card draw along with the silence. You mean to tell me with a straight face you think its suppose to feel good that this card's only advantage over a neutral common we've had for years is the fact it is better when it doesn't utilize the combo?
It's mostly that you have access to this thing rather than how Ferryman compares to Brewmaster directly. I won't argue that in some or more than 50% of circumstances you would rather have one or the other because that wasn't the reason to make the card.