[2020] Ask Us Anything! - Join the Overwatch development team for the celebration AMA

Hey there, we've definitely worked through audio occlusion bugs in the past, but unfortunately there's not a single problem that solves all of the issues. Any time you encounter this if you can post detail on our bug report forums it really helps us dig into that specific issue.

we're in a really great place to move forward right now. over the past year we were lucky enough to get jeff chamberlain to become our new creative director. jeff was the director of our original announcement piece for overwatch back in 2014 (as well as some amazing blizzard cinematics over the years -- WOTLK intro for example). we've also added some really amazing narrative designers and are partnering with our story and franchise development group (they make the animated shorts, origins movies, comics, statues etc) at an insanely intense level. there are some very cool lore developments coming.

Speaking for myself here - I play a lot of competitive Overwatch. I play 10-20 hours of competitive Overwatch every week, and I love it. I used to play flex. I truly enjoy playing a wide variety of characters. Lately I have been focusing on Tanks and Healers, as my friend plays DPS.

we've tried a number of things over the years:

• changed the capture points to work like control points (meaning KOTH points)
• added defenders spawn rooms closer to point a
• added a "middle" capture point between a and b
• moved the defenders spawn point back on b
• made it so that capture happens regardless of if defenders were on the point

those are just a few off the top of my head. lots of level design changes.

Actually, im not sure if people are aware of this even, but Sombra was initially prototyped as a Support hero. In fact, it really went far beyond prototype, she only became a damage hero in the stretch of her development.

She had some of the most significant changes while going through developement. Her whole kit used to be centered around staying invisible and landing hacks on the entire enemy team (you could hack more than one target at a time). But hack worked differently then: It used to last a lot longer and it reduced your healing and damage output instead of making you unable to use abilities.

This meant even though you had one less healer as a support, you theoretically made up for it by reducing the enemy team's damage. It was a pretty fun little mini-game for her, but people were ultimately very frustrated by being debuffed like that, even if they knew on paper she was reasonably balanced. In fact, we've tried a couple times to try a damage-dealt debuff (Wrecking Ball had this at one point too), but it hasn't stuck yet. That debuff ends up more frustrating to players than many other debuffs like Ana's grenade or Zen's discord even.

So we went with a more aggressive kit and moved her to Offense instead, especially because we were already feeling on the fence about non-healers being in Support (Symm was still there at that time). At this point in the game if a hero is in the Support category, they're expected to be able to heal, that would be a challenge here as well.

Hey there, the matchmaker COULD be aware of the level of a player but it isn't something that we currently take into account when matchmaking.

The reason we don't specifically take level into account is because it's an indicator of time played on a specific account and not the skill level at which someone performs. While it is true the more you play the more skill you're likely to gain, it's also true that skill gain will be reflected in the matchmaking values as well.

Cheers

From a tech perspective, I think the hardest might have been Echo. Her ability to duplicate an opposing hero brought a ton of challenges and bugs. In most areas of Overwatch, our ideas for what we would like to do are ever expanding in complexity. But we need to make it work with assumptions that we made years ago. Sometimes they don't mesh well. With Echo, it was a difficult clash to work out. In the end, she turned out to be a really fun hero, so it was worth the work. But there was a lot of engineering grey hairs generated around that one!

Sure! There were several factors in the decision to switch to the more flexible queue system we currently have.

1. There were concerns that a 6 stack only competitive queue would prevent some players from participating. While they do want to play competitively, improve their skills, and test themselves, there are many reasons they might prevent them to play as a team with others.
2. Six player only teams solved some matchmaking problems (you're responsible for finding teammates, not the matchmaker), but would create others such as making it more difficult to find fair matches at extremes of SR.
3. Requiring you to find your five teammates to play can be a logistical issue. Let's face it, queueing solo is just really convenient. You can do it at any time of the day, when your teammates are sick or on vacation, etc.
4. Finally, let's face it... If we didn't have a solo queue the first thing people would of asked when we introduced competitive play would of been, "That's great, but where's solo queue?" It's very popular for a lot of reasons.

We still have some long term cool (but far in the future) ideas around competitive formats using fixed 6 player teams, but it would likely take the form of short duration tournaments rather than month(s) long seasons like we have now.

This is always a possibility! Ana is an example of a hero that came from the story. She wasn't created initially to be made a hero but we saw a cool opportunity while making a new support to have it be her and it worked out great.

That said not every character/personality in the story will (or even should) be a hero, but we'll see!

the same team that works on and made overwatch is also making overwatch 2. we did find the need to grow the team quite a bit in order to do both and also because overwatch 2 is a much larger game than overwatch. the team that shipped the original game was approximately 70 developers. the team is closer to around 160-170 now. it's a very tricky balance and often very stressful working on two projects at once. the reason we chose to do things the way we did is because we wanted to maintain consistency of vision and direction over the game and the franchise and not make things feel too splintered. for an individual developer, they usually have a clear task that's either for the live game or overwatch 2. but we often find ourselves context switching between the two games.

Hey there, this is a good question. The matchmaking system gains confidence in its evaluation of your skill as you play more games.

yes the summer games are returning this year. information is coming very soon

Artist answer here - I really like the themes for the lunar new year events, and I look forward to working on those every year. I'm not sure what I would do if I got to create my own event, but the rockabilly idea and sketches are cool!

we have many years of lore planned for the overwatch universe. there are so many aspects that we're excited about -- far reaching plot lines and individual character arcs. there are also a lot of heroes, places and world groups we want to tell more stories about. we're also excited about telling stories in new ways and using new mediums.

I've heard that the French perfected scrambled eggs (via my eggs-tensive research on the internet), and so I think it'd be Widowmaker.

well there are about 200 pros and 50 million casuals.... but we both know you're not asking for that percentage breakdown =]

we listen to everyone's feedback. it's all important. when it comes to balance, we primarily balance the game around owl, gm/masters.

soon

This is actually really hard to answer actually.

Most of the time heroes go through a ton of iteration, and single abilities or weapons get cut. Often times we'll end up using an older cut idea on a new concept. For example, Orisa's first ult was an idea where she could "teleport" away (we'd probably kit this differently though), and then 'drop' anywhere else on the map via drop pod or something similar. This was back when she was extremely early and we didn't have a character yet, so I was sort of imaging a character more like Dutch from Predator (hence the mini-gun, etc). It ended up not really working for the gameplay, and really didn't make sense once we started developing Orisa's character.

Much later we were working on Doomfist and thought we could revist that old idea and thats how he ended up with his ult!

One of the things we're always looking at are ways to provide positive notes to players after they finish a game, even when they lose. The medal system was one of those, along with the card voting system during the end of round.

However, the medal system provides only view of the match results, when why someone earned gold, silver, or bronze can be way more complicated than just reflecting your skill. This is one of the reasons we want to look into systems that can help you understand key questions like "Did I play well?" and "Did I do something awesome?" whether you win or lose.

That might be an improved medal system, or something else. We'll see!

the most immediate priority is a balance patch aimed at the double shield meta. this will show up on the experimental card in the next few weeks.

we don't have any reworks immediately imminent for any hero, although we are discussing some.

2cp is more of an ongoing exploration for us. we've tried many, many things and have yet to be satisfied with any of the results.

paris has gotten some attempts at a rework but we weren't pleased with our results, so we paused. we haven't started looking at horizon again yet.

A lot has been added! Role Que, custom game modes, replays, new heroes (Sigma, Baptiste), and a ton of unlocks.

It differs a bit for everyone, but we usually start each day with a team playtest. Every other day we have a quick "team standup" where we discuss major initiatives we are working towards. And it's an opportunity to keep the team on the same page with where we are and where we're going and to celebrate achievements. For the rest of the day we are typically collaborating with one another in smaller strike teams, meetings, and individual work to drive features forward. At the end of each day we have another team playtest. Every day is different than the other and different for some individuals than others, but that's the very basics. The good thing is that it hasn't changed too much with COVID, though we do miss the face to face. Virtual face to face isn't quite the same. :)

pizza

proof: https://i.imgur.com/f0IGPu2.png

Sombra and Genji were the two hardest heroes to make, probably.

Sombra worked very differently and had to go through major changes pretty late in her development (as I posted about earlier), and Genji started out having only his sword for a weapon (he is still refereed to as SwordNinja in our internal files, etc). His main kit initially involved invisibility, wall perching, and having an 'execute' animation where if he got behind you he could stun you and start this long animation which would just instantly kill you if he finished it.

The idea was he could easily pick off people who were not with their team, but otherwise it would be very difficult to use execute. This kit caused all kinds of problems though, especially his sword being his only form of attacking. Once we moved away from invisibility and instead chased the agile ninja fantasy, and made the sword his ultimate, everything started clicking into place.

i play the game everyday. we have multiple playtests daily plus we play for fun on our own time. i love to play all of the heroes but some i really struggle on (doomfist, sombra, sigma). i play a lot of reinhardt, symmetra, mercy these days. i also love hanzo and mccree, although i'm not so great with them.

As a tank player I really like the option to charge or boop with Reinhardt. I believe the internet has recently deemed this 'Flat's cradle'.

we are very excited to explore the overwatch universe in a number of mediums. nothing to announce at this time but we would like this very much.

Great questions, here's an answer for a couple of them!

- An unexpected way covid19 has changed our day to day is the impact it has had on meetings. It took a while for people to become more comfortable in online meetings, but now they tend to be more focused and efficient. It's awesome! The engineer in me loves getting more time to write code.

- As far as tips for breaking into the industry in the current climate, I think it's a bit too early to tell how things will change long term. I hope to see a shift industry wide that is more accommodating for remote work. In general I think you should regularly look at open positions at studios you'd be excited to work at and apply as much as you can. Just experience interviewing will be helpful to land the perfect spot.

This is actually something that comes up fairly often. In general I like to try to keep 'tech' in the game if we can, but sometimes its some weird bug that can cause a lot of other problems, is super unreliable, or is just straight up overpowered.

For example, the Genji ledge dashing bug was cool but both extremely overpowered and also really inconsistent/buggy feeling (which makes sense I guess, as it was a bug). That one we opted to just remove instead of trying to work out way to implement it as a feature, mostly because Genji already has really strong mobility and we didnt want to add more onto that.

Mercy's glide jump though was something that wasn't really intended but we ended up having to fix for unrelated reasons. When we did that, we went ahead and implemented it as a real feature of the hero so it works a lot more consistently now etc. She also has the big vert super jump which is a lot more tricky to pull off but we haven't really adjusted or changed anything about that either, as its pretty consistent once you get used to it (though admittedly, a bit weird).

this is something we're passionate about and exploring.

I play nearly every day. I tend to prefer Zarya, McCree, and Zenyatta. I do a lot of Mystery Heroes with friends I grew up with. They live in Minnesota and I'm in California, Overwatch is one way that I try to keep in touch!

we've tried all sorts of balancing points, including 6 ticks. it solved some issues but created others. the more ticks, the more the defending team feels obligated to sit on the point....

we test the maps a lot. for a very long time. usually if we're patching a map into the live game, we'll do an extended ptr period for testing (same with heroes). for maps that shipped with the game, we tested some of those for years. we don't openly bring in outsiders to test maps early as we have a pretty comprehensive play test and map testing process. sometimes, depending where we're at in a dev cycle, we might bring a map to a trade show (i.e. blizzcon, gamescom etc.) before it goes live.

Golden guns are an ever-evolving topic. On some of the more original characters we ran into tech challenges when trying to add gold versions. One thing to know is that we came up with the idea of golden guns after some characters were already made. Example – Torbjörn turret was made to go molten core before we knew about golden guns. When we tried to apply the golden shader it broke molten core. As time goes on, workflows and tech improves, we get more and more flexibility of where we can add gold to a character.

the border system will go away with overwatch 2. we want to make sure to recognize your accomplishments but we'll be doing something different for the sequel.

I think we did do this at BlizzCon just before we shipped where we played some streamers. Back then it was great because we were the best Overwatch players in the world! :) I don't think we would fare as well, these days. But it would be a lot of fun. Fun side story: When we were in the final year or so of dev for Overwatch, we put in some somewhat long hours at the office. As a lark, our QA set up a multi-week tournament across the team to give us a moment to take a break, eat pizza and watch tournament games, shoutcasted by two teammates. The tournament games were SO much fun and drove us to push harder on spectator features and the plans for the league.

our first goal is to make the mode something that everyone enjoys. you can see by today's experimental card that we try various things to iterate on the game -- just some of them don't work.

for ow2 we have some ideas that will allow for more player agency but nothing is set in stone yet so i don't want to commit to any promises.

and in the spirit of being open, i think the criticism of 2cp is fueled a bit by the echo chamber of certain communities, when overall the maps and modes are not as universally hated as some would have you believe.

Ideas like this are what makes workshop so exciting. We've seen some super cool custom games that implement this concept.

Back when we are working in the office we have a hallway dedicated to fan art. It varies from entry level to professional level, and it is all inspiring. I regularly spend time looking at this art and seeing what the community is coming up with. We're always inspired by the community and it's art, but we don't have immediate plans for that kind of contest right now

the tab screen was an almost religious debate during the development of the game. i think many of us feel like there are other options for us to explore. a big fundamental change is more likely for ow2 than the current live game. but in general the team is interested in giving players more/better ways to examine their own performance. again more likely on an ow2 timeframe.

guilds is a feature that we've always been excited about. many of us worked on world of warcraft and specifically worked on the guild system there. so it's something that's near and dear to us. we have some really cool ideas of how something like that could work with overwatch. unfortunately, from a tech and design standpoint, the feature would be an extremely heavy lift (at least to do it right -- the way we'd want to do it). so not on the near horizon but definitely on our wishlist.

Our tools are pretty robust so its pretty easy for us to whip up some changes and run a playtest to see how they feel. We've done quite a bit of tests of all kinds of tweaks, some small some big, that come from the community.

For example, we tried Symmetra as a Support hero where her turrets healed instead of dealt damage. This was sort of fun but was kind of frustrating as a Symm player because Overwatch moves so fast. We started trying to tune the turrets to have super high range, or being able to have a ton of them out, etc. It ended up just feeling like a mismatch, but honestly i'd be up for taking another crack at it at some point. Maybe if she had some other primary way to heal and the turrets were just a way to provide auxiliary healing it might have gone better.

Its hard to think of other specific examples off the top of my head right now, but I know there are more things that we've specifically tried out and ended up not doing. There are some other popular ideas, such as Mei being able to slide on her own ground-ice, that we've talked a lot about but haven't actually tried in-game. For the Mei example, our big concern is giving her any kind of mobility would require some serious balance reworking to make her not more terrifying than she already is. I totally love the flavor of it though, maybe some day it'll find its way into the game in some form.

Try putting them on the left side of your screen for charge, and the right side of your screen for boops.

i think everyone on the development team would be extremely excited to do these sorts of things. they usually get bogged down by business/legal types....

but yeah, if they left it up to the dev team we'd be all over it

I touched on this in another answer earlier but the tldr of it is: We tried it once, and it didn't work out very well. That said, the door isn't fully closed on this yet, its possible we'll re-try her as a Support (healer) in the future.

Every day! I'm mostly playing whatever hero I feel like needs to be looked at more, or has recent changed we're testing etc. Though if i'm just picking to try to win, I find myself mostly going for Ashe/Widow/hitscan in general. Or Junkrat, I still have a lot of fun flying around with Junkrat.

Several cool questions here, so let's go?

Are there any current plans to continue to improve the player behavior experience for the game?

Funny you ask that, but we're going to make the penalties for leaving competitive games a bit more severe in an upcoming patch. (More details then!) We're ALWAYS looking to improve the player behavior experience, and there's several interesting initiatives for identifying bad player behavior that are currently in the pipes. We'll be able to talk more about them when they get closer to being real.

Have we thought about creating community moderation tools to allow players to review matches for suspected disruptive behavior?

It's been discussed, but we'd rather look into more internal solutions to try and identify disruptive behavior like throwing, feeding, etc. Community involvement is still super important though, as reports players who do bad things helps us tremendously. Please continue to do that, it really helps!

"Smurfs" are another topic we're discussing, but it's important to note that there's a lot of different issues here. Let me start by saying if you throw a bunch of games to derank so you can then rise back up to a higher SR, that is considered gameplay disruption and action can be taken against that account. Don't do that, folks.

Things get a little murkier when the "smurf" is a brand new account. We have multiple different systems in the matchmaker to try and identify new players skill and adjust quickly such as win streak bonuses, increased rate of MMR change for new players, and we still make some adjustments based on your personal performance until the player reaches a high SR. We can tune these more aggressively, but it can create negative consequences for players who actually ARE new. So we're looking at some new ways to identify skill so we can quickly identify the players who skill and SR don't match. It's a tricky problem, but we're working on it.

With regards to your second leaving question, I'm not sure exactly what you're asking. Leaving is disruptive, but there's always going to be a certain amount of it. Real life emergencies can intervene, internet connections sometime go down, computers crash, etc. What we try to do is to make sure there's never a gameplay incentive that encourages people to leave, and put in place penalties significant enough to discourage people from leaving simply to be disruptive and spiteful.

With regards to the Endorsement system and Looking-for-Group, both have had a noticeable long term effect on creating a more positive Overwatch community online. Has the positive effect lessened over time as the community has lived with the system for all these months? Of course, but that's not entirely unexpected. Endorsements weren't added simply to reward good behavior, but promote and recognize the players who make playing Overwatch a more positive place for everyone.

1. aaron did a lot of work on paris but we're still not happy with it. horizon is not on the...horizon (oof bad pun)
2. we always encourage players to spend their credits on things they want. there's nothing on the immediate horizon you need to be hoarding these for outside of event loot boxes
3. i'm not an artist but we like exploring alternate versions of heroes such as "Junker D.Va" sometimes
4. for maestro sigma in particular, that actually started with our composers and audio director. adam, derek and scott had all of this amazing music and we were brainstorming ways of getting it to players. we came up with the idea of "albums" that we were going to release (note: plural) and we thought that doing a mini-event around the first album would be really fun. then we thought about which character would make sense to have a musical skin (obviously lucio was discussed). our animation department had done the amazing conductor emote which was going to be a normal event unlock. we decided to use the emote, create a conductor skin and then release them both to promote the album. super fun process.
5. we've definitely thought about a solo queue but at this time we don't feel like it's necessary or best for overwatch. also, if i were to show you the statistics, you would see that it's an extremely rare case for solo players to be matchmade into group situations larger than 2
6. it's not that i think the system is problematic, it's just not very cool/interesting. we just feel like there are other opportunities
7. haha glad you liked. i have a ton. would be happy to share. for now, if anyone wants to see, just look on youtube for the overwatch archives panel with myself and arnold tsang

Doomfist Parkour

Edit: Adding a bit to this, here's the top 10 for the past week!

1. 1v1 Arena
2. Zombies
3. Genji
4. Thomas
5. Rein 1v1
6. Scrim
7. AIM TRAINING
8. Duel Arena
9. PRACTICE RANGE 2.0
10. Minigun

Hmm tbh I hadn't actually considered this until now. From a 'sniper' perspective, we had generally considered them to be more defensive oriented. But she isn't really a sniper, exactly. On just a whim i'd say Offense, but I could be swayed to make her Defense. I mean, this is basically a big part of why we merged them into one category, at some point it becomes fairly arbitrary feeling.

they are all super fun. it's seriously a dream job. most recently, i really liked being involved with sigma. josh and geoff drove the design of the character and qiu was the artist who drew him. but it was a lot of fun to help develop the character. i loved the idea of an absolute genius who is disassociated, who thinks we see what he sees, who thinks we understand physics and the universe the way he does. i loved the idea of a tragic victim who was being used for evil who was not evil himself....

I had another reply where I mentioned the challenges of Echo, which has, arguably, our most difficult ability from a technical perspective. But, to be honest, we knew that one was going to be challenging going in. However, what I have found is that even with the "simplest" of abilities, we often find that they are harder than we thought. Often this is because we expand on our original ideas as we develop the hero. Other times it just is unanticipated challenges. One example of the latter is Moira. Her animation was very challenging and took a lot of work from our tech art and animation teams to get correct.

As for funniest bug, I don't have one that is popping into my head that is actually funny. There are a lot that seemed funny when on very little sleep, but not so much after a good night's rest! :)

Hmm I've always wanted to do a grappler hero...

Please don't give Geoff any ideas! :)

tracer plz

Answering from an Art point of view here.. Brigitte Flail - This seems like a straight forward idea, but it turns out making a flail with a extending chain in a 1st/3rd person game can be challenging.

On question 2 - I have seen a lot of crazy bugs, even some from here on this reddit. One that comes to mind was the discovery of 'Depressionwatch' where many of the characters were exported without facial animations. Another OG bug that comes to mind that will always make me laugh was when Widowmaker's butt inflated on reload.

This is understandably super confusing, I totally agree with you. Its sort of came about this way from a chain of rules that really don't even matter anymore necessarily. But heres how it went:

Junkrat trap is made, it blocks movement abilities. Ok cool, lets playtest it. Oh, its weird on Reaper though... is Wraithform a movement ability? It technically makes you faster but doesn't really feel like a movement skill, and being able to wraith (even if you cant move) would feel a lot better. So we go with that.

Later, Zarya's ult gets the 'no mobility' rule added to it. We just follow Junkrat trap's rules, which means Reaper can now escape grav. Hmm maybe this is fine? Lets roll with it, its consistent.

Moira is made. Fade works very similar to Wraithform, we're consistent where can be with the two of them. However, Fade is clearly much more of a mobility skill than Wraithform... so it kind of breaks down in the junkrat trap/zarya ult case, which is where we are now.

This is something we're looking at right now actually, as part of the Moira Experimental v2 changes that will be coming somewhat soon. What we're currently testing is allowing Wraithform/Fade to be used while in those abilities (Sigma's ult too), but not let you get free of the effect. This means you can hit these abilities to get the immunity effect but you can't run away. We'll see how that feels when it hits the experimental card.

Our initial launch was a huge challenge. We wanted to release on all platforms (PC, Xbox One, PlayStation 4) in all regions at the same time. We put significant effort into performance profiling to ensure our servers would hold up on launch day. There was a huge amount of work involved between engineering and QA to get through the console certification process for both platforms. It was an exciting time, one I'll never forget, and I'm happy with how well it went!

i think the suddenness of transitioning to working from home was a shock for many of us. we had been tracking the pandemic for months and making preparations, but i don't think the reality had sunk in that we were going to have to close down as well. we have offices in shanghai and seoul that had been shut down, so it was something blizzard had been dealing with for months. but the call to close irvine down (that's where the ow team is) came very quickly.

of all things, team 4 -- aka the ow team -- was in the middle of a move when it happened. most of us don't even know where our office stuff is... in some box somewhere in a new office?

the first few weeks were rough. there were a lot of tech and communication hurdles. i am sure you all have dealt with that as well. we were really lucky to have amazing technical staff on team 4 as well as really great blizzard it staff to make it go more smoothly.

now, we're all adjusting in our own ways. personally, i like working from home. in particular i like being around my wife for more of the week -- it's really nice. but i miss the team like crazy. some members of the team feel really isolated. it really depends on the person. i'm worried as hell about everyone and hope everyone is staying safe and healthy.

as for the game, i honestly think some parts are moving faster and some are moving slower... overall, i don't think we're losing time and maybe have even gained some? but that's my biased view and others on the team might feel differently. as game director, i am extremely pleased with the progress on both games in the work from home situation.

these are cool ideas!

Bill hit some good ones. I'll add that another challenge is that we cannot just walk around the office seeing the cool things that everyone is working on or engaging in random conversations about what we can help with. We still have some of those over tele-conferencing, but it isn't as impromptu.

As to your "passion project" question, there are a lot of those on the team. From evolving the gameplay to introducing new, previously unplanned graphics features, to enhancing the effects in our game and even novel ways to reduce memory or load times. Every member of the team has their specific passions and areas of expertise and they often manifest in unplanned ideas that make the game so much better. I spoke about some of this at last BlizzCon, but it is an ongoing thing. We even have a week or two a year where we all think "outside of the box" about things we can add to the game that are awesome.

maestro

absolutely. ana, doomfist, sombra are all examples of this...

The one that immediately comes to mind is Symmetra's wall ultimate. When we (designers) first talked about it, it seemed pretty simple. Lets just take Reinhardt's shield and scale it way way up, that should be fine right?

Well it turns out... not exactly. There were two major hurdles with this thing.

1) On the gameplay front, a lot of our abilities need to know if they can hit something by checking if the object is just 'in range' of it. Well then it begs the question... 'where' is the shield exactly? For example, we wanted Sombra's EMP to destroy the wall, but it was having trouble finding it unless you were near the 'center' of it, since it was such a wide and flat object.

Gameplay engineering needed to come up with some new tech made just for her wall to allow certain abilities, and other targeting, to understand and be able to hit the wall correctly in different places.

2) On the engine side, having a giant effect cutting through your entire screen, that has transparency to it, is very costly for performance. We went through a bunch of different visual effects concepts to try to make it look good while still being cheap enough to not crush people's frame rate. We ended up getting to a good place, but there was still a lingering fear about 6v6 Symmtras, all using ults at the same time. This is why there is a concession in place where each team is only allowed 1 wall at a time. If a second Symm on your team uses an ult while yours is still running, yours will be removed when her's is created.

The biggest challenge, in that regard, was how we handle the budgets around the heroes and how we handle loading. It was tough, but we found a solution that would work (with some compromises). It should be okay for additional heroes, as well. But the imagination of our hero designers never ceases to amaze me, so perhaps I should knock on wood.

Scatter arrow was pretty straight forward, and then used simple geometry.

Making adjustments to achievements is extremely tricky, and we have to do it on a case by case basis. We've had to change a few due to some heroes getting larger reworks where an achievement became literally unobtainable, but making very difficult to earn achievements easier can feel bad for those who have put in the time and energy to earn them.

i just chalk it up to the way gamers talk and communicate.

as a culture (meaning gamers), we use a lot of "mean" language: broken, lazy, cancer etc....

i've just learned to filter that type of communication and focus on whether or not there is an issue. i can get sensitive and hurt... or i can just, you know, do my job.

Yea we've been talking a lot about this, especially as we're working on OW2.

Mei as a tank comes up quite a bit, its something we can maybe try. She would have to get nearly reworked though, for example her secondary fire is far too much dmg and long range to go on a Tank, but its a possibility.

Other ones that come up pretty consistently in conversations are Roadhog as a DPS, Symm as a Support, Brig as a Tank.

i'm so sorry this happened to you. please make sure to report all cases of this. it really does do something and we take this very seriously. we have some ideas we would like to explore to minimize the ability for this to happen. in the meantime, i strongly recommend leaving the profanity filter on.

Hey there, great question! Localization is an important part of Overwatch and audio is just one part of that effort. When we're discussing new heroes, new maps, and lore, we make an effort to collaborate with our diverse global team early in the process. A unique aspect of localization for us is empowering our regional teams to help select voice actors and provide early feedback when we're working with content that is from their region.

In general we try to have VO in both English and native language, and tend to alternate which we use based on if the hero is on your team or the other team. There are logistical considerations, for example not all voice actors are multi-lingual.

Audio plays a huge role in Overwatch gameplay. Scott Lawlor, our audio director, and the rest of the audio team are extremely passionate about Overwatch. Their hardwork and creativity result in not a great overall audio experience but also nice touches like the piano in Paris!

well apparently paris and horizon are the most difficult...

i think bringing blizzard world to life was particularly special for our entire team

There were many challenges, here are two specifics.

We had a hard time coming up with a look we liked for the chain. Most flail reference we found had the chains at a fixed distance. It was hard to find reference of something swinging like a flail, but also having an extending/contracting wire/chain. We ended up swinging an XBOX controller around, controlling the distance of the wire with our other hand.

Once we had a target look, getting it to move like that in engine was an animation challenge. We had to come up with a way to break up the animations and models for gameplay, but still get the look we liked from the XBOX controller video reference.

we really like the idea of cross-persistence. it's something we would love to see for overwatch. the technical and business hurdles are immense but we are very passionate about trying to make something like this happen someday.

<3

That's a great idea and something that we have thought about over the years. Unfortunately, it requires a decent amount of complicated technology changes as our engine and asset pipeline is primarily set up to deliver pre-made content. But it is something that we are continuing to consider.

Actually we've been talking about Tanks, as role, a lot recently. In the past we've had trouble trying to figure out how to make tank heroes like Roadhog compete with big protective tanks like Reinhardt. Lately we've been thinking, what if we just go the other direction? What if Reinhardt was more like Roadhog? What would that feel like for the entire game? Basically, what if all 'main tanks' became 'off tanks'?

This would mean Tanks would be less about pure protection, and more about brawly offense. Barriers would be less prevalent, damage overall would be higher (tanks doing more dmg, and reducing less dmg), which is kind of scary potentially. We've already been experimenting on this front and so far its been pretty fun, though the game does feel significantly different. Players have to play the map angles a lot more aggressively as you can't just rely oin your Tank/barrier for cover.

I'm not sure where these experiments will go but they're pretty interesting so far.

Doomfist could also work, hes kind of a fighting game hero already! He's certainly inspired by fighting games.

this type of question is very hard to answer because we all now have the benefit of hindsight.

sometimes i wish we had kept offense and defense heroes (rather than merging them into damage) and only allowing offense heroes to be played on attack and defense to be played on defense. i know that idea might sound crazy and stupid but i think it would have allowed us to more "all in" on some "niche" hero designs like Torb, Symmetra and Bastion. much of the issue we face with the hero balance on these heroes is our player expectations that these heroes should be fully viable on all situations, and that really wasn't our intent or vision for the game. we believed that with hero swapping, people would be more willing to adapt to "solve the problem". we're fighting some amount of human psychology and game design, i know.

the sombra arg is another standout... lol

This comes up a lot so i'd like to touch on this.

I would love to get the roster to 33% each, that's certainly a goal. That said, I'm worried about going a very long time without any damage hero representation. If you look at the release history, we've been trying to push it towards an even roster without just ignoring damage heroes entirely.

Not really... though as i've mentioned a couple times here already maybe we could have pushed harder to iterate on Symm as a healer instead of a damage dealer. Maybe we can still revisit her as a healer.

i think this is a good discussion. but i think a lot of the good points get lost in the echo chamber of the internet. i'll see a good thread, video etc about power creep but then there is a lot of "group think" that happens and you have a bunch of people who think they are agreeing on something, when really they are not.

"power creep is bad!"... "btw buff roadhog, revert rein earthshatter nerf, buff pharah, you should TOTALLY be adding way more new heroes... and they all better do new cool stuff...." ... "oh yeah and power creep is bad!"

personally, i prefer when the conversation is more focused. is time to kill to fast? do we feel like CC is too much? too much/little healing? and then apply all of these to individual heroes....

we've added 11 new heroes to the game since it launched. obviously, the game is going to change when you do that. we just need to evaluate each hero on their own and make adjustments.

my personal feeling is that i would like to see less barriers and cc and see the game trend slightly more in the fps direction than the moba direction

Hey there, I can tackle a couple of these!

We haven't seriously considered a public bug tracker. As you call out there are a few benefits that would come from having one, there is a major investment required to manage that system, and process changes for the entire development team. Our development team and QA folks are regularly reviewing bug report forums. It may not seem like issues are noticed but they're almost certainly being tracked.

We don't have a lot to share about OW2 today, but as a general strategy we fix bugs in the earliest version of the game that makes sense. We wouldn't intentionally hold off on a bug fix for OW2 unless there was some major architecture changes required to make the fix.

it's random

Hey there, that is also one of my favorite talks! There are a few GDC presentations just a bit after we launched Overwatch, if you haven't seen them all you should check them out! Right now we don't have anything planned for conferences, I think we'll need to see the impacts of covid19 further. Cheers

Some thoughts from me:

A passion for games (and Overwatch, in our example) is obviously a huge plus.

Experience is great, whether you gain that through your job or through hobby work outside of your job. That can apply to engineering as well as to experience working in a team. If someone has demos of their work, that goes a long way towards standing out.

For us, familiarity with C, C++, C# is big as is familiarity with some of our content creation packages (like Maya, Photoshop, etc).

And we're always looking for individuals that bring a different viewpoint to our team. We find the best ideas/solutions tend to come from a diverse team working together to solve a challenge or implement a feature.

We've certainly seen a lot of requests to reset MMR and SR for competitive play, but there are a lot of negatives associated with actually doing it. For instance, a hot topic amongst the community is Smurfing and how they can disrupt games by being much higher skill than everyone else in the game. Well, resetting everyone's MMR would basically have the effect of turning every Masters and Grandmasters player into a smurf with an incorrect SR. Match quality would be really bad for months, or possibly even longer.

Everyone has streaks of wins and losses, and most active players do hover around the same SR. Honestly, the trick to improving your SR is not actually worrying about your SR, but just playing games with a mind towards improving your play. If you have a favorite hero, pick an aspect of their kit to concentrate on improving. Play some games where you concentrate on more macro-oriented skills like making sure to attack together with your team. Think about the hard questions like "why do I typically die?" and make changes to try and prevent them. If you improve your play and skill, your SR will rise, too.

Good luck!

we read this subreddit all the time. if it's a great idea, it will get visibility

Hey there! Recently we added the ability to specify the server location when creating a custom game. Currently we don't have any plans to expand this to Quick Play or Competitive, but we do see people ask for this occasionally. We have some concerns with how this would be implemented, in particular trying to avoid win trading or match fixing by abusing low population regions.

If you're on PC you can select Americas, Europe, or Asia for your region when you launch Overwatch. That will limit you to servers that are part of that region. Middle East is currently part of the Europe region, you could try to use Americas or Asia to change things up.

Cheers

I think one of our core processes that helps us be successful is finding ways to fail early and fail fast. For us this usually means prototyping abilities with really rough blockout art until we have a kit we generally like. Keeping things in this rough state means it is a lot easier for us to change direction or iterate.

thank you so much for saying that. warms my heart. but i always remind people: overwatch is made by a whole team of amazing developers... i am just one tiny and equal part of that. but thank you.

Remote support is challenging. Our team regularly works together to support each other, we have chat channels for help, and people utilize that regularly for support. While engineers are often helping the team out with connectivity issues, it really has been a team wide effort to band together and solve problems.

This comes up fairly often, and in the past we have done a few things to nerf ultimates already. We've nerfed the ultimate costs across the board twice now, to try to make sure they aren't coming up too often. We've tried to adjust ults that feel like they don't have enough counterplay, such as Sombra's EMP getting a longer delay before going off.

Overall this is something we're very aware of and we're trying to make sure ultimates still feel like big moments within the game, but not just purely being an 'I win' button. Abilities like Baptiste's lamp (which btw i find funny the community calls it this, since internally we literally used a lamp model while testing it) were created to help open up more options to counter ults without requiring your own ultimate. Ideally it should feel great to land a big ultimate, but also feel great if you can work around or straight up counter an enemy ultimate, which is also why we made a change a while ago which made it so if you ever kill or stun a hero who was starting an ult, they always lose it (previously it didnt 'spend' thier ult meter until after the ult actually 'started' in a lot of cases).

In general this is a very tricky question and its something we've been constantly tweaking and iterating on over time.

One time I accidentally replaced Ana's dart projectile with Ana's ragdoll. That meant any time I shot Ana's rifle a dead Ana shot out of my gun hehe.

All parts of the team collaborate in the creation of the heroes. Heroes are usually available to play in team playtests months before they will come out. So we all get a look into them in an extremely early state.

The game engine is a good bit more flexible in development than it is in retail, so often the designers have what they need to make something work; we just need to make a pass to control the performance and efficiency of the heroes so that they will function well on all our supported hardware. On occasion, like with Echo, we need to do more significant rewrites of systems so that we can get things within a shipping budget.

creating new heroes is a lot of fun. our process is interesting in that new heroes can come from story (76, ana), design (pharah) or art (winston). we really inspire each other with ideas and we'll add different heroes for different reasons. we're always talking about "what does the game need right now"

Making Overwatch a better, more positive place to play for everyone is super important to us. While we've made a lot of progress with player reporting, thank you messages for reporting bad behavior or abusive chat, endorsements, etc., we're always looking to do more.

We definitely know it can be difficult for women of all ages to use voice chat, as the abuse and harassment takes away your enjoyment of the game (and as you noted, your SR!). I hear this from the women on the OW team, across Blizzard, and in our community.

There are some initiatives under development that we hope will directly address these issues, but I can't quite talk about them yet.

In the meantime, everybody please continue to report players who are being abusive through the in game reporting mechanism, whether they're being abusive to you personally, or anyone else. The reports do matter, and as a community we can make Overwatch a more positive place for everybody.

On the lethality/ttk front:

Its hard to know exactly if lethality has crept up in the actual hero data, if the meta and/or player skill has pushed it up further than it was, or if its all just perception, but we've tried a couple internal experiments where we've tried lowering dmg and healing across the board to see how it feels. That experiment didn't quite work as a simple experiment, because there are a number of heroes that have certain thresholds that break if you lower damage etc. We'd have to do something more methodical to test it properly but I can tell you for sure the game doesn't feel very good when lethality gets too low, even if healing also gets low.

On the healing front especially, once health bars are moving up and down a lot more slowly, it makes it really hard to feel like you're "saving" people in clutch situations (unless you use an ult or something of course).

Interestingly, on our list of experiments to try is to make the game significantly more lethal than it is right now. I'm not sure how it will play out (we haven't tried it yet), but the goal would be to allow for more 'carry' potential, as a great player would be able to more easily 1v2 or 1v3 enemies. Again this would just be an experiment, but its interesting to think about and potentially try.

All of this isnt to say we couldn't try to make changes to lower damage across the board somewhat, but we'd have to be very careful not to push that too far as the game really does feel and play a lot worse when the lethality drops too low.

This is a reasonable request. I'll look into it.

Sorry no this would be part of other changes, like Fade still causing your team to phase out (like the last experimental card).

It was a fun experiment, but it was too good as a dispeling/cleasing tool and far too good at countering zarya/sigma ults. We're going to try a version where she can still phase out her allies, but she doesn't remove debuffs (like ana nade) and it wont break people out of these ults. You can, however, learn to time it vs. sigma ult to try to cancel the damage part of the slam, or just reduce incoming damage while stuck in Zarya ult, which can still be very powerful (maybe too much still? we'll find out)