The Art of LiveOps

IO Interactive: Hitman

December 12, 2019 James Gwertzman and Crystin Cox Season 1 Episode 13
The Art of LiveOps
IO Interactive: Hitman
Show Notes Transcript

In today's episode we talk with a group of folks working on the game franchise, Hitman at IO Interactive. IO Interactive is on a similar journey, taking a hit franchise with a traditional box distribution model and moving to a live model, that many game studios are on today. Listen in for some key takeaways about how to make this transition as smooth as possible.  

Support the show

James Gwertzman (00:05):

Hello, I'm James Gwertzman.

Crystin Cox (00:06):

I'm Crystin Cox. Welcome to The Art of LiveOps podcast.

James Gwertzman (00:15):

So Crystin, I was looking at who we're talking to today, and it seems like the entire Hitman team practically.

Crystin Cox (00:20):

Yeah it's a big group of them. So it's a bunch of the people who've been working on Hitman from IO Interactive. They've been working on Hitman since the reboot in 2016 for that series. So it's a couple of designers and one of the game directors.

James Gwertzman (00:33):

And they're based in Copenhagen?

Crystin Cox (00:34):

It's a lot of Scandinavian names that are on today.

James Gwertzman (00:37):

So we're not going to try to introduce them ourselves.

Crystin Cox (00:38):

Yeah, we're going to let them introduce themselves.

Mathias (00:40):

Hello, my name is Mathias.

Cecile (00:43):

My name is Cecile [Hermatson 00:00:44].

Eskel (00:45):

Hey guys, my name is Eskel [Mul 00:00:47].

James Gwertzman (00:47):

But it should be a really fun podcast because they're going through the same journey I think a lot of game studios are going through, which is they've been building, for a long time, a hit franchise with traditional box product distribution and now they've really moved, I think very successfully, to a live model.

Crystin Cox (01:02):

Yeah, this is pretty exciting. I'm a big fan of the Hitman series. I love Hitman. So, I was really interested in doing this because when they made this transition in 2016, they really went from this long... Hitman has been around a long time. And they went from totally traditional, to an episodic content release schedule with live events.

James Gwertzman (01:24):

Wow.

Crystin Cox (01:25):

Yeah, it's going to be really interesting.

James Gwertzman (01:26):

I can't wait. Okay, let's go chat.

Mathias (01:30):

Hello, my name is Mathias. I have been at IO Interactive for five years. Most of that time I've been a Senior and Principal Level Designer. Also a Tracker, which is a person who is basically a mini-director of a location. Meaning a Marrakesh or a Hokkaido or Sapienza. So a person who keeps track of everything that goes in and out of a location, an ultimate arbiter for that. And now I am a game director. That's my role here so far.

Cecile (02:23):

My name is Cecile Hermatson. I am a senior game designer at IO Interactive and I've been here for a little over four years. And these past four years I've spent most of my time working on all the non-main mission content in Hitman in season one. It was a lot about escalations, elusive targets, challenge packs, and so on. In season two I spent a lot of time making ghost mode, the first multi-player game mode in Hitman and also worked again on escalations and the other live stuff. So all the non-main mission stuff.

Eskel (03:06):

Hey guys, my name is Eskel Mul. I've been at IO since 2012. I came in roughly a year before we shipped Absolution. I started with a title called AI Dialogue Consultant, which ended up being creating the AI dialogue in the game because there wasn't any. And from then on I kept doing that, being a part of the systemic dialogue in the game and I also started doing game design and game features that I thought could be interesting to do. And after doing that during season one of HITMAN I became Associate Game Director on HITMAN 2.

Crystin Cox (03:59):

Great.

Eskel (04:01):

Yeah. Yeah, that was awesome. That's what I said. And I've sort of been a squid pretty much because I still kept doing some of the AI stuff and some animation and ran around trying to be the glue between different departments. And that is pretty much my role in IO.

Crystin Cox (04:25):

Nice. Well thank you guys very much for joining us. I think I told you guys this when we started talking about having you guys on the podcast. I'm a huge fan of Hitman. I love the series. I was really excited in 2016 when you guys released HITMAN, which was just HITMAN was the name. Even though Hitman had been around for a very long time before that. There've been many entries in the Hitman series. But HITMAN in 2016, you guys really tried something different with Hitman. At the time it was released episodically, so there were... It sort of started with just a couple of locations and you were adding more locations over time. And I was really excited about this, I though it was a really interesting, different approach to releasing a game. Can you guys talk a little bit about what led to that decision? Because before that Hitman had been a very traditional box product franchise where you would make a game every X number of years and ship it and move on to the next one.

Eskel (05:29):

I wasn't part of the decision making back then but I can make an educated guess that we needed to release something pretty fast and Absolution had taken well over six or seven years. And so we need to get out there with something and so we thought about this release model, and that could be... We could release something and get feedback from that, from the community. And then gradually release more and more labels. And I think that was the main reason in the beginning to do this.

Crystin Cox (06:08):

And then when you guys started doing it, because I've actually worked on a game that also was an episodic release, and so I know how challenging that can be. Once you guys actually started getting into it, how was that? What were the challenges you guys ended up facing? Either expected or unexpected?

Cecile (06:28):

I think challenge wise I think it was just production. Like the pipeline. It was a new thing to make a live game. So everybody who was already working at IO had only released boxed product before. So getting all the pipelines up for all these multiple releases that overlapped each other. That was a challenge. To get that system up and running.

Cecile (06:52):

But when we had that up and running, it felt smoother and smoother. As an ongoing thing. It was kind of a journey through season one.

Eskel (07:01):

Yeah, there was a mentality change that needed to happen and then I had to say that once we got into that, as a developer I can only recommend this type of release because before that it was very much crunch at the end and then just get it out there and it's done. With this sort of approach we got into the mindset, of course with the first release, Paris and the Prologue, we were super... We also working really, really hard to get that out. But then after a while, we started getting into this rhythm. This cadence where okay, there's always stuff that, "If only I had more time to fix this or fix that." And with this... We started realizing, "Oh we can release something and it's okay it's not 100% perfect, because we can actually go back and fix some of these bugs." Unfortunately I always break new stuff, which you always do, right? But that was, as a developer that is so nice to work under these conditions because you feel like you have a second chance or third chance [crosstalk 00:08:09] it's a race.

Mathias (08:09):

I think that's exactly what happened. All the games I worked on before, you release it, maybe there's a patch or two, and then you're kind of like go and think about the stuff that you never got to do. And here I actually got to fix that stuff because we did the live game with constant updates to the game. So that was a really, really nice experience for me, going into this.

Cecile (08:33):

And it was really beautiful the way we get these reactions from the community. From the players, while still making the game. And we spent a lot of time on the forums and social media, seeing all these reactions, really engaging with the players. And it just felt really amazing to be able to actually react to these things because a game is never perfect. But actually having the time and possibility to go in and change stuff, that was just beautiful.

Eskel (09:03):

Yeah, I remember certain things that we... Because certain players had played Paris for instance, and there was, I think, this team on YouTube who had done this crazy puddle kill where 70 people died. And actually on one of the later levels, some of the guards are talking about, "You hear that thing in Paris where 70 people dies?" "Yeah, that's horrible." So we could actually respond and add content that came as a response from the community, which was beautiful.

Mathias (09:38):

I think going back to the technical stuff, which was super revealing for me. It's the, "Okay, we're releasing Paris, and so we're branching." And I'm like, "Okay, so we're locking content on main, right?" "Nope, we need to release the next one in a month." So I was like, "What?" Like blowing my mind that we're like we're going to continue work on one branch and then as we branch off and continue to fix that branch for the release and then all of the integration requests and all that stuff is jogging that around. That was like an eye opener for me and a crazy thing to do. Having done the complete opposite for 10 years. That was-

Cecile (10:23):

And also a learning curve for the whole team, right?

Mathias (10:28):

Exactly. Like now we really need to be careful here what we do.

Cecile (10:32):

I think towards the end of season one, we kind of mostly had a, "Okay, we know how to do this and we have the mindset." But it definitely took-

Mathias (10:42):

I feel that we can be better at the branching and integrating stuff, but pretty much have a smooth process right now, which is cool.

Crystin Cox (10:53):

Yeah, that's a part that I think a lot of developers don't really think about is that, "Okay, if we're releasing content really frequently, how do we actually get a content pipeline that works for the team?" Because you can't really go, in my experience, you can't really say, "We're going to release this release and we're going to finish it and then we'll start the next one." Because it's just not enough time. You have to really layer them in on top of each other.

Mathias (11:21):

There's no pausing, it's just going to keep going. And that was also a big... In terms of workload, where I as a Tracker on one of the locations and that was for me, achieving a box product. Like that was I really went into we really need to hit the date, we need to release this by the end of May and that's the set date. And then you do that and then, if you do a box product, it's like, putting this big ship through the water and it goes away. And then after that you're kind of swimming around in the back water of that big ship. We're floating around and that's nice. No, that's not happening, because now we are like releasing this big thing and then you go on to the next thing. Because even more stuff coming in, so that's like there's no-

Eskel (12:13):

There's no end.

Mathias (12:14):

There's no rest. It's like it's a marathon. That's a huge, don't spend all your energy on the first thing because there's a second thing and a third and a fourth. And that was a little rough [crosstalk 00:12:26].

Cecile (12:28):

Speaking of that, the finished locations, they're finished, but there will still be features coming in right? Game play systems that might change because of other things because of bugs and because of new decisions, which might break stuff in Marrakesh. Work on release. So we might have to go back several times to fix new stuff.

Mathias (12:46):

And bugs coming in because the template was updated. Like, "Oh shit, now this doesn't work anymore. Guess I have to go in and do that on top of releasing the new stuff." And it's tricky, it's hard to do.

Cecile (13:02):

And we still have to do that. I mean I still have to go back to very first escalation I made in Paris and fix somethings once and a while. I mean it's because it's a living, breathing organism we're working in. It's not a done building that just stands there [crosstalk 00:13:20].

Eskel (13:22):

I think all of the release managers said it at one point, for every four bugs that we fix, we break something new.

Crystin Cox (13:29):

You know, that's a good ratio actually.

Eskel (13:31):

Yeah, it's not bad. But some bugs are... Some of the things we break are more severe than others.

Crystin Cox (13:40):

So my co-host has joined us.

James Gwertzman (13:42):

Hello, James Gwertzman here.

Cecile (13:44):

Hello.

Eskel (13:45):

Hello James.

Mathias (13:45):

Hello James.

Crystin Cox (13:48):

But we can just keep the-

James Gwertzman (13:49):

No let's jump right in. I'm hearing you talk about these problems and I think it really talks about the need to be able to role back quickly and how do you... One thing we hear a lot of people talking about is A, how do you know when you're having problems? And B, how quickly can you role back? We've heard tales of economies gone muck because some drop rate increased too much and by the time they figured it out the entire game was ruined because the currencies were flooding the market.

Eskel (14:16):

Okay. [crosstalk 00:14:17].

Mathias (14:16):

What is the question?

Crystin Cox (14:20):

Yeah, I was mouthing, sorry, I'm not officially part of this, but I was mouthing, we don't role back.

Mathias (14:27):

We never did.

Crystin Cox (14:28):

We never role back, we only role forward. [crosstalk 00:14:31].

Mathias (14:30):

We actually never did role back. That's part of the big update or something.

Cecile (14:36):

Actually moving forward and then fixing in the forward direction.

Eskel (14:41):

There's no going back.

Cecile (14:41):

There's no going back.

Crystin Cox (14:46):

I think that's a mentality that a lot of-

Cecile (14:48):

[crosstalk 00:14:48].

Crystin Cox (14:49):

That's a mentality that a lot of teams take, right? James said some people need to be able to role back, and some people do. I've done it. It's really painful. But you guys have the mentality of no, we just got to keep going forward.

James Gwertzman (15:01):

Forward, yeah.

Mathias (15:04):

In some sense we do role back because we can introduce a feature or a change that is not welcomed by the community and though that doesn't mean we role back to a change list but we role back to a design we had before.

Cecile (15:23):

But technically-

Mathias (15:23):

But technically there's no role back. But we do change features or tweak a vision cone or whatever to make things work as they were before because that was actually better. So things like that we do.

Crystin Cox (15:40):

Do you guys have an example of that? Because I think that also speaks to, you were talking about the great benefits of having the live game, is you actually get to hear from the players. You actually get their feedback and then react to that feedback. So do you actually have some examples of times where you put something in the game and you were like, "This is going to be great and everyone's going to love it." But then maybe there was a different reaction?

Cecile (16:01):

You want to talk [crosstalk 00:16:02].

Eskel (16:01):

I have one. And it's my fault. Might as well. We had a vision cone thing where it turned out that the vision cone was placed in the chest of the MPCs. Which meant that we had all these idle animations on all our different MPCs and when they would turn their head and look straight in 47's face, even though he was trespassing, they would not react. So we... That didn't make them see super smart. [crosstalk 00:16:30] it was confusing for game play because why did this guy, totally looking at me and he's not seeing me?

Mathias (16:37):

And we had designed the game around that, being spotted on locations.

Cecile (16:40):

Being spotted, yeah.

Eskel (16:44):

So we moved the vision cone up into the head, because that's awesome. It sort of mimics real life better but that opened so many problems because suddenly the player got spotted constantly, everywhere because the vision cone was just around all crazy, because they have all these idle animations on them, and even though they're bending down, the vision cone would look backwards and stuff like that. So people got severely punished and [crosstalk 00:17:12].

Mathias (17:11):

It was introduced. It's introduced in a state where a lot of our players had already mastered [crosstalk 00:17:20]. They want to master it even more and now it's everything's random because the idle animations are random and that doesn't fit with the kind of mastery we go for when you actually get to [crosstalk 00:17:31]. There was an angry group of people out there that were like screaming for a longtime. [crosstalk 00:17:42].

Eskel (17:42):

It was kind of tricky to fix, but eventually we got the time to actually fix it.

Mathias (17:46):

We made some tweaks on it.

Eskel (17:48):

And I do believe, I still believe it was the right decision to do.

Mathias (17:48):

Of course it was, it's way better now.

Cecile (17:48):

But it was kind of an, "Oops, what had happened?" But we fixed it. And now it's awesome. Now it's pretty intuitive. [crosstalk 00:18:05].

Mathias (18:05):

No one can complain.

Crystin Cox (18:09):

We just don't listen anymore. I think it really speaks to my skill as a player that I don't think I noticed. I think I was just like, "I'm just bad, I'm sure it's my fault if someone sees me."

Mathias (18:22):

Yeah, well that feels fair that thing, then it's okay I guess.

Eskel (18:29):

We actually have some fans that are very outspoken.

Crystin Cox (18:33):

Yes, of course. Of course.

Eskel (18:35):

And it was mostly the speed runners. It's a better experience. It's not as punishing as before now.

James Gwertzman (18:42):

And for changes like that one, do you actually have data in your analytics system that helps you see the impact? Other than the loud minority flaming you on forums and message groups, do you actually have... Is that an example where you actually saw the effect in data, where player death rates are spiking, or other events coming in or telling you, "No, no, this really is a massive problem, not just the people complaining?"

Cecile (19:07):

No.

Eskel (19:07):

We do have a lot of big data but the problem is we have limited resources plowing through all that data. So this was mostly from our self testing the game a lot, and also looking at the forums.

Cecile (19:22):

And that's also, again, where following the community is super important for us because we get a lot of really, really helpful feedback from these very committed players that really spend a lot of time playing the game. So we use that feedback a lot.

James Gwertzman (19:47):

But that-

Mathias (19:47):

[crosstalk 00:19:47]. It wasn't as big as a problem that was experienced in the real world, but I think we felt it too when we played it. So we played it and we felt, "Okay, this is something we need to deal with."

James Gwertzman (19:55):

That's a really good, actually, segue. Because one of the things we've talked about on this interview with other teams before is when do you use data? How does data get involved? And I think a lot of teams are in a similar situation, where they gather lots of data, but they don't have the resources to necessarily get into it or analyze it as much as they would like. And so I'm actually curious, how do you use the data that you are storing? And when do you, for example, is it just for dashboards or when would you actually put the resource to play to really dive in there and do a deep analysis?

Eskel (20:27):

Well I remember these hit maps that were sort of interesting, where we could tell where players died, especially in the first level, on the Prologue, where they gravitate towards for whatever reason, which was super interesting, and then I think we used some of that data to tweak the experience or tweak the game playing. Because we could simply tell that everybody, why are they dying here all the time?

Cecile (20:53):

And we also used it to look at the different... We have so many different game modes, look at retention and stuff. Like how much do people actually engage with this kind of content compared to this kind of content. We've done some analysis on that.

Mathias (21:10):

I think we've taken some business decisions around what are people playing, what do they do on the maps? So we have been [inaudible 00:21:20], but it's on a pretty broad. Like it's not a super detailed... We've never had the time or the man power or the actual skills to ask the right questions, to get validity or can use it in the right way. I don't think we have that skills yet, to be honest. It's super scientific thing. I think you can use data extremely wrong if you don't know what it says. Like you can find any type of fashion you want and motive yourself into doing something that you just want to do anyway. We never really got into that, to be honest.

Mathias (22:06):

We did use, for one other thing, about the tail end for HITMAN, we released a difficulty mode, a harder mode to play through the entire campaign, and we did look at how players looked around, and where the hit maps were going, and where they failed and where can we make it harder, and where's kills happening of the targets, and what can we do on this, trying not to do something data driven either.

Crystin Cox (22:37):

Do you guys think that's something you'll want to expand as you go forward? Like are there things you really wish you were looking at? Like what are you curious about?

Mathias (22:49):

It's a good question. I think it's super strong tool. I personally need to know more about or get to learn more about where it can be used in the most effective way. I thinking if we were doing a PBP or some type of data experiments it would help tremendously in looking where players are going. Where we are right now, I think it's more about how are they navigating the menu? Can the find the [crosstalk 00:23:27].

Eskel (23:30):

Super interesting place to see how are people and can they navigate the menu. Because it's a complex server [crosstalk 00:23:41]

Mathias (23:40):

Tool, which I know myself is really, really big on. That would be a strong way to use it.

Crystin Cox (23:49):

Yeah, UX is definitely one of the first places people often go, because as you said, it's... Games are just complicated. Like you have a lot of options in menus, it can be really hard to know, are people finding things? Do they know where to look for stuff?

Eskel (24:06):

Well that was actually one of the challenges we found out by releasing staggered season one because we didn't have a lot of content to begin with and we wanted it to feel huge and worth it and all that. So the menu became, it looked super complex, or it looked like there was a lot of stuff, which there actually wasn't really at the point. But now this is one of the game that has so much stuff. That the same menu needs to be kept alive. So it's a problem. We need to see if we can simplify it because [crosstalk 00:24:45]. Yeah, we're seeing a lot of people, I've noticed a lot of people on the forums where they get the game and they're asking the forum, "How do I start? Where do I actually begin?" There's so much stuff to do.

Crystin Cox (24:59):

So we talked a little bit about how you guys changed a lot of the way you did things when you moved to HITMAN in 2016. You guys released HITMAN 2 in 2018 and you didn't go episodic for that one, but you did keep a lot of the live content plan. Can you talk about what you're guys thinking was going into the second game and how you made the changes?

Eskel (25:20):

Yes. The thing was that even though, as we talked about before, that we really, as a development team, really love the episodic release and we also saw a lot of players actually... Well in the beginning they were skeptic because it was sort of a new thing. People had to sort of get used to it. And I think a lot of people who got on board from the beginning, they found out that they really enjoyed it, this episodic release, because our levels are so dense and there's so much stuff to do. So re-playability is a huge part of a Hitman game. So it really fit the IP really nicely.

Eskel (26:06):

What we did also find out though were a lot of people were saying, "Well, I'm just going to wait until the whole thing is released, then I'm going to get it then." And then, but that's a year down the line and people have moved on to something different, something new. So that part didn't really work out for us.

Eskel (26:25):

So the decision was made to, even though we think it fits the game and it's wonderful to develop, I think. Under these circumstances, let's give them everything at once, sort of what we had was certain series also they came out on Netflix or whatever. Where all the episodes they release them at once. [crosstalk 00:26:48]. And then you can binge if you want to. And then we said, "All right, we're going to double down on the live content."

Eskel (26:54):

But this was of course a huge undertaking that we had to do, because we had to go back to the old mentality of releasing everything at once. And we had to also keep season one alive. And we had to do all the new live stuff. That was [crosstalk 00:27:14] pretty tense two years.

Mathias (27:18):

And I think part of the tension was because we saw that releasing the locations one by one was obviously biggest fight in terms of activity and retention. But it was also doing this weekly, daily thing, that we added in between the locations that kept players in and kept players going. So if we can still do that... The thought process was, "We can still have the retention going, without actually... Still binge releasing everything." And so that was the... And I think that worked out well. People are still engaging in the content we're releasing right now.

Cecile (28:06):

And again, that means that like...

Eskel (28:09):

Don't you have a funny story Cecile?

Cecile (28:10):

Oh yeah. I read recently on a forum, this guy, he was just like, "You know what, I really like the episodic model. I actually played HITMAN 2 episodically." So he only just finished it now. He got in on release, and then he made this timeline for himself, like I can play Miami for a month. Or, I don't know exactly what the structure was, but he loved it so much he kind of enforced it on himself. Because he also thought we kind of think too still, that the levels are so nice to experience like that. Having a lot of time for just one location. The problem for him though, the sad part of it, was then he of course missed some elusive targets. Because he couldn't play them. So we have [crosstalk 00:28:57]. Very cool guy, yeah.

James Gwertzman (29:04):

So for people who maybe haven't played Hitman as much, can you explain what is your live content? So you binge released all of your locations, what are the types of content that you do release on an ongoing basis or maybe your limited time events? How do you choose what to put in there?

Cecile (29:23):

Well, I mean, we have all the different game modes. We have elusive targets, which are these one try missions. If you fail, you can't replay. If you die you can't replay.

Mathias (29:48):

Oh the elusive targets.

Cecile (29:48):

Yeah. So you only have one shot, kind of. You can replay before the target is down, you can play three tries to preplan, but you only have one go.

Eskel (29:59):

Once you press start, you are stuck.

Mathias (30:02):

Have you guys tried one of the elusive targets?

Crystin Cox (30:07):

Yes, I have. I have... I think I missed one this season but I think I've gotten all of the other ones.

Cecile (30:16):

That's impressive.

Crystin Cox (30:16):

Yeah, I definitely... It's actually a super great, I really love it as live content because it really forces you to plan, because you only get the one try. And so I'd really like sit down and be like, "Okay, I really got to make sure I actually do this correctly. I got to make sure." I would actually replay the map, at least once, before I would win. Because I don't remember where everything is, so that I know I can absolutely get this done. Which is really great because it's super engaging, it gets me back into the game, and it's a really wonderful tension. Like you know you're only going to get this one shot, so it really makes you... You're not like, "Oh I'm just grinding through a level I've done before." You're really super engaged. Even when it is a level I've 100% and I should know it really, really well, it adds this extra tension which keeps you more interested.

James Gwertzman (31:12):

And we've talked about the importance of predictability in live content, and the importance of giving your players predictable content they can schedule around. Is that something you've plant in your design? Is there a regular cadence to these missions that your players can really build their schedule around?

Eskel (31:30):

Yeah, we had to try for that also, where in the beginning I think they were 48 hours.

Cecile (31:36):

Very short.

Eskel (31:37):

And left on the weekend. I think 48 because it just looked cool that this counter would go to 47 immediately. But it turned out it was too short of a time and also in the weekend, a lot of people are doing stuff besides playing video games and they didn't want to miss out. So we had to take that feedback in and then prolonge it. I think they're a week now, generally?

Cecile (32:06):

Yeah. But we experimented a lot with this feature in season one. It was 48 hours, then it was longer, then it was shorter, then it was kind of testing all the way. It's changed a lot. But I think here-

Eskel (32:19):

The rules also sort of changed a little bit maybe never on release. But I think the idea was in the beginning, "All right, when you press start, then that's it, then you don't ever have a chance again." And we had to go back on that because right now you had to do something illegal and kill someone and then now you cannot restart. But the problem was people lost the connection to the web, or something like that. It felt super devastating that, "Okay, you've failed it."

Crystin Cox (32:47):

Oh yeah.

Eskel (32:50):

That was... I get so engaged myself, actually sometimes I feel the tension is too much. Like I'm shaking afterwards. And sometimes I could have just restarted but I'm so involved in it that I forget that that's an option.

Crystin Cox (33:04):

Too right.

Mathias (33:06):

I failed my first one. [crosstalk 00:33:12]. I know it was a mission I worked on. Like I knew exactly how. Like the elusive target itself but the actual, the location was Marrakesh and I, "Okay, I'm going to go in an do it." And I fucking fried myself. [crosstalk 00:33:27] I stood in the pool and I was electrified and I knew that [crosstalk 00:33:30] I wasn't so nervous.

Cecile (33:32):

We just talked about, just before, like you can remember all the times you succeeded and when you succeeded really well and found success. But you can also remember all the times you failed. And that's sometimes even more of a funny story to tell your friends or your colleagues. Which is kind of amazing.

James Gwertzman (33:53):

I think that's another example of why this is such a great type of live content, is that it does enable that kind of story telling among your community. Your fans will talk to each other. "Oh man, how do you do this one?" "I did this one this way." And that is exactly the kind of engagement it does create that really tight community that you're obviously going for.

Cecile (34:12):

Yeah.

Mathias (34:12):

Yeah, definitely.

Cecile (34:12):

And I think even some people they're so scared of elusive targets that they'll wait, and then they go and check out some of the good players, their play throughs because they're so scared to fail. So then they actually go watch other people play before to not fail. Like completionists are really, really scared of failing.

Mathias (34:34):

We also have a reward for the player, and they want to get them all. So if you miss out it's going to hurt.

Cecile (34:41):

And just that red tile in your menu, for some completionists, that's just like, "No, no, no. Can't happen to my beautiful, perfect game."

Eskel (34:49):

And then there's two experts who go in and brag by doing the most... Like that guy who did this thing with Sean Bean. He managed to get Sean Bean all the way down to the side of the tracks and puts a laser trip wire so when Siarra's car drives past, by, Siarra's car actually triggers the trip wire and he blows up. Crazy stuff.

Mathias (35:14):

They do amazing, creative things with the targets on the elusive. That's very cool to see.

Eskel (35:24):

What else?

James Gwertzman (35:25):

Well you mentioned completionists. I'm curious. So obviously we know there's many kinds of players, there's the completionist, there's the adventure junkies, there's the killers. How much do you think about that when you're designing other missions or content? Do you think about, "Okay, I'm trying to make sure I've always have something for everyone." Or is that not something that you actively put in your minds as you're designing your live content?

Mathias (35:49):

As a location owner, I would say we try to identify the key archetype players and then make everything for everyone in there.

Cecile (36:03):

And I think as a live content creator, because one of the other things we make as a live content mission type is the escalations, which are like these small, very gameplay centric, mini-missions on the existing locations and the design of those, we try to ensure that there's a spread of different types. There's some that are super, ultra stealthy and really hitman-y and some of them are totally crazy with axes and bombs and all that stuff. Some of them are super easy, some of them are silly. So we try to make something for all because we know there's room for so many different player types in this game, so we want to make content for all of them.

Crystin Cox (36:54):

You guys have talked a little bit about it already, but this game has an incredible amount of YouTube videos, of people doing really crazy things, which is one of the wonderful things about a sandbox. But I think also because you've done the live content, there's a constant stream of more and more of these videos. How much do you guys think about that when you're designing your content? Do you think, "Oh man, people are going to make great videos for this?" Because I have to imagine it's been very good for the game because I can go to YouTube and I'll see videos, I mean at this point the game has been out for quite a while, and I'll still see videos, every week more and more videos.

Cecile (37:35):

I think, before we released HITMAN season one, we didn't think about it as much. But then as it was coming out and we saw all this content being created by all these players, we realized, "Oh wow, we really have something that is YouTubable and whatever." So now I think we actually think... I mean, it's not all we think about. But we do have these conversations about, "Oh, this thing would be awesome because I can't wait to see what the players are going to do with this item or in this mission or in this context."

Mathias (38:10):

It could even be something that emerges, that you see that looks like a bug its a bug or something that we didn't want to happen, and then we go, "That's an awesome YouTube video. Let's keep it in there and try to make it happen." Just being open about it. Maybe it's something that looks ugly, but it's fun. We own it, and we say, "Hey, let's do it for the YouTubers and the engage the community with it." So we are very aware of the reality factor of those things.

Eskel (38:48):

[crosstalk 00:38:48]. Sorry.

James Gwertzman (38:49):

Is the suitcase still...

Mathias (38:51):

Yeah the suitcase [crosstalk 00:38:52]. I don't think we actually... That was not intentional.

Eskel (38:59):

That was absolutely not intentional. When you threw the suitcase in the game-

Cecile (39:04):

If you locked on to a-

Eskel (39:06):

If you locked on to a character, it would just follow that character around.

Cecile (39:09):

Around the corner.

Eskel (39:09):

Around corners and stuff like that. It was so bizarre.

Mathias (39:09):

The throwing animation was pretty slow, so if the character is running, and has time to actually run around a corner, but the suitcase is locked onto the character, so the suitcase would just bend around the corner and it was amazingly awesome.

Eskel (39:32):

It's always for the benefit of the player. [crosstalk 00:39:36].

Cecile (39:37):

And that video with somebody standing in the church tower in Sapienza and locking on to somebody on the ground in the graveyard with the suitcase and them throwing it and then crawling all the way down on a drain pipe and getting to the guy before the suitcase. I mean that's great YouTube. That's great YouTube content.

Mathias (39:52):

And the community were actually like, "Please don't fix this. Don't fix it. We want this."

Cecile (40:07):

You know it's not intentional but I love it.

James Gwertzman (40:10):

What about your key YouTubers, the influencers, is there a small subset that you've really built a deep relationship with yourselves? Besides putting out content to see how they react to it, actually chatting with them or trying to precede content?

Cecile (40:22):

Yes, and actually we started for HITMAN 2, we started making... Oh, we actually did it once in season one. Making installations that were designed by some of our top players. In season one we did it with the guy called TheKotti, who is like one of the best speed runners of all time.

Eskel (40:44):

47 also. Where the guy, he spells out-

Cecile (40:49):

He hasn't made an escalation yet. But bigMooney, in season two, we made one with bigMooney and there will be more coming out. And they're super awesome, these guys. I mean, they're really engaged.

Mathias (41:02):

They're really good at thinking [inaudible 00:41:05].

Cecile (41:04):

Yeah, exactly. And they actually sometimes they know our levels and our systems better or differently than we do. So they have these awesome ideas. [crosstalk 00:41:16].

Mathias (41:19):

They have different perspective. They have some really deep knowledge on how the systems work.

Cecile (41:23):

So we really want to do more of that stuff. We actually had a little, when we were doing difficulty mode, we had a little summit, at Copenhagen, at our old office.

Mathias (41:35):

We flew in some key fans and had them try the difficulty mode and talk about that.

Crystin Cox (41:44):

That's cool. We are just about out of time. This has been an awesome conversation. Thank you guys so much for taking the time. This has been really interesting.

James Gwertzman (41:54):

Thank you.

Mathias (41:54):

All right, thank you guys.

Crystin Cox (42:01):

Thanks for listening to The Art of LiveOps podcast.

James Gwertzman (42:03):

If you liked what you heard, remember to rate, review, and subscribe, so others can find us.

Crystin Cox (42:09):

And visit playfab.com for more information on solutions for all your LiveOps needs.

James Gwertzman (42:14):

Thanks for tuning in.