The Art of LiveOps

The Future of Cloud Gaming w/ Kim Swift: The Art of LiveOps S3E8

June 13, 2022 James Gwertzman and Crystin Cox Season 3 Episode 8
The Art of LiveOps
The Future of Cloud Gaming w/ Kim Swift: The Art of LiveOps S3E8
Show Notes Transcript

Kim Swift is Senior Director of Cloud Gaming at Xbox Game Studios Publishing. She is best known for her work on "Portal" with Valve. She has touched a lot of different technologies across the industry, having worked at Amazon, Twitch, E.A., Motive, and Google. 

Kim talks about her hopes and dreams for the future of cloud gaming and what it means for LiveOps; how players can ascribe meaning and lore to benign design elements; and how even the simplest of design intentions can have unforeseen consequences. Crystin and Kim also discuss machine learning algorithms and how they might be used to create even more interesting interactions for players like advanced matchmaking or tailored experiences.

This episode is brought to you by Azure PlayFab.
Visit https://www.playfab.com for all your LiveOps needs.

Support the show

00;00;05;00 - 00;00;06;14
James Gwertzman
Hello. I'm James Gwertzman.

00;00;06;14 - 00;00;09;23
Crystin Cox
I'm Crystin Cox. Welcome to The Art of LiveOps podcast.

00;00;14;20 - 00;00;23;13
Crystin Cox
Hi, everyone. Crystin here and welcome to another episode of The Art of LiveOps. Today I'm joined by Kim Swift.

00;00;23;22 - 00;00;43;06
Kim Swift
I'm a senior director of cloud gaming at Xbox Game Studios Publishing. I actually work with Crystin. My background is pretty varied. I am best known as that like "Portal" lady. So the first ever game industry thing that I ever shipped was a "Portal" back in 2007.

00;00;43;07 - 00;00;58;00
Crystin Cox
She's going to come and share some of her experience in the industry, which has been largely on the design and creative side. But man, she has touched a whole lot of interesting different technologies and initiatives across the industry.

00;00;58;03 - 00;01;21;24
Kim Swift
I actually went to Amazon for a little while. I worked at Twitch. From there I went to E.A. and I was a design director for the studio Motive. Then I went and I worked on Origin and then went to Google. Working on Stadia. As of last year, started here at Microsoft trying to figure out what the future is for cloud gaming here at Xbox.

00;01;21;25 - 00;01;45;10
Crystin Cox
So she is really following a lot of the interesting technological trends in the industry over the last 15 years. So we're going to get to chat with her about her experience and some of her hopes and dreams for cloud gaming and what it means for LiveOps. So let's dive in. And today you're going to tell us what that future is, and then we'll all know exactly what the future of cloud gaming is.

00;01;45;26 - 00;01;46;20
Crystin Cox
Problem solved.

00;01;46;21 - 00;01;57;18
Kim Swift
Yeah, exactly. And then and then I'm just going to solve it for everybody and they're going to jump the gun and I'm just going to be S.O.L. because I told everyone at the secret and I'm just going to be really sad.

00;01;57;22 - 00;02;22;04
Crystin Cox
Yes. The secret of new technology. But in all seriousness, your career, you've been a designer your whole career, but you've always been so close to technology and to new emerging technology. And you had stints working on platforms. You worked at Twitch when they were really just starting to come out into big, big. in a big, big mainstream way.

00;02;22;28 - 00;02;51;25
Crystin Cox
And now you're working in cloud. And I think that in a lot of ways you've really followed the journey of Western game development as far as connected technology experiences. So I'd love to hear you talk a little bit about the changes you've seen from your own perspective from those days at Valve when you know you were shipping "Portal" and then moving on to "Left for Dead" to now working on like what is the future cloud?

00;02;52;01 - 00;02;53;20
Kim Swift
Yeah. So I mean.

00;02;54;03 - 00;03;27;23
Kim Swift
I'll sort of answer your question. To start with, and then hopefully for realsies answer your question. But I mean, for me personally, what I've always found interesting about the connection between games and technology is that games are really sophisticated interesting way to prove out technology that is different than any other industry out there because of the creativity that goes into you know, making that game, appealing to players and then also to as well because of that interactive component from players.

00;03;27;23 - 00;03;53;15
Kim Swift
It just like opens up so many different permutations of complexity that I find super fascinating. And for me, I actually have a computer science degree. I was one of those DigiPen kids. And so I graduated with real time interactive simulation bachelor's degree, which is their fancy way of saying this is computer science degree with the focus on, you know, real time interactive simulation aka video games.

00;03;53;15 - 00;04;26;18
Kim Swift
Right. And so I've always kind of chased that intersection between the creative and artistic and technology and, you know, over my career, which like at first blush, makes absolutely no sense. I've always just sort of looked at it from the perspective of what is an avenue in the industry that I haven't done yet. That I can further understand better because I want to see the future.

00;04;26;24 - 00;04;56;15
Kim Swift
Right. And so it's really difficult to try to predict what's new or interesting or next and what's going to have legs for a long period of time without having a better understanding of the different aspects of the industry, including, you know, what is the player journey too as well, which is why I was so interested in Twitch, because we have this new way of interacting and talking and communicating around video games.

00;04;57;02 - 00;05;17;07
Kim Swift
And at the time it was like, oh, well, this is just this nice marketing thing. I guess I'll throw it in at the end. And it's like, no, this is a whole new way that players are expressing their community around the franchises and IPs and games that we're making. Why are we not taking this more seriously? Because this is the future.

00;05;17;09 - 00;05;48;23
Kim Swift
Great. Like the Internet is no longer a place where we can look at it and say, Well, it's just this other thing. It's a digital space. No one really cares. Like, no, like, these days, it's just as important as the real meat space, right, to people. And so we should be looking at the ways that that people are communicating and thinking and sort of keeping the mindshare around the creative things that we make, which is a very long winded way of like not completely answering your question.

00;05;50;07 - 00;06;16;27
Crystin Cox
Well we can dive in, I think a little more deeply to there, that's a lot of really great insight and diving a little more deeply to say like, you know, when you were at Valve sort of earlier career in the mid-Aughts, you're making games and then you worked on some games that for the time. Very strong communities. Very strong feelings about those games that have carried on for years and years.

00;06;16;27 - 00;06;36;18
Crystin Cox
And years. And it's interesting to me to hear you talk about going then next to dive even deeper into that community layer and like spend even more time in that space. I am curious to hear you give a little bit more detail around how that relationship with players has changed for you.

00;06;36;26 - 00;06;55;11
Kim Swift
I mean, so when I first started my career, I'll be honest, I didn't really think about community because at least with "Portal", you know, is the single player game we were, the team and I were like approaching it from this perspective that probably no one's going to play this. And if no, it's going to see it.

00;06;55;19 - 00;07;17;18
Kim Swift
Let's just do whatever the heck we want to and be silly. And, you know, at the time, we were all in our early 20s, just fresh out of school and so we were just like, Let's just be ridiculous, right? Because no one's going to see this thing. And then it blew up in a way we had not at all predicted in any way, shape or form.

00;07;18;07 - 00;07;40;17
Kim Swift
It was very jarring. Yeah. And it threw us all for a loop. But I think realizing like that was like a dawning realization moment for me that, oh, wow, this thing that I made. That again, I didn't think anybody would notice or care about it, and we were just doing it kind of for fun. And no one really was paying attention to at Valve.

00;07;40;17 - 00;08;17;16
Kim Swift
They just hired us like, go, go, kids, make this thing. We're like, All right. Seeing how much of an impact it sort of had on the God, I hate this word. The Zeitgeist of Games. It was. Was this really like wake up moment for me of like, oh, wow. You know, the things that we create really do matter. And seeing how people took the lore of the game or tried to say, like, I saw like frickin like PhD papers on, like dissecting the meaning of the things that we did in this game

00;08;17;16 - 00;08;41;22
Kim Swift
that on a surface level, were just like, not that deep, right? But seeing people really kind of ascribe meaning in ways that I hadn't really thought of was sort of like, oh, wow, these things matter. We should pay attention and start to think about from the design perspective, of what we're creating and how that is going to spawn behaviors in a community from different people.

00;08;42;13 - 00;08;57;04
Kim Swift
And so sort of like the very beginnings of me thinking about this idea of pro-social gaming, right? Which is now becoming a more increasing important topic in our industry. But that was sort of that wakeup moment for me back in 2007.

00;08;57;10 - 00;09;23;13
Crystin Cox
Yeah, yeah. I know. It's interesting to hear you talk about that because I think it's so typifies this idea of creating in a liveOps space, right? Like I think that when we have the most success, when we create in a LiveOps space, we open ourselves up to the idea that players can have big ideas too, or that the community can drive a design in a particular way or a vision in a particular way, and we get a little more comfortable.

00;09;23;13 - 00;09;40;23
Crystin Cox
Hopefully with the idea that something you just said, which I think is amazing to be able to say like, hey, we did actually this stuff wasn't that deep. Like, I didn't have every answer. Yeah, you know, let's be honest, like some of these things we just threw against a wall and we saw what stuck. But hey, but players might come and put a lot more meaning into it.

00;09;41;01 - 00;09;46;26
Crystin Cox
That can be wonderful. Like, that can be beautiful. And we can then carry that forward with them.

00;09;47;07 - 00;10;19;16
Kim Swift
Yeah. Yeah. It's both wonderful and dangerous, right? Because what seems to be a very small thing for us could be huge from a community or even just like a co-op group reaction. So, like, the example that I use all the time is that, hey, let's say you've designed a game and it's co-op and you get a loot drop thing at the end after you killed the big boss but the drops are just like one set of drops for everybody, and it's not done per client.

00;10;19;20 - 00;10;40;13
Kim Swift
Right? And so. Oh, so now you've introduced fighting infighting amongst a group of people that maybe don't even know each other when what you're actually trying to do is have people come together. So why did you introduce that one element of conflict that was ultimately not at all necessary for the situation?

00;10;40;13 - 00;10;41;15
Crystin Cox
And unintentional.

00;10;41;21 - 00;10;49;08
Kim Swift
Yeah, unintentional. But like to us, as like designers it's like well, I mean, yeah, sure. Like, what's the big deal? And like, oh, no, it's a huge deal. Oops.

00;10;49;09 - 00;11;07;00
Crystin Cox
Yes, that is a good way to put it because that doesn't I mean, like on the surface, it doesn't seem like a big decision. Yeah. Like, is loot shared or not? And at the end. But really that group dynamic creates a situation, right? Well, now you've introduced politics. Now this game is about politics. Did you want the game to be about politics?

00;11;07;00 - 00;11;09;26
Crystin Cox
Because it is now you can't really change that.

00;11;10;01 - 00;11;19;18
Kim Swift
Yeah. And now there's consequences. And now you got to go deal with that as the designer of this, the unintended consequence of what seemingly is a very small piece of the design.

00;11;19;28 - 00;11;35;10
Crystin Cox
Right? Yeah, man. You know, I would say like, how do you manage that, right? Like it as you went forward in your career and you're just, you know, designed many more games since then. How did it change your approach to making those decisions?

00;11;35;22 - 00;11;59;25
Kim Swift
Yeah, I mean, I think it's in part try to ask yourself as many different difficult questions as possible and internally get as many different eyeballs on the thing and as many different perspectives on whatever you're trying to ship because, you know, hey, I unintentionally did a thing where I've now alienated colorblind people,

00;11;59;25 - 00;12;00;07
Crystin Cox
right?

00;12;00;07 - 00;12;04;23
Kim Swift
Oh, we could have caught that if we showed it to the color blind guy on our team.

00;12;04;23 - 00;12;39;01
Kim Swift
Oopsies, right. There's a lot of things that we can sort of catch internally in our teams and then secondarily is planning on the community that you want to have from the get go and being intentional about the design and building the trust sort of before you hit the ground running and seeing if again, getting more eyeballs on the thing, can you select a representative group outside of your team that you're trying to attract and appeal to and get their eyeballs on it too as well through play testing through UR.

00;12;39;01 - 00;12;51;19
Kim Swift
just try to see as many different avenues of getting different perspectives. So that way hopefully you can catch the problem before it comes becomes something really big.

00;12;52;00 - 00;12;57;04
Crystin Cox
Yeah. And now as you're sort of looking at cloud gaming.

00;12;57;04 - 00;12;57;14
Kim Swift
Yeah.

00;12;57;14 - 00;13;13;03
Crystin Cox
You know, where do you see LiveOps sitting in cloud gaming? You know, sometimes I talk to people about cloud gaming and it's very focused on technology. They're like, it's going to deliver the bits in the cloud and that's really what it's all about. And in some ways that's true.

00;13;13;03 - 00;13;26;22
Crystin Cox
Like, you know, at some level yeah, it's going to deliver bits in the cloud. But it's always interesting to me to talk to creatives about this because whatever new technology comes into play, you're asking like, What is this? What do we think this is going to unlock for us?

00;13;27;11 - 00;13;56;18
Kim Swift
Yeah, I mean, so I think when it comes to Liveops and Cloud Intersection, I think you're going to see some of the most interesting gains in the fact that it is a ubiquitous, quote unquote experience right where I can take a game and have it, so that way it is pushed to my desktop via browser, or I can play it in this app on my PC, or I can go play it on my Xbox or I can play it on my phone.

00;13;57;18 - 00;14;38;03
Kim Swift
Caveat being that you have a strong Internet signal right? But the fact that I have that ubiquitous experience means that there are avenues to design your games and products to think about what modality is the player in on those particular devices. So how is a player going to play my game on a phone versus playing it in a browser? Maybe the experience that they have on their phone is pretty much purely social, right? And so what are the social avenues that I have in my game and how do I continuously sort of deliver against that over time and expand those capabilities on each device?

00;14;39;00 - 00;15;12;13
Kim Swift
And then from a developer perspective, you know, you're seeing a lot of ease of development and continuous development because you are only having to deliver one build. You don't have to split your mindshare between different SKUs. You don't have to test against a hundred Android devices when you're trying to get a new pack of content greenlit. Right. So I think there's something really nice there for development stories of being able to just ship. Ship that sucker once.

00;15;12;14 - 00;15;12;24
Kim Swift
Right?

00;15;12;24 - 00;15;13;03
Crystin Cox
Right.

00;15;13;03 - 00;15;18;12
Kim Swift
Build your updates for one platform as opposed to, Oh, I have to build it for six platforms.

00;15;18;19 - 00;15;22;05
Crystin Cox
Right? Yeah, it is moving that off. Move that problem off to the cloud.

00;15;22;11 - 00;15;49;23
Kim Swift
Yep. And then also because of the ubiquitous experience, I do think there's something really exciting to explore when it comes to ongoing community too as well, because I don't have to download something for an hour in order to play. So if my friend made a really cool UGC level in a cloud game, they could just send me a link and I can bring it up on my phone and immediately dive in at least through Xcloud on our end.

00;15;49;23 - 00;16;16;04
Kim Swift
You know, looking at Xbox. Dive in through Xcloud immediately play the thing my friend sent me, right? So that automatically drops down a lot of boundaries to allow players to be able to share and connect to one another. And I have a lot of ideas about how we can take advantage of that and excited to work with developers in the future to figure out how we can really lean into that experience.

00;16;16;12 - 00;16;55;11
Crystin Cox
Yeah, I think it's fascinating and I think there's a dream that some people in LiveOps have. Like it's not, you know, not everyone does, but there is a strong contingency in Liveops and dreams of sort of like perfect personalization. Like the, I can make a game and then dynamically tailor it exactly for you. Right. And it would be different than the way other people play it, but it would be exactly, you know, for your taste and that there would be a world someday where that personalization could exist in a larger context where we're playing the same game, but we're getting that kind of personalization delivered to us.

00;16;55;16 - 00;17;05;24
Crystin Cox
I'm sort of curious your thoughts about that, especially like with your background looking at things like Twitch, where that's a communal gaming experience

00;17;05;24 - 00;17;47;26
Kim Swift
Yeah. I mean, I think that personalization to me and I sort of see this within the bounds of cloud too, as well, at least in terms of like how I've defined my mandate at Xbox is there's a lot that we can do with cloud AI and machine learning to better understand our players and I think too as well combining that with the potential for increased horsepower because we've got, you know, one or many different pieces of hardware mapped to one person that things are being streamed to or hey, maybe we split off components like the A.I. or the physics, and then we push that to the cloud and have rendering and everything else done on, you know

00;17;47;27 - 00;18;31;12
Kim Swift
the local client, like what can we do by unlocking all of these things? But I think to me at least having sort of been at Google and been at Amazon, I'm really excited about the space of machine learning and what can we look at from a technology perspective of stuff that we use every day like Siri when I talk to my iPhone or Hey, the types of algorithms that are run to predict what I'm going to buy on Instagram and what it's going to serve me based on my customer profile, can we leverage those things to make a better experience for our players?

00;18;31;13 - 00;18;52;12
Kim Swift
So something that I've been thinking about for a long time is how do we leverage those types of systems for matchmaking? Like how do we allow you to, you know, instead of it just being matchmaking, just trying to solve for very basic things. What if we let it go a little bit deeper in terms of like, what is your playstyle?

00;18;52;12 - 00;19;11;23
Kim Swift
What are the modalities of play that you seem to like? Are you an aggressive player or are you a support player? Hey, if you're a support player, maybe we should pair you with somebody that's a bit more like run to the front lines immediately so you have a good person to support. Hey, what are the what do you consider a success criteria for a game?

00;19;11;23 - 00;19;42;12
Kim Swift
What makes you happy playing games cool? We should match you with somebody that we think you would have that success criteria beyond just like, Hey, are they online? And are they playing this game and are they in the same general region? Cool match made, right? Like, are there are there different ways that we can dive and a little bit deeper to understand our players and facilitate those connections to help you maybe find a cohort or a community online?

00;19;43;00 - 00;20;04;15
Kim Swift
That was something that was really interesting to me while I was working at Twitch and observing the behavior of these communities. Around, you know, typically like one person who's the streamer, but watching how those communities get to a point where sometimes it almost wasn't about the streamer it was like, let's coordinate to go play games and watch movies together.

00;20;04;22 - 00;20;28;10
Kim Swift
We came together as a cohort because we liked watching this one person because we felt like, oh, maybe their personality matched ours a little bit or that they're playing the games I like. But hey, I also, in addition, found this community that I could go be friends with and so how do we start to replicate some of that in our games and our platforms?

00;20;34;10 - 00;21;10;04
Crystin Cox
[ad music] The Art of LiveOps is presented by Azure PlayFab. PlayFab is just one of the many Microsoft tools that game developers use to build games on any platform. From the industry leading development platform, Visual Studio, to the award winning physics of Havok, Microsoft makes the tools that power game creators. To learn more about all the ways Microsoft can help you on your game development journey, head over to MicrosoftGameDev.com to find documentation pricing and other developer resources. That's MicrosoftGameDev.com. [end of ad]

00;21;10;04 - 00;21;34;05
Crystin Cox
Welcome back. You're listening to the Art of LiveOps. And today we are talking to Kim Swift who is working on cloud gaming direction for Xbox Game Studios Publishing right now. And we're just chatting about her experience in the industry through the state of technology and ubiquity in gaming right now and what she sort of sees for the future.

00;21;34;10 - 00;21;37;10
Crystin Cox
So let's dive back in and continue that conversation.

00;21;42;04 - 00;22;05;03
Crystin Cox
Yeah, this is a fascinating topic to me. Like, I'm really interested as well in seeing what could happen with machine learning in games. And part of it is because I as a consumer on the Internet, have I feel like seeing this degradation of like suggestion engines, like they've gotten much, much worse in the last ten years.

00;22;05;12 - 00;22;29;05
Crystin Cox
And like as someone who has like some understanding of the way data processing goes, I get it right. The goal for a lot of those things, like you're saying like if you want to serve me ads on Instagram, the goals that are at play there and the very limited behavioral data you have about me, which is very, very, very small, means it's mostly just a regression to the mean, right?

00;22;29;05 - 00;22;49;09
Crystin Cox
So like, you know, I look at it as like people sometimes joke like, oh my gosh, it's like they know everything about me, but like I would say nine times out of ten, I'm like, they know nothing about me. None of these are relevant to me in any way, shape or form, which makes sense because as I said, it's mostly a regression to the mean, right?

00;22;49;17 - 00;23;26;00
Crystin Cox
It doesn't actually know that much about me. It knows, hey, are you this age, this gender, and live in this place? Here's the most average thing. And we're right enough, right? We're right enough. I'm fascinated to see what might happen with machine learning in gaming, where we actually have incredibly rich behavioral data. And yeah, so I'm really interested to see, but I'm also curious to hear your thoughts on can we avoid that massive regression to the mean because it's not like we're immune to that in games like I've certainly seen in design and systems design, we can sometimes fall prey to that very easily, right?

00;23;26;00 - 00;23;31;02
Crystin Cox
Like, well, this works for the average player and guess there is, but no individual player is an average player.

00;23;31;03 - 00;23;53;16
Kim Swift
Yeah, for sure. I mean, I think ultimately the barrier for us right, is we tend to segment by game, right? And we only really think about it in terms of our game, which is again, why I think the industry overall has been behind the curve in terms of adapting to things like Twitch, right? Because we're like, oh no, it's just our game.

00;23;53;26 - 00;24;16;29
Kim Swift
And that's the only thing that we care about. And we sort of heads down on that game and we need to actually be thinking of things at the platform level and coordinating across games and like what does that taxonomy look like? How do we start to define things? What is the commonality Venn diagram across all these games where we can judge? Do we segment by genre?

00;24;17;05 - 00;24;30;11
Kim Swift
Like, I think it's a huge like massive data analytics problem at a high level that we tend to be terrible at in the games industry. Like we're just bad at it.

00;24;31;11 - 00;24;50;22
Crystin Cox
So it's a little bit. So again, I think it's a really hard problem Sure. I mean, I can use like music as an example, right? And say like, okay, we have all this data for people listening to music on Spotify, but there's a joke, right? Like go start on any song at Spotify. Count how many songs it takes you to get back to.

00;24;51;07 - 00;25;21;10
Crystin Cox
Just one of the songs is in the top ten of all listened to songs. It has nothing to do with your taste. It's in the algorithm. Like we're in this weird, uncanny valley to some extent with learning algorithms where the algorithm can't help itself. It's like you can start on the most bizarre, you know, you know, instrumental, you know, African drumming, but you're still going to get to the latest hit songs from the Weeknd in about five steps, because the numbers are there.

00;25;21;10 - 00;25;28;29
Crystin Cox
Most people are listening to that song because it's the most popular song. So I will forgive us a little bit and say, like, it's a very hard problem, right?

00;25;29;06 - 00;25;55;03
Kim Swift
Yeah. I mean, devil's advocating that, though, too. There's also an incentive for those platforms to show you the most popular thing because they know they're going to make the most money off of those popular things because of the marketing deals that they have behind the scenes. Like there's always it isn't just purely the algorithm necessarily. There's always an incentive of, I want to show you the popular stuff because we're probably going to make the most money off of the popular things.

00;25;55;23 - 00;26;19;07
Kim Swift
I know that this this is like there's like a huge drama in Twitch-land right now regarding this particular thing. Because they're trying to give suggestions of who you should watch next. And so the drama right now is that the folks that are, you know, getting those pop ups at the end of their stream to go suggest people go send them to somewhere else.

00;26;19;16 - 00;26;51;20
Kim Swift
It's the most popular streamers, right? It's not anybody from their recommended list. And so you're seeing a lot of very upset people because they're like, I would never recommend this person to my group. They're actually the antithesis of my stuff. They're toxic there, etc. Like I would never recommend you to go watch them watch the stuff. Why are you making money sending my audience that I worked really hard to build over to this person that I like despise?

00;26;51;21 - 00;27;16;01
Kim Swift
Like what the heck. And so I think, you know, going back to what I said in terms of like consulting the community in which you are trying to make a product for, I think there's a lot to be said there of how we are going to continually iterate on a particular system. So that way we make sure that we're training something with a wide breadth in mind.

00;27;16;11 - 00;27;35;23
Kim Swift
And also keep in mind. Yeah. Like at some point we do need to make money from this, but how is this potentially damaging our user base by trying to direct them to the thing that we're trying to make money off of? You know what I mean? It's this like back and forth dance of do we actually understand our community well enough?

00;27;35;24 - 00;27;37;26
Crystin Cox
Yeah. To really get to.

00;27;37;26 - 00;27;38;20
Kim Swift
Make money off of them.

00;27;38;21 - 00;27;59;07
Crystin Cox
Yeah. Yeah, it is a excellent question. Right. Do you think games can help themselves? Right. Yeah. I always when I talk to you about my monetization designer and I was like, make sure your customer is the customer right? Like you're going to have an easier time if you know the people you're trying to please are the people who pay you, right, because that's going to, you're going to be aligned a little bit better.

00;27;59;07 - 00;28;26;26
Crystin Cox
But but yeah, I think it's basically what you're saying to kind of bring it back around to your thoughts about social matchmaking. Is it that... that is so exciting to me because it is an idea of like using machine learning and AI to take into consideration the player preference. Right. Instead of just... like you said, Yes. There are some some incentives at play around music suggestions, of course.

00;28;27;27 - 00;28;44;17
Crystin Cox
But then there's also just you don't have that much behavioral data. Like all I've got is you listen to this and then I've got to map that. Right. And in games, there is an opportunity potentially to get so much more and even just ask for more for sure.

00;28;44;25 - 00;29;08;17
Kim Swift
But I also think the algorithms are you're like pooh poohing, like saying, oh, well, they don't have as much data as as we think they actually...they have some data, right? Like so for example, like proximity, not just like, hey, what region are you in? But like, hey, if you've got multiple Facebook users in a household, it actually judges the proximity between the humans in the household, too, as well.

00;29;08;29 - 00;29;24;06
Kim Swift
And so, like, it increases the likelihood that you are going to see something like if your partner is browsing, looking for something on Amazon, it is likely to show up in my suggestions because we share the same household like that. That's a thing

00;29;24;06 - 00;29;27;08
Crystin Cox
There is... Yeah, that's a good point. There is maybe like.

00;29;27;08 - 00;29;28;20
Kim Swift
There is some [insidious] stuff going on.

00;29;28;20 - 00;29;54;06
Crystin Cox
I'm simplifying it probably way too much. But but for sure, I think there are some fascinating opportunities in games around like about that social like, you just said it. I think like asking the question why would you want to play together? And then actually trying to answer that question and say like, well, maybe they have the same ideas about what, you know, what makes a good time or maybe they have the same goals that they want to achieve.

00;29;54;06 - 00;29;55;12
Crystin Cox
That's exciting, I think.

00;29;55;16 - 00;30;18;09
Kim Swift
Yeah. Yeah. I think, you know, one thing that we can start being a little bit more diligent of in the game industry is really starting to understand, you know, the psychological modalities of our players and why they're doing things You know, one of the first questions that I, I started to try to answer while I was working at Twitch by like interviewing a whole bunch of content creators, I was like, why do people watch?

00;30;18;22 - 00;30;39;06
Kim Swift
Why do people play? And, you know, why do you pay money to continue to watch? Right. And I think we can do a better job of asking ourselves those questions from a LiveOps perspective and figuring out. Whoa, is there a way to custom tailor this a little bit more based on those motivations?

00;30;39;06 - 00;30;56;13
Crystin Cox
Yeah, that's a good call. You know, like just even asking that question, because sometimes it's easy to get caught up in and many game teams, I don't blame them. They're just trying to survive. They're just like, can I get the game out the door? Can I get enough attention to exist? But it is very easy to fall into this trap and be like, What are we trying to do?

00;30;56;24 - 00;31;18;15
Crystin Cox
Increase engagement. Okay. That's such a blunt instrument. Like, why? Why are you increasing? Why would players want to spend more time in your game? What are they doing when they do it? And I actually think this a lot of the things you're investigating around cloud and ubiquity and even some of the machine learning stuff I think opens up the door to much

00;31;18;15 - 00;31;40;14
Crystin Cox
more interesting questions like that, I guess, because it becomes like, well, how are you engaging and what exactly are you doing and why are you doing that? And how does that feed into a larger relationship you have? Because we're not fighting so much anymore to be like, can we get you to spend that precious 2 hours in the evening on the couch with us?

00;31;40;14 - 00;31;43;04
Crystin Cox
It's like, nah you can be engaging with us any time.

00;31;43;13 - 00;32;02;29
Kim Swift
Yeah. And then also you're probably engaging with multiple things at the same time too, as well, right? Like, hey, you know, from a cloud perspective, if you're playing on your phone while you're also watching TV, like, that's cool. Yeah. What, what does that player scenario look like? And, and how should we design knowing that that is the use case for phone?

00;32;03;00 - 00;32;26;27
Crystin Cox
Yeah, that's a great question. It came up for us in "Flight Simulator", right? Like a lot of "Flight Simulator", people who haven't really experienced "Flight Simulator," is just flying a plane and a lot of flying a plane is you moving in the sky at the same altitude over long distances. And so we did think a lot about the scenario like, well, you're probably doing something else you know, you're probably not just doing that.

00;32;26;27 - 00;32;28;11
Kim Swift
Yeah. You've had autopilot.

00;32;28;11 - 00;32;35;07
Crystin Cox
Yeah. But people do it and they spend lots of...there must be something rewarding and engaging about that experience.

00;32;35;14 - 00;32;48;17
Kim Swift
Yeah, for sure. I think the interesting there too is like never discount what is rewarding for players too as well because you will always be surprised and be like, Oh, you like that? Okay, I guess we can design for that.

00;32;49;02 - 00;33;06;05
Crystin Cox
That is a good point. And that's something I love about the video game industry, honestly, is I think we're better than a lot of other people about not being judgmental about that. We're not going to come here and tell you how dare you want to sit on autopilot for 6 hours flying over the Atlantic we're like, Oh, interesting, you want to do that?

00;33;06;05 - 00;33;09;08
Crystin Cox
That sounds, why? We're legitimately curious.

00;33;09;21 - 00;33;12;29
Kim Swift
What is the reward cycle for that? Can we make it better?

00;33;13;03 - 00;33;26;02
Crystin Cox
Yeah, that's kind of nice. Oh, I know. We're this has been a really fascinating conversation. We're getting close to time, so I want to make sure I ask you our favorite question, which we ask everybody, which is,

00;33;26;02 - 00;33;26;15
Kim Swift
okay,

00;33;26;15 - 00;33;30;05
Crystin Cox
do you have a LiveOps disaster that you can share with us?

00;33;33;05 - 00;33;57;17
Kim Swift
[sigh] I'm probably really boring. I don't know if I have like an absolute disaster. I mean, okay, I can talk about this one a little bit. So I mean, at E.A., I came in sort of last minute on "Star Wars Battlefront 2". I would like to caveat that our team worked on the single player.

00;33;57;27 - 00;34;00;08
Crystin Cox
Right. Gotcha.

00;34;00;24 - 00;34;35;15
Kim Swift
But there was, you know, a bit of a kerfuffle around the monetization scheme for how that they were running multiplayer. And for that title, you know, we there were actually three teams that were working on the title. There was Motive that was based in in Canada. We were working again on the single player. We had a team that was in the UK working on the flight experience and like piloting the ships.

00;34;36;17 - 00;35;10;23
Kim Swift
And then there was DICE that was working on the multiplayer scenario and so we had sort of some visibility of what, you know, DICE was planning on, you know, their LiveOps plan, how they were tying monetization to their characters in multiplayer. And like for us, we were worried pretty early on, but it was it was interesting being a developer to see this happening in slow motion and having absolutely nothing that we could do about it.

00;35;11;24 - 00;35;12;09
Crystin Cox
Yeah.

00;35;12;09 - 00;35;47;07
Kim Swift
Because it's like, well, you worked on the single player, you have nothing to do with this. Like just go back to what you were doing or like, Oh, okay. But can we be worried? No, we can't. All right. Okay. And, and like from that perspective, I know it was, it was definitely an emotional hit for the team because we were really proud of what we created and seeing the narrative of this game where all of the components were actually really well done and sort of seeing the narrative becoming only about the monetization.

00;35;47;07 - 00;36;13;23
Kim Swift
Right? That this game is a rip off and how dare the greedy devs try to be making money. What is wrong? I paid my money. I should get all of the things immediately you know? I think that was it. Again, it was just this, like, really interesting and also depressing moment of watching what a community can do to the narrative of the game.

00;36;13;29 - 00;36;36;01
Kim Swift
Like my dad, like I gave him like an EA hat or something like that. And he was like at a feed co-op store in the middle of frickin nowhere because they live on a farm and had someone come up to him and be like, EA, that's a terrible company. Did you hear what they did with "Battlefront 2? God, they're the worst.

00;36;36;02 - 00;36;39;29
Kim Swift
My dad's just like, I just wanted by some hay.

00;36;40;25 - 00;36;59;29
Crystin Cox
But it is really fascinating too because I know it is really interesting to see and I think a lot of people have spent a lot of time like asking like, you know, what could have been done differently in that situation? Blah blah and that's you know, I don't think, you know, it's really hard. Those are hard situations. It's from the outside.

00;36;59;29 - 00;37;02;02
Crystin Cox
And at this point, it's like beating a dead horse too.

00;37;02;02 - 00;37;26;18
Crystin Cox
Yeah. Oh, but I think that angle of like saying like it's really interesting. You just have to sort of reinforce something you said earlier like getting those perspective and then like listening to those perspectives and, and being willing to hear sometimes what you don't want to hear. Right? Like the harsh reality of like there's going to be stuff that you do that people aren't going to like.

00;37;26;21 - 00;37;46;26
Crystin Cox
It's just better to go in with an open even if you decide, you know what we're doing. And anyway, which I look at like I'm not saying be totally reactive and like oh, you know, go check with your most loud forum or Reddit users before you make every move. But just being able to have that perspective is so useful because then you can prepare for that a little bit better and maybe not, maybe not get as caught off guard.

00;37;46;26 - 00;38;18;24
Kim Swift
Yeah, and, and I think it's also to like don't underestimate the momentum that a community can have around your game. Right? And if you're going to take that risk, be ready for it. And how what are you doing to prepare? If you still want to take that gamble, you're like, no, no. Like based on feedback that we're getting, you know, hey, maybe this is not the best thing we should do, but we have to roll ahead for X, Y, Z reasons.

00;38;18;24 - 00;38;37;10
Kim Swift
Okay, we got to do it. All right. What how are we mitigating this? Knowing that this might come back on us? What are we doing to prepare ahead of time? Because we got that warning. And instead of just, like, crossing your fingers it doesn't happen. When all signs point to. Yeah, that's going to happen.

00;38;37;10 - 00;38;58;27
Crystin Cox
Yeah, exactly. Like and it's hard, like, I will say, it's not always easy to have that perspective right. It's you get a lot of feedback. I will say that too. Like when you're making a video game, everyone's got an opinion and all over the map, those opinions can be, but it is, you do want to try to be realistic about what's going to happen.

00;38;59;00 - 00;39;00;07
Kim Swift
Yeah for sure.

00;39;00;07 - 00;39;17;22
Crystin Cox
And I often fall back on asking like okay put myself in the player's shoes. Like what am I going to think? You know, because they're never going to give you credit or cut you slack for a problem that's your problem as the developer. You know they're not going to go, I don't like this, but you know, I bet they're under a lot of budget constraints.

00;39;17;22 - 00;39;19;18
Crystin Cox
That's not the player's problem, right?

00;39;19;18 - 00;39;25;03
Kim Swift
In fact, most players don't believe what the budget is that you're like, oh, this is a really low budget. They're like, what?

00;39;25;06 - 00;39;25;28
Crystin Cox
I know.

00;39;25;28 - 00;39;30;22
Kim Swift
You could get like an amazing game. Just add multiplayer for like a dollar like what? No.

00;39;31;14 - 00;39;40;13
Crystin Cox
And like that makes sense, right? It's not players jobs to know how to make video games, right? Like, that's not what they do. So you can't count on them to understand.

00;39;40;13 - 00;40;00;04
Kim Swift
Yeah. Yeah, for sure. I mean, at the end of the day, yes, it's a creative project that you've poured a lot of heart and soul into, but it is a product. At the end of the day. And people are going to take it, you know, probably more as a product than this amazing creative endeavor at the end of the day.

00;40;00;04 - 00;40;02;19
Kim Swift
And they expect to get their money's worth, right?

00;40;02;19 - 00;40;24;05
Crystin Cox
Yeah. Well, and I also the community aspect is at play. Having to remind people that a lot of reactions from players are actually pretty rational if you think about them from their perspective. They love these things that we make, like they love them and their identities are often tied up in them and they have no control over them.

00;40;24;20 - 00;40;42;24
Crystin Cox
They're at our mercy. We can just make decisions and ruin a thing that is a huge part of their identity. So they do operate from a place of kind of fear that's a bit understandable. So these are the things I tell myself when I try to really be a little more forgiving and understanding when the players explode.

00;40;43;00 - 00;41;02;01
Kim Swift
Yeah. Yeah. And then I guess, you know, that comes back on us to like what are the avenues of communication that we have with our players or some sort of representative of our players, particularly from a LiveOps perspective, because you're just going to keep delivering on this thing, right? It's not like you're going to put it in the box and then you know, like forget about it.

00;41;03;02 - 00;41;22;23
Kim Swift
Like, no, you just have to keep going and going and going. So how do you continue to have that relationship with your players? So hopefully you're at least going in the right vector. Where it's not going to blow up on you and if it does blow up on you. All right, what's your mitigation strategy? So that way you retain your player base, right?

00;41;22;23 - 00;41;43;11
Crystin Cox
Yeah. Sometimes I think it's a little easier on LiveOps games, though. You get a lot more reps, you know, and at some point you learn most things blow over. And if you can build a back and forth with your community, you can build some resilience in them to understand that we are not going to be perfect. Like actually we're going to make mistakes.

00;41;43;11 - 00;41;45;19
Crystin Cox
Sometimes it can actually help in the long run.

00;41;45;23 - 00;41;58;27
Kim Swift
I think that's after you get over the six month hump, right? Because like if you fail enough times in the first six months of your product being live, you've basically nailed your own coffin. right.

00;41;58;27 - 00;42;17;11
Crystin Cox
That's true. You lose everything lose everything. Yeah, this is right. It's been fascinating watching Triple A kind of a step in to the LiveOps space because for so long, the LiveOps space was sort of trying to mitigate that by being like, well, yeah, that's why we don't put huge amounts of expectations on the launches, right? They sort of slowly roll things out.

00;42;17;20 - 00;42;33;16
Crystin Cox
We do these soft launches we do a long betas. We do testing. As that's like slammed into triple-A mentality. That's like, yeah, but the expectation is still that everything is going to be amazing. On day one. That's been, I think a challenge for people to navigate, for sure.

00;42;33;19 - 00;43;09;18
Kim Swift
Yeah. And then also the impact of, yeah, this isn't going to be a one time loss of money. If maybe you didn't like nail the first outing for your title, this is going to be like probably a year and a half of losing money. Before you start making money, at least, you know, from what I've seen, that tends to be the bad taste in triple-A's mouth is like, Oh, so I'm just going to keep losing money after this thing launches for like a year and a half and fingers crossed it starts making money, right?

00;43;09;18 - 00;43;28;29
Crystin Cox
Yeah. Yeah, no. Getting it, getting people to sort of understand that like growth over time model and investment with long tail payoff. It's hard, like, legitimately tough. And honestly, I think like at this point, like I kind of feel like having seen both sides of it, I'm not going to sit here and tell you like one of them is right.

00;43;28;29 - 00;43;34;13
Crystin Cox
Like, you know, a lot of it just depends on like your strategy. Neither one of them is a guarantee they're there is risky.

00;43;34;13 - 00;43;35;12
Kim Swift
Yeah, for sure.

00;43;35;12 - 00;43;38;04
Crystin Cox
Well, this has been fantastic. Thank you, Kim.

00;43;38;09 - 00;43;40;16
Kim Swift
Thanks for inviting me. This is fun. Yeah.

00;43;40;22 - 00;43;45;24
Crystin Cox
You know, hopefully we'll see some amazing things coming out of

00;43;45;24 - 00;43;47;22
Kim Swift
Yeah, fingers crossed.

00;43;47;22 - 00;43;50;21
Crystin Cox
when when you do let us in on the secret of.

00;43;51;01 - 00;43;54;09
Kim Swift
Oh, yeah, definitely. Yeah. I'll just map it all out for you.

00;43;54;11 - 00;43;54;26
Crystin Cox
Great. I'm looking forward to it.

00;43;55;03 - 00;43;56;10
Kim Swift
Maybe I'll write a book.

00;43;57;27 - 00;43;58;22
Crystin Cox
Thank you again.

00;43;59;02 - 00;43;59;12
Kim Swift
Cheers.

00;44;02;03 - 00;44;04;17
Crystin Cox
Thanks for listening to The Art of LiveOps podcast.

00;44;04;17 - 00;44;09;23
James Gwertzman
If you liked what you heard, remember to rate, review, and subscribe so others can find us.

00;44;09;23 - 00;44;14;22
Crystin Cox
And visit Playfab.com for more information on solutions for all your LiveOps needs.

00;44;14;22 - 00;44;15;17
James Gwertzman
Thanks for tuning in.