The Reinvention Of Tomb Raider

Okay, all right. We'll kick it off. So, this is about an hour. We might have some time for some Q&A. We'll go through at a high level how we took Tomb Raider, started it, looked at reinventing it, and all the way through the last 4 years to what we have released March 5th just recently. My name is Darrell Gallagher. I'm the studio head for Crystal Dynamics. And to my right here is Noah Hughes, creative director at Crystal Dynamics on and on Tomb Raider. So, we just off the back of a a very successful launch of Tomb Raider. How many here have actually played it or seen it? Some hands. Good. Awesome. Nice. Nice. That's a good start. All right. Thank you. How many here have got a hangover from last night? That's why we're So, excuse any stumbles. Okay. So, thankfully we we had a great great launch in March for this game. And we had this talk prepped prior to launching the game on the reinvention

and reboot of Tomb Raider. And it was successful, so we didn't have to rewrite the talk, which we're very happy about. It could have been a what went wrong. Thankfully, it's not that. So, it really is a broad case of here are some things that we thought went well. Um some pro tips around the project. Um some experiences we had from start to finish that we hope that we can share with you today. So, we're going to go through quick outline here uh from a franchise perspective, how we went about reinventing an icon such as Lara Croft and Tomb Raider. Goals and methodology uh in terms of our high-level goals at the beginning of the project, some of the things we looked to want to achieve. Some production techniques in terms of how we made the project and how we went through each phase of the project at a high level. Creative techniques with Noah in terms of how we went about making a very narrative-driven production. Uh and it involved changing major parts

of how we did things at Crystal. And last but not least, some pro tips that we hope we can leave you with today and maybe take back to your studios and learn from. So, at a high level, it was roughly a 4-year period from start to finish on this project, and it wasn't always perfect. So, while this PowerPoint is nice and slick and we've rehearsed and we'll go through it in an hour, the actual production of 4 years you're going to have bumps along the way. Um so, I just wanted to say that that these are some of the good things we've learned, but any project you do that is special, and we wanted this to be special, is going to be tough and it had bumps along the way. So, this this is uh you know, a consolidated version of that. Digging into the project, it's the biggest game we've ever done at Crystal. Um it's two teams that have worked on this. We did the single player here, uh close to here, 25 miles away in Redwood City, and the multiplayer was done with our sister studio Eidos Montreal in

Canada. We were lucky enough to get uh a huge amount of uh recognition along the way as we were making this game. Uh we had over 50 magazine covers uh and two week three stage demos and over 150 awards and nominations along the way, which on a long development of of 3 or 4 years is is very encouraging and definitely put some fuel in the fire that we were doing something right. Uh it's it's been the biggest week one sales in franchise history. Uh we're only a few weeks into that launch right now, and it's been the biggest opening so far of 2013. So, we're in uh we're happy with the the outcome. We're certainly in a in a place where we feel like we're on our road to achieving everything we wanted to. I'm going to play a quick video right now, which is just a short 60-second sort of recap of where we landed. Mayday, Mayday, there's a lot of cops. Every survivor has a story.

You've got great instincts, Conrad. You just have to trust them. Every mystery has its truth. What the hell is happening on this island? Every saga has its beginning. All right, here it is. So, we're going to talk about how we got there from beginning, middle to end. It was an epic adventure, much like one of Lara's. Uh and we're going to walk you through that at a very high level. Some of our challenges, you know, it's great to be working on a franchise like

Tomb Raider with the recognition and the popularity it has, but it it came with challenges. And with 17 years of history since 1996, um it left us in a position as we finished up three games at Crystal feeling like we had to make some changes. Uh and that's natural when you're dealing with a long running franchise, but we were already encountering this for the first time. You know, as a result of the ups and downs of that long history, you have sort of mixed feelings and loyalty around sort of a fan base. Um we had new contenders at that time in 2008, 2009 when we were starting this in the likes of, you know, Assassin's Creed or Uncharted or God of War 3. There was new generation of hardware, and there were new contenders to the action adventure crown that were bringing their new angles and and really doing some great stuff around action adventure, too. Also, we had a generational change not only in hardware and but it was also in in terms of our fan base and our audience as well. Obviously, you know, 17 years later, the people that may have started playing Tomb Raider in '96 may be different

people today. They may be never played it before. Some people will have and some people won't. So, we also had to to think about how we spanned from one generation to another and that was some of the challenges we were dealing with going into this project. So, these were some of the things we did coming off the back of a trilogy we created in Legend, Anniversary, and Underworld. And as we finished that up, that was very much based on a formula very much from Core Design and and we were successful with that formula, but going into this new sort of new phase of Tomb Raider, these were some of the challenges we felt like we had to deal with. From a goals perspective, ultimately, we wanted people to fall in love with Tomb Raider once again with this next game. And this is what we started with way back when. Um we had a real goal to make uh Lara somebody you could get to know again, to humanize her, for her to be less of this superhero that you knew because you played n number of games and seen the movies, but somebody that we actually got to know once again and fall in love with over again like it happened in 1996.

So, therefore, we had to challenge the expectations around Lara and the brand. We had to make sure that we were embracing this new audience that maybe hadn't played Tomb Raider before or maybe they were old fans that were revisiting that we'd lost along the way. And to do that, we really felt like we had to make some big changes. Quality and innovation was synonymous with Tomb Raider in 1996 and along the way. I mean, it was one of the pioneers in action adventure. So, we wanted to put that stamp back into this game. We wanted to be right up there as the best of the best. So, these were high goals that we had at the beginning of this project. And ultimately, we still had to thread the needle because it's a tightrope to walk between old and new that we had this franchise with this fan base and all this rich history, but we wanted to take it to a new place. And that that that tightrope is a very very narrow one to walk because too much old and it feels like it's the old game and it may as well be a sequel. Too much new and maybe it wouldn't be Tomb Raider anymore. And whilst we've sort of approached this

with a a blank canvas, the actual reality was it was a very very narrow razor that we had to walk. Um so it's tough doing these these types of games as reboots. Originally, it actually didn't start as a reboot. It was almost a pseudo sequel. And as ideas were being kicked around in the early stages of development development in concept and in pre-production, it became clear that to achieve the goals that we just set out in slides before, that we we ultimately had to wipe the slate clean and we needed to make a fresh start with this character and a fresh start with no preconceptions built in for the audience. So ultimately, it became an origin story and we were buoyed uh and encouraged by other franchises around us outside video games where we looked and we felt that they had actually taken a similar approach. So here we have Bond and Batman that are two amazing examples outside games where Bond has been around 50 years this year and has just done a billion billion dollars at the box office. So

it's as fresh as it it was 50 years ago today. And Batman is in a similar place as well. And I think the key thing about this and why we looked here is because in games having 17 years of history, uh ultimately we were one of the first to have to go through this. You know, it's a it's a a shorter time time scale in games in terms of um people having franchises because game space is much narrower and obviously uh younger than things like movies. So we we looked outside and what we felt was the these these franchises have done such a great job in retaining the DNA of their characters and their franchises, but bringing into new generations and really making it feel fresh. And they managed to span generation to generation but retain their franchise integrity. So, we were very much inspired by this and and this origin story. Um but we really had to find our own slant on that. And ultimately, we knew we wanted to make an origin story. We knew that we had these high goals. But we were looking in early

stages around how we could hang our hooks on something that made it feel fresh for Tomb Raider and also fresh in the industry. So, many ideas were were thrown around and and explored in the early stages of development. And ultimately, we landed in a spot where we we found something we felt that we could hang our hooks on, which I'll hand over to Noah to discuss. Yeah, um it was uh important to understand that in in humanizing Lara by rolling her back to an origin story, we were making her uh closer to you or I and hopefully a more relatable character. Um but the most exciting part about it was that we could then take her on that journey. We could take the players on that journey to becoming the Tomb Raider. Um but we needed a a sense of identity beyond just origin story. So, um we looked at a lot of different things, but ultimately we came upon the idea of survival stories. And for us, we were really inspired by the the real world um tales of of survival. We have uh Aaron Ralston here who um featured in 127

Hours ultimately had to cut off his own arm in order to survive. Or uh the rugby players in the Andes who resorted to cannibalism in order to survive. Um these are at sort of at the surface very dark stories. Um but uh what was interesting to us is there there was a sense of uh of inspiration to these stories that there was there was a memorable journey here that showcased the human capacity to overcome. And and ultimately, through that journey, these people were changed as well. So, that idea that we could take Lara's origin story and tell it through the lens of survival, um, was very compelling to us and we felt would help humanize her as a character. Um, but we had to zoom in a little bit and understand what that meant. Um, I think what we found was that survival can be many different things to to many different people. So, we tried to clarify what it would mean to us in this game. So, first of all, survival action. Um, and that's to sort of balance out

this sense that there's a simulation version of survival as well, but because pace was incredibly important to our game experience, we really wanted to capture that survival action. Um, but it would still be brutal, perilous, raw. This would be the harsh environment that would forge Lara. Um, it would ultimately ask her to fight to survive. That we wanted to showcase her determination as a character, her endurance, her will to live in the face of death. Um, but also as a character who's, um, sort of an intellectual hero, uh, we love that aspect of resourcefulness that we saw in these stories. Here's people who had nothing and they had to make do with what they had and it was really that that ability to adapt to their situation that let them survive. So, these were all things that very much resonated with, uh, the franchise for us. So, just to capture some in a sentence, the idea of the determination to go beyond human endurance, taking the moment of raw instinct and stretching it. It's about overcoming death near-death experiences time and time again.

So, we had, uh, now an inspiration in survival. And at the end of this concept phase, we sort of summed that up with a survivor is born. And ultimately that became our tagline, but internally it was also a guiding vision. It captured that idea of Lara's origin story seen through the lens of of survival. So, with that we felt we had a very clear sense of a high-level vision for the game. So, I'm turning it over to Darrell to get back into production a little bit. Thanks, Noah. So, now we had our identity, at least a hook we felt we could work with, and a high level of what we wanted to do with the game. And ultimately going into the high level in production, I guess I wanted to sort of carve out what we use, which is the rule of thirds. Um this at a high level is what we've done on on the last couple of projects at Crystal, including Lara Croft and the Guardian of Light. And essentially, it's just breaking through production from a high level into 1/3 for pre-production and and what we call vertical slice, which

is the end of pre-production for us, 1/3 for production, and 1/3 for post-production. Some commentary around that, I mean, typically what we've seen, uh you know, we've we've all made lots of games here, is that that first third and that last third tend to get squished uh and short-changed in lots of ways. And And we felt that it was important that we actually maintained sort of that an uncompromising view on what we wanted in each third, uh and therefore shaped the production around this rule of thirds. Uh we wanted to give as much time as we could in post-production to polish and refine what we had. And we wanted to make sure that every step of the way in pre-production and vertical slice, it was clear what game we were making. So, there's always a lot of pressures in each one of these phases. Um even though it was a 4-year period, essentially once we found our hook of survival in early exploration, really that pre-production to vertical slice phase was 1 year, production was 1 year, and post-production was 1 year. And we're going to walk through each one of those phases at a high level and show some videos of where we were at each stage.

So, some quick stats around it. Um we made a 12-hour plus experience. Uh 10 to 12 hours is what we were aiming for. We actually found that once people played the game, we saw them play for 14, 18, 20, 25, and more. So, we've been uh very pleased with the feedback. Um so, we absolutely built a very, very big production in this game cuz it was a complex, big game to make. So, we had to focus very much on our tools and pipeline. I know Jason has given a talk this week around our tools. Also, there's some other talks around our process today as well from some other folks at Crystal. So, please go see those. But, tools and pipeline because we weren't going to be a Excuse me, an army of hundreds making this game, we were going to be under 100 ultimately making this game happen on the single-player side. So, we we wanted tools and pipeline to really be that driving force to help us iterate quickly, to help designers and artists really have the power in their hands.

So, there was a big investment in the tools that we made to make a level editor that really put the power in the content creator's hands with live editing, live updating, and a very flexible, powerful system to create content. We totally turned our production methodology on its head. Uh, and in many ways in this production, we did things very, very differently. And because we were building such a an important origin story, we actually made it a beat-driven development process. Noah will actually go into a little bit little bit more detail there, but essentially in previous games, it was saying, "We wanted to do 10 or 12 hours. What's the footprint for that? It's about this big. Let's fill it with stuff." Um, this one, it was we still wanted 10 to 12 hours, but we wanted every moment to feel like it fitted into a larger picture. So, the the scheduling and production was built around beats and moments that we wanted to deliver from an experiential standpoint. Uh, and that's how we costed things. That's how we, uh, produced things, and that's how we, uh, scheduled out the whole

production. So, the actual footprint from the game was built from beats, which we'll go into again a little later with Noah. We also had a very modular level design process, uh, especially on the art side. That ultimately we had to build big worlds, biggest worlds we've ever built for Tomb Raider, and we had to do it in a way where we kept the fidelity, but we kept flexibility in building them, too. So, we had a very instance-driven art approach with the art team and the design team, where there was a lot of repetition, a lot of kind of clicking together LEGO pieces to really build an organic world. And that was a complete 180 on how we built things before at Crystal, where we had a you know, and still do have a very crafted look, but ultimately we wanted that crafted look, but we wanted flexibility with it. So, it was really about embracing this modular development process and pipeline as well, and that allowed us to ultimately get to the point where we could bring on partners to work with us. We had a robust outsourcing pipeline that by the end of production, we were able to actually send out a rough

version of a level in gray box that had been designed with modules over to partners overseas in Shanghai, and we'd get back an alpha quality level at the end that required some touch-up and finishing, but ultimately was produced, you know, with external partners, which was an amazing feat when we have a you know, a very high bar for our game. We had multiple cinematic production companies through production as well. So, we we worked externally with a number of partners on over 90 minutes of story content for cinematics. So, it was a very big production overall. I think one of the key things that we did in the last couple of games, including this one, is this philosophy is know as you go, as I put it. And essentially, we put quality as the premium, which makes it very difficult to schedule. So, ultimately we put the emphasis on knowing every step of the way whether we are actually achieving what we wanted to achieve from a quality perspective. And this allowed us to scope along the way when we were

off in time or off in target, and shape and change our production in order to maintain quality and deliver against our high goals. Ultimately, one of the problems we've seen in the past is if you deliver things at 50% or 60% at alpha and don't take it all the way up, it can be difficult to schedule that that post-production part of the project. And quite often we see and I've seen projects really get to the point where they've actually taken too much time in production. They haven't really understood how long it takes to get to the quality bar they're aiming for and ultimately there's a block of time which is your alpha to bug fixing to shipping and that is relatively immovable at that point because it's very late in the production process. So what we wanted to do on this was essentially not wait to find out exactly what game we were making but be sure every step of the way. So it had to be great at every milestone and that was really our philosophy on the last few games at Crystal that ultimately if you're not great every step of the way, then don't expect to be great at

the end. So we wanted every milestone to be great and we wanted to know whether we were on or off and ultimately even though we use production schedules to drive our metrics and uh we use production to really calibrate what we're doing and how fast we're going, the sticks and the screen really are our ultimate production schedule and it's about picking up the controller, playing the game and saying is it great? And if not, scope accordingly, change, fix it and do whatever we can to get that right. So we were iterating to make sure that we held up that bar every step of the way. It's important you you don't drink your own Kool-Aid and you're not the one saying it's great but other people are too. Ultimately we're making this for an audience. So it was important we also brought that audience into Crystal and we got validation along the way that what we thought was great was actually working for them too. So we have a focus testing lab at Crystal. We had focus testing from pre-production all the way through to the end of the production roughly every two weeks early

on in pre-production through production going down to every week and almost every day as we got through to the end of the project. And that was key in terms of understanding whether our choices from a design standpoint were actually working when people got their hands on the sticks. I think one of the other biggest things we did was build our game as episodic chunks. And this was a very different way of building. We had built our game in contiguous chunks of content that were almost products in themselves within a larger production cycle. And so they were developed to an alpha quality before we moved on. And ultimately, if we didn't hit what we felt from a quality standpoint within these chunks, they were roughly about 3 hours each. Then we would iterate until we did and we would scope the remainder of the project to make sure that we still hit our timelines that we needed to. And this allowed from a if we were productizing along the way and essentially making nuggets of product that were 3-hour chunks, what we

could do is really very clearly see where we were on or where we were off and and make choices along the way. And as we said, you know, we're we're it's game development, so it's never perfect and you're always making choices. So the more knowledge you have and the more data you have about whether you're on or off from a quality standpoint, if you're putting that as a premium, it allows for better choices down the line. So we made many scoping decisions along the way around the content at the end of each one of these chunks, which we're actually going to show so shortly some videos of. We could review it, we could refine at the end of every episode and we could come back to the next episodic chunk that the whole team was on at any one point and say how do we do it better next time. And through that process, it really did get faster and quicker to produce the game. So moving into pre-production, again, that sense of it being a product really existed in our first playable. And ultimately, even though we spent quite a long time looking at different flavors

of the game and and before hang our hooks on survival, ultimately when we came to doing our first playable demo, we we wanted to make that feel like a product in itself. And ultimately that had a beginning, middle, and end. It was a 30-minute experience, and we wanted to feel like you could put it in somebody's hands and it was as close to a shipping game as we could make. So, we started that um in 2010. It took about 5 months to develop our first playable demo, and I think a testament to that is out of that first playable, it was taken to a as I said a 30-minute sort of chunk of shippable game almost, we actually presented that to to Game Informer and and they were we were lucky enough that they put it on the front cover and really embraced what we were doing because they could see very clearly at that stage what we were making. So, it was definitely one of those things internally it was a great calibration tool for us to work with. In terms of approaching it, Noah's going to take a a look at the pillars, and then we'll show some short videos as well around it.

Yeah, so as we step through each of these phases of production, I'm just going to give you a few glimpses of some of the key documents we used to to capture the creative intent, and also um give some snapshots of actually what we were putting on screen. Um so, we had this idea of the survival-informed origin story, but it was also a Tomb Raider game. So, as we obviously as we start to set out to make the game itself, understanding how that applies to the gameplay experience is important. So, we've always had um the sense of our core pillars in the Tomb Raider franchise. So, traversal, combat, and smart uh puzzle solving are really our core gameplay mechanics. Um and as we discussed before, we were going to infuse them with this flavor of survival action. So, um in the case of traversal, that meant, you know, treacherous, dangerous world, and um ultimately less canned uh controls, less canned jumps, so you really felt at any moment you were you could you could fail, um and you were in connection with the character. Um, in terms of combat, we moved from, uh, a lock-based system to a

full third-person aim system, but also really tried to capture that sense of desperation. We were rolling Lara back to a point where she was far less confident and far less competent than people knew her as, and we wanted that to show in in her animation and her in her mechanics. And then in the case of, uh, puzzles, we really embraced the idea of physics, um, both for its simulation aspect, but also the idea that it it it helped players with an understanding of cause and effect, that by using fire and water and mass and pulleys, we could create environmental challenges that really asked the player to, um, like we saw in the survival stories, use what was at hand in order to overcome their obstacles. Um, so this is very simple, and you'll see with a lot of these, um, it is important to be able to boil things down to a very simple view, um, and and sort of guide your vision with with very clear and precise concepts. Um, this is a a little more complex, but it's really just speaking to a philosophy we had, which was the idea that these gameplay, uh, the spikes of

different colors are really our different gameplay flavors, but the dotted line represents this experience. So, in a lot of ways, what we were creating was not a a sequence of of game design challenges, but really an experience made up of those challenges. Um, so that's a philosophy, this idea of variety and intensity as a a function of pacing is something you'll see through all of our docs here. Um, uh, we're going to take a a quick look at actually some of the content that we were putting on screen at this time. So, this was in mid 2010. This was, um, cut from, uh, probably about 30 minutes of gameplay, and this we cut it together at the time. So, this was sort of our trailer packaged up, um, and this helped us both internally see our concept in practice, but also, as Darrell mentioned, starting to to generate, um, support externally both within the organization and and within the press.

You think you know me. You think you've seen me before. The rich princess. Rich girl with a pretty face. Shiny guns. And a flare for gymnastics. I think you've got me mixed up with someone else. I'm not perfect. I'M NOT BULLETPROOF. I STRUGGLE.

I bleed. I get lost. I know pain. I know fear. One day I will become a legend. A WARRIOR. A RAIDER. BUT FIRST, I WILL BE A SURVIVOR.

STILL THINK YOU KNOW ME? All right. Pro tip, don't put your date on the end of your first playable trailer. Uh and announce it to the public. Uh but ultimately, actually, I think the thing about the date, I mean, we really were aiming for the end of 2012, uh but the know as you go um philosophy really actually showed us very early in production that we were going to struggle to meet that. Uh so, it meant that we could actually communicate to Square and to the wider

group that based on what we wanted to achieve, we did actually have to move the date. And it wasn't a surprise. It wasn't like it came at Alpha. It came fairly early in production. Um so, even though the date was changed, uh there was no surprises with that, uh which I think is sort of the good moral of that one. So, this wasn't the end of pre-production for us. That was first playable. But it really had that identity that we wanted. It was absolutely playable on the sticks as a 30-minute experience. Going into the our next phase is what we call vertical slice, which is the last phase before production. This is a 7-month period where the team worked on really taking that skeleton and fleshing it out and putting the meat on the the to make playable four playable expressions of all the pillars. Uh This amounted to about 1 and 1/2 to 2 hours of content at alpha quality. All the base mechanics were were in and implemented. We didn't think it was going to be needed to be 2 hours. That wasn't really the aim. The aim was to make sure we had full playable expressions of all the gameplay patterns

that we needed to go on and replicate in production. So, it was a case of making sure we had one of everything. And because experience was really what we were aiming for rather than components, it had to be strung together as a complete narrative start to finish. So, it was contiguous. It represented every part of what we need to build in production. And it ultimately came as as a 1.5 to to 2-hour experience, which was was larger than I guess we planned originally. I I guess the good thing from that, again, putting quality as a premium and not moving on until we actually achieved it, is when it came to doing things like demos and actually getting the word out there, we didn't have to do very much, which was a good byproduct. One, the team knew internally what they needed to build, but two, when we got asked to be on stage in 2011 from Microsoft, we said, "Let's take 5 minutes or so of our vertical slice and put that on stage." So, literally the work involved to do a demo for a Microsoft keynote was a small

team of a few people for a few days to really just cut down the content and say, "Here you go." So, it was not custom crafted. So, that was a great byproduct of of taking quality and pushing it hard early. Another aspect of vertical slice is one of the common challenges in game development, which is sort of planning at the same time you're putting that hour and a half on screen that's a demonstration of your vision, you're planning the whole game. So, for us that was it was important for us to come out of that pre-production phase knowing exactly what we were going to build. And and in some ways that's a challenge just from a documenting perspective. So, I think one of the ways we approached this game was to to look at it as an an experience-driven game, which meant it wasn't really purely a story-driven game, it wasn't purely a gameplay-driven game, it was really all of these things coming together to create an experience

because the players really don't experience them independent of each other. So, this document is not necessarily important to see the details, but the horizontal separators are levels. So, between that would represent about 20 minutes of gameplay. So, each segment was about 2-minute granularity at this point. But looking at the columns, the center column was story beat. So, that was much like just a treatment that you'd have for a game story. But then as you sort of expanded on that with gameplay concepts on the right-hand column, you have Lara's character arc because obviously that was very important to us in an origin story. And then to the right of that, her gameplay progression. So, where gear and abilities came in. And on the left-hand side, you can see the colored bands. And the color the one color band represents the different flavors of gameplay, and then the grayscale represents the intensity of that. So, through this documentation process, we were able to visualize a little bit more richly than just a level document or a story document might do.

And so, we were able to map out and create a framework for the entire game. Um in our case, we were also trying to tell a fairly linear story in a fairly non-linear world. So, this is just an example of then mapping that experience to our spatial layout. So, we we chose a hub-based system, which ultimately uses gear for the player to unlock new spaces. So, down the bottom of the page here, you can see the gear that's acquired over time, and each of those is a key for one of the locked areas. So, by getting new gear, you can see you you can progress through the various areas of the game. So it was just tools like this that allowed us to sort of zoom out and visualize the whole game experience. Um but one of the great things about having that sense of of a full framework, um all of the disciplines can work within that. So in a lot of cases you might have art understanding what the region is and being able to concept against the basic themes, but because we knew Lara's journey, art could do exercises like this doing a great job of really capturing not only uh the

physical requirements of the space, but really the emotional context of where it is in Lara's journey. So having that sense of an entire framework and coming out of pre-production with a fairly clear vision of the whole game was important to us. Um but as Joel mentioned, we also had to put it on screen. So what we'll show you now is probably a 5-minute video cut together of um an hour and a half of gameplay content. This was done probably at the beginning of 2011. So this is still 2 years before we shipped, but it represents the first shippable essentially shippable content if we had to. Jesus Christ.

GET OFF ME! BROTHER. STOP IT. THAT WAS CLOSE. OH [ __ ] THAT WAS CLOSE.

ROSS? ROSS! ROSS, I'M COMING! THANK GOD YOU'RE ALIVE. THAT GOD'S GOT nothing to do with it. Good to see you, too, girl. Sorry, they did a real number on your

leg. Oh, no. Looks worse than it is. Something to pry this open.

LET GO! GET OFF ME! IT'S LIKE SOMETHING OUT OF A DREAM.

NO!

OKAY. SO, WHAT WAS GREAT ABOUT THAT AND IT'S EVEN amazing to look back on it in some ways that that was 2 years before our game shipped and before we even started

production. But, we had 1 of everything. We had a hour and a half to 2 hours in the can and the key thing about making it a product every step of the way, this was going to ship. And it had to be treated like that. So, we then had to just move on and make more. So, that was the next phase of production, which was a year. So, there was 12 months in production. We had about 9 hours of content to be produced in that period. We had sort of an hour and a half to 2 hours in the can from what you just seen. And we were making them in episodes, as as I mentioned earlier. So, that represented one episode. Holistically, the team worked on each episode start to finish, including all the sound, the effects, the the usual finishing things. And each episode or vertical slice chunk, as we called it internally, really comprised of that same sort of pre-production, production, alpha, bug fixing, sort of almost a product in itself from start to finish. I think it's the most important thing is

to look at the ratio of time per episode and know that practice makes perfect. So, if we said that we spent 7 months on that first chunk of an hour and a half, and then we had 12 months to do the next 9 hours, that doesn't add up on paper, as as as we knew internally. But, what we do know is we get better with practice. And what we had to find out was how quick we could go and how much practice would make perfect. And ultimately, going into production, really we did. And and one of the key things to that was building expertise around each discipline and around each pillar. So, we had these pillars of traversal, combat, puzzles, and some other ones outside that, such as hubs and effects, music. But, ultimately, the key pillars we actually built teams around that were multi-disciplinary, that had their own leads, and they had their own ownership over a piece of the experience that was mapped to that larger, broader strokes in the documentation that we were going through. And ultimately, as each team worked on a pillar, they became better

and better and faster and faster. So, over the course of production, the first hour and a half to two hours was about seven months. The second one was about six months, five to six months of that year. The third, and these are all three-hour increments moving into production, the third one was around three and a half months, and the last one was about two and a half months. So, by the end of production of that year, the team, along with all the external partners we had brought on to help us through production, was moving at um three hours of content in in around two to three months versus seven months for an hour and a half of content in vertical slice. So, what it allowed us to do, as I said earlier, at the end of each each episodic chunk, was actually make decisions and scope, and really learn how much game we could produce in the time whilst maintaining quality. And ultimately, production was comprised of what you just saw four times over. We had 250 cinematic moments um and 90 minutes of custom uh storytelling cinematics. So, it was a large cinematic pipeline that we produced in that year,

as well. And ultimately, we had teams around the world working for that year, producing content for uh the production phase. So, teams in San Francisco, obviously with our sister studio in Montreal, Shanghai, India, LA, and uh a number of other places, but uh you name it, we brought them on board and said, "Let's get this done." So, it was done in in a year period, and that really brought us to alpha. Some quick talks, I'm just going to flick through them um In terms of scheduling, because we were building in patterns, right, ultimately, we could actually schedule out, you know, what are the patterns and what do we need to repeat. And each level was comprised of those patterns that we could then take from vertical slice and almost sort of map to production and say, "It it's kind of like this, but with a different wrapper." And ultimately, those patterns we could then cost against. So, I'm not going to go in detail through this, but again, similar to Noah's document from a production standpoint, it was we we added up the costs associated to each beat or each moment we produced, and we broke down each beat and moment throughout the

whole game, and then applied a cost to it. Ultimately, you're never right with documentation when you start out, and we refined it over time and became more and more accurate about what we could produce in the time we had. The other key thing is is understanding where you're at at any one period in time, and having a team all understand it, too. And one of the things when you've got a lot of people with their heads down producing content, you know, just saying here's when we need to be done and being clear about it is is very very important. So, we actually lined the internal team up all around the same milestones, rather than have a lot of um I I guess uh cross-hatching, if you say, in a schedule where people are finishing things at different times. We really tried to finish everybody at the same time on the end of each vertical slice chunk of content. So, everybody knew when done was and what done looked like. It had to look like our first vertical slice repeated. Yeah, and as Darrell mentioned, um working with those patterns was

important. So, uh from the creative perspective, we had the Uber Doc, which mapped out the whole game, but it was uh a higher level of detail than you would want to actually produce the content. So, as we zoomed in, um this represents a beat doc. So, for every level we would do, we would do a beat doc for about uh 20 minutes of gameplay, um you'd have about 10 plus pages. So, the the first page is something you find in a lot of docs, probably the setting, uh premise. We're tracking Lara's emotional arc as well, cuz that was important to us. Gameplay mechanics, uh enemies that you'd find in the level. So, a fairly standard overview. But then on the lower uh right-hand side of the the screen, you can see sort of the first beat laid out. And um it was important for us to actually from on a moment-to-moment basis describe what the player was doing, rather than just frame the high-level goals. Um and and that was a bit of a culture shift for us as well, because in a lot of ways, um developers like having a bit of a a blank canvas, but the good news with

putting something in place is that you know you have an answer, and that answer can always get better. So, in in the beats we'd also put goals or conditions of satisfaction. So, not only did we have sort of our starting point, but we understood what this beat was trying to achieve. So, that as we iterated and as we improved upon it, we could make sure that we were still achieving those goals. We had to map out levels spatially. So, obviously paper maps are fairly standard, but really pairing design with art as well. And Darrell talked about those cross-disciplinary teams. It's great to get in and really think about the same thing at the same time. So, something on the left is an example of a key art image that would really help capture the person a moment of the level, but also the personality. So, that as we're talking about what's going on, we can all visual visualize ourselves in that space. And then on the right-hand side you see sort of our beat boards, which any moment where we had heavy character performance, scripted action, cameras, that we tended to get much better results when we took the time to sort of

go through the process of of doing these beat boards. Sometimes it feels, especially when you only have two months left to to do three hours of gameplay, that you know, skipping some of these steps is a is an efficiency, but in the end for us it really wasn't. Speaking of cinematics, just a quick glimpse at our cinematic pipeline cuz that was slightly different from the in-game performance moments that we did motion capture. So, once we had that full Uber doc describing the game, we could write the script. So, that looks like any TV or movie script. In our case, we had to extract any props, any set design elements. So, in this case we're looking at a scene that takes place in Himiko's tomb. This is the first tomb that Lara raids. We would build that out, measure it out, because when we got down to LA, this is the capture volume. There's actually a grid on the floor, and all that junk around Lara really represents elements in the 3D world, so she could be aware of that environment. Um you'll also notice that she has

cameras mounted in front of her on her helmet. So, we didn't actually capture face data, but we got frame-by-frame reference for the animators. So, we were capturing the full body data, and then we were using that face information for animation reference. You can see in the background, there's some monitors. So, as Lara's acting out a scene, we can see something like this on the monitors. It's extremely rough. Um but the other great thing about this is it becomes a previs for that scene. So, even out of our rehearsal shoots, we would get something like this, and then could plug it into the game, and then now that that story element is is starting to be represented in the game experience. Um once we get effects, lighting, sound, face, hands, um all of the additional layers we can put on, the cinematic should look something like this. I hate tombs.

A sun cup. So, as Darrell mentioned, we went through basically three uh uh versions of that production, and we really did hit our stride. But part of being able to hit our stride was sticking to those patterns that when we visualized a new scene, we needed to see it through the lens of something we had already done. Okay, so that brings us into our final third. Uh we're going to run through that very, very quickly here. Uh so, we had a 12-month period where we had those four chunks of content in the can, and we could play it start to finish, and it was a already at a a very good quality level. But we knew that we wanted to edit what we had on paper and make it better once we actually saw it in the flesh, so we built time in for that. And that 12-month period was basically split 6 months into editing and 6 months into bug fixing. Uh and ultimately that 6-month period of editing is gold, right? I guess I I would just just say that that without that we would not be able to realize the game that we wanted to because going from your idea of what a game is at 10 to 12 hours to reality,

it's always going to be slightly off when you actually see it and play it through. So it gave us a chance to actually go through, play it ourselves, use usability to actually give us feedback, and it essentially go through and edit that content. And then the last 6 months really were polish, polish, polish and heavy play testing, heavy fixing of bugs to make it a polished gem that we wanted it to be. Um and it was a great opportunity to have that time. A lot of times it's all we can do to just fix the bugs. So to have some time to actually sculpt the experience. And I would say that even though we tried to make each vertical slice as finished as possible, there's certain things that you can only sort of identify like overall pacing and character arc once it's all in place. So that inherently there's some improvements that can be made on any game at this phase if you leave yourself the time. But we wanted to I guess make the best use of that time. So here's just another quick document. And this should look familiar cuz it's very similar to the Uber document. But whereas the the first document we looked at represented our intent, this is actually our actual. So every square is

a minute of gameplay. And again it's color coded by the type of gameplay and the experience. This was valuable because games are fairly intangible and the script doesn't really represent the game. So when you're sitting there and you're talking about how do we sculpt this experience, it's really nice to have something a tangible representation of that experience. So this is zoomed way out. You can see that's our 10-hour experience or our 10-hour core experience laid left to right with all of the I guess progression along the bottom there. Now, that was just basically assessing where we were at. But in terms of informing iteration, we really relied heavily on play testing throughout this whole process. So, you qualitatively asking questions about do you enjoy this? Is this too difficult? Or in this case, story comprehension. So, we had a multiple people come in, play through the entire game, ask them questions at various points. And in this case, we identified some issues. The lower comprehension at the beginning is fine because Lara's thrust into a crazy

situation and she doesn't even know what's going on. But that dip at the end was a problem that we didn't want people to not understand what was going on at the end of the story. So, ultimately we did a lot of rewrites at the outset of our post-production process. But then we would validate those changes. So, a subsequent playtest gave us this graph and we're like, "Excellent. We definitely made some improvements and there's still a couple points that we could then identify and say, 'All right, now let's let's go address that.'" And so that the idea of informing your refinements through playtesting and validating your changes through playtesting was essential. And just one other component with that process, which is in addition to asking them questions, we were also capturing data over the network. So, regardless of what they said about a level, we could see things like how much damage they were taking in a space or how many deaths. And And that really helped us with the with the tuning and balancing aspect of it. So, I guess the key is leave yourself that time to iterate and make your game really what it what it

wants to be in the end. So, finally just wrapping up. I guess this is great to give you a broad view. There's some more details talks as I said uh around Tomb Raider today. Um so, please go see those. Um I want to leave you with some pro tips, but one of which is that it really is a team effort and Noah and I get to represent the team up here, but ultimately it was a team that made this happen and the team, you know, really had the the drive and passion to to go through all the different uh challenges you get in development to produce the game we produced. So, it is not as effortless as what you see in this PowerPoint um as I'm sure you all know. Um but the pro tips really to leave you with really that know as you go. I'm going to show you one last video just after this to wrap up, but knowing at every step of the way what game you are making and whether it's good or not, whether you're hitting your goals, your quality bar was critical and has been critical for the last couple of projects at a crystal in the way we produce them. So being great at every step is critical knowing that you're great at every step,

knowing you're getting what you want and not waiting till alpha to find out and find out you haven't got enough time to make it great. Uh the cross discipline teams around pillars was a key change that we made in terms of how we produce the content fitting into a larger macro map of the narrative and gameplay experience and with this ownership and this expertise, it really gave us the quality but also the speed to produce content that we need to in in that production phase. Lastly, we went through the post-production phase, but you can't shortchange that. You're never going to get the game you want when you see it at alpha. So build time in to get the game you want post alpha. Don't just bug fix and put it out the door. So that's really a a key tip for us and ultimately that bug fixing and playtesting all the way through to the end. I mean it really is what makes a polished refined experience. So don't shortchange that as well. Lastly, I'll just leave you with a short video which is really the culmination of all the work which is a trailer we put out just before Christmas, but I think the key thing to look at with that know as you go mentality is at the end, it's

very close to what we had three years ago and I hope you can see that through this talk and through the production of this game that every step of the way it really wasn't too far off. You could squint a little bit at the beginning in the first playable, but from then on it was the real deal and ultimately that really allowed us to scope and manage our project as we went. It wasn't on rails, but it gave us the feedback and the input we needed to make the right choices in order to hit our dates and our quality. So I'll leave you with that and say thank you for joining us and and I don't know if we have time for full audience questions, but Daryl and I will hang out so if anybody has questions about the process or Crystal or anything. So much has happened.

Our captain is injured. Have you heard from any of the others? Nothing. Our crew is scattered. I have a bad feeling about this. I've got to find them. We're hunted by the ISLAND'S INHABITANTS. THEY'VE GOT ME LOCKED UP. SAM, LISTEN. I'm going to get you out of there. Ah! Watch out! Ah! Ah!

You can do it, Laura. After all, you're a Croft. I don't think I'm that kind of Croft. Sure you are. You just don't know it yet. Yes! No! No! All right, let's do this. Sometimes you've got to make sacrifices, Lara. You can't save everyone. There's no other choice. I've got to do this.

If I don't survive, none of us will. All right, thank you. Okay. And please stand by for