Beat it: A Retrospective of Need for Speed: Hot Pursuit

[Music] okay I'll uh I'll get started firstly hello hi everyone uh my name is James senson and I'm a producer at a wonderful British Studio criterian games uh over the past decade these are some of the games that we've made you might have played a burnout game or even black or then more recently uh some of our Need for Speed games I joined Criterion just after the launch of Burnout Paradise in back in 2008 uh where I cut my teeth as an online engineer working on that year of paradise where we just kept on adding new content uh to the game after its initial release since our most recent game we've been keeping busy working on a huge new and exciting multiv vehicle game this one's going to be the biggest game that criterion's ever built and it's not something I'm going to talk more about today but I will promise one thing and

that's that it's going to share something with all these previous titles and that is our Criterion DNA so these are our core pillars these are the values that we hold up to ourselves they Define the games we've made and all the games that we want to make so first up accessibility criterian is a studio that makes arcade games we want our games to be easy to pick up but difficult to master with the depth of Mastery then there's intensity intense visceral experiences Define our games we really want to get the players adrenaline pumping and then quality we hold ourselves as a AAA studio and we want to be up there at the Forefront of Technology making great games and great software that our players will be uh will love and be free of frustrations then there's this idea of connected so what we've always been trying to do with our games is push the connected experience because we think playing with your friends that's the best gaming experiences that you can have that one's quite simple and then

there's Innovation we always like to do things differently each game we make there's always something that we're trying to do differently and new and not rest on our Laurels so today I'm here to talk about our 2010 title Need for Speed Hot Pursuit and I want to talk about autog specifically I really want to drill into AOG um partly because that's the part that I made and here is the autog team back in 2010 with me looking a little bit more uh fresh-faced and floppy head uh than I am now I suppose that's what becoming a dad does to um but then so 2010 that's quite a long time ago most of the time we see post-mortems on games that have just come out or these games that have are fed by Nostalgia so but I think it's really important that we take time to actually recognize our history and keep looking back at games throughout out to

so that we can spot those Trends and spot those changes but also Hot Pursuit it's one of countless Need for Speed games but it's really important to talk about Hot Pursuit because this is the point for Criterion where we really recognize what it means to be constantly connected with your friends and it really changed our way that we think about making our games also it's a great core game experience I think there's a lot of big AAA development happening today where uh where the ambition and the huge spra scope is is contending against the the ultimate quality that they can achieve with um such split and segregated development whereas with Hot Pursuit that's a title that the the game experience it's it's really well honed it's really simple it's really straightforward and it's a great core game and I still think that the actual experience of playing that game is still one of the highest quality funnest experiences that you can have in an arcade racer then there's autog so at first wants um it might be thought of as just

friends leader boards and some mechanics to uh some mechanics to sort of push that rivalry but it really came to define the game it made a genuinely social experience where friends could play together and against each other even when they're not online at the same time and those lessons still apply today but before diving into the making of autog let's have a look at a a trailer from the game just before it shipped just to refresh everyone's memories as to what that game was [Music] [Applause]

strong you can out have the G [Music] around the corner coming

[Music] uh so that was Need for Speed hop Pursuit exotic cars epic drives cops the uh the name hop Pursuit it's about the thrill of the chase this was a game of two halves it had two separate progressions two sets of super cars to unlock you a player could choose to play as the races uh complete completing races trying to set records or you could play as the cops now the cops their super cars were tuned and upgraded and they had cool te and weaponry so that they could they could give Chase and catch catch those races and then take them down but ultimately whether you're playing as the cops or whether you're playing as the racers it's a game about speed and at its heart it's a pure arcade racer so throughout the whole game it was really pure and simple the motivation was to be the fastest and that competition like any good racing game defines that core

experience and because we nailed it down to this very very tight core experience and didn't diverge from that it meant that auto could be there to amplify that core experience and it focused it directly on competing between you and your friends the development of it started a little over six years ago so jumping back to 2009 which in the world of tech feels like a lifetime ago this was criterion's first Need for Speed game and we wanted to keep keep it simple so we drew inspiration from the original 3dio need for speeds exotic cars this was the first time we had licensed cars and we really wanted to celebrate the heritage of those cars and also it was the first time we were working with the data from manufacturers and we we kept true to that so we we redid our physics model we redid our handling model we did it again and again and again until we had it perfect so we could take all of those talk curves take all of those aerodynamic coefficients and all that good accurate not to 60 times have the

accurate top speeds in fact when we were making the game there was a point in development where I think it was the Mercedes SLS uh it was sort of 5 mes hour above the actual qued top speed and we were scratching our heads a bit and then we realized that the cad model the the tires on it weren't were slightly wrong there were the wrong wheels and just that tiny little difference made that difference in the um the simulation so it's good to just have that data and and be true to that cuz it's a really nice physical underpinning to your cars and then the next was epic drives so our ambition was to make a world that was stitched together from all the best of the Epic West Coast drives that there are here in the US so that was um stitched together with these huge big Vistas and every time you came around the corner we tried to put as many big Vistas in the game as possible whether it be a Mountain Vista a big open stretch of desert these gorgeous uh stretches of Coastal Highway way suddenly see a big bridge or

something um but then also stretches of rural areas and farmland and then giant redwood forest then our third pillar players the cops Advanced Weaponry battling highs speed Pursuits felt like this was a great opportunity for us to get our Criterion spectacle into the game when we had these things we just slather over at the top of it Criterion special Source this is what Criterion is known for as I said before we're an arcade racing game studio and we've got a great heritage of AR of arcade races through burnout and these are underpinned by spectacular physics but also we always try and have accessible handling over the top of this so as well as boost we've got responsive controls and the world was made with these long sweeping turns so actually that road network uh a lot of time goes into the the angle of those turns the angle of all those corners and um in terms of having intuitive drift it's not a drift button there's you can either tap the break you can hold the break you can tap

the E bra all of these different ways so that whatever the player's intent is we can pick that up in the game and uh translate that into the drift that you want to see then crashing sense of speed intensity cinematic spectacle is where Criterion really feels at home and from our previous games from burnout from black we know how to craft these experiences the other thing we hold really dear to us is is friends and putting friends at the heart of the game so Burnout 3 was when we started experimenting with this it was EA's first online title and then Burnout Paradise this is where we um where we really pushed our open world to be able to have have that seamless online so you could jump into a game jump into a friend's game then matchmake into a public game all without leaving the world and constantly driving around and we started really beginning to experiment in that year of paradise when we were developing free bound challenges which for Burnout Paradise fans that

still stands as some of the best C player you can have but this friends at the heart of the game it's not a solved problem but we know that it's really important to us because when you're playing a game with friends then that makes the game endless your game becomes a vehicle for your friends to get together and hang out and interact together and just have those interactions between real people and real people gamers are social people and having those interactions with friends that you can attribute um um that you can attribute those actions to to a real person just makes it a lot more a lot more interesting and a lot more meaningful and it makes it more memorable and then those interactions and that fun that you have in the game that can spill over into your real world conversations and in that way the game just grows and grows and becomes a really great experience to share but then the problem with trying to get real friends into a game together especially back in 2009 real friends lead real lives and it's really hard to

coordinate to get online at the same time it's um it's hard enough to get like to get colleagues out to the pub after work on a Friday night let alone everyone online at the same time with the same game when they're at home and so this this is something we actually knew from our Telemetry on Burnout Paradise um a really great part of being part of EA is that we have access all this great data across um across EA's portfolio and so what we saw here this is looking at the average Peak simultaneous users uh weekly Peak simultaneous users back in back from 2010 across a representative set of vaa titles and we were seeing that more and more players were now connecting to our servers but then only a smaller proportion of those people were playing in online multiplayer um so while we were having more players connected there aren't as many people as we wanted getting in to have that connected experience to have

fun with their friends because they're just not going into online multiplayer and I know from my own personal experience online multiplayer it's not always the most welcoming thing especially as you start getting older your reactions are slowing down and I'm there's only so many times I can get schooled by 17 year olds uh so then we really thought rather than trying to push players into online multiplayer how can we create these meaningful and ongoing social player experiences for those millions of players that are connected but don't want to play online againsts so this felt like a really great opportunity to create play with friends using asynchronous interaction so just to recap synchronous that's a phone call asynchronous that's a text message but the thing that's great about asynchronous it's things like broadcasting you know you get notifications Twitter Facebook and WhatsApp these are these are communication tools that allow people to share much more with more friends regardless of of uh living different

lives and this felt like a real opportunity for the game the space in which we could innovate and bring friends together by focusing on those asynchronous interactions and then we could really make it about real friends so just to um put your minds back to 2009 this was when we started making the game so in 2009 there's a picture of the um of Xbox Live at that point or of the dashboard and it was a very different place so we knew from other Rea titles we had Telemetry showing us that um the majority of players so on Skate 2 54% of players had only one friend or zero friends so friends it was just a very early time in in um how people were using the social features of Xbox and and PSN they were just really getting off the ground

but at the same time we had social networks and they were now mature because Facebook had been around for a few years at that point and so we wanted all of these things that you had on on Facebook and on on other social networks uh so we wanted to be able to find friends of friends to be able to to grow our friends Network by seeing what friends of friends had got up to and also to just have those things like sharing like commenting like profile pictures which we now just have built into our consoles and take for granted so that was step one step one for for our our social play for Hot Pursuit we didn't quite know what we were making but the first thing was hey let's build a website and then we're going to build a web browser and put that into our game and that web that website in the game that was our game front end and then hey you know that was the year I got a uh smartphone for the first time and that felt really cool and exciting I could put that website on my smartphone it was the same website in the game and is out of the game so um it felt very new and exciting and different but the other thing that we learned very

quickly was that actually working in web for a front end is really powerful because that rapid iteration speed that you get with with web is just worlds ahead cuz there's no build time you know you're not waiting for anything it's just um you make a change and you see it straight away so we could try out lots of ideas very very quickly so that meant it was great for Discovery uh but then the problem with this is this is this is not a console ux experience and text entry that's a problem for trying to make these social features uh so writing out comments and posts using an onscreen keyboard that's no way to spend precious console time so version two this is what we really got to to this was the exciting part we automatically generated stories we did the work for the player and it worked it gave people a real sense of playing with other people the game had to track everything and all of these actions became stories and they were then shared back

to me and my friends and it became the way he played the game there was a feed for every game mode and so there was no longer an event list there was no uh set way of playing through the events you just played the events that were based on what your friends played and so you got a sense of what was popular and playing the events became this sort of back and forth conversation with your friends but the real promise there came from when we started tracking leaderboard changes and then we could build these stories from those leaderboard changes to make them as rich as possible and also it's that point of rather than just showing a leaderboard you really by writing a story about it you're really focusing in on where that player is on the leaderboard and what they need to do to take that next position so just break down that task of getting to the top of the leaderboard into one step at a time the other thing we had was a call to action so at every single opportunity throughout the this website front end that was the first version of

autog we gave people the opportunity to jump straight into the game to play what was being talked about to play what was posted to a photo feed to to play what was talked about on the wall and we put these call to actions at all points so wherever you were in the game you could just jump straight to the event and play it and that to us felt really really interesting at the time because it meant that it was totally Dynamic whatever I was playing whatever I was being fed was what the game told me about out through this social network through what my friends were doing to show an example of that here was our rapid response feed at the time so those events were recorded and every action was tracked and then every a comparison was made for every single time one of your friends attempted the event and so yeah as I said there was a feed per event type split by CPP or racer and that was the main way to play the game so if your friend tried in event it was in your feed ready for you

to try and at that point we were like excellent now you're really feeling like you're playing with your friends whenever you played the game uh but the the problem with this was you know Need for Speed social that doesn't sound like it might have been engaging it was it was interesting you know you got a real sense of what your friends were doing and you could have taken that further to start you know if you got into it then you could really choose how you played and and sort of set up your own progression with your friends but uh ultimately it wasn't fun it hadn't honed in on this on this core experience and there was also this other problem of we've been building in web for a while now and we had really bad feature creep we were writing checks that we wouldn't be able to cash in production because we had all these features we had we had a timeline feature where all of your um all of your actions got Amalgamated and aggregated into these timelines that you could compare against other players and we were just getting ahead of ourselves

we had all these multiple feeds and ways of managing those feeds ways of looking at your friends feeds and really we we'd gone a bit too far with it and that meant we had this terrible problem with signal to noise ratio and it was really hiding the fun of the game so what what were we going to do so this got us us scratching our heads a bit and the answer was in beat it so we were asking ourselves how are these features in Serv to the game and we took all of those automated feeds and then we curated them automatically to be one feed of recommendations because throughout the whole time of playing with your friends it might seem obvious now but the most compelling motivation Way Beyond anything else is someone beat you ultimately when you look at popular events and things like that they're just not as interesting because they don't change over time necessarily or or um yeah yeah whereas someone beat you that means hey one of my friends they've

played my event and they they've beat me at it but you know I I didn't really try so I can go back and I can I can beat them I can I can be that one faster and it also made us change all of our events because uh before that point we also had events which were uh based on points and we had a few different event modes and but if they didn't sort of if if they didn't support this idea of being the fastest then they they fell to the side because they weren't as engaging they weren't as natural to talk about an event you know when you're comparing points it's it's one step removed from this idea of be the fastest and then we used all of these other Rich stats and comparisons that we had and that we were generating they were then layered through the whole game to flavor every choice that the player made as they um as they played through it so then our previous recommendations or our previous stories that were generated in the previous feeds that we had became this one dedicated

recommendations feed so we took all of our recommendations and worded them as a beat it that was really important so if we had an event that was recommended because it was popular or because it was a way of uh pushing the player through the progression we rewarded them we we took the player that was one above you on the speed wall which is our friend's leaderboard and then we just worded it as this player is 0.2 seconds ahead of you beat them to take second and this with this that that just meant that um every single time we had even if we had run out of betit of those places where a friend had beaten you were're able to just give that nudge for competition uh we always like to say it's not really we don't want to push the player through through competition we don't want to push the players into rivalry but it's more just like the game is there it's just giving the player a nudge and being like you know what this guy I'm just saying but but he's beating you and we felt that that was really working

um the other thing we did was we balanced that signal to noise ratio so we tuned it back so there were never more than five cop and five racer recommendations uh and also at that point for going into production we recognized we recognized from a while back that we wouldn't be able to make the responsive ux experience that we wanted on a website uh especially on previous generation consoles you just needed too much memory to to have that sort of web EXP experience so we re rebuilt our front end um as a web driven console oriented ux so that meant you know it suited up down left right forward and back so then to uh to really break it down and give those examples I'm going to have a look at a session with recommendations just to talk through what we shipped we had a Auto for auto and that was connect compare compete so this

first section that's about the front end it's about connecting me with my friends when you booted the game you got this live connected front end so it had a load of content that was delivered by by web um and it set the scene by giving you recent updates of what you'd been recently up to and it also gave you refreshes of what your friends had been up to so the first of those social features we still had those social features that made it through was photos whereas today we've got Xbox record that we've got the share button on PS4 they do a lot of capturing of in-game content but by building a dedicated photo mod mode by giving the player the tools to take the time to compose a shot that's still really powerful and I think that's really important to remember because it's a great way for people to spread their passion with their friends but also to it's it's got that in common with uh tools that you want to build communities and that's what we started to see with

our photo mode so just to show you these are some of the photos that uh that came up through through dedicated blogs that started popping up about the Artistry of CEST County and about the great things that had been or the the great sites that could be seen in the game and then some people take it further so a lot of the slides or a lot of the photos in my on my slides are from the website deadend thrills. comom where you've got this guy dunking Duncan Harris and he he runs the website and he goes around taking screenshots from games that have good screenshot capabilities he doesn't touch any of them in Photoshop and he just makes makes an art out of um out of taking screenshots and so then by having these tools you're allowing your community and those in those outstanding individuals in your community to to do this do this stuff and make your game look fantastic and then share the best of your game with everyone else so that was really cool it was um nice to see

those and then we had our wall feature and as I said before type uh using a onscreen keyboard might not be much fun and and it might seem like a bit of a niche feature but Al it got really really heavy use in in our PC communities especially in the hardcore PC community in Scandinavia and Eastern Europe and in Russia because PC players that was their main tool for friend Discovery so they were able to go there and because on the wall you could see posts from your friends of friends and replies from your friends of friends so we had these huge huge wall posts of people just introducing themselves to one another and getting acquainted um to all share in that autog experience the other thing we built that was in our front end was friend recommendations so as I said earlier more than half of all players had only one or not friends and this was a game about playing with your friends so so if you've got no friends now you can get

friends but back in 20110 when we were getting ready to release this it's a sign of the times that we really had to push with first parties uh to get this feature released because they were concerned about the the Privacy setup for being able to see friends friends of friends lists whereas in today's world that's just sort of that's very very common place and it seems weird why would you not want to see your friends friends um but in fact I can't remember it only shipped on one of the pl forms I think because of that so if that was connect then the other thing throughout the front end and throughout your event select we had compare so we we put friend comparisons at every opportunity to really drive home the drive home the sense of playing with your friends so here on our map screen a player Works through both cop and racer events through this one map screen and in the bottom right we've got

this speed wall um and this is sort of a cycling wall it goes through the speed wall which is your friend's leaderboard and then it goes on to show you a comparison right where the player needs it and it influences their event choice so all the speed walls and storage were pre-downloaded uh so that they when you were picking an event they were right there and you you should have never had to wait for leaderboard fetches and it was a expensive Tech investment but for a game of about friends leaderboards that was really really fundamental to it and worth it because we needed that sense of every single time you're flicking through the events trying to pick something you can see your speed wall and you can have that influence on your choice based on that so we also put a face to the name we shipped a feature to be able to get profile picks in the game again this is now provided functionality by first parties but we really want to humanize it we're also really about this idea of trying to get your real names into the

game because especially on PSN like I'm still embarrassed about my um my PSN ID that I set up all those years ago but I can't change it but now at least they provide the functionality to put your real name in the game and then there we go with the story so by automating the story you get a refresher of what you did last time out and uh you get that friend compare I said earlier about the idea of taking Rich stats to really flavor the stories to to be able to build up more of an interesting story and attempts was a really big one for NE speed top Pursuit so by tracking the number of attempts and presenting that back to the player it made that friend competition really interesting so looking at this example we've got Matt Webster who's our studio GM so uh DIY rasal 22 that's a group of all of RIT guys that's the account that they Ed to play in their lunch break and they took second from Matt but Matt couldn't have that so he

went straight back in there and took second back from them and at that point the competition got really serious so 1,375 attempts that equates to about 20 hours of dedicated Play Time on on just this one event which is ridiculous I mean well done them for trying but ultimately they didn't get them anywhere so jokes on them so then even with a context of friends a simple select now has real meaning to me so I can see what my friends have chosen to set their personal bests and then also when we see this car so Matt fet he was our vehicle designer and holy crap he's amazing at racing games and he was at the top of every single speed wall so when I see Matt fet use this car to take first I'm going to take this car and try my best to uh control it personally I found it a bit backend heavy but by knowing that Matt used that to take first um it'll make me

try that again it'll make me replay and try the game in different ways so connect compare compete that was compare and then the last section compete so this was taking that nudge taking that taking that nudge for friends competition but seeing it through the whole way through to close out that game Loop so when you're on the map screen you can get an autolog alert this is to say since you were last here someone's come along and they beat you and then when you follow that autog alert that button prompt you get taken through to your dedicated recommendations feed and we find that Matt fet's beaten me again so as I said before the signal to noise ratio that's had excessive tuning to surface up all of those beat you recommendations first they were the they were the stickiest ones they were the freshest ones they were the ones that really got the competition going then we also that up with some of those popular um those popular events to show

what what your friends have been doing which was sometimes really interesting if you got um a run of friends trying to compete on one event and then we finally if we didn't have any recent beatus we followed that up with uh these events that we recommended to push you through push you through the progression to so that you could play the whole game through the recommends feed if you if you wanted to and keep just sort of hoovering up those events you hadn't done yet so once once I choose to play a recommendation uh it tells me about the recommendation it tells me about the recommendation through the loading screen and then again in the pre-event screen and then the whole time I'm racing rather than racing against gold silver or bronze I'm racing against that recommendation so it's something that's there for me to focus on the whole time I'm playing and then when I finish the event hey I came first and I beat clean Pete so uh it's great cuz with that I know

that my recommendations are updated but I also know that clean Pete when he logs back in he's going to have a recommendation sitting there for me to say hey you know what just saying Sven just beat you and then just to rub it in we get that automated story and we give the player the opportunity to post it to the other players's wall to just really rub it in because then everyone can see it and the game takes me back to the autog recommends feed ready to go again and that's the loop closed so that was a breakdown of a session of autog uh but next I just want to talk about the outcome so overall the game did all right I'm quite pleased with uh with its reception my particular favorite here is Keith Stewart from the guardian so um The

Guardian if you if you don't know it it's sort of like the British New York Times and that's a review I can show my mom so uh she'll be really proud of me so way to go but the thing that he experienced and recognized is that the game creates a beautifully seamless sense of competitive multiplayer action even if you're not online at the same time and that's exactly what we set out to do it was also at launch the highest rated Need for Speed game and then we won a uh baa for it which was really nice because that's that's the industry it's voted for by others within the industry it's sort of like the British Oscars um and that was awarded to us for Innovation and making great social gameplay they really honed in on this idea that we we had found a new ways and and really got it about playing with your friends this idea of the social asent gameplay

and the great thing about it being a game that then launched was real people could now play it and we could see how our design ideas actually paid off and that's the great thing about having uh Telemetry and having this time passed since it's released is we've got a real sense of how people played it over time so we know that 45% of all the events that were started were because of a recommendation and a player took on average five attempts to beat a recommendation so you're you're seeing that we were getting real replay value through the players that engaged with it we had real replay value for them and also 17% of event starts were through that dedicated recommendations feed and that's actually a really big deal so 17% like one in sixish um but those people are playing purely for the competition they're not playing through the progression and they're not picking their own events as to what they're playing they're not they're not looking at them map thinking oh I want to play this event today they're just playing whatever they

whatever the game tells them this is where you're going to beat your friends so that was that was really neat the most competed over event was vanishing point and that's the same one just by coincidence that Matt had that competition with the it guys uh and it was a great event this was this was a pure expression of of great arcade racing it had a low barrier to entry it was early on in the progression and it had this really straightforward simple track that ran through the middle of that desert um and it only had two turns but you had so much depth in how you placed that car to shave off those microsc and how you when and how you kicked it into drift um and also it looked great it had the Pagani z in it it was an exciting event to play just because it was really straightforward straightforward but also very fast and then we've also got this Telemetry through looking at um this graph is actually generated

after three years of play so it's a sort of uncannily straight line it's hard to believe I had to check the data myself but I don't want to push the idea too much that um of the relationship because as everyone knows correlation isn't necessar necessarily causation but you could say that hey the longer players played the more Need for Speed friends they had so need speed friends that's friends that they added to the um friends that they added to their friends list after they started playing the game that also had Need for Speed so as they went through the game we saw people adding more friends or having more friends and then so but then you could also look at that from the inverse of the more Need for Speed friends a player added the longer they played because they were getting more replay value out of playing these against these different people that they were adding and to put a slightly different view on that same idea is when you plot

the total number of Need for Speed friends um against the rank that they achieved in the game so the game's progression how you unlocked cars was by earning XP to go up a rank and so what we saw here was even after reaching that top rank players kept on adding friends to squeeze further playthrough autal so it really meant we were beginning to see or what we've seen is that the people that engaged with the friends play really squeezed a lot of extra a replay value out of the game they were doing that thing of using the game as a means to keep playing with their friends rather than trying to play for uh for sort of reward in-game rewards you know intrinsic rewards and as I said earlier what that leads to is it spills over into the real world and we can see just from looking at Twitter some of the things people just keep those conversations going you know that they're tweeting to each other saying like hey man I've just beaten you um and it's really nice I don't know this last guy asynchronous gameplay is my

ideal method of multiplayer console games need to get on that tip uh hop suit was a nice start # autog # dog I have no idea what # dog means but we'll take it and to date uh the game's had over 10 million players so that's 10 million unique players have logged into our servers which is cool so then fast forward to today and again being part of EA we can look at the trends across uh a wider portfolio of games and we see that that connected Gap so here we've got the number of players that are connecting to our servers and the number of players that are playing synchronously online so in online multiplayer and we see that they've both grown but that connected Gap it hasn't gone away so even though we've got more and more online games and very heavily online games we're still seeing a lot of players that are connected but aren't

choosing to go into online multiplayer so that's really important it's saying that the opportunity that led us to make Hot Pursuit back in 2009 or make those decisions that led to autog those conditions haven't gone away and it's still this big opportunity the other big Advantage we have today is we've got just more and more Academia and games and and psychological research into player motivation so in 2015 so last year there was an independent study that showed that 17 78% of American teenagers agreed with the statement they feel more connected to their friends that they play games with and also from this study in 2014 that was commissioned by twitch we see that gamers are really inherently social people so for those people that identify themselves as Gamers we've got uh 57% of them agree with the statement my friends the most important thing in my in my life whereas for non-gamers that value

is 34% so there's a big difference there and what that says to me is that oops sorry jumping ahead yeah what that says to me is that Gamers want these social experiences and we need to give them to them we need to find ways to to make genuine interaction between players and then today admittedly there's more advanced friends networks you can follow players you can have much bigger friends lists it's no longer limited to 100 players but it's still hard to all be online at the same time with your mates especially now that every time I seem to boot up my console I've got another couple of patches or updates to download before I can even get into the game and all we can see from Most Wanted that the which is our last title that the majority of players when they're matchmaking online they're going just against strangers and not playing with

friends and the thing is playing with strangers it's easy to forget that and make games that are predominantly expected to be played with friends because as developers we often sit with our consoles all lined up next to each other and sitting alongside each other playing our games but that's not real world conditions and it's easy to um to forget when you're when you're making your games how important it is to find ways to put friends in the game when you're playing away from each other when you're playing at different times so that opportunity that's still there and since since NE for Speed HP suit there have been some other really good um examples of asynchronous social play but it's not become this dominant Mainstay so in racing our next racing title which was Need For Speed Most Wanted we had uh we had autog again it was in the open world but we had Billboards because we wanted to be able

to put the players face in the game so we had your friends the Billboards were where you would try and go for um go for a long jump and so the billboard was to say this is this is your friend that got the longest jump on that on that Gap and it meant that you could you know you could have that sense of competition with your friends in an open world game and that was that was fun that was something different but then driver Tas I'm a big fan of for's driver Tas system system because every time I go in and play play that game even if even if there's friends that haven't played the game they're still their name is above the car so I can attribute I can attribute the actions of that other driver that AI driver to someone that I know and it makes it just a little bit more fun and and um personable the the problem for me with driver tire is There's No Loot back so it's not sort of a closed game Loop of you don't get to see what your drivert did when they're out there I mean and I'm sure that be a big challenge but uh yeah there's there's a great job done

there but there's still so much more that they can do Drive club now Drive club's a really interesting game because with their development having had all of these ongoing iterations I'm making big assumptions here I don't know any of the drive club team but by seeing how it performed in the wild they've now got to this point where they've got this really great system where you can you've got this great async play because you join up with a club and then you work together against challenges to uh to get rewards as a club and you're competing against other clubs but again it's all asynchronous you don't have to be online at the same time but you get you get that shared progression uh and then actually just jumping this is not a console example but Real Racing I love the fact that they've made a physical ghost system so you can take your friend's time and then you can actually race a real car that's that's physical that you can bump up into to uh to try and beat it

so then mobile as I was saying with console development it's really easy to fall into the Trap of just sitting there playing a game together with your friends and thinking right we've got our connected experience whereas on mobile so much more of the iteration is done through this sort of um these quick release Cycles so I think it's naturally a much better fit for making a synchronous and making asynchronous decisions that really support that so async on mobile is a space where I think there's a lot that we can really learn from as console developers and start thinking about how we get more opportunities to make social play through connected players that don't necessarily want to be online at the same time as others especially in a competitive space and then now for us at Criterion this is a picture of our board game Corner we're a big fan of board games and as I'm sure the uh table that's upstairs

is quick to point out we're going through a golden age at the moment with these master of async social mechanics because these games are all about building trust and creating deception and then using peer pressure to and collaboration to make make new game mechanics it's really interesting what they're doing there so there's a lot of these great inspiration points for how we could build new async mechanics but again it's something something where I don't think the examples we should be looking to for building new types of ASN play come from console cuz they're just not there yet so my advice to everyone would be to play as many different things as possible for that opportunity of trying to get players to play with their friends well connected but not playing online together we should be trying to invent new ways to play and and just be thinking about um well just sort of be inspired and be iterating and trying trying things in

real world conditions so that you can actually get a sense of async Play Because async play is different to synchronous play in that you can't play test it in one session you can't sit down and be like right we've made this change let's have a go and see how it how it feels it's something that emerges between the sessions it's something that you only get a sense for once you've once you've tried it and then tried it a few times and tried it in real world conditions where you're not playing at the same time so yeah as I said that connected Gap that's still there and I feel like that represents this huge space for Innovation so I'd love it if um if we started seeing more and more console games try and take that on and try and make good asynchronous play for players to have that social experience have that online experience so then just to summarize autologue so hot pursuit was a game about Arcade racing and it was this

pure arcade racing game and it really focused and intensified that core we had this sort of very overworked social network but we paired it back to this idea of beating real friends everything was about that b it and that Amplified the core racing experience and the outcome of that was that people added more friends they kept adding more friends and they were playing with their real friends and out of that they got much more replay value out of out of the game and then fast forwarding to today we can see from our Telemetry that that connected Gap is still there so we've got more players online but they're still not playing synchronously so there's this great async play out there but it's just not there on Console yet so I invite everyone in this room to get Social start trying to invent new asynchronous mechanics because I I really believe that this is a big opportunity and it's primed for break breakthroughs so take what you've

learned today go out into the world and beat it sorry um so that was my talk but we've got a couple of other talks from Criterion going on at the moment it's actually in a different room it's at the same time which is really annoying but you can catch it on the vault is um Pete Lake who's been talking about rapid uh idea visualization techniques that we use at Criterion and also with his talk you get some juicy behind the scenes uh in production pictures from and videos from black and from Burnout 3 in Revenge uh but then also on Friday at 1:30 we have our wonderful tech director Molen who's talking about how we've changed as a studio how we are um in this new phase of development where we're where we're building our new game which is a really exciting um Team to be part of so uh that's it for the talk but does anyone have any questions feel free to come up to the

[Music] mics hey did you hey how's it going uh did you notice any major differences in Behavior between players that play synchronous multiplayer and the players that engaged with the asynchronous stuff a lot like differences in how much they engage with the game or anything like that sorry what was that last bit uh just differences in how they engage with a game how much they play or um not so much for Hot Pursuit because with Hot Pursuit the synchronous play um didn't get as much pickup as we would have wanted to but then at the same time we didn't focus on it as much and and there was there was great play and great fun to be had with it but um it's hard to really have enough data to draw comparison and it's not something that I looked into in a lot of depth before this talk

because it was so focused on autog so um any data that we might have available I don't have fresh in my head I'm afraid okay that's cool thanks thanks he well first thank you for that uh very good talk thanks [Music] um sorry just give me a second uh that's okay I have a few questions um first uh did you keep the web browser on the console after your first uh two experiments so the question is did I did we keep the the web browser on Console after our first two experiments yeah I know you you changed the way the user interface was but did you actually keep the the underlying web browser so we didn't keep the underlying web browser in the client um we didn't use it for anything but we we did use we kept a lot of the same back end um some of the backend we rewrote because we' been writing all of our back end in PHP just first speed uh whereas then that then was harder to manage for going live

so there were some aspects that we rewrote but um that was the back end in terms of the front end the front end was all custom Tech and was just driven by Json Snippets that would come down from the web uh and also um yeah some other bits of binary blobs and things okay um did you try any other aspects other than the speed for the Beat It components you know like I'm going to beat that guy in a racing game of course the speed makes the most amount of sense but did you try anything else uh yeah so the question was did we try anything other than time for those comparisons um we actually yeah underlying that the under the underneath system can generate those recommendations based on speed Wars that go either way we played a lot with we did have events that were based on points I think the rapid response actually maybe it was the Interceptor was a point space game where your time was just one factor that came into it as to how quickly he took someone down I think we also had

like how many weapons he used and and other things to so that we could try and shap the players play in a particular way but then ultimately it was just that Simplicity that came back of by keeping it just about time that's a natural thing you don't have to explain how those game Mechanics Work everyone just implicitly knows it so it became just a much more natural important thing to be able to talk about and it short cuted a lot of the um the difficulty of trying to explain the competition before you before you said it down okay my last question uh so I've worked on a game where we had a similar system and um generating a graph between all the players and the relationships and all the processing to generate that um how did you guys handle that cuz 10 million I mean obviously not 10 million simultaneous players but I mean you must have had a tremendous server load for that how did you guys deal with the complexity of that uh so the question was with all of that real time sort of uh comparison and just churning

through all that data how did we deal with the backend is that right in terms of the technology platform um with a lot of unexpected life support would be one answer so as in we um we did stress test and we stress tested as much as we could um and then also it's amazing there's because of because the amount of investment that Facebook and Google have done into web backend services there's so many great tools for scalability and also it's something that lots and lots of different there's a lot of people outside of the games industry that know loads about how to um how to scale up web very efficiently and so we lent on that because there's a lot of great knowledge and how to how to do that um but then we could have done it more effectively and efficiently than we did because we still actually shipped with some of our back end in PHP uh which would have been better probably if we had sort of gone and and baked that down into into see I'm not sure though actually so I I changed over to become a

producer to avoid questions like that all right thank you very much yeah thanks hi it was a great talk thanks uh regarding the auto version three how long did it take you to to balance it to to your satisfaction so how long did autog version 3 take to balance to our satisfaction uh good question so it was lots of iterations and the it was quite late in the development that it changed the order to have all the the beatus up front um and then it was even later that we had the prompt to really push you through the whole way from the map screen back into the game um and it was that was part of just the Insight we had from some of our from Matt it was Matt Webster who Who Matt Webster and Steve C Who are the two guys that really drove autog and they recognized that and they felt very

strongly about it whereas on the team we felt like hey we've done all this work you know we've done all this work to to track the popular events we've done all this work to uh track those recently played events try and make some intelligence about what you've been playing and how you want to play but ultimately all of that just fell away next to the this idea that someone's beat you do something about it so yes so it was this gradual process and it was as we shift did that mentality from pre-production and discovering what what we could do to then actually production and shipping thinking okay what's the most important thing what's the best player experience so um yeah initially we had loads of ideas we put a lot of effort into trying lots of different things out but then as we went through the development it all just got slowly paired back and back and back just to that Simplicity of the your friends beat you right uh and my second question is uh regarding did you employ any automation to to test the correctness of this system sorry so the

question was did you did you do any automated testing for uh making sure that the system is working correctly yeah so um the question being did we do any automated testing yes we did a a chunk of automated testing we have a a team at at EA that helps with our automated testing and so we had to work very closely with them because it was something where it was so driven by how players played the game and how we expected players to play the game so it wasn't that it wasn't straightforward toward toate you know it's not like just saying this is a game where we're going to have have many players simultaneously playing on servers uh so we need to stress test those servers it was a lot more complex than that because we've got to sort of Imagine well how are the players going to be playing the games when are they going to complete the events and how many friends are they going to have um so it was tricky to do the testing but did have a a big automated pipeline to that we could write scripts against to um test against

our infrastructure thank you okay you're welcome any other questions okay so um after this talk I'll be hanging around outside just in in case anyone wants to have a chat about it or find out more and I think Pete Lake who's another producer who's just finished his talk he'll be out there as well so if you want to just come hang out and talk about any of criterion's games then we're always very happy as well okay thanks very much