Changing the Time Between Shots for the Sniper Rifle from 0.5 to 0.7 Seconds for Halo 3

[Music] all right everybody can I have your attention so you are in room 130 it is 1:30 and I have about 130 slides so we're gonna go really fast [Music] welcome to design in detail how to subtitle your talk with an entire paragraph and make sure you don't forget to mention halo 3 before we start is anybody thirsty can I get you anything okay I'm gonna talk very fast like I said let me tell you first about myself I am a gameplay design lead at Bungie Studios I've been there for 11 years worked on all three halos and did the webbing balance for all three halos except for the halo and pistol that was somebody else and it was at the last minute and they didn't even tell me so when you curse somebody for the battery of a spread that's me

there's a face in feelings last year I was at the artist city of Chicago that's where I grew up not at the Art Institute but in Chicago and I saw one of the most famous paintings of all time and I took a picture with my iPhone isn't me want to guess what this is all about so many got it good this is a Sunday afternoon on the island of the ganja it's by Georges Seurat and this painting and started inspired my talk and it also inspired my very long title you probably will recognize it better from a little farther back what I showed you was the lower-right by the monkey this painting is not famous for how it looks or what it shows this is famous because of how it was made Seurat living the 1800s and he was very interested in how we perceive color scientists of the time we're just discovering that we see one color but what we're actually seeing is a mixture

of several different types of color and to demonstrate that fact he invented the art school of pointillism which basically takes dots of basic colors and puts them together to produce any image with a huge variety of color so I started thinking what happens if we take hey on three we point lies it we break it into these tiny details and we sorry I'm pop and pop it a lot and then we take just one of those tiny details and explore it exhaustively specifically the time between shots for the sniper rifle so this is my actual talk before we get started I have a couple caveats my opinions do not represent Bungie Management they rarely represent reality this is the past as I remember it and I'm sure that I'm glossing over stuff in fact one of things that we're talking

about is how unreliable your brain is your memory I have some examples from games they are all from games that I like just to avoid controversy just to avoid fanboys I'm not ripping on any game that I criticize and this talk was really hard to write like I said it's in like 1:30 now it got to 300 slides at one point because when you are focusing in on one little detail you actually have more and more to say so I'll per be presenting the full talk at sniper rifle con 2011 thing so first we're going to get some contacts from Halo 2 and what I'm going to show you through halo 2 is sort of the four or filters that I'm going to look through this detail through peyote was pretty popular we released in 20 2005 and according to Major Nelson we were the top apply of game of 2006 and 2007 2008 2009 there

was lots of months in there where other games came in a win but if you look at the uniques for the whole year it was it was halo - it's going to shut down in April and I'm assuming gonna be the top game of 2010 and then because live for Xbox one is shut down forever sniper rifle remain balanced all the time and what this does is it gives us a kind of a practical definition of balance which is often argued about and I'm going to say the only definition of balance it actually matters is that balance is longevity if it doesn't last wasn't balanced so we're gonna do a little audience participation balance this equation and just as soon as you get it shut it out that's fast you're a programmer aren't you so now balance this stack of blocks anybody okay

so my point here is here that balance is not a process it is not something that you do - game is a state of the game and so while you're trying to get to get lunch evety is a static situation that is balanced rule number four of Jenga code it with me yeah your turn ends ten seconds after you stack your block so even in a game like Jenga they understand that balance requires lasting requires longevity let's talk about the halo - crunch there's been talks about the halo - crunch I'm not gonna rehash him it was less than ideal my post mortem is don't try to fix bugs in the tutorial and balance the weapons at the same time find something else into the tutorial production is always worried about a repeat of this situation so they're constantly asking me when I'm gonna be done they kind of have a

love-hate relationship so I invented the balancers paradox balance can't happen until the end balance can't wait until the end you can answer any question with this with one of these two points but the promos omit to the end and you have to resolve the bouncers paradox so the technique that I came up with is to balance and passes so what does this mean it means at the end of each pass through the balance the game is static the game is it has that longevity to a certain level then you don't backtrack once you achieve balance don't lose it you can try to balance at a finer degree of detail but but don't ever upset the applecart halo 3 head 6 bounce passes roughly I mean they kind of blur together but will flow strength

limitation exceptions and perceptions in nowhere I'll get to these in a or detail so after all that stuff about able to dominated life he was not balanced we ship a o2 and the desires after waking up from the coma started to play it on live and felt wrong it was different than how we played it in the office in a lot of ways it was broken and we sort of looked at each other and we knew this is not gonna have legs he's not gonna last how do we know this well during the development we had fine-tuned our sense of balance Malcolm Gladwell lays out in his book outliers the 10,000 hour rule 10,000 hours of practice is required to achieve the level of mastery associated with a world-class expert on anything so this means I am an expert in

trying and trying to balance a sniper rifle it was more offers an alternative definition an expert is a person who has made all the mistakes that can be made in a narrow field and that also means I'm an expert at trying to balance - and in driving too fast for conditions but I asked about that so in order to establish the sense about you you need to understand how your brain works so I'm going to talk just really briefly about the neuroscience of intuition you have this thing called the orbital frontal cortex easy for me to say it's called that because it's located sort of behind your eyes but we call it the gut and it goes through a learning process what it does is it builds a model of the world tries to make predictions and anticipate what's going to happen and if it's right it releases this chemical called dopamine and that kind of cements

the neurons and into that configuration and this is over and over and over again about everything or may even you don't even care about like ordering fast food like it it tries to predict how long it's gonna take before they bring your money back and you can't turn it off but you can sort of guide that process and what you're trying to do is you're trying to get the game inside of your head and that's what happened to us in Halo 2 we had the game so internalize that Demma be played it was different than the game that was inside of us and we knew for the way it was different that we're gonna have problems down the road so back to here - nobody wants a patch chooses patches hard luckily we had networking bugs so we never could have to action anyway choosing what to patch was even harder some things we may melee do a little bit more damage grenades smaller radius we were just trying to make it so that the game didn't degenerate into dual

wielding all the time but choosing what not the patch was artist of all and we didn't patch a sniper rifle and that was like the top of the things that we were going to change but we ended up not it's just didn't have a Spartan impact and other changes and our studio head used to be a test lead and so he's pretty harsh about stuff like that so my final theme for this talk is make the hard choices don't put them off don't let them happen by chance don't let something else do it make it yourself what do I mean when I say hard choice is why these shape choices hard again answers in your brain but this time it's in the neuroscience of reasoning this thing called the prefrontal frontal again easy for me to say cortex and I don't know why it's called that but I call it the brain your brain is a poor tool it can only handle

a very small number of factors it is easily confused it's subject to all kinds of errors and fallacies and biases you're not going to be able to reason about your game and it's irony but what you can do is reason on a very small detail scale so radio lab great podcast slash radio broadcast New York describes this psychological experiment where they bring people in they give them a number to memorize some people get a two-digit number some people get a ten digit number and then they say ok we need you to walk down the hall and go tell this number to our assistant in other room on the way and it's not clear that's part of the test because that's how all good psychological tests work somebody offers them a smack and they can either have an apple or some cake and what they find is

that the people with the short numbers she's apples because they're thinking calories got too much weight you know their reasoning about their choice whereas the people with longer numbers we're not talking the first 100 decimals of pi we're talking about a 10 digit number you know you have a social security number or something they choose cake like my huge margins that's because they're rational brains oh 10 numbers is all it takes to fill your brain and so don't think you're gonna look at a weapon tagged with 40 values and understand it all in your head at once at least not rationally so my four themes balance is lunch everything balanced and passes develop your sense of balance and make the hard choices we start with Hale three papers design are there any producers in the audience a couple don't worry I don't bite

this lies for you papers not happens in pre-production you need to start before you have 40 artists sitting around with nothing to do and don't exit it too early that is actually what it lets you halo to having such a rotten crunch every eight-game clean object needs a paper design as a desire it's your responsibility to be disciplined and this is going to save you time in the long run so this has never been seen outside of bunchy I think probably most of the current bunching employees have never seen this I didn't ask if I could show it but that's okay this is the paper design for the original sniper whoops skip it sorry this is the paper design for the first sniper rifle you can see it's four lines long talks about

the range five round clip which is not where we end up shipping it's got a lot of stuff about the flood in it so spoilers sorry about that and the the second trigger actually we originally planned that the second turn controller is going to control zoom so some important things to note about that spec it's pretty flawed there's a lot of things that were wrong it's not five shots the contrail is not small or short there's a lot of mechanics that are unclear the zoom is totally it says to be decided in the spec headshots are very vague and many decent details are missing maximum ammo cab body shot damage you couldn't hand this off to a programmer and get anything back they'd have too many questions that's okay this is paper design paper design is about role which is an abstract concept so now let's look at a go freeze paper design it's a little bit longer a little bit

more organized it ought to be this is the third time we did it I'm gonna I'm gonna break it out in further slides so you don't need to worry about reading at all but I'm just kind of get a feel for for how big it is and how detailed it's still pretty abstract but the role is even more clear there something about fiction in it there's something about enemy characters very little bit about art so the very first line role long range into the kill sniper rifle the reloading makes it difficult to use and really the only reason why reloading is in there is because there's a coven of sniper rifle where he makes it difficult to use talks about range that's what the damage talks about its limitations that's it clear leave specify the role in your paper design well let's look even closer it has a bunch of supporting mechanics so two zoom levels magazine four shots with a slow reload it talks about that fielding we want to bring out does massive damage use good descriptive

words like that it also points out critical assets this is not so a producer can go and make a schedule out of it it's so that you're thinking about the sniper rifle you have to be thinking about that special death animation every single time you can't do one without the other and it's got a couple little special details like there's a delay you fire your last shot really what it should be doing is unzoom in reloading immediately right but then you're not seeing what happened with your last shot so it waits a second half second actually include what is necessary nothing more so how do you develop your sense of balance for paper designs it can be done I think I can look at a paper design and have a pretty good idea of how it's going to work mostly you're looking for the role and a couple of key factors that are the results of the role so you're looking for a sense of depth you want to balance between simple and

complex if you make a role that's too simple people are gonna get bored if you make a roll of this too complex people are gonna get boggled so what you want is a barely manageable number of choices you want people to look at your weapon and be thinking of all the crazy stuff they can do but really hard to execute on most of that stuff this is where role comes in because game theory sorry the sniper rifle is sometimes the best weapon and game theory tells us that if strategies have the same payoff they'll have a numerical equipment playoff then people are just going to pick randomly it doesn't matter other hand if you have roles you have asymmetry and if you have asymmetry that means sometimes you're going to get moved to a certain strategy or another and that's sort of the core of depth so roles require real differences and that's why rock-paper-scissors though it makes a great t-shirt is bad game design every

choice in rock-paper-scissors has the same payoff so you just pick randomly it's not even it's not even really a game it's a it's gambling so the sniper rifle is the only very long range weapon what does that mean well technically there's being rifle but those are on appearance' missions so one for the purposes of argument let's say there's only one means that for every will there is one weapon it's unique you you don't want to add multiple copies in the same role because you're just gonna add complexity you're not adding depth it's really hard to add weapons to Halo because most of the roles are film there's a limited number you're gonna have to force a little bit of space in there or rely on AI or something and so so rules must be unique rock-paper-scissors- lizard-spock a funny joke not a good game it's worse

than rock-paper-scissors because it's more complicated but it's not actually more design more depth Congress of what I just said is also true the sniper rifle is a necessary role so the ratio of weapons to roles needs to be one-to-one as well if you have a game without certain roles then you're gonna have an incomplete game and people are gonna feel it this is why almost every first-person shooter out there has the same set of weapons it's not because we're not originals because these are the weapons you need to make a game like this rules must be filled rock paper that doesn't even make any sense I should admit halo 1 that close to shipping without a shotgun I know but we didn't think that there was a necessarily and we were proved wrong so

we had to Dennis ground to the last second so the last way that rules provide depth is you don't have a crappy sniper rifle you want to avoid strict dominance what is strict dominance okay who's gonna pick the health potion and who's gonna pick the bees apart the health potion is single-use it just with punishes your health you can go buy it in the store peace of heart is totally unique it makes your health bar longer and you can never buy it yeah everybody picks peace of heart sorry if you didn't you were probably eight it's okay so no crabby sniper rifle because you want to avoid strict dominance and then you want to practice iterative deletion which is when you go in and you remove all the dominated strategies so you remove all the crappy sniper rails and then you remove all the strategies that were only good against these strategies that you just eliminated so for example if you have an anti-tank mine and then

you cut the tank you should also cut the anti-tank mine otherwise you have a strategy that doesn't dominate anything so the second way that you develop sense of balance about paper designs is to look for anticipation and anticipation is not the right word but there isn't really a word maybe in some other language what you're trying to do is pick between chaos and certainty if you have a game that's too chaotic then everybody is just guessing all the time if you have a game that's too certain then it's predetermined got to be careful because certain parts of your audience want the game is totally chaos because the number of good of your game and they want a chance to win and some elements of your audience want a game that is totally predictable because they're really good at your game and they don't want anything to get in the way of them winning we really need to lie in the middle and that means that balance is probable but not inevitable future events may be one way I think about is some a bad form of anticipation which is I am NOT a great fool so I can

clearly not choose the winner in front of you you all know the scene I'm sure Zini starting himself in circles right and if they were having a sword fight while this happened he would be doomed so there's a guy in audience David Stirling who calls this kind of anticipation Yomi which is Japanese for reading the mind of your opponent Yomi comes in layers oh no I'm out of power we're back on so the first layer is I know what I will do second layer is I know what he will do third layer is he knows what I will do in reaction to what I know he will do this one is great if you're a professional street fighter player a championship level it's not so great if you're a frat boy and you're just wanting to play while you have a beer so for a game like halo stop stop at three

in fact stop at two and a half not beers Yomi layers the sniper rifle has one purpose if I see you carrying it you know what you're gonna do I don't I guess great so you want to limit them out of second guessing and that helped people anticipate that helps your balance successfully sniping sniping confers no benefits how many of you have heard that feedback loops are always good in games it's not true if you have a positive feedback loop do well you get rewarded by being able to do well more you're gonna win you had a kill streak I mean you've all been on the receiving end of a kill streak it's not good balance if you have experience points I kill an enemy they a a Warcraft mob say and I get experience points that makes

me better at killing him right it's great for a single-player game it's great for a game that's built on investment and progression but if you're really just pulling back to balance how's it a feedbacks in this venue out of control same thing with negative feedback so if I do bad I do worse that's why the best new features of modern warfare is that you if you do bad they let you do better right so if I miss with the sniper rifle now I'm outta ammo now have a sniper rifle or if I miss four times and I'm gonna reload now I can't even fire my weapon feedback loops push you away from balance they're great for a lot of other system in games they're not great for balance so the last-minute rules do is they let you bite off more than you can chew sniper rifle is never your only option I'm going to hurry it up here you want to use multiple gameplay channels so that you're never relying on just one weapon being balanced sniper rifle takes

up a very specific inventory slot so if you have monolithic inventories like cars in a car racing game those are really hard to balance you gotta bounce every targets every their car on the other hands and slots are easier weapons in Halo only have to be balanced against other weapons that ought to be balanced against the Warthog everything's don't have to be balanced against equipment to use bounce groups whenever you can the sniper rifle is not in every map so you can kind of limit the interactions in that way our choice is about a paper design picking the right mechanics we left off the sniper rifle night-vision scope in this paper design because it just wasn't that important to become the right mechanic you want to capture the move so engage the animation the imagination don't use sterile language when a focus on your quote first night your spiced sniper rifle was done before the over shield should be obvious but a lot of times you have some little weapon that's kind of TWiki and weird and still getting you

use one place you spent too much time on it well a very important mentor my life said once you're done cut hack this is always good advice and never sounds like it but this can't describe the role cut it if you think it's too similar to something else cut it if you can't sell it to the team cut it you're gonna want to cut so that you can focus on the things that remain so the second pass in the sniper rifle setting is is setting up the initial values and what you're trying to capture here is flow again producers this happens really early in production as the objects come online you want set the in front you gotta stay ahead of everybody else and you want to give your designers lots of space in this stage a feedback is not important when you're setting up the initial values it's the fact it's best done in private you wanna do it on your machine you don't you want to check it in sometimes you guys heard of Csikszentmihalyi his book flow by XM

behind my exam I like it's like flow right there right Department zeros book in the 70s it's about that kind of state of consciousness that you get in when you're really grooving we're in the seventies you didn't mention videogames in it so how do we apply it to videogames almost there's a lot of research that says your physical aspect influences your mood so smiling makes your happy laugh and makes you happy sniper rifle fires with a very specific timing because your makes your fingers happy if you have a set you set up our sniper rifle so it's too slow your fingers get bored in between shots they're waiting if you send it up to it's too fast then the shots blur together you're not really seeing the results of every single shot so you want to you want to focus on cadence when you're trying to capture flow verisimilitude which is the quality of seeming to be true sniper rifles on a trigger for weapons verisimilitude is

really easy for others it's not some further mechanics it's not so easy so what you want to do is you want to make sure the finger movements match the action in some ways you're doing on our trigger a punching it's on a button you don't want to set the opposite because then people can't get comfortable with your control scheme they can't enter that flow state so if you have punching and kicking on a stick say right like that isn't analogous to the to the action that you're taking switching the inventory on a d-pad that's not the way you do it any other interface right and some people aren't used to that and it's gonna they're just never going to learn to do it naturally without popping up out of flow virtually anything on a stick click like I said we were wanting to put the zoom onto the trigger for a good reason like nobody gets a stick like if you play a lot of Halo eventually you'll get used to it but it's not it's not easy so you want to play in your spectacle the sniper rifle has a delay on the sound contro hangs in the air for

a really long time like I said there's a special animation when you get headshots and the reason for that is if you if you are in a flow state the spectacle keeps you there it stops you from being distracted from what's going on off your TV you take a story from this screenshot red guy shoots a blue gland with a sniper rifle misses then there's this crazy ricocheted insanity and the man can't get to the trail so how do I tell the story from the screenshot that's not me that's not I'm not that right guy that was the blue guy that dodge cause alley basically you want to pass the YouTube test and the sniper rifle looks really good on YouTube there's a giant contrail there's a special headshot animation you want to exaggerate your causality you want to delay it artificially stretch it out so who's you've seen that chain explosion of grenades in Halo there's like a point one second delay between when one grenade goes off and the next one goes

off the next one goes off so you can see the whole process unfold it's also one of the reasons why halo feels a little bit floaty is because we want it to be a little bit slow so you can watch the causality it also makes it a really big game to for other people to watch causality preserves flow because you understand what was happening the whole time and you can kind of predict it along with with what's happening so how do you get your sense of balance how do you know what initial settings should be lower your flow barrier so sniper flow is very fragile it's really easy to pop out of it and what you want to be able to do is make yourself easy to entertain you got to train yourself to ignore framerate problems you want to ignore bugs and glitches you want to excuse me practice filling in details with your mind when you use your imagination and see the game as it's gonna be when it's done and not how it is right now make mouth

noises and it sounds ridiculous right don't do it at your desk like if you have a shared space but you know you peep you that that puts you in flow that gets your whole body connected what's what's going on the screen and then play the B games with a with an open mind the last game you play with avatar I had so much fun does he just move the game where it is another hand sniping flow has a high ceiling meaning that if you're really in the flow state you can really get in the flow State so you want to also raise your flow barrier you want to never be satisfied you wouldn't get in a mood where you can get irritated with every single bump gotta unleash your inner critic you take the best game to come out last year and you just rip it apart point out every flaw like I understand everything that's that's preventing you from reaching that maximum flow and I'm warning this is gonna wreck your ability to play games for fun that's okay because now you're

making it for fun this is why your designer those of you in the room that are designers I really think this is the essential design skill don't expect other disciplines to have a programmers and SC bugs are a certain to see kind of still-frame and compositions producers are going to see inexplicable delays and you're just like constantly doing what looks like nothing over and over again but you need to learn how to apply conscious control you know know when you're having fun no when you're fooling yourself this is really hard to interview for one thing that I have noticed is that philosophy majors tend to have this more often than the general population I think it's because in philosophy you're being taught to analyze your own thoughts and your own opinion sort of why you believe what you believe and that really helps for controlling your own flow state most sniper rifles aren't fun I'm sorry they don't find that flow or they had some of their goal like they had a mission they had it support or

something the reason why maybe because they're not searching broadly enough they're doing science they're making one change at a time they have a control group and they make a change and then they compare it can't do that you got a this stage you got to go drastic you got it let's let's try a sniper able they'd only fire once every 10 minutes okay well it turns out that's not fun but that's okay because we're trying to it we're trying everything I don't tell that engineers that you're trying all this stuff try to get a tool you can do yourself so thank you crazy my engineers think I'm crazy so imagine that your turn to learn how to drive right but nobody's trying to teach you accelerator doesn't do anything until you find a mission the blinkers are pointless unless you're in traffic don't tell me why if I said that but they're pointless they don't do anything unless you're in traffic and sometimes when you release the emergency brake you roll down the hill and you crash you don't understand why that's okay you got a change in groups you're gonna change multiple things at once don't just well I'm gonna iterate on this one number until it's perfect

because you're never going to get there because maybe that number doesn't do anything until you change something else this is a great book embracing the wide sky it's but and by Daniel Tammet he has Asperger's and he's just numeric genius and he sees prime numbers as pebbles and all their numbers are these Jaggi things he literally sees it that way when he looks at it on a page and you need to develop that same sense for me it's a clique like what I know I've set the right value and cliques this picture has nothing to do with anything I just thought was awesome so once you find these flow knobs write them down for shot per clip zoom time total mo is enough to fight for an entire counter you don't want to change these later except under extreme circumstances Merle I should leave this step doing a backtrack so know what you

cannot change and then have the discipline not to change it once you find these flow knobs crank them up crank them to all the way sniper rifle is perfectly accurate it travels instantly it seems almost instantly it always kills with a headshot like don't don't go half way with these flow knobs because the more you loosen them up the easier and the more people are gonna reach this flow state you're also going to notice the things that aren't allowing you to get in this flow state if everything else is loose then the one thing that's tight you're gonna notice we're really easily alright so you got a you got the flowing sniper rifle it's fun by itself it's probably really powerful how do you put it together with other weapons this is one there's for the engineers in the house I know you like to fix bugs last I know good coding practice is to optimize at the end but you can just start doing it earlier because it's impossible to balance without a solid build right

bug-free code that's that it's that easy right you need to fix the gameplay book bugs right away a lot of times this isn't a priority you got a bunch of features and probably features I asked for but I can't really do my job until there's enough bugs out of the way that I can get a sense for this balance and optimize there unless you're going to ship your game at sub 30 don't let your desires play sub 30 get a special mode or something you know the test map or something where they can run the game and have good lighting sounds and effects and animations as early as you can this is joke but kind of talk about where if we're going to do a play test we have to make sure that there's light Maps because if there's no light maps people think the AI is dumb I don't know why but it's true it's like a 15-point difference people will think you're a eyes done you don't have good light and then sometimes place over R squared so strength what does that mean I'll put the slides up it's okay

how does strength help you put all your weapons together so Gladwell in another book what the dog saw talks about why are there 50 times of mustard and only one kind of ketchup I don't know I comes up with these topics it's awesome so he says Heinz ketchup the taste begins the tip of the tongue where you have sweet and salty receptors it moves along the sides where you can taste sour the strongest and hits the back of the tongue for mommy and bitter in one long crescendo how many other things in the supermarket hit all of the kinds of taste that you can have the answer is none Heinz ketchup it's got all the flavors they're all really strong flavors and they're all perfectly balanced against each other tasters call this amplitude halo is like ketchup it's everywhere no it's it's got a lot of flavors he's got close range it got

long range you got aiming you got running vehicles grenades millions just tons of flavors all of them are very strong right it's okay it's not overwhelming because they're all the same strong designers call that balanced and it requires drinks one of the ways you can tell if something is strong is through affordance which is a visual clue to the function of an object hider Heidegger calls this ready to hand this is a great phrase this is where sci-fi games like Halo run into trouble what does a proton phase amplifier do I don't know it's not important you need to give me a clue you have to explain a strength to people it's not a strength okay this is a hard core textbook and it rocks like like it's dense and it hurts you when it hits you kind of rocks then if and I think probably there's a couple of the authors of this book in the audience buy it and read it it's great they do

find confident competence this way it's a condition of quality of effectiveness ability sufficiency or success and they show through experimentation that it is a major human motor fire people want to be confident they want to show that they're competent they want to feel competent sniping starts off hard but the more you learn the more confident you get why is this important strength some things that you can get good at and competence is an implicit motivator instead of an external motivator what's the difference external motivators evaporate as soon as they're not there anymore so if somebody's paying you to play Halo since I stopped paying you you'll stop playing on the other hand internal moderate motivation intrinsic motivation the more you get of it the more you want it so the better you get it's never the more you want to play with the sniper rifle and again balances longevity so this is a great way to make your game have longevity there's so many of these

screenshots and like masters you slipped in more cause that's great so how do you recognize strengths when you see it how you train yourself to appreciate these strengths people hate it when you change the sniper evil on them just when I was getting good they say why because it resets their confidence but what you're trying to do is you're trying to prevent optimization brains are lazy and they will optimize out all of the the interesting parts they won't both stop savagely thinking some people speculate this is what beginner's luck is like the first time you play darts you're really focusing in you're like calculating all the winning variants and everything and then as you start getting used to it you bring it to lazy it's like yeah whatever I can do this and you can get worse right until you until you get the muscle memory and you get better again so the other reason why you want to change constantly is preventor should bias that's the way it's always been it's perfect why are the designer still making changes me again ship we could

call this deadline by us I guess but you want to give them something different with every play test you don't want anybody used to the game how it is because it's gonna change especially in this stage you also want to change it passes lots of Impaler weapons kill in one shot and so strength is relative it's only strong when it's compared to everything else and you'll find that things get weaker as time goes on so you want to go through all the weapons and start back at the beginning and see if the first weapon is still as strong as what you made it if can't make everything strong you can't have anything strong so it's kind of a good real film this guy's really good at halo I don't even know what button that is pro players often complain that the guys making decisions about halo stock at their own game it's true I'm having good to do it I'm really not that good it's not be I'm not that good at anything in Halo except for

having a war dog and I'm not that good on purpose I like to think it's on purpose of it the problem is dopamine is not specific when you get don't mean feels good your body saying yes keep doing that success feels good to success in the game it gives you dopamine just like success at balance in the game gives you dopamine so if you're good at the game you're gonna think it's balance and it might not be you've got to keep your neurons plastic don't play too much this goes back to competence confidence is a great motivator if you play too much you're gonna get invested in your confidence you're gonna be one of those guys saying don't use the sniper rifle it's perfect don't forget your paper designs get it you got to constantly go back to you know what was my original intent here and don't specialize I love the word hog and I think it shows it's my baby it's my favorite but don't pick favorites if you can help it don't you're gonna

neglect everything else I'm gonna spread yourself out so what is this not true so when do you want to play a lot remember you're trying to train your brain so spend more time playing what does work and what doesn't work is not it's kind of if it drives people crazy I spent hours and hours driving the war dog around and it's fine and then I spent no time driving some other vehicle that's not fine it's because I know it's not fine I don't want to imprint on that I play a sniper rifle encounter over and over and over again once I know the sniper rifle is good or I find some game where it is good and I play it over and over again because then I can recognize it once at work she got to get it stuck in your head did you notice that the strength is increasing as we get to these five hard choices all right producers this is your most important slide after this you can leave you want this is an actual bug sniper rifle reload it's too long it's bug

12:00 we were in this camp like three days and somebody started giving me feedback in raid as a bug sniper rifle real it is too long don't let people file bugs against balance during this during this rough adage of balancing in fact don't let people file bugs against balance period what'll happen is you know your brain likes to optimize so instead of thinking about what should I do you think what is this bug tracking program tell me I need to do don't do that listen to complaints and I got to tell you don't listen to complaints but but really you know that on the foundation yet right if anybody's gonna have enough of the game internalized to make complaints can be you and you don't have it at this stage so chances are nobody else is gonna have it so don't be reactive don't design by committee or else you'll end up seeing each other in a big circle other hard choices there are certain things about the

sniper rifle that make it strong damage this is accuracy what you have these things random doubt it's just like fun knobs you can write it down you didn't know which ones not to touch later because you don't want to backtrack no one make things weaker again don't change it once you know that much have these strength knobs you want to crank them to 11 make everything overpowered to sort of paraphrase The Incredibles if everything is overpowered nothing is so you've got your flow knob set you've got your strength knob set turn everything else is zero if it's not important for flow and it's not important for strengths not important you need to remove that complexity because it's just gonna get it given the way you don't want to obscure the role that you initially decided on and anything other than those tours there noise we balanced the sniper rifle for most player we've talked about this before once you have things working

in multiplayer it's pretty easy a retrofit single player can make the AI accommodate in a bunch of different ways again you cannot make things too powerful so moving on you go to the Polish pass how do you find Kuna sniper rifle the answer is not by giving it weaknesses if you read a game preview it'll say oh don't worry it's not very Apple is not too powerful cuz it's got some huge negative like your head is ten times bigger when you're using it or something you don't want to add weaknesses right because there you talked about why you want to make things up everything overpowered so what you have to do is you have to have limitations on that strength there's been lots of talks about our play test process John Hopson gave me great one last year it's online or look at it but here are some quick points start before things settle you know somewhere toward late late production do do a regularly

scheduled you want to have a really quick build process can't take longer than a hours so you can make a change in the morning make the bill do plate doesn't afternoon and don't tell people what's changed because you don't want to you don't a bias them you want their natural reaction and then after we focus groups that know what to call it that but sort of the popular name you want to get the feedback immediately well it's fresh for people you actually want to you want argue with them right you want to kind of let them argue at each other and don't spend any target time talking about solutions get together with your designers afterward figure out the solutions you know even talk about it in a focus group so limitations if your game peanut becomes predictable it probably means that you need to limit strengths because there's something that's so powerful in so many situations that people are always using it on the other hand if your game is too random that probably means your strengths too narrow and

whoever happens to be in the right situation for the strength of their weapon is going to win there's nothing really that other people can do game theory tells us that the ratio of people playing the strategy is about the same as the ratio of their effectiveness so if you have a lot of people doing one thing it probably means it's more effective this is really good slide I skip right past it taking team feedback and not giving them a boot to the faces there are three kinds of feedback the first one is very desirable it's I don't like this everybody everybody on your team should be telling you I don't like this this is what I don't like what this is what I like usually this if I don't like there's a even better kind of feedback which is thoughtful I don't like this because I don't like this because it feels bad is even better than just I don't like this but I don't like this because it makes it too long is the recovery is too long it takes you know too long before I can acquire it and

then shoot again or whatever that as long as it's thoughtful the undesirable feedback to is you should change this to this because they're not telling you they're telling you what you what they don't like they're not i've been prescribing how to fix it most of the people on your team really need to pay attention to what they don't like but it's another job it's how you had to fix it and so you should I mean listen to them sometimes they have great ideas but that should be your primary goal of like please tell me how to fix this trust their guts but not necessarily their brains let us feedback me what do people trying to tell you when they say they don't like it they're saying the game in my head is different than what I'm playing so you got to ask once again you're in their head is that the game that we're making is that the game we decided to make in pre-production even people with the wrong game in their head maybe they really like some other game or some other game came out and they're really in hammering with it right now even if they have a different game in

their head their feedback is still useful but you got to understand where it's coming from then you gotta watch out for a development bias people know that all you have to do is go back your desk and change something and it'll be different so they're much less likely to overcome the barriers to entry get confident with something they're more likely to complain but once you ship and it's burn onto a disk people are much less likely they're much much much less sensitive much less likely to complain about stuff like that and I'm gonna I'm gonna skip through the types of players what you want to do is you want to figure out what kind of people are giving you the feedback figure out kind of what style they are if they're an optimizer and they're only using one weapon it's probably overpowered if they get you know really angry really easily then watch them because as soon as they start to simmer that's when you know something is disturbing their flow we've got some guys it doesn't matter what game type doesn't matter what the state of the bounce is they they want to be a sniper they want to

the guy who runs around with the shotgun invisible let him role play it that's great they're gonna be trying to think that nobody else is trying cuz that's what makes them feel powerful are you the guy in halo one would call them not when you use around your mom because we would say hey is this too hard to learn well I don't know let's see if your mom can get it yeah true story I used to I used to balance easy by playing with my nose and nothing else right and this guy still couldn't beat it look for griefers if if Nathan Walpole I mean if Nathan is having fun segment I mean if somebody else is crying and then look for pros these guys want everything to be perfectly predictable they don't want anything to get in their way and that can help indicate to you when when there's too much randomness in your game

and when there's not enough don't spend all your time gonna play test plane you want to watch the people that are playing differently than you but sometimes you need to lay your head drive wow this is really good but I think I'm gonna have to skip it okay so basically everybody has this thing in their head that tells them Oh moving in monkeys from million years that can make anything that makes Shakespeare it's totally not true it's the plan right the Planck length is a 10 in the day 20 30 centimeters the fastest speed is a speed of light C that means it takes less than 10 to the negative 4 3 seconds to travel any distance there are 10 to 80 particles in the observable universe anything that's not observable it means it can't affect us because that means we can't see it that means lights not getting here so we can discard anything that's not observable and there's been 10 to the 16 seconds since the Big Bang so if you take the smallest distance in

the fastest time times the number of discrete objects in the universe over the lifetime of the universe you get 10 to the 139 possible events that means 10 to the hundred thirty night conceivable trials so this laptop has a higher Keys including capitals this PowerPoint deck has 21,000 letters including spaces this talk has tended the 22,000 possible configurations so if every particle in the universe was a monkey and every possible event in the universe was a talk and they started at the dog time there's some lonely one out of a hundred and fifty eight probability of this talk being created randomly my point don't relax random changes you're gonna underestimate how long it's going to take you to figure out anything by just changing things for handling it's really tempting you want to change physics over math players don't perceive math

directly so you're better off changing the rate of fire which is something they can't perceive then damage per shot or you want to fix the base damage you don't want to add random variance in there and again players don't perceive averages right so you want to fix that instant snapshot you don't want to fix it by changing it over time it's really tempting to say allison i pref was too powerful but we only give it six shots so it's not actually gonna kill a larger percentage of the people but people don't they don't understand that oh well that guy killed me five times the mukil you can only kill me one more time you need to balance with players can see be wary of fairness it's not balanced you're gonna have a temptation to go into symmetry because that's fair right but Hale is not fair especially the sniper rifle but on the small scale that's fine but you don't want to end up a tic-tac tie that's bad game design you want to polish off your edges in this

stage all right finally with five minutes ago we're at the detail this is the point in development where we actually made the change from 0.5 to 0.7 seconds sniper rifle is overpowered it's ok remember we've made that way pom is is making all the other weapons feel weak the optimizers those those guys thing reason they always find the perfect weapon to use we're using it exclusively the depth was dama the game was simple it became halo : snipers and the role players who refused to be sniper abs because that's not who they imagined themselves as there didn't Crete that's been used at close quarters that's like directly opposed to the role that we intended in the paper design so it's outside of its role it's bleeding into into the other weapons roles making them dominated and it's contrary to all the effects and the expectations that we're setting and thing has this giant

scope on it and if you can use it at point-blank range that's not what you expect it's gonna be what we call a newbie hammer and I had no effective counter strategies so sniper could acquire you zoom in on you fire him in the head before he could even pin him with a battle rifle and pop them at a zoom even after a body shot which is a sort of a failure case for most snipers you couldn't get back a newcomer I was hitting there in there but they hit you right away and most importantly this goes back to kind of training your sense of balance every time I saw something use a sniper rifle I got nervous the gap said if I was getting an ulcer but really it's just my my neurons were sensing that something was clashing with their representation of the game and they weren't giving me dopamine my internal model was wrong my space sense was tingling I was reaching out with my feelings although you know the classic lines so use your sense of balance to

feel when something is wrong because you're never gonna figure it out by looking at graphs look at a data you're not gonna figure it out that way so we decided to make a change you know we couldn't touch the strength knobs right you can't change the damage because that's what makes the sniper rifles good is that you can kill somebody in a headshot and kill somebody in two shots of the body we couldn't show the range because that's what makes the sniper rifle the sniper that's the rule it's really seductive to add weaknesses at this point I'm going to ask you not to do it don't do it it'll ruin your game you can't add air to a sniper rifle right under any circumstances it just makes it so it's it's not a sniper rifle anymore it's a sniper rifle but we couldn't remove the melee attack couldn't give it some kind of conference sound so everybody always knew where you were when you're stamping we don't want to make you move slower

one of the one of the big problem of adding weaknesses is that people will figure out ways to get around it so we have to touch the flow knobs remember we've removed everything that wasn't a strength or a flow table and so if we can't touch the string top sometimes we have to mess around with the flow downs turns out cadence is sort of the most flexible flown up especially for the sniper rifle it's the most prominent prominent part of the sniper rifles flow is that really beginning to when you're sniping when you know you're I don't have to be on his head until right now because that's when they can shoot and it has a large impact on strength but it's not a strength knob explicitly you can make the sniper rifle fire slower and it doesn't actually change the core of the sniper able it just changes how effective it is so we're using our rational brain at this point now once we've used motion to figure out what's wrong we're using our rational brain to

decide how to change and so we're going to limit the strengths we're going to figure out what change fixes the problem what has the least impact we consider changing the shots on the clip but the problem is that three shots it's not usually enough to kill two people and what's gonna happen is you're gonna kill one guy and you're really gonna feel like you need to reload because you only have one shot left or maybe two can't miss we could have increased the length of the reload but it would have to be increased a lot to fix this problem and that felt like adding a weakness we did play around with changing the time to a full zoom after you hit the button for until your until you're fully zoomed it fixes the problem with acquisition speed like I can stick my head out by the time you've finished zooming in I can run Mac but it doesn't fix close range juice and it also encourages people to fire without zooming and that breaks the fantasy of being a sniper we tried making it so you couldn't get headshots when you're unzoom but that really limits a strength and it only not only fixes the close-range problem don't

fix the long-range acquisition speed problem we thought about limiting the maximum total ammo but like I said before that fixes the average it doesn't pick the instant it doesn't make you feel any better and you get killed that eventually the guy's gonna run out of ammo hopefully he's not gonna run out before you kill him and take it back and use it yourself so time between shots fish it fixes the acquisition speed because if you do miss it takes a while before you shoot again and it fixes the close range because what was happening at close range was people just patting the trigger and shoot shooting it like a shotgun almost like trying to get as many bullets out of it as possible it's not a negative feedback loop it prolongs the fight of so many misses and I run back in to cover the fight is longer didn't it didn't end quicker so this is one we chose it was our choice gotta make it carefully it change from 0.5 to 0.7 seconds don't make changes less than

10% it's it's not a hard rule but it's it's it's a good rule of thumb it's not gonna have a noticeable effect and you can twiddle these numbers forever and one game in a work and one game it'll won't work you don't need to get that specific what you're trying to do is you're trying to reduce the balance to the point where it's overcome by noise in the battlefield try to overshoot and come back you want to make sure that this change actually made a difference right so you make a big change and then come back I'm almost done and don't break the flow anything higher than 0.7 we would add a change something else because I just broke the flow too much so verify the change it's still over power that's fine the other weapons don't feel weak optimizers they're doing other things now it's not being used in close quarters it's not a predictable outcome at long range somebody's going to get back into cover and my emotional state no longer nervous so you use both your brain and you've got to verify change into the sniper rifle is balanced and

I'm going to skip the very last section so thanks for my talk [Applause]