"Game Balance Isn't Real"

Let's go. So, I was sent this from Keats a few days ago, and for those of you that don't know, uh worked with Keats on KI. He, uh, was the lead gameplay designer, right? I was doing video story [ __ ] and he was out here making all the characters and stuff like that with the team. They went through a lot on the KI side balancing and going through uh the checks and balances of making like a fighting game. And this is like the biggest one where you really have to like respect a lot of fan feedback and how that game was was a lot. Boy howdy. Ki fans with balancing was definitely something. Me and Adam sort of vibe a lot on how [ __ ] goes in fighting games. Like what makes fighting games fun, what makes them enjoyable. Some of that stuff is missing from a few modern fighting games in some way for the sake of like chaos and overall player balance, but he has a point in this as I opened up the beginning of the video that I kind of agree with that like game balance is not a real thing. It's a thing that kind of happens like and it's the same thing that I've talked to people about where if somebody comes up to you and they're like, I'm looking to get a fighting game. Which one is the most balanced? What does that tell you about that person? You should know at that point that they don't really understand much about fighting games. So, you just want to get the one that's like the most fun with the most characters, which is fine because that comes from a perspective where they've never played enough fighting games to know that almost none of them are really balanced. Dude, what makes fighting games always interesting to me is the fact that they're imbalanced. It's up to you to counter that balance with other characters and [ __ ] like that. You can't just take character A and fight character F and just expect it to be a five5 every single matchup along the way. It doesn't work that way, dude. Let's see what Keith has to say about it. What is game balance? If you're into competitive games, I'm sure you have a pretty decent answer. But can you answer me this question? How do we know when a game is finally balanced? Let me take you back to the early 2010s. That's a good question. League of Legends is one of the most popular competitive games of all time. And on February 16th, 2011, the developers released a patch containing the following change for Vladimir's Sanguin Pool. removed oncast bonus movement speed. The player sentiment after playing the update was rough. Users rushed online to express how bad this change felt. Vladimir's play rate plummeted and his win rate also decreased. But here's the twist. The developer had forgotten to submit the files for that particular change and so it never actually went out to players. Could you believe how much competitive games in general are based on vibes? something that didn't even happen, like a placebo effect that didn't even happen, had a tangible change on how people perceived the game. Vibes were down. Nothing actually changed. Crazy. Crazy that that all competitive games are like this [ __ ] dude. If vibes are down, it doesn't matter how good things are getting. Vibes are down. If vibes are up, it doesn't matter how weird and dumb [ __ ] is. Vibes are up. Game balance isn't real good intro. Okay, I know that might sound ridiculous and dramatic to some of you, but give me a chance before you rush to the comments to light me up. Let's define some terms. Game tuning is the act of adjusting a game to achieve a goal. Right? Tuning is a very real action that developers take. Game tuning is real. Game balance, however, is not an act. It isn't something you do. It's a goal. Generally, developers and players alike want to achieve the goal of game balance. when players give feed. Very good visual for anybody that knows where this comes from. This episode of The Simpsons, very good use of this visual representation. Back to developers, they often do so with game balance in mind. Developers tune the game to try to achieve balance. Again, tuning is the act, balance is the goal. So, while tuning is easily defined, we need to ask, what does game balance as a goal even mean? If, as we just learned, simply telling players that something in the game has changed when it has not can so deeply affect their perception of the game's balance. True. Game balance isn't real. That's right. Game balance is mostly just vibes. Yeah. It's It's not like like the game becoming balanced is not like a tangible like thing that you build towards. It's a it's a hope that you eventually get there, but not like we did it. Like, you know, a balanced game is just five fives across the board for the most part. But they're vibes that really matter. Whether you're a developer tuning a competitive game or a player interested in playing or discussing these games, let me explain why game balance isn't what you think and share some of my techniques for tuning competitive games. But first, why should you care about my view on this topic? Let me introduce myself. My name is Adam Hart, but you might know me as Keats in the fighting game community. As early as age six, my dad and I would compete with each other and create new games together. I started playing fighting games when Street Fighter 2 hit arcades in 1991 and I got the competitive Tommo and James Goddard who Keith's worked alongside in KI season 2. Shout out to James probably being uh 17 years old here or something like that. Jeez, man. James was the dude that I was up late at night uh in Chicago and he was he was trying to get ultras and I was sitting here trying to get them in the game in the video uh and recorded and cool looking and we were literally like burning the midnight oil to get these things functioning. He worked on um he's he's actually the creator of DJ in Street Fighter Chat around 1997. The lack of nearby tournaments led me to running my own local events, which got me hooked enough to start traveling the country to attend regional and major tournaments, which led me to evolving my own events into the regional major series ultimate fighting game tournament or UFGT, which popular fighting game commentator Ultra David has called one of the most influential tournament series in FGC history. Thanks. during my travel. Uh Keith, what was the first U? What was the UFGT I went to before it all became Combo Breaker was when we KI was there for like the first time and I think that was like 20 14. It was 2014 because it was happening like right around the same time frame as TJ in the works or some [ __ ] I'm trying to remember back when all this was going down. TJ was hot in the oven. It was UFGT9. Okay. Okay. Okay. Okay. That was a wonderful tournament. It was It was small, but a wonder It was the most grassroots [ __ ] ever, dude. I scored an EVO top eight medal in 2011 by getting seventh place in Tatsanoko versus Capcom Ultimate Allstars. I was the editor-in chief of top fighting game website sherukin.com for many years, writing about news, events, and I have to give uh shout outs to Keats and the SRK crew at the time. greatly beneficial in my expansion as a content creator because there wasn't many fighting game content creators um around that time frame and the majority of videos that I was putting out for the games at the time were making front page of SRK and like Event Hubs and things like that and that was a big part of me like becoming you know even a notable person at all co-hosting their weekly podcast wake up SRK. I got into game development when I created Divekick with my wife and friends and then I joined Iron Galaxy Studios to release a commercial version. We also fought each other at E3 2010 in Ultimate in in Marvel vs. Capcom 3 vanilla and we both won shirts, right, Keats? I mean, you were you were definitely there. You won a shirt because you had to like beat 10 people in line or some [ __ ] and me and Keats were one of the few people that did that. And I still have the shirt. It's dilapidated at this point. Yes, you still have that shirt signed and framed. It's sick. I also helped Iron Galaxy Studios with Street Fighter 3: Third Strike Online, Marvel vs. Capcom Origins, and Darkstalkers Resurrection. I then served as lead combat designer on Killer Instinct, the 2013 version season 2, and for the rest of the game's development cycle, and then as lead designer on Rumbleverse. On Divekick, Killer Instinct, and Rumbleverse, in addition to my many other duties, I was in charge of game tuning, and I interacted heavily with community. In other words, trying to achieve competitive game balance has been a major focus of my career as a game developer since 2012. Get a lot of it. Seeing a lot of Rumble verse fans in my chat is going to make Kates feel very good. Chat, during this professional experience, I came to view Game Balance very differently than it did at the start and very differently than most of our industry. The proof of this view is in the pudding. Endgame Killer Instinct is still considered by many to be one of the greatest fighting games of all time. With major tournaments still ongoing, a lot of players feel its balance is among the best in the entire genre. But before I jump into the meat of this talk, I want to share a quick word about the examples I'm going to use in this video. I lived through nearly all of the events I'm going to reference, but the ones I was less close to required me to do some research. I did my absolute best to represent the facts and feelings of the time. And I want to make it clear that if I use an example in a negative way that I'm not disparaging the people who made those games, made those design choices, or players who love those games. I want to acknowledge that all of us developers are trying to do the best we can at a very difficult task. We've all made mistakes, me very much included, and I'll probably make more mistakes in the future. The point of this video and its examples isn't to point fingers or say I'm the best or something like that. It's to use our history to observe lessons to help you understand my perspective on game balance. Cool. All right, let's begin. Before we get started, chat, hold on a second here. Do me a favor and uh go ahead and follow the channel. Here's Keith's video. Go ahead and follow a channel. Leave a like and stuff like that. All right, hold on a second. Let's get into it. Happened to you. You're at the GameStop reading the back of the box for the latest competitive game, and the bullet points say 16 playable characters, 12 fighting stages. The game play is very fair. I bet you couldn't wait to get home and play the game so you could experience the fairness firsthand. Let's be real, fairness is not a selling point, and people don't generally buy games because they're fair. Why do people buy games or download them in the case of games that don't require a purchase? Here's the crazy part. I I I think I disagree with this a little bit. I think fairness is a selling point to to extreme casuals. Not like people that understand things and play things a little bit like the most extreme casuals of just like I just want a I just want a fighting game to put on the shelf, man. Like I just want one to be there. I just don't I have a sports game. I got a shooting game. I got another shooting game. a new fighting game, which is a large percentage of people that buy this [ __ ] In all honesty, like though, that's the audience that fairness and balance is important to because they they feel that way, you know? Well, the game looks fun. You can do a bunch of cool stuff in it. Explosions, style, mastery. Every person's definition of fun is unique, and the type of fun they seek is their primary motivator when deciding what to play. People also buy games because of FOMO. their friends or their community are playing and they don't want to be left behind. Gamers will often play a game we don't love just to keep up with our friends. Even outside social pressure, such as everyone's playing it, can trigger the desire to not be left behind. Back before postlaunch live service update cycles were the norm, the point of a sale was all that mattered. As a developer, if I could get you to buy your Super Nintendo copy of Street, the relationship ended there with you ended right then and there, wouldn't have mattered to me how long you played the game or what you thought of the game balance or the fairness or any of that. Even if the game was deeply unfair, you'd likely only be playing it against your local friends who were good enough at the game for that to even really matter. So, as a dev, I just hope that I did a good enough job to make you want to buy the next version of the game in a year or two. Now, game development and marketing costs have skyrocketed. Development timelines are only getting longer, and the way consumers interact with competitive games has completely changed. After your initial purchase or download, it's important for developers to keep players playing. If they aren't playing, they won't be around to buy DLC and fill the online matchmaking cues. So now it's become just as important to keep players from quitting as it is to get them to buy or download the game in the first place. That's a very good point. Yeah. And it's funny is that KI was kind of at the epicenter of that. It was the the the game to start that [ __ ] where instead of waiting like a year or two for a new title update, Ki was just like, "Here's characters a year. Get ready. Here's a new season. Here's a new season. Here's a new season." You know? especially if the game is free to play. If you want to stop players from quitting, you need to fully understand why they quit. Players quit because sometimes it's not about the game itself, and there isn't anything you can do. Maybe a new game came out and pulled their attention away. Perhaps real life happened and that drastically lowered the players free time. But there are reasons that players quit that you can at least somewhat control. If a game becomes more frustrating than fun, players will quit. If a game becomes more thoughtless than engaging, players will quit. So to summarize, games no longer are sold in a box and left as is. Competitive games need to last for years just to justify the rising cost of development. Developers are expected to keep players engaged in playing. That means frustration is a real problem that must be dealt with. Boredom is a real problem that must be dealt with. And so developers decide on goals and then they tinker with their games. And this gets called rather interchangeably game tuning or game patching or game balancing. But each of these terms has their own definitions, and we should not use them interchangeably. Let's reiterate the terms as I've defined them. Game tuning is the act of making adjustments in your game in an effort to achieve a goal. Game patching is the act of deploying updates to the game. And game balance is merely one possible goal of game tuning, right? But what is a balanced game really? How do we know when a game is finally balanced? Five fives across the board. Most boring [ __ ] in the world. Is it when every weapon or character is used equally or wins equally? Is it when the players have nothing left to complain about online? Are these things even possible? When we tune a game, we're looking to reach some specific measurable outcome. So, a problem we're about to run into is that game balance doesn't have just one definition that everyone agrees on. Each part of the audience for your game is looking for something different based on their own ideas of what successful game balance looks like to them. Most players primary driver for wanting game balance is the idea of fairness. They think they want a game in which all options are equally viable and they can rank up or win a tournament with any character or weapon or team. Yeah, I understand why this is the way to th this is the reason these these it's not like these feelings don't exist. They exist. Like a lot of people feel this way about this [ __ ] but there's a reason behind it and it simply becomes it comes from inexperience, right? just not really knowing like how the genre works or how competitive game works in general. You you would expect there to be a fairness to everything like across the board that they choose. They want a game where the tools they use are objectively powerful but are perceived as fair by others and that those tools are rarely used against them. If the game doesn't feel fair to an individual, that individual gets frustrated. Viewers matter a lot to any modern competitive games ecosystem. And the viewer's primary driver for wanting game balance is variety. They want to tune into live streams and see like a lot of different characters and team compositions and weapons represented to keep things interesting. And this is also important nowadays um because a lot of fighting games lately aren't super entertaining to watch. They're really engaging at like high level. Don't get me wrong. the decision-m and all that kind of stuff and even execution in some ways is really profound and mad cool. But there's a weird thing about older games and it's that the the way you can play them is so varied that just through the act of doing like one thing like moving around you can greatly tell the difference of a good player to a bad player. Um, and there's a game that's happening with that right now, which is where I'm having a lot of fun, and that's with Invincible Versus. Dude, the high level of Invincible Versus looks like such a goddamn different game than any other level of the game being played right now. Same thing with like Fatal Fury. The high level of that game is crazy looking. Want to see lots of play styles interacting with each other, especially when viewing tournaments. It's a lot less fun when the top eight character selections look like this. If the choices being made on screen are too repetitive, viewers get frustrated. And this third one, yeah, we'll talk about this one later. So, how do developers figure out how to tune the game? People hate saying the same [ __ ] over and over again. More variety. One common strategy is to survey or talk to players to see what it is they want changed. Another common strategy is to use telemetry to look deeply at data that represents real in-game outcomes, such as use rates or win rates, to try and make it so that things make sense mathematically. Tekken uses this a lot. Yet, despite trying all of these things, competitive games so often struggle to achieve good game balance. Why is that? Let's break down each of the traditional methods for making tuning decisions and see where they tend to go wrong. Let's start with telemetry. Data lies. On the first day of my high school statistics class, our teacher shared with us a quote from early 20th century Austrian psychiatrist Wilhelm Steele. Statistics is the art of lying by means of figures. That quote really stuck with me. As it turns out, if you're clever, you can use data to prove almost any point you want. Or worse, data can convince you of a false truth. Data can be very useful in convincing your development teammates of a design direction, but it's much less useful for finding the right tuning decision to make for your audience. Checking telemetry from League of Legends, you might find that the item Maji's Soul Stealer has an absurdly high win rate, often close to 80%. But this statistic is highly misleading. The item is rarely purchased unless you're already winning, which means the player purchasing it was already very likely to win before making the purchase. Yeah, if it's just not around as much. It's like it it it dude this is the the same [ __ ] where people were telling me um Lars has a has a has one of the highest win rates in Tekken. You stupid bastard. It would usually be accompanied by you dumb [ __ ] You stupid bastard playing one of the characters with the highest win rate in Tekken. And and I'm like mother like the only reason that is is because he is like the bottom third of used characters. like he's not popular. So the amount of people that are playing him are significantly less than the people that are playing the characters more like the more popular characters. Like 75% win rate, 16% pick rate. It's like, dude, come on. The developers understood why the win rate for this item was the way it was, and that kept them from making the mistake of trusting the numbers alone and making a poor tuning decision. This means that in order to find guidance within the numbers, you need an extraordinary amount of context. Let's say that you look at your character pick rates and one of the characters is very under represented. Do we have any idea why just from the numbers alone? Is it because the characters too weak or are they difficult to use? Or do they lose matchups to some of the most common characters? Or are they just kind of ugly? Without the reason why, this data is effectively useless as a means to understanding what changes to make in your game. Early on in Overwatch's lifespan, developers saw the use rate of the support hero Symmetra was low. And so on December 13th, 2016, they overhauled the character, changing the way her move set worked. Developer design notes said they wanted to make Symmetra feel more active and give her more meaningful contributions to the team. Overall, Symmetra players like the changes, but the use rate barely went up. The players were somewhat happy, but the goal of normalizing the use rate was not achieved. She still wasn't very popular. On June 26th, 2018, a year and a half later, Symmetra was reworked a second time. This time, moving her from a support role into a damage role. Many Symmetra players lost their fun in this overhaul, and some physically disabled players could no longer play the character at all. The use rate dropped below the original version while simultaneously making the player base upset. without understanding why the use rate was low to begin with. Tuning the character, or worse, reworking them entirely is like throwing darts while blindfolded after being punched in the face. Was the pick rate low because Symmetra was weak? Or was it because her toolkit didn't fit in well with the dominant strategies running the game? Or was it because she was only fun to a certain type of player? And if I may ask, is it even a problem that her pick rate was low or that the game has niche characters that aren't tournament viable? Well, is it? Without context, data is often worse than useless. It can make you think a problem exists where there isn't one, and it might lead you to making decisions that remove the fun of some of your players. Data without context is dangerous. We should also consider that data is almost never qualitative, and it's easily skewed. Developers that use telemetry to observe the win rates of various characters and weapons will find obvious and large differences when they start to sort that data by player rank. But can we really trust a qualifier like player rank to sort and understand this data? Ranking systems are only good at sorting people who use them a lot. Consider that there are tournament level players who do not play ranked mode often at all. And when they do, they may be much lower ranked than their actual skill level due to their lack of ranked play time. You must also realize that not everyone in the top ranking bracket is of equal skill. And so this can affect a character's win rate rather drastically even within that segment of players. And within that segment, you may find that certain strategies or characters are barely used at all. Think about what else goes into the win rate. Is this a 1v one game or a team game? Is the character difficult? It's funny cuz at like these levels things like scrub killer characters almost don't exist, you know, characters that will actually like ream platinum and below because their the counters to their strategies is not immediately obvious. that [ __ ] just goes away. Like because once you know what to do and you're refined and toned enough in the game to have an understanding of how to counter it, it's like, oh, well, now that's just not even a viable strategy at all. You know, certain strategies or characters are barely used at all. Think about what else goes into the win rate. Is this a 1v one game or a team game? Is the character difficult or easy to use? How do we interpret the stats of a character who thrives in online matchmaking but struggles to place well in full tournament brackets? How close are the game's best character specialists to reaching each character's known potential? And what don't we know about a character's potential? Yoshi perfectly understand the quality of our win rate data. We simply can't learn enough from it to identify if we even have a problem to address, let alone possible solutions to that problem. Maybe we can spot a lopsided matchup with data from time to time, which could be really helpful in knowing where to look, but it isn't at all helpful in figuring out what to change. Remember, data will lie to you. Using Street Fighter 6's publicly available telemetry, we can see such a case in action. Doulam, Lily, Honda, Rasheed, Blanca, and Kimberly are the bottom six characters in terms of use rates and ranked, both across all ranks overall and when isolating the master rank players. And yet the characters with the highest win rates in master rank are in order, Kimberly, Dawson, JP, Blanca, Rasheed. And I'm assuming this gets to the point where the reason that is isn't because the characters are inherently like better. It's that they're being chosen less and the people that are good with them are good with them. They're character specialists, the majority of them. So they cook like they will cook the majority of players because you just don't know how to fight against this [ __ ] So this is almost the exact inverse and player conceptions of these characters place in the tier lists bear no relation to any of this data. Most players think Blanca and Rasheed are high tier. Honda and Lily are low tier. And ideas about Doulam and Kimberly's strength vary wildly. This data could easily lead you to believe that these six characters are weak due to underuse or make you think that they're too strong because of their high win rates. Data without context is dangerous. Yes. Yeah. It doesn't even mean anything. But even trying to apply proper context to your data is almost impossibly difficult. So if we can't really have much value, we should probably talk to the players in our community, right? After all, they play the most. They're often extremely skilled and they know the game pretty pretty damn well. Well, unfortunately, even though this is so, bro, you Chads, look what this says. Players also lie. If there is one thing that you do not want is to tune a game based on like the high level of the game. You do not want to do that because the intention behind it like downplaying is at legendary levels lately in like all genres across the board. But we saw this happen like intimately with KI at many points, right? Like this this was a big deal uh at several points of like to try to make it seem like that your character isn't doing good so that the other character can get nerfed. and don't mean to. Players frequently lie, too. If you've been in or around gamedev long enough, you've certainly heard the phrase, "Players don't know what they want." This isn't entirely true in my experience. There's a relatively well-known clip of me from the wonderful fight on Killer Instinct documentary by Hold Back to Block in which I tell an anecdote about a time when the KI community rallied together. I think what he I think the better way of describing it is that of players not knowing what they want, people don't know how to describe what they want, right? And that's the big challenge is that people have a tough time communicating, especially over the internet. People let their emotions get the best of them. Like the aggressive thing comes out or the overtly positive thing comes out instead of finding an eloquent way to describe what you're feeling and what should change as a result of it without affecting other stuff. That's hard. Like that's just that isn't just a balancing thing. That is that's hard in general for people to communicate even what they want from [ __ ] and really wanted us to nerf JGO. Specifically, his ability to heal himself by landing fireballs during his instinct mode made them feel like they had to kill him three times instead of two. But nerfing his healing, as they had requested, would have cut away at the character's core identity, and a large part of what makes him unique and fun to play. The community was laser focused on seeing the healing as the problem, and their view of the game was that JGO was either too strong or too frustrating to play against, depending on who you asked. Whether this was the true state of the game or not is not relevant. They believed JGO was a problem and their perception, no matter how flawed, is the reality they live in. To tune games properly, you must accept this. A player's perception is their reality. So instead of succumbing to the community pressure campaign and nerfing Jgo's healing, we took a different approach that allowed us to maintain his fun and unique healing gimmick while weakening some of the more brainless parts of his kit that allowed him to negate his intended weaknesses too easily. The community cooked us pretty hard for a few weeks, insisting that this wasn't the change they asked for and wouldn't make a difference. But a few weeks after that, complaining about JGO had all but stopped. Players knew they wanted a weaker JGO. It's just that their idea of how to do it would have had way worse knockdown effects on the game and JGO player base than what we landed on. And of course, a very small number of JGO players were unable to adapt to the changes that we did end up going with, and they either switched characters or quit the game entirely. That sucks, but it's a natural part of the tuning process. You can't please everyone, but you can limit the negative blowback of a nerf if you're thoughtful. Was JGO too strong on paper? Did he even need to be nerfed in the first place? Believe it or not, it doesn't actually matter. Ultimately, players were frustrated and something had to be done about that. The something we chose weakened JGO, which is what the players wanted, but the way we did it strayed pretty far from most of the common complaints. Yeah, this is the uh this also is related to the same [ __ ] that happened in Guilty Gear where no matter what you do to make adjustments, the goal is to create like happiness is the goal is to make like people enjoy something to to to expand more. It's good for the overall longevity of the game, but you might have to make a sacrifice, right? Like some hardcore like JGO players in this case are probably going to drop off, which is honestly fair. This was the same [ __ ] we were describing before Guilty Gear Strive came out because it was so different of a game. And I was like, chat, this is the whole thing. They're they're redesigning the way this game works so that they can build a new audience and like for lack of a better word kind of sacrifice some of their core fandom that has been that fandom for a long time but not a sizable amount to get a more broadened amount of happiness to get more customers to get more people to get more like folks that'll engage with the game a longer period of time like and yeah I felt like I'm I I was hopeful for this but I was definitely one of the ones that just found way more enjoyment out of the previous games like and that's do I feel betrayed or something like that? Like well no the old games aren't going away and they to be real they re-release the old games and [ __ ] like that with better net code so that they can always be there which is I think the best way that you can compromise and allow for that satisfaction. So anyway, yeah, you have to make a tuning and adjustments so that the health, the longevity of something can be good, but create like a whole bunch of people that might build upon this while not having the core audience that you maybe had before for like a character or a game. So, in a way, players do occasionally know what they want. It's just that they don't usually know the least destructive way to get there. Most of your user base will lie to you, but I'll state again, they usually don't mean to. They're generally well-intentioned and they just want the game to get better. But better for who? Even among the most well-meaning users, you'll find that the overwhelming majority of them struggle to see the big picture or consider the viewpoints of others. Their idea of a better game is very commonly one in which they with their character or weapon of choice can win more easily. things that beat their play style get weakened and their play style gets stronger and no one complains about their play style so they can feel good about using it. I admit it. Yeah, I I you if you don't think you think this way, you're lying, right? Like you you're just lying. And I I have identified those things about myself where instead of like trying to get to the goal and create a narrative around the things No, I'll just say that's like I really love the way this character plays right now. like I I really dig this [ __ ] instead of like having it get derogatory and all this kind of stuff. No, it's like I just love the way this character plays. Um if if that got worse then it might make me play the game less, you know? Like for example, if they release a version of Third Strike and Ken in Third Strike was severely nerfed and lost all his cool hit confirms and all that kind of [ __ ] I probably just wouldn't play him. Probably be fine. um like and that that's fair as long as I can tell you that like yeah, I'm attached to this character and the the fact that they're all so good at the same time. Like it's important to realize that like you have those opinions. You know, you really like the way this character plays for a variety of reasons outside of just they look cool. Like the gameplay is also vibing with you and you're winning and all that [ __ ] I used to think this way myself, so I'm not trying to cut you down if you currently think this way. It goes without saying that this isn't a good way to make a more fair or less frustrating game. So, you need to take every piece of feedback that you get with a massive grain of salt and try to apply as much context to it as possible. Developers can be easily misled by data. Sure, they can even more easily be misled by players. Data can be convincing, but players can be very, very loud and hard to ignore. It is impossible to know what you don't currently know. That won't stop players from insisting that their perspective is the truth. After all, to them, it currently is. You know what the You know what the craziest Keates, I'm just going to say from my perspective of being like a mini dev on like MVCIB, you know, was the craziest [ __ ] thing that happened on that whole project was at the end of it and you release it and the biggest change to the game is mostly visuals and also new menus and shiny [ __ ] right? And guess what the perception was of the game when that happened? Actually, this game was always awesome. And actually, the gameplay was always good. And what was the biggest thing we changed? Just the coat of paint, dude. The vibes were up. That was it. It was It was just better vibes. And yeah, we did gameplay adjustments across the board, but nothing that like anybody that played this what back then was going to notice. It was only for the hardcore audience. So, we said like we made gameplay changes, but no one really knew what they were. And then everything looks pretty and sounds cool now. God damn, this game was always cool. I've been saying it's always been cool. It's like, what the [ __ ] dude? That that's what I say where it's like just changing the UI, just changing the interface of a game will bring up the vibes so much. Just changing the presentation a little bit in some small way will do a a substantial amount to bring up good feelings about how this game is. But their perspective can shift quickly. Players are very easily influenced. Extremely easily influenced. Influencer types can often without even meaning to construct completely false narratives that large swaths of your user base will bandwagon on and help amplify. I still don't get the rage quitting [ __ ] in Invincible. Dude, like I I I do not understand it. Like I have had the equivalent, if not same amount of rage quits that I've had in Invincible that I've had in Mortal Kombat games for the past like 10 plus [ __ ] years or uh or Tekken. Like I don't get it. Like it's and you get points for it, you know? Like you get the win. It's just rage quit. You win. Like I don't understand how the there became this entire rage quitter thing around like this happens in everything dude like most things I don't get it. Suddenly the desires or misinformation of just one popular player can start to sound like hundreds or thousands of people who all agree about what the problem is and how to fix it. Group think is a major issue that puts This is why I don't like I do tier lists on dumb [ __ ] I I don't I I do not like the idea of doing tier lists from from a content perspective of like making tier lists on how characters play. I just do not it just I do not [ __ ] with it. I feel like I have too much influence on [ __ ] to to to to manipulate people in that space. You're not uber competitive. I mean, I'm talking like 10 years ago when I was actually playing games like a fuckload, right? I would not do tier list. I would casually talk about it, but I would not create things to try to build a narrative like, "So, let me try to tell you why Full Gore is mid-tier, right? Let me let me try to build this up. Let me try to tell you why Saber Wolf in season 1 actually wasn't good. Sadro was always the best character." Like, no, no, no. I I never wanted to do that [ __ ] A lot of social pressure on devs that can be very difficult to navigate. And it isn't just influential players or malicious players who can influence a crowd. It's often tournament results themselves. Early on after the launch of Injustice 2 in May of 2017, generational talent Sonic Fox won nearly every tournament they attended using Captain Cold, a character that many at the time thought was on the weaker end of the roster. Sonic Fox's dominant win streak resulted in a community that was forced to reassess the actual strength of the character. Sonic did this with [ __ ] MKX as well. The same [ __ ] Nobody was playing Aaron Black, dude. Nobody. And then he came in and started beating the [ __ ] out of everybody with Aaron Black. It was the same thing happened again. And eventually developers that nerfed the character despite Sonic Fox's insistence that Captain Cold was indeed weak and just counterpicked certain characters. Well, but you don't even have to go on a dominant winning streak to change people's views. After Super Street Fighter 4 launched in April of 2010, right before tournament season really kicked off, many of us looked to our strongest players to assess the power levels of the 10 new characters that were added to the roster. Going into EVO 2010 on July 9th, the widespread sentiment among top players was that Adon was among the weaker characters added to the game. Many of us went into that event writing the character off completely. Gamer show Gamer B took fifth place out of about 1,700 players with Adon that weekend in the largest open bracket tournament we had seen to date. Totally upending how we had viewed the character for months without a single balance change. I remember uh Ari Flo Flo did this with Jury. That was like the moment that Flo became like a household FGC name. You remember that [ __ ] I can't remember what tournament it was. It might have been a UFTt, but he ended up doing mad good with jury and it was like, whoa, this character's Whoa. The jury jury forums were on fire, you know. Rest in peace. For all you people out there don't believe that can take it to the top eight. In an even more dramatic example, let's look at Super Smash Brothers Melee. One of the most popular competitive games of all time and a game that has not been patched since the 1.02 players choice discs were released in 2003. Melee has a very long and welldocumented history of community agreed upon tier lists to rank the strongest and weakest characters. By all conventional wisdom, Yoshi sat as the 21st best character in the game in a roster of 26. Not good. And then Amsa appeared in the tournament scene and started getting results with Yoshi. Yoshi's rank went from 21st to 12th. It's just perception to 2015. And after's win at Big House 10 in October of 2022, arguably the hardest melee major of all time, Yoshi now sits as the 10th best character in the game on community tier lists. despite nothing changing. That's it. Yoshi rose 11 spots in the ranks. That's the that's the whole [ __ ] thing about tier lists is that like it is literally just building like a narrative. That's the whole point of them is like to build a narrative. This it once again ties into the entire aspect that like there's this data that all these players think this thing, but does it actually like is it is it exactly tangible? It's all just vibes, dude. is the best of the 987 competitors. Again, we simply cannot know what we do not know. Whether it's Sonic Fox thinking their tournament winning character is weak, or a consensus of top players thinking that a character they don't even man is weak just before a breakout performance, or AMSA believing in their characters power despite ridiculous execution requirements and being the only one in the world who can do it. It just goes to show you how easily players and developer opinions can be swayed by tournament results and influencers. When group think and player pressure campaigns arrive at your doorstep, it's incredibly hard, maybe impossible as a developer to know what situation you're actually in and guess right. This is the one of the reasons why weak or strong. Without many years of evidence, you'll have to make an assumption. You need years of evidence. You'll be right sometimes and you'll be wrong other times. And it's a huge part of why tuning a competitive game is so difficult. And of course, this is why one of the only things that I usually gauge a fighting game on nowadays in terms of just like how it is, what the competitive vi viability of it will be in the end. I'm I'm like that that doesn't matter. It doesn't matter how competitively viable this is going to be. Cuz you know what does matter? How much fun it is. Is this just enjoyable to play? Are there things getting in the way of you getting to the enjoyment? Does it have enough characters and things that you'll like about it? Does it is is there a controller issue, right? Is there enough that you can go does the net code have problems? Is there enough? Like to me, the more you focus on like is this a thing that you can see a bunch of people getting into and playing? That's what I usually try to represent a a fighting game on nowadays. Like you first play it, you first touch it. Is it really fun? Does it have a good feedback loop? Is there enough in it? Is it specifically for hardcore? like like those things are good, but does it mean it'll become like a competitively viable game in some way? No, it doesn't. And most of the time the games that are like the most competitively viable are the games that are just fun as [ __ ] and build an audience. And people love that [ __ ] so much that eventually sometimes they get pissed off that someone else is saying they're good and they're better than you, which creates a competitive environment. It's just the natural way of it, dude. So, as long as it's just fun, like there's just cool [ __ ] happening and it's it's a great feedback loop, as long as you treat it like it's a normal video game and not something that you have to accept, right? And nowadays, it's like I don't I don't feel like I have to accept a lot of fighting games out there. Will they become Alzheimer's for me? No. But I can say that like a lot of this is doing good. Like a lot of this has a has a feedback loop that I really understand why people are going to vibe with this. Like I think Guilty Gear Strive was the best example of that where it's like I'm enjoying parts of this. My perspective is [ __ ] because I've played the previous games a bit and I know how much fun they are. But now it's like, yeah, I I I get less fun out of this, but I can see where a lot of people are going to this is going to hit for people. Dude, when talking about our inability to know what we don't yet know and how that influences the way we think about our passions in the world, I need to remind you, dear viewer, of the Dunning Krueger effect. We can be extremely confident that we know everything about the state of a game when we in fact are only seeing a small part of the big picture. Even the strongest players in the world who are regularly winning tournaments may only be aware of a tiny fraction of what's possible within a play space. And while balanced discussions and tierless theorizing can be incredibly fun, it is for these reasons that players make misleading and unreliable sources of guidance when deciding how to tune your game. Okay, so we can't rely on data or players for guidance. Uh, this is this is also the thing where it's like I kind of want more fighting games designed by people that like big ones that have folks at like the the highest level making decisions about how things how cool things can be not being fighting game players because you usually get a perspective of people that don't know how this [ __ ] works and try to actually like innovate at times and try a weird [ __ ] idea that a fighting game person that would never know how that works. And then suddenly it's like, "Hey, fighting. We need to get somebody that knows how this [ __ ] works so they can put in my wild ass idea that seems like that might be really cool, but you would never do that in a fighting game, right? You would never do that [ __ ] Why would you ever do that shit?" And it's like, "Oh, well, just to get the people that would know how to make it work in there because they have a lot of experience in this stuff. If you get it all designed by only people that are just on fighting games, guess what? It's going to play like a fighting game. It's going to play like many other of the fighting games that exist. But what about the goals we'd set out to achieve in the first place? Let's examine our list of goals and try to understand if we're It's very Gate says, "It's very easy when designing fighting games to optimize the fun out of them if you're a strong player. Maybe a topic for a future video." Dude, I watched it happen in Street Fighter 5. Like, I don't know what to tell you. Like, I've I've watched it happen over and over and over again. I'm not even going to I'll use like old games as examples, but literally watched it happen to several games where it's just like why are we just removing what makes these games fun for the sake of like balance, for the sake of competitive integrity? It was like Jesus, dude. This is the opposite of the way things should be. My god. Even attempting to steer in the right direction. Goal number one, a favorite goal of players is fairness. Fairness should not and cannot ever be a goal. Why? Because no two players can ever universally agree on what is fair. It's an emotional metric that changes by the minute depending on who you ask, the results of their latest ranked match, what salty player showed up on their social media algorithm that day. It is simply not an achievable goal. Kate says, "Old games have a looseness to them and a sloppiness that makes them more interesting. Games today uh because you have to like problem solve, right? Like a lot of things aren't handed to you as a player." Yeah, I agree. Uh games today are a bit too tight, have too much like rigidity. Yeah, I mean, yeah, it's true. They're they're they're a little like overdesigned. Yeah, I think it's true because game balance is mostly vibes. You cannot measure your success by measuring fairness as the only way to do so. Data or player surveys are frequently misleading. You know what that You know what that's crazy to think of about that the majority of people that are like big into fighting games now only understand the rigidity, right? Like especially over the past like precoid like right co pre-COVID like most fighting games are sort of that way like the majority of them. So, a very large portion of what the fighting game audience now is, when a game doesn't do that, it's described it, it usually is given very uh very negative descriptors because the game does not line up with the conformities that have sort of been built up to, you know, it's described as jank or broken or kusogi or whack or that kind of [ __ ] right? and like immediately not given a chance because it does not like line up. So yeah, it's very true. Remember, as a developer, you also don't know what you don't know. Things you think you know may be false, and you are just as easy to manipulate as the players are. Isn't that a good thing? So, it it's not a good or a bad thing. It's just if a game plays the way it's supposed to be played and there's very little ways to make adjustment around it, then you know what you're getting into. Like that's it. You you know how the game is going to be played eventually. Usually a lot of those games are going to boil down to something. It's going to be a guess, right? It's going to be a very frequent guess that will even out the playing field for for many people. And how the whole point is to get you into those guessing situations over and over and over again, right? And that's just kind of what happens. You know, this is this is the nature of grounded strikethrow fighting games that aren't like Virtual Fighter. Remember that League of Legends example from the start of the video about Vladimir Sanguin Pool? That wasn't the only time something like that happened in League. When the champion Riven was released, she was viewed by the community as weak. Riot devs apparently agreed and hotfixed in some buffs for the character. Shortly after posting the hot fix notes on the forums, the character's play rate spiked and the feedback was very positive. Players reported how good the buffs felt and what a difference they made. But the hot fix hadn't actually gone live yet. The character was unchanged. Game balance is mostly vibes. The idea of fairness is so individual to each player that it can never be an achievable or even measurable goal. Even when group think makes it seem like thousands of players have the same view on an issue, their overall feelings of fairness will differ greatly. Only way to tune your game into a state in which no player is complaining is to make them all quit. Because of these reasons, I strongly recommend that we stop using the idea of fairness as a tuning goal. Goal number two, a favorite goal of viewers is variety. But variety is not the result of mathematical balance or fairness. It's the result of enabling diverse player expression. Let me explain why. Each individual player has got to have a way to play something differently on what aspects of your game. Can't play the way it's only meant to be played. This includes the type of risks different players want to take when trying to win. Some players want to play it as safe as possible and focus on reactions with a low-risk, low-reward strategy. Other players may want to gamble, taking big bets and trying to get maximum rewards. And everything in between those two extremes exists as well. If a character is harder to learn and pilot and optimize, should that character be equal in theoretical power output with a character that's trivial to learn and pilot and optimize? An easy character could be considered low risk, a difficult character could be considered high risk. If both risks result in the same reward, which character will most players choose? Or to put it more plainly, one character's low-risk, high reward, and the other one is high risk, high reward? What I've observed over the years is that when a game is tuned toward a more flat mathematical balance, play styles start to converge. When this happens, players gravitate towards one of two types of characters. They'll either play the characters that are easiest to win with, thus lowering their risk, or they'll play the characters that look the coolest or sexiest because the vibes matter a lot to us. And watch out if your easy characters are also your coolest ones. We won't often achieve tournament variety by flattening win rate and use rates because doing so flattens out the number of reasons different players gravitate toward different characters or play patterns. It's the quantity of different motivators that you offer to your players that breeds variety. Despite not being widely considered to have the most balanced roster, the original version of Guilty Gear Xer sign would frequently see large variety of characters in tournament top eights. Why? Because the characters so different and diverse that they appeal strongly to different player types. That was that was always the uh that was always the great benefit of Guilty Gear, like especially older Guilty Gear games. Like the the system mechanics really never got in the way of how you could express yourself with a character and you would just get some specialist with some insane ass character in old Guilty Gear and it would just be like, "Holy [ __ ] this is so crazy." Because the characters are so different and diverse that they appeal strongly to different player types. Conversely, in the Tekken series, many of the characters share variations of the same game plan and have similar riskreward values. But in Guilty Gear, almost no characters play even remotely alike. This is AI character diversity was modeled after and inspired by Guilty Gear. Yeah, it was. Especially with Zak, right? Um, yeah, a big part of him being on board. Why? When Tekken misses the mark, it's much more common to see something like four bobs in a Tekken 6 top eight like at EVO 2011 or six Leroys in a Tekken 7 top eight like at EVO Japan in 2020. If the Guilty Gear developers worked hard to flatten the game using data, they'd end up destroying many players favorite play styles and the result would almost certainly be a game with less variety for viewers. And we did in fact end up seeing this in motion at the launch of Guilty Gear Strive. Top eights dominated by many appearances of soul bag. a character boasting the trifecta of easy to use, powerful, and super cool. Because at launch, Strive was more simple, and many of the characters felt more sy than in past iterations of the franchise, players gravitated towards this very easy and powerful badass. Down the line, as new mechanics and extremely unique characters were added to the roster, tournament variety improved a great deal. As a goal, variety is an admirable and understandable one. It's actually measurable with telemetry or by looking at tournament results. And if you aren't hitting a good level of variety, your player base will absolutely let you know. But unfortunately, I believe that you'll have a much harder time achieving variety by making it your end goal. Variety is a symptom of a healthy game ecosystem, not the result of a mathematically balanced game. Variety has another enemy, optimization of play patterns, which I'm going to talk more about later. And finally, the third goal that we skipped over at the beginning, the only real and true goal worth pursuing with game tuning efforts, player retention. Retention is all about making your existing players or at a bare minimum preventing them from quitting. Fairness is a cute idea. Variety is nice to have if you can get it, but if players aren't playing, there isn't any point to anything you might change in your game. We tune the game and engage with players so that they'll continue playing. That is the goal. We already discussed why players quit and among the reasons within our control are that they get too frustrated or the gameplay stops being thoughtful and engaging so they get bored. Using this, we can break down our main retention goal into smaller parts. We need to a reduce player frustration, b increase and reinforce thoughtful play patterns that result in engagement, and c do all of that while trying to enable and maintain a high level of diverse player expression. Easy, right? Naturally, doing all of those things while keeping a community happy with the direction you're going is what you do. You know what is the most engagement you can possibly have from this? To have a sweeping swave balance patch that makes adjustments to every character and including system mechanics. You know what you do? Just say a lot has changed. Have fun. And you don't document any of it. You let them document and figure it out. Don't say [ __ ] It'll take them a year to find the actual changes to weird characters and stuff. It would take those dudes a year. Just vibe patch the [ __ ] out of it. And like in actual put in changes, but don't say anything. Just say nothing. See how it goes or false document it. Yeah. Completely different things, you know. so easy after all. But in my experience, there are strategies you can use to make it work. So, here's my approach to game tuning with those goals in mind. To the developers watching this, the tuning strategies and methodologies I use work for me. They might work for you or they might not. My goal is to share with you what I've learned so that you can use the parts that make sense for you and your project. Just remember to always be critical about what you use and why. To players watching this, I hope you take away an improved vocabulary about gametuning and a wider view of what the goals of gametuning really are. And I hope that you can use this to improve the way you communicate with developers and each other. The classical approach of using data as a metric to measure the success of your tuning efforts is what I'll call tuning for flatness, wherein you try to achieve fairness and variety by normalizing telemetry and survey data. But when tuning, you need to approach the problem less like a banker balancing the books and more like a beekeeper making sure that the hive is healthy and thriving. Imagine that you're responsible for balancing the environment of Earth and you get a data report and discover that there are 20 quadrillion ants on the planet but only 3 trillion trees. If you believe this is an imbalance, you aren't in the right mindset to tune a competitive game. A version of Earth with 20 quadrillion trees would have no space at all for anyone to exist. A version with only three trillion ants is going to cause serious ecological disasters as the food chain collapses. Good game tuning has a lot more in common with musical composition than it does. God, patch this right now. Requires harmonious choices, purposeful and intentful imbalance of volume and frequency and tempo to achieve the desired emotional effect. What works in one song won't necessarily work in another, which means one game's right answers might be totally wrong for another. All aspects of your game need to harmonize and come together to be stronger than the sum of all of its parts. I call this tuning for harmony. When tuning a game, you're the conductor. Your job is to lead players emotionally where you'd like them to go, not follow the players whims or manipulate telemetry numbers. A game that is tuned for harmony is trying to achieve retention by creating a game state in which the highest number of players can each have their own unique brand of fun, express themselves, and stay engaged. This is a harmonious environment where joy and passion can thrive. Any imbalance that you put into place with intent is based on your vision for the game there to make it more interesting and exciting. Dude, you know what? This is the best example of this cuz I see tabs in the chat chat. You remember how in the first Avatar beta vibes were really good, right? Like vibes were up there. We all we all can agree on this. I don't think it's even just me. It was like, "Wow, this is just such a crazy game." I was unexpecting it to play like this. It was just really enjoyable. The vibes of Avatar Beta 1 were crazy. That game was so [ __ ] up. Like that version of the game was so [ __ ] dude. Like every single character had wild and weird crazy [ __ ] long as [ __ ] combos. Like the game was so [ __ ] up. Like in any in any other sense, like anybody that's like a competitive mindset would have looked at this and been like, "Oh, this game's going to be [ __ ] dude." Like, but vibes were good. Like the vibes were so good that it didn't matter that Qatara was infiniting you from every single touch anywhere. And it was like, "Oh, this is actually whack as shit." But the vibes are good. It isn't an accident. Mathematical imbalance in your telemetry is not seen as a problem under this lens. It could be a symptom of a problem sometimes, but it could just as well be that the imbalance actually makes the hole stronger. How will you know the difference? Well, in order to tune for harmony, you're going to need a few things. First, you're going to need to put someone on your team in charge of game tuning. And that person needs to be able to play at a high enough level to feel the real game. a good point of that Nenet impact best example. When were the vibes ever good on Nen Impact? Even before it came out were the And I Yeah, it came out as like a crazy mess of a game, right? But I'll be real, dude. When were the vibes ever good? When was it when did it ever seem like a super appealing thing that this is what this is what everyone really wants? Actually, I I was trying people tried to tell me that, but then when it came out and it had like im initially funky ass net code before it had roll back before it like looked as really rough as it did for a very long time before it was still a fullpriced ass game. It was like I don't know if the vibes on this were ever good. So regardless of how like good or bad the game is, it doesn't matter because the vibes are bad. Doesn't matter, dude. They don't have to be tournament champions or the best in the world, but they should be strong enough to play alongside players of that caliber. This player needs to be under a few mandates. Number one, they can't play in tournaments. So, they can't be selfish about tuning. Not saying no one on your team can compete. I'm just suggesting that this one person should not compete in the game that you are making. Yeah, that's fair. And feel free to compete in other games to scratch that itch. Number two, they shouldn't even have a main character or weapon. They should spend their play time using the characters or weapons the community think are the weakest and against strong players using the characters or weapons people think are the most offending to see if they even agree. Number three, they should try to think of the game as a whole. Play the whole roster relatively well and understand what aspects of each character are the most core to their identity and fun factor. A player who say only uses grapplers or only uses shotguns and can't get out of their comfort zone is going to struggle in this role. Number four, they should try to understand the feelings of players at various skill levels and keep each bracket of players in mind when tuning. You don't tune to give weaker players a chance against more experienced players, but you do need to tune against inexperienced players unique frustrations if you want to retain them. This can be a challenge. That's the thing like you have to the teams need to be tuned not just for the high level, okay? Because there could be really frustrating things happening at a mid and low level that don't happen at high level at all that just suck that like and you had to learn through that [ __ ] You know, chat, you you've played a fighting game like this before, dude. Mortal Kombat 1 exists, my dude, where you have to get through this swath of weird [ __ ] and like understanding and like, oh god damn, like what the [ __ ] weird what what is this? And then you break through and you're like, oh, okay, I'm kind of having fun now. I think in in terms of a recent example, Mortal Kombat 1 is one of the best examples of that [ __ ] dude. Where it's like this game's actually really fun when you get there, but when you get it took a long time to get there, dude. Like, a lot of commitment to eventually get there. A lot of patience to eventually get there. Set of asks for a member of your team, but I believe it's crucial. This person is your lantern in a dark cave. Without them, you'll be lost, and you'll spend most of your time guessing your way back to the light. If you don't have this person on your team, you're going to have to find them externally and hire them. Next, listen to your players. Seriously, players lie, but you still need to hear them. Players are often right about the symptoms, even if they don't understand the best ways to fix the problems. If you can find a few players who love your game and have a better big picture sense than most, and talk to them a lot. These players are your canary in the coal mine. Don't take what they say at face value. Always think critically. I'm gonna I'm gonna I mean Keats knows like Keats knows they they did Fogore pretty rough in the the the 10th anniversary patch and every time I've shown interest in KI I've actually wanted to play other characters like and that's fair that's fair I I I understand that I understand that it was for the health of the gaming type stuff right I mean the only argument I had was that it was like it's a few people that are really good. It was a few, but still I I I understood why they made the changes they did. Who's this [ __ ] But at the same point, it that that did create a healthy thing where it has uh it has created me wanting to play other characters more than I ever had before than just sticking with like fog all the time. You know, a discussion with these key people can lead you to the right choices much more quickly and accurately or keep you from making major mistakes. on KI. Of the hundreds of players I spoke to, I trusted exactly two to give me the kind of feedback that helped me validate or make tuning decisions and maybe 10 more to give me thoughts that were helpful. I learned who was who by chatting with the players a lot and being observant. Even with their help, we didn't make the right choice 100% of the time. And you won't either, but you should try to find your key players and make your success rate stronger. As you get information from telemetry and player sentiment reports, filter it and test against it rigorously. It's critical that you do not knee-jerk react to the things you hear. And it can be extremely difficult not to when influencer group think doorstep. You need to deploy changes quickly enough to minimize frustration. But you need to be measured enough to know when a problem is being overblown or a solution that works for your game has not yet been found. As an example, in Killer Instinct and Rumbleverse, I found that an average of 3 to four weeks work for non-c catastrophic issues based on those games audiences and goals. For catastrophic issues, every day counts, and you need to get those fixes out in 2 to three days, if possible, while communicating often with the community. In most phases of gamedev, there's a mountain of work to do and not enough time to do it all. So, how do you prioritize your work and attention? During development, before you even have players, you need to spend time and energy on features that will let you deploy quick game-tuning changes without full and expensive title updates. These hot fix systems could literally save your game's life. Once your game is out, stay alert for mega frustrators, things in the game that are so far out of bounds that players won't tolerate them until your next major patch. Use your hot fix system and nuke these from orbit. There's a lot of competition for players attention, so don't give them an excuse to quit and play something else. Early on in Rumbleverse, players discovered a bug when double bouncing using a stop sign weapon that allowed them to safely drop a city block sized instant kill damage zone with zero counterplay. Our next patch was 5 weeks away, and our user base would have shrunk to near nothing in mere days if we'd had been unable to fix this issue. Thanks to our hot fix system, we were able to get a fix deployed to players in a few days instead of the five weeks it would have taken if we had to wait for our next major patch, ensuring that most of the frustrated players retained. Nailing this built trust between us and the players so that if something like that ever happened again, they'd help calm each other down knowing that we'd likely fix it quickly. There are, of course, some common mistakes to avoid when tuning. You should avoid focusing on individual character versus character or team comp versus team comp matchup tuning. Conceptually, these things are big buckets of variables. Way too easy to get lost and cause unexpected knockdown effects in other areas of your game if you take on something so broad. You should instead spend time on much smaller and clearer details, which I'll discuss in this section. You also need to avoid overreacting or reacting too quickly if the severity of the issue isn't high. Judge severity by the level of frustration it seems to be causing. Validate it yourself by playing against it and see how pissed off you get. So after all that, what exactly is my method? First, I start by thinking about the big picture of risk versus reward across the entire game. So he was a part of the Rumbleverse devs. Uh he was the main designer of Rumbleverse from what I understand. I think Keith is in the chat. Uh yeah, he was the I've been hearing about Rumbleverse and Keith's talking Rumbleverse for a hell of a long time before it came out. And yeah. Yeah. tuning decision made if you're able to correctly assess both the risk and reward of common actions in the game. Any new character or move you put in the game has to match this framework that you've conceptualized. Let's see some basic examples of risk versus reward and assess them. Think about head shot and shooters. If you aim for the body and you miss left or right, you hit the arms. If you miss low, you hit the legs. If you miss high, you might get a head shot. Your risk of missing when aiming for the body is at its lowest, and the reward is usually standard damage. If you aim for the head and you miss low, you hit the body and you get standard damage. But if you miss left, right, or high, you don't hit the target at all and deal no damage. Reward for a headshot is vastly increased damage output if you can land it. As a system, the riskreward of headshots makes logical and practical sense. This is a good example of an extremely common and effective riskreward system that breeds thoughtful play. As both aiming for the body or head are effective, but carry different levels of risk and reward that change as you switch weapons, distances, fight targets that move at different speeds, or even increase your own skill level. Now, let's look at a riskreward problem. Take for example Ultimate Marvel vs. Capcom 3. If you use Magneto's heavy attack, which is slower and harder to land to start your combo, your combo damage will be high enough to kill even the sturdiest characters in the game. However, if you start your combo with a quick light attack that comes out fast and hits low, and it's one of the easiest moves, you got to get creative. Your combo damage will still be high enough to kill even the sturdiest characters in the game. The lower and higher risk options both have the same reward. So, in this case, skilled players will almost unanimously gravitate towards using low-risk attacks since the outcome is the same no matter what. Let's compare that to another game's take on this. In Guilty Gear Strive, using a quick punch or kick attack has less combo options and scales the combo's damage and hit stun duration more harshly. Using a slower slash or heavy slash attack gives you the highest damage combo starters and lets you combo more generously. The lower risk option here has a lower reward. This gives players a reason to use more of their options in various situations rather than falling back to the easy and quick ones. The next thing to look out for when hunting riskreward problems is unjustified inconsistency. Let's check out Killer Instinct. In KI, every character has a five frame quick attack, but not every character has the same range or properties on their quickest attack. Not all of them get the same level of reward. Is this a problem? If the designers made these choices randomly, it might very well be. But if you deliberately design weaknesses into the character kits to compensate for their more extreme strengths, which is usually where their identity and fun lie, it could be perfectly okay to have a character whose quickest attack has poor range or one that can't be rapid fire canceled or one that can only successfully combo into one specific unsafe opener special move. Context is everything. Equality will flatline the play styles. Riskreward should make sense within each character. This is like the the biggest thing that's happening to Street Fighter right now because it was a it was a thing in Street Fighter 5. When when a new Street Fighter 5 character came out, you kind of you like you would do a thing and you're like, "Oh, cuz that that leads to this, this leads to that, and then you could do this." Like, you can almost figure out how they play with barely any actual play time. You know what I mean? So, and it's not that much different in Street Fighter 6. this except the the timing the the the frame data itself has changed, right? SF6 as a game follows a very similar sort of like I would say backend of how frame data is. So you kind of know this is going to lead to a plus two or this is going to lead to like a minus whatever, you know, like that kind of [ __ ] And then if it does, that means your drive rush version of that is going to give you extra frames here. Then you can do this after that. Then you can do this. like like the conclusions apply to everybody, you know? So that sort of leads to a game where a lot of the same [ __ ] happens and you see a lot of the same things over and over again. Luckily, Street Fighter has a lot more stuff, characters have a lot more things that they can do, which does make it generally successful, right? But because of that, like as time goes on and not a lot has changed, you see a lot of the same [ __ ] It's a very much a frame data game, you but also across all characters. Consider the characters Saberwolf, who is one of the easiest characters to play as, and Arya, who is the most difficult character in the game to play as. As I mentioned before, high difficulty is a risk, as the player has way more mistakes they could conceivably make in a match, and their options may be much more difficult to optimize. In a game tuned for flatness, perhaps Saberwolf and Arya should have a similar potential at top levels of play. But when balancing for harmony, we must acknowledge that Arya's high difficulty means she should have a higher number of top level advantages than other characters. Her higher risk should be justified by a higher potential reward. If an easier character could match her potential, almost no one would invest time into the character. Imbalance can absolutely ruin a game. Or imbalance could create harmony. How? If each character's moves are not built equally, there's an imbalance in their strengths and weaknesses. Without the imbalance of high-risk, highreward characters living alongside low-risk, low-reward characters, player identity gets stifled. You need to find the right mix. So long as the riskreward values across your game make sense together, player identities will shine through and create variety and reduce repetitive play patterns. For instance, low or no- risk tactics that have high rewards, such as safe jumps in fighting games, can be simultaneously boring for the attacker and very frustrating for the defender. Try to use this line of thinking to identify outliers and then study the characters or situations these outliers belong to so you can understand if they're a problem or not. Be purposeful with your design. It's the It was the kind of the problem that happened to Street Fighter 4 in the end was that you could just option select too much, right? Like it's it's such a cool game, but you almost like optimize any of the interaction out of it. Like and that's just what it that's why Daigo historically said what he said about Street Fighter 4. It's like you can be good at Street Fighter 4 in training room only. Like you just you don't have to really fight many people to be amazing at Street Fighter 4. If you just get really good at option selecting a bunch of [ __ ] like from training room, it doesn't matter who you're playing. It doesn't matter because you're optimizing out all the interaction. build in deliberate strengths and weaknesses and commit to them. This is not scientific. The developer responsible for game tuning is going to need a strong vision, a clear understanding of their target audience, and the ability to step back and try to feel all of this from a distance and see the big picture. The next thing you need to stay on top of is frustration. The opposite of frustration is fun. And so, these exist on a single scale together. Some frustration is okay, and it can even be good for your game, so long as frustration doesn't overtake fun for your target audience. Imagine leaping off a building in a game with fall damage versus one without. A game without fall damage is definitely less frustrating, but it also lacks the adrenaline rush of attempting the leap and finding a way to land safely despite the danger. Neither of these approaches are wrong. It's just that each This is the reason Souls games have stamina. Like there's the reason the majority of Souls games are designed around a stamina system. Just the reason because like it's it's the adrenaline rush of managing all these things at the same time. It's like yeah, would it be more inherently fun without it? Sure, it would be. But like it's why better suits a different audience. When an individual's frustration outweighs their fun, they quit. In a single player game, this part's really easy. In Zelda Breath of the Wild, the fun is in exploring and traversing this massive handcrafted world, solving puzzles, getting into battles, and then you have to interact with the very frustrating menus to cook food and manage your inventory, and then all of your weapons break. Some parts of the game are fun, and some parts are frustrating. They pull on each other within me to help me decide if I should keep playing or quit. So, let's explain the fun frustration scale. I quit Breath of the Wild. Nothing was as fun as that first day. Is my own personal fun versus my own personal frustration. This scale represents me, not the game itself. This scale is not about truth. It's about my relationship with the game. If the fun side weighs more, I keep playing. I'll place weights that represent my personal enjoyment from exploration, puzzles, and combat on the left side. The fun side of the scale is heavier now, and so it moves down. Fun weighs more. And now I'll place weights that represent my own personal frustration from having to cook or navigate menus or when my weapons break. Maybe the weight of your personal frustration from say weapons breaking is less than mine. And that would be reflected on your own personal version of the scale. As you can see, the total weight of my fun in Breath of the Wild is heavier than the total weight of my frustrations. If we make the cooking interface a little better, the weight of my total frustration lowers, but the weight of my fun doesn't increase. Because improving the cooking interface doesn't improve the exploration, even though it does improve my overall relationship with the game. Remember, the lower side of the scale is the one that's winning out. Even in competitive games, some frustration is totally unrel Bro, don't get me started. That [ __ ] is deeply frustrating. They'll just have to wait for [ __ ] ever for anything, dude. MK1, Tekken 7, uh, and yeah, Guilty Gear Strive at times. You have to wait for [ __ ] everated to gameplay. Why do you think it's so Why do you think it's so easy for me to just jump in and play Street Fighter 6 at like any point and just have a genuinely good time? I get it. If I play it was the only thing I played for like a week straight, I would have frustrations with like gameplay and guessing and all that kind of stuff. But if it's just an act of like jumping in and enjoying and playing it and trying the character or something like that, dude, there are so few things getting in the way of me just enjoying that game online, right? Of just jumping on and just fight fight fighting people. There's so few things that get in the way of that happening. Like if it takes you 3 minutes to simply…

Transcript truncated. Watch the full video for the complete content.