on Hero design, Risk, Reward, Skill.
The video discussing this post can be found here
This one has been a long time coming. For decades before Overwatch I've always half-joked about skill, noobs, pros and generally have held a bit of an elitist view on video games. As I've gotten older and study game design more I've refined it into an understanding of skill, when something requires too much and when something requires too little. Please stick around because this is going to be a longer post that covers some parts of game design and some of balance, I promise it all comes together, eventually.
I want to start off with a statistic from Jaffe Alex's discussion "Cursed Problems".
This one has been a long time coming. For decades before Overwatch I've always half-joked about skill, noobs, pros and generally have held a bit of an elitist view on video games. As I've gotten older and study game design more I've refined it into an understanding of skill, when something requires too much and when something requires too little. Please stick around because this is going to be a longer post that covers some parts of game design and some of balance, I promise it all comes together, eventually.
I want to start off with a statistic from Jaffe Alex's discussion "Cursed Problems".
"Players are said to believe a game is fair if they win 70% of the time "
Keep this in mind when we talk about balance and how the community perceives it, whenever someone talks about "balance" 70% win rate is what they probably wanting. Right now the big complaint in OW is Genji, who is seen as a high skilled hero will often lose to Moira, which is seen as a low skilled hero. I'll address this issue directly but understand the overall goal of this example is to better understand risk, reward and what role skill has in the entire exchange. My goal of this post is to make everyone believe that players are not upset at Moira, but the concept of Moira. Which means that unless something drastic happens to her, Genji players will always be upset.
The first step to understanding the concept of Moira is to look at how Overwatch was designed. It was created to be an easy accessible game that brings new players into the FPS genre. Please refer to my first post back in 2017 for a longer explanation. The main idea is to promote a "flat" design so that anyone can pick the game up.
When I say flat design I mean that there are very few moving parts and variables in Overwatch. If you're looking for an answer to a problem, you simply pick the hero that solves your issue. Most players would suggest that most games do this, but let me clarify by showing you what isn't a flat design, MOBAs.
I'm about to get a little out there. I apologize for any confusion'
I'm about to get a little out there. I apologize for any confusion'
In my head, a game like DOTA or LoL is designed for players have a specific goal or a win condition in mind. Now picture that goal in space, make it a ball for simplicity. Now picture a tree attempting to grow towards this goal. The trunk of that tree is the largest and most stable growth towards your goal but cannot be changed once started. This is you picking your heroes. They set the stage for the game and how you're looking to achieve your goal. Remember that the enemy team is doing the same, they have a tree and are reaching towards their own goal. These goals often conflict and that's where we have tactics and strategies clashing together. An example of this clashing is one team losing a couple of heroes early on. After a clash it's often that the teams' goals will be affected. In such a case as taking an early loss, let's say your goal is now farther away.
But you already had a plan! The majority of it was set in stone with your hero picks! How can you deal with all of these variables that can change every game? You need a mechanic that allows customization, in my tree metaphor, these are the branches. The branches represent items and talents. Items/talents allow you to specifically customize how your hero operates based on the situation. While they don't have as large of an impact as hero picks, they do provide a flexible way to play the game so players can adapt on a per game basis. Items are designed to fill a specific role in the game, meaning there are items for pure damage, damage and health, movement speed or magical damage and utility. An item's role almost never overlaps with another. MOBAs have to provide this ability to customize heroes because once you pick a hero, that's it, you're stuck with it. If a hero cannot deal with a specific situation then the opposing team has an easy way to win, which makes for a poor game on both sides. A game that's too easily won feels like rock paper scissors and since these games take time, the players feel helpless the entire time they're losing.
Eventually teams will need to have many branches to reach their goal. Each of these branches that allow players to get to a specific goal are a use of game mechanics. Mechanics that players have to learn, mechanics that the developer has to balance. Not only do these mechanics have to interact with each other normally, but they have to interact with everything at different times of the game. This is immensely complex and has a serious learning curve. This is exactly what Blizzard wanted to avoid with Overwatch.
So how did Blizzard take Overwatch in the opposite direction of MOBAs?
But you already had a plan! The majority of it was set in stone with your hero picks! How can you deal with all of these variables that can change every game? You need a mechanic that allows customization, in my tree metaphor, these are the branches. The branches represent items and talents. Items/talents allow you to specifically customize how your hero operates based on the situation. While they don't have as large of an impact as hero picks, they do provide a flexible way to play the game so players can adapt on a per game basis. Items are designed to fill a specific role in the game, meaning there are items for pure damage, damage and health, movement speed or magical damage and utility. An item's role almost never overlaps with another. MOBAs have to provide this ability to customize heroes because once you pick a hero, that's it, you're stuck with it. If a hero cannot deal with a specific situation then the opposing team has an easy way to win, which makes for a poor game on both sides. A game that's too easily won feels like rock paper scissors and since these games take time, the players feel helpless the entire time they're losing.
Eventually teams will need to have many branches to reach their goal. Each of these branches that allow players to get to a specific goal are a use of game mechanics. Mechanics that players have to learn, mechanics that the developer has to balance. Not only do these mechanics have to interact with each other normally, but they have to interact with everything at different times of the game. This is immensely complex and has a serious learning curve. This is exactly what Blizzard wanted to avoid with Overwatch.
So how did Blizzard take Overwatch in the opposite direction of MOBAs?
They removed all of the complex mechanics simply by allowing players to change their hero or their "trunk". Why worry about items and how they interact with each hero. Even before that, why worry about building an entire system to get and spend gold to get those items? Why worry about talents and how they have to be balanced to allow heroes to be relevant during different parts of a 30 minute game. Screw that. Let's remove all of these and make it simple. If you want an answer simply pick the hero that can reach out and grab it. If you need a long ranged option then pick it. You need a close ranged solution then go to spawn and get what you need immediately. That's why it's so important that you can swap in Overwatch - because originally heroes weren't meant to deal with every situation, to reach every goal. (This is also why I don't agree with hero bans)
Heroes ARE items, they ARE talents/perks/passives all in one. With a flat design you only need to design heroes to fill a role. This has its own set of problems but solves the immense issue of getting players into the FPS genre, which is Overwatch's primary goal.
Another part of Overwatch's flat design is that heroes don't have any complicated classes or natural bonuses/weaknesses. They fill their role naturally though obvious and often visual indicators. In other games such as an RTS tracer would be given a flanker title, which makes her do +50% damage to supports but take 50% more damage from explosives.
All heroes are balanced for a role and their abilities and weapons support that. As an example let's say Blizzard wanted to create a flanker. A long range flanker doesn't feel balanced, as a long ranged flanker would just kill you head on from afar. So to support the hero being close ranged they give them a gun with heavy fall off and a wide spread. Now this leaves them vulnerable because close range targets are more easily hit, so how do you address this? Give them an ability to get in and out of close range. As an example, blink or dash. I'm sure during a couple play tests they learned that too many blinks or dashes felt unfair or unfun to the defender. So how do you give this hero the ability to take risks but not feel heavily punished? You have abilities to mitigate risk, such as recall or deflect. Designing these heroes required no special mechanics to make them a flanker. They use simple concepts such as fall off, weapon spread, HP and model size to naturally make them a flanker. Realistically, design could have started from any one of these points to get to the released version of Tracer or Genji.
This is how we have themes with specific types of heroes. Talking about flankers, we also have Doomfist, Genji and Echo. Notice how all of these heroes have weapon mechanics that force them into mid/close range. They all have abilities to get them into a fight to take a risk and then abilities to get out of the fight. Of course there's different variations to them but that's what gives us a nice diversity. It's why in most cases when someone doesn't like a hero, it's not specifically about that hero, its about the role that hero fills. Doomfist is a highly controversial hero, and rightfully so. Does there need to be a flanker with options to "flank" head-on? Notice how I didn't mention he has burst damage? It's because it's inherent with being a flanker. You can't nerf him to not to do burst damage because he stops being a flanker at that point and his entire skill-set needs to be redone. This is what we call a re-work, where the hero is removed from their role and put into a slightly new one. Another good example of this is how 250hp heroes are designed, see this post for more on that.
As MOBAs create items and perks to deal with different situations so too must Blizzard create different heroes to fill these specific issues you can face in Overwatch.
Rightfully so, to avoid role overlap new solutions need to be created for old problems. We've seen Blizzard attempt to solve problems by introducing new heroes multiple times before. This isn't a bad approach at first but if you create a single hero to deal with a problem of a specific team composition then that hero will be strong enough to deal with 6 combined heroes. Every time Jeff says a hero is "meta changing" I get worried because I know an overpowered hero will be released. A single hero can't nullify an entire team and still be balanced. I believe this was noticed during the time period of Doomfist , Moira and Brigitte. Ever since then the direction has been more generalist than specialist.
Another part of Overwatch's flat design is that heroes don't have any complicated classes or natural bonuses/weaknesses. They fill their role naturally though obvious and often visual indicators. In other games such as an RTS tracer would be given a flanker title, which makes her do +50% damage to supports but take 50% more damage from explosives.
All heroes are balanced for a role and their abilities and weapons support that. As an example let's say Blizzard wanted to create a flanker. A long range flanker doesn't feel balanced, as a long ranged flanker would just kill you head on from afar. So to support the hero being close ranged they give them a gun with heavy fall off and a wide spread. Now this leaves them vulnerable because close range targets are more easily hit, so how do you address this? Give them an ability to get in and out of close range. As an example, blink or dash. I'm sure during a couple play tests they learned that too many blinks or dashes felt unfair or unfun to the defender. So how do you give this hero the ability to take risks but not feel heavily punished? You have abilities to mitigate risk, such as recall or deflect. Designing these heroes required no special mechanics to make them a flanker. They use simple concepts such as fall off, weapon spread, HP and model size to naturally make them a flanker. Realistically, design could have started from any one of these points to get to the released version of Tracer or Genji.
This is how we have themes with specific types of heroes. Talking about flankers, we also have Doomfist, Genji and Echo. Notice how all of these heroes have weapon mechanics that force them into mid/close range. They all have abilities to get them into a fight to take a risk and then abilities to get out of the fight. Of course there's different variations to them but that's what gives us a nice diversity. It's why in most cases when someone doesn't like a hero, it's not specifically about that hero, its about the role that hero fills. Doomfist is a highly controversial hero, and rightfully so. Does there need to be a flanker with options to "flank" head-on? Notice how I didn't mention he has burst damage? It's because it's inherent with being a flanker. You can't nerf him to not to do burst damage because he stops being a flanker at that point and his entire skill-set needs to be redone. This is what we call a re-work, where the hero is removed from their role and put into a slightly new one. Another good example of this is how 250hp heroes are designed, see this post for more on that.
As MOBAs create items and perks to deal with different situations so too must Blizzard create different heroes to fill these specific issues you can face in Overwatch.
Rightfully so, to avoid role overlap new solutions need to be created for old problems. We've seen Blizzard attempt to solve problems by introducing new heroes multiple times before. This isn't a bad approach at first but if you create a single hero to deal with a problem of a specific team composition then that hero will be strong enough to deal with 6 combined heroes. Every time Jeff says a hero is "meta changing" I get worried because I know an overpowered hero will be released. A single hero can't nullify an entire team and still be balanced. I believe this was noticed during the time period of Doomfist , Moira and Brigitte. Ever since then the direction has been more generalist than specialist.
Now that I've briefly talked about OW's flat design, what does this have to do with genji vs. moira? Flankers were an issue at the start of Overwatch, specifically for supports. Supports could sometimes handle them but it wasn't consistent enough. Blizzard needed to introduce a new hero to solve that old problem. It's tricky but it needed to be a healer that could deal with flankers but isn't strong enough to easily deal with tanks or other DPS heroes.
This is the basis of Moira. Her primary attack gives that need for constant damage on an evasive flanker, while still giving them time (3-4 seconds) before killing them. This is pretty fair if you consider Overwatch's low time to kill. If you decrease her consistency then your evasive flankers will end up dodging the damage and it becomes just like the current supports - with the issue not solved. For a higher consistency weapon you need lower damage to make it feel fair to both parties. If the damage was higher and the consistency lower you'd have higher SR players crush flankers and lower SR players lose to flankers.
You're now starting to see the issue with any type of minor change to Moira. Either you lower her damage to a point the flankers always wins or you make her less consistent but stronger where flankers have even less of a chance to win at higher ranks. In either case you make the hero feel less fair. While players get upset about specific abilities, what they're upset about isn't that Moira is strong. Its that a consistent anti-flank support exists. They can't complain about it like that because if they did it would be pretty obvious that they're upset that there's a support they don't have the advantage over. Moira in any form as long as she keeps to her role will be a thorn in a flanker's side because she's designed to be. The real answer to the issue is that because of flat design choice you need to change the tool you're using or greatly change the method that you're using with it.
You're now starting to see the issue with any type of minor change to Moira. Either you lower her damage to a point the flankers always wins or you make her less consistent but stronger where flankers have even less of a chance to win at higher ranks. In either case you make the hero feel less fair. While players get upset about specific abilities, what they're upset about isn't that Moira is strong. Its that a consistent anti-flank support exists. They can't complain about it like that because if they did it would be pretty obvious that they're upset that there's a support they don't have the advantage over. Moira in any form as long as she keeps to her role will be a thorn in a flanker's side because she's designed to be. The real answer to the issue is that because of flat design choice you need to change the tool you're using or greatly change the method that you're using with it.
So where does this leave us? Both side of the genji/moira argument can read this and still not feel any better. Why?
Skill... or at least the perception of it.
First let me define how I see skill. To me skill is consistency. Something that is more consistent is easier than something that is less consistent. In the same manner, someone who is more consistent is better than someone who isn't. This is why I prefer best of 5/7 matches or games with a longer time to kill. The better player should win in a longer scenario. While abilities may be balanced so you get less value from something that's more consistent and more value from something that's less consistent, that has nothing to do with skill itself. If you were to play widow, what makes a headshot harder than a body shot? Consistency. Headshots happen less and are therefore harder. Of course there's more hitbox available for non-critical shots but that's the mechanic that leads it to being less consistent. The fundamental idea is to introduce mechanics that make players less consistent to make an ability harder. The problem now becomes that skill is now subjective, which it is. If everyone in the world could get to a 50% headshot rate with widow, we'd have to change her because it's too easy. In our moira/genji example, it's because moira is meant to be anti-flanker that she needs a large leeway on her attack. This makes her more consistent - to do otherwise removes her from an anti-flanker role.
Players believe that when they do something that is difficult that they should be rewarded for it, specifically when against something that's not skillful. At the surface level it makes sense but completely clashes with the idea that different heroes are there to respond to different situations. Remember how heroes are designed with inherent abilities to fill their role. By definition, "inherent" means you did nothing to earn it, which is contradictory of skill. This is such a predominate theme in overwatch that any hero that has an advantage over another means that the stronger hero takes less skill to deal with the weaker one. It is easier for a winston to deal with a genji than it is for the opposite. The same can be said for phara and junkrat, hit-scan and phara or flankers against ana. The latter always has to be more skilled to deal than the former. Just because you use a hero where it's improper doesn't mean you should be rewarded, being able to use your hero in a poor situation is the reward for mastery.
Again... this still doesn't make anyone else feel better. Logical talks about an emotional response hardly end well but it does bring up an issue with a flat design. If players still feel upset about these match-ups then maybe the natural advantage one hero has over another is too much.
It feels like rock, paper scissors. The player feels like no amount of skill could help them thus feels frustrated and hopeless. This is the exact opposite of what an FPS should feel like (The skill part)
These are some of the challenges a role-based flat design in a FPS have. Winning by picking the right answer doesn't feel fair because FPSs are inherently based on mechanical skill.
At first the community was fine with niche heroes and specialists but they were quickly outed as weak heroes. Sure if i wanted to fill a specific role, I could just reach out and grab that answer but the counter answer is just as easy. This is why the enemy team reaches out and grabs a simple answer to deal with your specialist. Overwatch's objectives can be won in a single fight and a specialist is so specified in what they can do that you can easily counter-pick it; resulting in your immediate loss. Specialists were quickly faded out in the competitive scene because of the risks they had. This starts to make sense realistically instead of just on paper, it's what was missed in Overwatch's design from the start. If a specialist is too good in their role it feels like they're overpowered and they didn't work for it. If they're not extremely good in their role then they have no use in the game at all. This is why we saw reworks for torb and sym while we got large changes for bastion, junkrat, and reaper. Blizzard had to change their balance philosophy because of how easily some heroes were dealt with, not because they were weak but because of the role they filled. It starts the idea that all heroes need to become more generalist to survive being counter picked or more heroes need weaknesses that all heroes can exploit. It drastically changes the concept of balance to "X hero is OP or X hero is weak" to "how many options are viable against a hero". Identical in my definition of skill, more options means the hero is easier to deal with. The fewer options the more difficult the hero is to fight.
Which brings us to the most recent genji changes as of june 2nd, 2020. Genji is getting quite a few buffs, Deflect is now cancelled and duration increased to 2 seconds. His spread with his right click is lower and he gets a small damage increase (8%).
Notice anything odd with these changes? They won't help against Moira at all, in-fact he'll still lose against her on average. These are changes that change the focus of genji from an anti-support flanker to more of an anti-tank/dps flanker. They're trying to tell players to use genji to deal with other issues other than diving the back lines. This could have been solved by picking a different hero but OW is about diversity. Everyone wants more heroes to be viable.
This it where we're at today with Overwatch. If we want heroes to be viable, first we need to examine if their role is still needed. If it is not, then rework. Reworks often pivot skills to fill a new role. If their role is still needed then next we need to see how many options are available against that hero. If too few options then we identify a general weakness and look to expand it, if too many options then identify the most common role that exploits it and bolster it. Lastly we start to look at how we actually change a heroes kit. Make sure it lines up with the goals we defined above. If we look at multiple heroes and see the same type of weakness, maybe it's ideal to weaken an entire role of heroes. No idea is too big or small if its the right answer. It doesn't matter if people have muscle memory, if a change is needed then Blizzard must implement the change.
Originally I mentioned that not liking moira is more about not wanting an anti-flanker support but maybe the role that wasn't needed originally wasn't moira's, but instead genji's. A flanker that had an advantage over every support was a bit too strong in the early days and as new heroes get rolled out to deal with his niche he's become less of an option. He's still strong, its just difficult to play him consistently. I believe the proper response is to rework genji instead of moira. After all a hero that's designed to get into the back lines and easily kill the support heroes doesn't sound fun for anyone. Especially if support heroes keep trending to being more self-sufficient.
This is all i could coherently write but I'm sure there's more, I have more ideas that I'll keep blogging. For me the first step to understanding an idea is to write it out, it's only then you can start correcting it. I know designing from this point of view seems bland but most of my ideas are a way to organize, understand, and develop a system. A lot of the old games I grew up on in the 90s started from a cool idea and making it a reality. Its important to not lose sight of that, most of our favorite heroes didn't start with balance and skill, it was more of a fun idea that was ran with and then balanced over time. Blizzard and the community just need to stop sometimes and take a look back and understand how it works and how we got here. Its the only way we're going to move forward in a proper manner.
Skill... or at least the perception of it.
First let me define how I see skill. To me skill is consistency. Something that is more consistent is easier than something that is less consistent. In the same manner, someone who is more consistent is better than someone who isn't. This is why I prefer best of 5/7 matches or games with a longer time to kill. The better player should win in a longer scenario. While abilities may be balanced so you get less value from something that's more consistent and more value from something that's less consistent, that has nothing to do with skill itself. If you were to play widow, what makes a headshot harder than a body shot? Consistency. Headshots happen less and are therefore harder. Of course there's more hitbox available for non-critical shots but that's the mechanic that leads it to being less consistent. The fundamental idea is to introduce mechanics that make players less consistent to make an ability harder. The problem now becomes that skill is now subjective, which it is. If everyone in the world could get to a 50% headshot rate with widow, we'd have to change her because it's too easy. In our moira/genji example, it's because moira is meant to be anti-flanker that she needs a large leeway on her attack. This makes her more consistent - to do otherwise removes her from an anti-flanker role.
Players believe that when they do something that is difficult that they should be rewarded for it, specifically when against something that's not skillful. At the surface level it makes sense but completely clashes with the idea that different heroes are there to respond to different situations. Remember how heroes are designed with inherent abilities to fill their role. By definition, "inherent" means you did nothing to earn it, which is contradictory of skill. This is such a predominate theme in overwatch that any hero that has an advantage over another means that the stronger hero takes less skill to deal with the weaker one. It is easier for a winston to deal with a genji than it is for the opposite. The same can be said for phara and junkrat, hit-scan and phara or flankers against ana. The latter always has to be more skilled to deal than the former. Just because you use a hero where it's improper doesn't mean you should be rewarded, being able to use your hero in a poor situation is the reward for mastery.
Again... this still doesn't make anyone else feel better. Logical talks about an emotional response hardly end well but it does bring up an issue with a flat design. If players still feel upset about these match-ups then maybe the natural advantage one hero has over another is too much.
It feels like rock, paper scissors. The player feels like no amount of skill could help them thus feels frustrated and hopeless. This is the exact opposite of what an FPS should feel like (The skill part)
These are some of the challenges a role-based flat design in a FPS have. Winning by picking the right answer doesn't feel fair because FPSs are inherently based on mechanical skill.
At first the community was fine with niche heroes and specialists but they were quickly outed as weak heroes. Sure if i wanted to fill a specific role, I could just reach out and grab that answer but the counter answer is just as easy. This is why the enemy team reaches out and grabs a simple answer to deal with your specialist. Overwatch's objectives can be won in a single fight and a specialist is so specified in what they can do that you can easily counter-pick it; resulting in your immediate loss. Specialists were quickly faded out in the competitive scene because of the risks they had. This starts to make sense realistically instead of just on paper, it's what was missed in Overwatch's design from the start. If a specialist is too good in their role it feels like they're overpowered and they didn't work for it. If they're not extremely good in their role then they have no use in the game at all. This is why we saw reworks for torb and sym while we got large changes for bastion, junkrat, and reaper. Blizzard had to change their balance philosophy because of how easily some heroes were dealt with, not because they were weak but because of the role they filled. It starts the idea that all heroes need to become more generalist to survive being counter picked or more heroes need weaknesses that all heroes can exploit. It drastically changes the concept of balance to "X hero is OP or X hero is weak" to "how many options are viable against a hero". Identical in my definition of skill, more options means the hero is easier to deal with. The fewer options the more difficult the hero is to fight.
Which brings us to the most recent genji changes as of june 2nd, 2020. Genji is getting quite a few buffs, Deflect is now cancelled and duration increased to 2 seconds. His spread with his right click is lower and he gets a small damage increase (8%).
Notice anything odd with these changes? They won't help against Moira at all, in-fact he'll still lose against her on average. These are changes that change the focus of genji from an anti-support flanker to more of an anti-tank/dps flanker. They're trying to tell players to use genji to deal with other issues other than diving the back lines. This could have been solved by picking a different hero but OW is about diversity. Everyone wants more heroes to be viable.
This it where we're at today with Overwatch. If we want heroes to be viable, first we need to examine if their role is still needed. If it is not, then rework. Reworks often pivot skills to fill a new role. If their role is still needed then next we need to see how many options are available against that hero. If too few options then we identify a general weakness and look to expand it, if too many options then identify the most common role that exploits it and bolster it. Lastly we start to look at how we actually change a heroes kit. Make sure it lines up with the goals we defined above. If we look at multiple heroes and see the same type of weakness, maybe it's ideal to weaken an entire role of heroes. No idea is too big or small if its the right answer. It doesn't matter if people have muscle memory, if a change is needed then Blizzard must implement the change.
Originally I mentioned that not liking moira is more about not wanting an anti-flanker support but maybe the role that wasn't needed originally wasn't moira's, but instead genji's. A flanker that had an advantage over every support was a bit too strong in the early days and as new heroes get rolled out to deal with his niche he's become less of an option. He's still strong, its just difficult to play him consistently. I believe the proper response is to rework genji instead of moira. After all a hero that's designed to get into the back lines and easily kill the support heroes doesn't sound fun for anyone. Especially if support heroes keep trending to being more self-sufficient.
This is all i could coherently write but I'm sure there's more, I have more ideas that I'll keep blogging. For me the first step to understanding an idea is to write it out, it's only then you can start correcting it. I know designing from this point of view seems bland but most of my ideas are a way to organize, understand, and develop a system. A lot of the old games I grew up on in the 90s started from a cool idea and making it a reality. Its important to not lose sight of that, most of our favorite heroes didn't start with balance and skill, it was more of a fun idea that was ran with and then balanced over time. Blizzard and the community just need to stop sometimes and take a look back and understand how it works and how we got here. Its the only way we're going to move forward in a proper manner.