Wyatt Cheng recalled his previous experience when developing Diablo 3 Vanilla.
6 Likes
Meanwhile the D4 team preparing the â10 years in developmentâ Q&A.
11 Likes
Negative Nancy is at it again 2016 - 2022 = 6.
3 Likes
Meh Iâd rather look ahead to Immortal and see what they learned from the plethora of D3âs mistakes.
Well, at least he acknowledged âdesigning to a numberâ is a bad idea and results in, often, poorly conceived and forced mechanics much like âdesign by committee.â Iâd agree that the rune unlocks are possibly one of the worst features mainly because only 2 or maybe 3 are even remotely useful/used. Making the legendary item have these effects instead makes them much more interesting and desirable.
1 Like
This is actually a pretty interesting interview, and good to know how itâs influenced DI. I also feel itâs more communication and feedback on D3 than weâve received in ages so itâs pretty welcome.
I wasnât at all interested in DI previously either, but I may just check it out now after reading these thoughts. Good stuff overall.
1 Like
I love Wyatt for interacting with players like this. Other game devs should learn from it.
But damn, I cant get around the fact that I disagree with so much of what he says 
- The RMAH was designed to help players trade in-game instead of using external websites or tools and was not designed as part of a âgame as a serviceâ.
Surely nobody believes that?
Not saying the former wasnt part of the reason, I very much believe it was. But then it turned into a game as a service system with heavy P2W elements.
- Initially, Diablo 3 difficulty was influenced by WoW raiding.
Not a bad idea at all. As long as it is in a soloable context.
They should keep aiming for this in D4.
- Weeks after launch, the development team recognized that it would be more rewarding for players to farm challenging content efficiently repeatedly instead of overcoming extremely difficult content.
Why does it have to be either/or? 
I very much agree that A-RPGs should mostly be about becoming more efficient over time, rather than a binary check on whether you can do the content at all. That is one area where D2 did it really well, with its low difficulty (albeit, too low difficulty).
But you dont even have to choose⌠Have a bunch of content that is reasonably doable, where it just becomes faster to do over time (but not a lot faster! dont have silly power creep), AND have some raid-style challenges (though once again, it should all be soloable, the raid-style only refers to the kind of challenge offered).
- The Paragon system was not part of the original D3 design idea.
And it never should have become part of it.
- PvP didnât make it into D3 because the team had problems to balance the complex interactions between skills, skill runes, legendary items, and visual noise.
I dont really care about PvP, but maybe the issue here was the attempt to balance those complex interactions?
As for visual noise⌠eh, that is also a big problem outside of PvP⌠and the fix ought to be the same. Scale down all those absurd spell effects.
we determined during internal playtesting that players were not getting enough reward for levelling up. In Diablo 2 you got Attribute Points and a Skill Point every time you leveled up - this made gaining a level super exciting. In Diablo 3 when you leveled up you got - nothing.
The solution to people being disappointed about not getting skill points and attribute points when leveling up, seems so obvious that I wont even spell it outâŚ
Anyway, there was nothing wrong with the rune system as such. Albeit it seemed better during development than how it ended up.
Furthermore, the effects of skill runes overlap heavily with a design space that is much better occupied by Legendary Items.
Noooo Talk about learning the wrong lessons.
Not every skill matches nicely to having 5 rune effects. Some skills benefit from having more! I call this âDesigning to a numberâ and it is a common pitfall in game design
He is not wrong about the Designing to a number issue. BUT it is a self-made issue. Just because you tie skill modifications to runes, nobody is saying you must have 5 runes per skillâŚ
That holds true if Runes are droppable, as during D3 development, and it holds true if the Rune modifications are part of a âSkill treeâ or whatever.
Legendary Items solve all of these problems!
Maybe, but they introduce their own.
Absolutely nothing that Legendary items solve, couldnât be solved through having those skill modifications tied to the character instead of items (nor does it have to be either/or!!!)
In Diablo 3 the player has huge control via customization choices (gearing, skill choices, etc.) how much of their character is focused on Damage and how much is Survivability. This is by a factor of 10x or more. In order for the game to be fun, and for itemization to feel fun, itâs important that both categories are valued by the player. What generally happens is players prioritize Damage as much as possible and they donât think about Survivability until they really have to.
Have you ever heard about death penalties/survival bonuses?
At least he acknowledges that offense and defense should both matter. Will actually be interesting to see how D:I performs there, because in D3 the balance is a disaster.
Even though players CAN spend 30 minutes fighting something, theyâre not going to if itâs not the most efficient way to get gear. Morgan Day over on the WoW team passionately tried to tell me this before ship but I foolishly ignored him.
And for the very same reason you can get rid of the GRift time limitsâŚ
Efficiency already rules supreme, no need for hard caps on time spent.
Which is not to say you cant have enrage effects. WoW does that too. Totally fine. Enrages can be âsoft capsâ rather than âhard capsâ, depending on their design.
For situations where players are proving and testing the limits of their character timers can still be appropriate.
Test those limits by having a cap on how many times your character can die instead for example. Or soft caps. Like D3 Butcher filling up the room with fire is a soft cap.
The Paragon system was not part of the original D3 design and we didnât realize we needed it until after ship. One of the changes we made between D2 and D3 is lowering the level cap from a near-impossible-to-attain level 99 to a very attainable level 60. I donât think we fully appreciated at the time the gravity of that decision. Once you hit level 60 suddenly the game became WAY less rewarding.
Yes, that was a bad decision. But there is a few miles between realizing that lvling should not be over too soon, and ending up with endless Paragon 2.0.
The original Natalyaâs 4-piece bonus that gave 2 discipline a second. Due to a bug it went live with 2 discipline/second instead [âŚ]
Ultimately we kept it because some players would have spent real money at this point to get the items and it would have been disastrous to nerf an item in that situation.
Might that be yet another reason why P2W is a bad idea?! Yes⌠yes it might. SeriouslyâŚ
One of the game design lessons learned is that when you are providing difficulty settings itâs important for players to be able to pick an appropriate difficulty setting for themselves. This is obvious when stated plainly and with the benefit of hindsight. Your lowest difficulty needs to be for the least skilled player in the worst gear. Your highest difficulty should be for a highly skilled player in great gear. Hope we can agree on that.
While I at least consider it somewhat understandable why people might have the above opinion, I also cant agree with the first half of it.
Always having a perfectly appropriate difficulty is bad. It makes things less interesting, similar to auto-scaling seen in various games. It removes the ebbs and flows that are central to most game design. It greatly worsens the feeling of running in a treadmill in an A-RPG. A constant difficulty, a constant speed on the treadmill, perfectly adjusted to your current situation.
It is a horrible decision imo to have freaking 150 difficulty settings. Even the 16 torment settings are way too much (though in reality, most of those can be ignored due to power creep).
Going from Nightmare to Hell in Diablo 2 for example represents a change of pace. Because it takes you out of your comfort zone for a while. It isnt just constant difficulty forever after.
Turns out the correct number is however many you need so that a player doesnât feel STRANDED between two difficulties.
As above, feeling stranded between two difficulties is good! Anything else is handholding game design taken to an extreme. Even more so in the context of the game Diablo is (player vs the demonic horde). It should feel hopeless, it should feel like you are fighting an uphill battle sometimes, and then, as you gear up, it gets more and more tolerable.
The moment you think that having 150 difficulty steps are fine, why even bother to have difficulty steps. Just automatically, perfectly, scale the game to peoples gear. Also just cancel the game at the same time
We experimented multiple times with an âAngelic/Demonicâ influence system both at Blizzard North and again when the project was rebooted in Irvine. The idea was that narrative choices you made would influence your character and your characterâs development. We were inspired heavily by Star Wars Knights of the Old Republic. The idea that players could choose to walk down a light path or a dark path was super compelling to us. There are a TON of positive cool things that came from this idea and we were super excited. Three reasons it got killed (not exhaustive):
One - Not all players were interested in reading all the dialogue required to make the good/evil choices meaningful. Two - Players often identified often as âalways goodâ or âalways badâ, so you didnât end up with a game with a lot of nuance, you had players just pushing more and more towards one side or the other. This was fine for a single playthrough but not great for repeat playthroughs
Two - Players often identified often as âalways goodâ or âalways badâ, so you didnât end up with a game with a lot of nuance, you had players just pushing more and more towards one side or the other. This was fine for a single playthrough but not great for repeat playthroughs
Three - We tried to tie mechanics to this system. For example - you must have 50 Demonic alignment points to unlock X skill or equip X item. This put player build choices at odds with the player fantasy. When your head cannon runs up against the build you want to play the type of player who plays Diablo a lot tends to prioritize the build they want to play. This causes uncomfortable dissonance as you make dialogue tree choices that arenât what you love
Now that was good read. I can only agree with the pros and cons of alignment systems. As cool as they are. HoweverâŚ
Three - We tried to tie mechanics to this system. For example - you must have 50 Demonic alignment points to unlock X skill or equip X item. This put player build choices at odds with the player fantasy.
That is not inherently a problem. It is no different from having to pick 1 skill over another etc.
Build choices are meant to be at odds with your desire to âget everything all at onceâ. That literally is what makes them choices.
The game being âhardâ was influenced heavily by the popularity of raiding in WoW. It wasnât until after ship that we realized the outdoor world (Act 1-4) wasnât the right place to put âsuper difficult contentâ
Makes me so sad to read. The best realized game worlds, especially semi-open worlds, are great exactly because they have all kinds of difficulty mixed together. It is the same argument as with the 150 difficulties. Running into a challenging mob walking through the area is fun, it is cool. People remember getting destroyed by the Fel Reaver giant in the first area of WoWs TBC expansion etc. The idea that everything must feel the same is so weird 
And this is not about whether the content is easy or hard, at all. It is about the lack of diversity in the challenge, the ebb and flow, the pacing.
Moving the âhard contentâ into GRifts creates the exact same issue as having the easy content in the overworld. That âhard contentâ in a GRift is also constant. Boring.
We realized the extreme difficulty scaling of Inferno wasnât making the game more fun. The fun comes from running the content efficiently and repeatedly, not from overcoming a very difficult challenge once. (both can be fun, just speaking in broad terms here) We quickly realized after launch that players had beaten Inferno, it just wasnât fun to do.
Glad he says that both can be fun at least. And with that clarification, I do agree that an A-RPG should focus somewhat more on the former. But again, diversity is good. Mix both of these together in the same overworld. And in the same dungeons.
Anyway, beating Inferno at release was more fun than anything in D3 today. That doesnât mean it was great or anything. Just less bad.
Conventional wisdom at the time was that repair costs keep your gold economy in check. This is false, and good economy design is much more sophisticated than that. (and the game industry as a whole has advanced collectively in this regard as well)
Yes! It is difficulty to name a single game ever created, that had a good gold economy. That is difficult to achieve (but should of course always be attempted). Repair costs are rare more than a silly annoyance.
I wonder why he believe the game industry has advanced on this issue. It seems as bad as ever.
- Yes. That said for a game like Diablo 3 supporting mods is a HUGE commitment and risks fragmenting the community among many other technical and design challenges.
Interesting that they (according to Wyatt anyway) considered mods. Mods would have been great. The risk of fragmenting the community is a complete non-issue, based on the false premise that the community is worth preserving in the first place.
I donât know if thereâs a single /most/ critical thing but Iâll just throw out something meaningful. Players donât like it when things get nerfed, and they like it when things get buffed. Everybody says they are okay with nerfs until it happens to your beloved build. Even putting something on PTR (public TEST realm) and then ânerfingâ is seen as a nerf and hurts.
Okay⌠but the players are wrong, and a responsible game designer should not give players what they want in this regard.
People might dislike nerfs, but you do them anyway, for the betterment of the game.
Drop rates were never throttled low because of the AH. Drop rates were designed to be fun for a player playing solo.
Hereâs a thought experiment. Choose any drop rate you want. Now introduce an Auction House that puts the player at the 99.99% percentile of loot.The problem is only solved by adding friction to trade or removing it completely.
While I honestly dont buy that droprates were not affected by AH, he is 100% right about his point and his example.
And yes, the solution is immensely simple; do not ever allow trading.
- Some player types donât like doing math, they just want to pick what sounds cool. Attack, Precision, Defense and Vitality do not sound as cool as Str/Dex/Int/Vit
Maybe dont design A-RPGs for the players who dont like math?!
It wont even matter, no matter how much you dumb down the game, those people who dont want to think about decisions will google an answer anyway. Nothing was gained trying to design the game for them.
(now, Attack, Precision, Defense and Vitality still sounds like a bad attribute system, so not arguing in favor of that, however names are just names, if you genuinely think Strength sounds cooler than Attack⌠then call the attribute StrengthâŚ)
The problem is when you have stats the player invests in and grow in value for a percentage effect you have to run them through a conversion formula. Precision was effectively âCritical Hit Chance Ratingâ and Defense was effectively âDamage Reduction Percent Ratingâ. Ratings are harder to understand and value as a player. When a player is looking at 2 items and thinking âWhat do I want, Attack or Precision?â or âDo I want Vitality or Defense?â youâre left scratching your head and the answer is always âIT DEPENDSâ. So we scrapped that and pivoted to the system you know today.
Oh noes. The horror of having meaningful decisions where the outcome depends on the inputs. Totally understandable that such design was scrapped.
/end sarcasm
Itâs like triage at the ER. Sure players are farming pots in Act 4, but these other folks canât even log in. How do you juggle these priorities?
2 Likes
I have always liked him; he seems like a down to earth guy. When some of the people said they wanted him dead after the Immortal announcement I was like, chill out dweebs.
Exactly some players in D3 do not even care about defense that might be why they die ever 30 seconds. That is why I created a thread some time ago about this for D4. I wanted to know if death should have a penalty in D4. It appears that many players do not want any type of a death penalty in D4. They do not want to be punished for doing stupid things. They typical response was âgo play HC if you want a death penalty.â In otherwards just remove the defensive skills, passives and gems from the game altogether. Because some people never use them and dying all the time does not matter to them.
1 Like
Or donât have spells at all. Ez. Not that that would make PvP fun in an Arena, but in its own open world survival mode it will work.
And it should be fun, not efficiency, when in a non-competitive dungeon. Thatâs why D3 is a ghost town.
Having perfectly adjustable difficulty is better for content like Rifts that you repeat over and over daily. For open world having a few difficulties/zones is better. For bosses and challenging content it needs to be earned aka GR steps/levels.
Itâs interesting to read this manâs thoughts on D3:Vanilla. On some things, heâs very much correct, and on others, he gets things critically wrong. As I read through, my impression was that he was trying to make a game people would like. In the end, he ended up designing what is effectively a drug: a highly addictive interaction designed specifically to keep feeding dopamine to players in order to boost sales and keep them playing.
The paragon system was a perfect example. He says he didnât like the D2 finite level system because it didnât provide enough reward. It took too long between levels, and when you did level up, you were afraid to spend your points for fear of making a build that was suboptimal and ruining your character, so youâd bank them, thus delaying your reward even further. So, they came up with paragon to keep feeding you rewards even after you hit âmax level,â so you kept feeling rewarded for playing. Instead of simply allowing there to be a point where you had accomplished the goal, he put players on an infinite treadmill. By refusing to allow a player to âbeat the game,â he kept them playing and kept them happy by giving them more power the longer the played.
Now, to be clear, I donât think this was nefarious. It was done in the interest of keeping the game fun - a designerâs imperative. But, like the doctor who overprescribes pain pills, he either failed to think about the unintended consequences or ignored the risk. So, he ended up encouraging obsessive game play and botting. And he turned the entire leveling process, that whole aspect of the game into an extended tutorial - a chore to be completed before playing the real game which started at max level. It was a terrible design decision driven by the cowardice to say ânoâ to players.
We see the same mind set with the choice on âalways buff, never nerf.â In fact, he still defends the core decision that touched it off. He had a situation where an off-by-one bug ended up shifting a decimal point, creating a 10x too powerful bonus. Itâs quickly exploited by players because⌠itâs 10x more powerful than other options. Instead of fixing the issue and restoring the balance of power, he doubles down on it out of fear of angering players who had potentially paid real money to buy broken items to exploit. He was afraid of not being liked, so he started down a path of power bloat that now has players literally capping 64-bit variables. Theyâd have to reprogram the game from the ground up to accommodate more power growth - all so the player gets his dopamine reward.
From all accounts, Mr. Chang is a nice guy. Heâs clearly intelligent and capable. But, what he seems to lack (at least at the D3:V stage of his career) is the self-assurance and discipline to say ânoâ and stand by it. Like a parent, this game was his baby, his child to raise and nurture. Sometimes that means disciplining the child and not giving him what he wants when he cries or throws a tantrum. It requires a lot of personal integrity to make a âtough loveâ call, be honest with the reasoning, and stand behind it knowing that some of his players are not going to like it, while hoping that with time, they will respect it. Itâs a level of discipline that I think any professional in any career needs to learn and practice, IMHO.
1 Like
D3 because the team had problems to balance the complex interactions between skills, skill runes, legendary items, and visual noise.
Visual noise eventually completely went out of control, lol.
And it seems D4 devs are struggling with balancing too, taking into account terrible d3 like items and insane dps numbers shown already.
1 Like
Actually, if you look at D3 Leaderboard, those players were actually care about defense. They even swapped their normal main stat gems to defense stat gems when pushing high GR.
Players only care less about defense when they can kill things before the things strike back, hence the best defense is offense or strike first which is something very common in regular contents such as overworld zones.
I canât read the interview as WoWHead is blocked where I live, and seeing this quote here made me smirk. They gave the same reasoning for WoW Tokens.
D3 is in Classic Game status now.
The fact Blizzard cares enough to keep working on D3 is worth kudos when they only have to do bug fixes and basic patches. They could just focus on DI and D4 but they are still doing developments in D3.
Thatâs my take on it.
If you want to read the whole thing, you can read the original on Twitter. He was answering questions people asked him there.
WoW raiding in regards to bosses with phases, and such. Definitely was something for the better.