Cain on Games - Tim Cain's new YouTube channel

[Tyranicon]

Endgame slides should compliment you as a person and give you a BJ.

 

[NecroLord]

Endgame slides should compliment you as a person and give you a BJ.

That one of your endgames in your RPG?

 

[Infinitron]

I talk about all of the sections and subsections that you would expect to see in a complete game design specification document.

 

[Roguey]

Disagree with Cain's take that a game should have a number of acts in-between Fallout's three and Arcanum's 27. Fallout having only three critical goals was great. I like having an RPG where 90% of the content is optional but it's a good idea to do enough of it so you have the equipment and experience to handle what you're required to do.

 

[StrongBelwas]

Cain is using a fantasy RPG as an example
Setting and then story and system mechanics. The setting should describe the world the game is set in, and what the major areas are. History of the whole world, particularly where you have the game in.
The story is the main acts you are going to go through, Fallout has three acts (Find Water Chip, Stop Mutants, Stop Master), Arcanum has 27 acts. Cain feels the best number is somewhere in between.
Should also detail key characters that move story along, provide most of the main story quests. You may consider some Essential and make them unkillable, in Cain's games he prefers to just consider Essential to mean you have to prepare a plan B if the player kills them (Side NPC steps up, you find their journal, etc.)
Then you go to the system mechanics, which will need many sections and subsections.
Character System Mechanic like attributes skills spells feats backgrounds races flaws, etc. You probably won't have all of those, but you'll have a bunch of them, and you'll need to describe what they are how they're obtained, how they can be raised during the game, expected range, and whether you can respec.
Than you give a mechanical idea of what character construction is, what the player is given to select, do they have points to spend or do they roll, then you consider the exact mechanics of what happens when the player levels up.
Consider all the status effects, some of them could come from perks or hazards in the game. Could easily be 4/5 dozen of them, Cain has had games with over a hundred status effects. Must be exact when describing these and how they interact with other status effects.
Then comes inventory, encumbrance, different equipment slots on the player and NPCs.
That leads into combat, which is going to be a lot. Is it real time, turn based, what is the combat like on a high level, what form does it take (turn based, real time), the rules for hitting, missing, critical hitting. Melee vs. ranged, are the range weapons projectile or hitscan. Have to go into excruciating description. What kind of armor is there, how many armor slots are there.
What creatures are in the game, what kind of abilities do they have, what attacks and status effects could they use.
Then another big section is dialogue, have to explain exactly how it works, how you want it to branch, what restrictions there. Cain personally restricts it to no more than 5 nodes displayed at any one time. How do skills work in dialogue, how NPC reaction affects dialogues.
Now in stealth section, how hiding works, does it work with lighting and/or cover, how NPCs detect you, how does picking locks work, what does a stealth path for a quest look like.
If you have companions, you're going to need a whole section for that, how players get them, how to keep the companions when they leave, how many there are, how many can you have on the same time (Does it depend on skills/abilities/perks?) Do some companions have restrictions, how do they interact with the main quest and what are their personal quests if any.
Economy gets it's own section. Loot, loot tables, vendors, probably want to throw in a crafting section with everything that entails.
Going back to areas in the game, now you consider how big they are, what towns are there, what dungeons do they have. What encounters are guaranteed, what kind of random encounters could you find.
User Experience section can be broken up into three parts. Pre-Game, Out of Game, and In Game.
Pre Game would be anything the player sees before actually playing. Main menu, how that is designed. Consider options such as visuals/audios/accessibility. Some of these you're going to want set before game, others are set during the game (Cain calls this out of game, such as game saving.) Have to be careful because some options apply to every game, others apply the options just to the game currently running.
Than you have to consider all the in game UI. Level up screens, journal screens, inventory screen, all the works. You'll need looting, bartering, local maps, world maps, the notifications the player gets when they are set on fire or damaged.
A few considerations, some of these may need their own sections, maybe you could attach them to the other sections.
PC Vs. NPC. Some attributes may work the same, others work differently. Can NPCs knock out players like players can knock out NPCs? You may not want the player to fly, but maybe you have dragons or other flying enemies and you can control where they can fly. Don't have to worry about a dragon flying into a house, but if you let the player fly that's the first thing you need to consider.
Considerations for mutliplayer. If you have multiplayer, some game features may go, maybe you don't want the turn based option in multiplayer. Maybe in multiplayer people don't make custom characters like they do in single player.
More recently, Cain has begun implementing Design Fallbacks. These are more simplistic designs that can be used instead of your main plan if it turns you lack the time/resources to implement it. They used this several times in Outer Worlds. One of them was grenades, Cain wanted throwing grenades, but they could never get the arcs working right so they went with the Grenade Launcher you see in the game. That caused a throwing skill in the Melee section to be removed.
Fallbacks are important because you can rarely sacrifice something without consequences, i.e they lost grenades, now the throwing skill has to go.
Cain shows the Arcanum design specifications printed out, a meaty 148 pages. There is a digital version, but Cain believes it is still covered under NDA and won't upload it.
Codename for Arcanum was Epic, also what they called the post apoc demo.
A surprising amount of the design specifications actually are the same as the shipped game, there's a map of the game shown at 17:15 and it's pretty close to the final version. Designs are incomplete, lots of interface stuff at the end left unfinished. Was used to get the contract with Sierra, where they were told they had 90 days to make a prototype if they wanted to keep the contract, they used a small subset of that design specification to create the prototype.

 

[The Wall]

The gnomish quest was written by Leonard Boyarsky who is Jewish himself and not by Tim.

Gnomes Jews love to flex

 

[__scribbles__]

I talk about how to fix design problems, using an example of such a problem that shipped in Fallout.

 

[StrongBelwas]

Cain considers the problem of being able to tag Energy Weapons at the start of Fallout without getting energy weapons for a long stretch of the game.
First you have to notice the problem, they didn't notice the problem.
Vast amount of player builds you may play without noticing the problem. QA may not report it as a problem because they think energy weapons not dropping is working as intended. This is why Cain likes Telemetry, so he could see where energy weapons are placed.
Second step is confirming the problem is a problem. Think you've noticed a problem, but maybe it is working as intended. In the case of energy weapons, check if the energy weapons are actually not possible to get early or if it's just bad luck at the loot tables or vendor randomization. Repeats his wish for telemetry in Fallout 1, even with all of their QA fallout was a 40 hour game and QA could often only play the game once per build.
Now you go look at the design docs and consider if the lack of energy weapons is expected, and if it is a design goal that you can get all skills before they are useful. There were no goals in the Fallout design doc, Cain considers that a mistake.
Cain compares the lack of energy weapons to putting lockpicking in a game and not placing any locks into the game until the final level.
Now you reach the question of where to place the energy weapon. Still quite possible for energy weapons to not appear for most people if you place them somewhere hard to get like a drop from an NPC most people won't kill, or a rare chance to appear from a vendor. If you introduce a lot of energy weapons early on, you'll have to make them weak, which may also make some people question tagging Energy Weapons.
Another solution is to change your system design, such as locking the energy weapon skill until a certain level/stage, or energy weapon is a specialization of the general weapon skill you can put points into after getting a certain amount of points in the general skill, which is what they did in Outer Worlds. Wouldn't require changing loot tables or reitemizing the world
Fourth step which people sometimes forget is test the solution. Give QA instructions to make a note of when they find energy weapons and if the energy weapon skill is useful. Now look at all of the other skills and see if they have similar issues, are there enough traps, are Doctor and First Aid both useful, etc.
You can have a perfect design document implemented just the way you hoped for and still see the final execution and find it lacking.

 

[NecroLord]

Cain considers the problem of being able to tag Energy Weapons at the start of Fallout without getting energy weapons for a long stretch of the game.
First you have to notice the problem, they didn't notice the problem.

Good thing the Fixt mod, in a display of common sense and gentlemanly behavior, gives you a Laser Pistol if you have tagged the Energy Weapons skill.

 

[Alienman]

How is it even possible to miss that issue, and why would telemetry be needed for that? Let's add a very specific weapon unique to the setting, but somehow totally forget about the addition.

 

[NecroLord]

How is it even possible to miss that issue, and why would telemetry be needed for that? Let's add a very specific weapon unique to the setting, but somehow totally forget about the addition.

'Muh Balance, right?
LIke I said, good thing the Fixt mod, even with some of its faults, gives you a Laser Pistol. It's not too bad, but also not great, just the right starting weapon for an Energy Weapons focusing character.

 

[Alienman]

A simple solution would have the overseer ask you for what you want before leaving. Surprised that it was never brought up apparently.

 

[NecroLord]

A simple solution would have the overseer ask you for what you want before leaving. Surprised that it was never brought up apparently.

 

[Roguey]

More from Tim. I don't see the problem in having a delayed gratification skill you can tag at the beginning that will pay off much later (and it certainly does pay off).

The only change I would make would be to change the description to note that energy weapons are incredibly rare.

 

[NecroLord]

 

[Alienman]

Talk about circumventing both the weapon and ammo issue.

 

[NecroLord]

Tim should've also addressed the issue of Laser weaponry being nearly useless against anything better than leather armor.
Really, you have to rely on critical hits in order to do noticeable damage against armored enemies (Metal Armor, Combat Armor, Power Armor, Super Mutants).
Compare with Plasma which remains consistently strong against pretty much eveything the game throws at you.
And those Plasma death animations...
Just awesome (and gruesome).

 

[StrongBelwas]

Thinking about writing a book on game development, without the memoir stuff that would aggravate people (Don't hold him to this.)
Starting at Cybron. Wishes they had a logo, he has a Pegasus shirt but they changed their name right after. Shows a few high school pictures.
Moving onto Interplay, a 1994 picture of everyone who worked there, and then another picture of just the programmers. Then a 1997 picture of the Fallout team, a picture of a super mutant head and another head he thought was Morpheus but apparently wasn't sure. Shows a picture of Nicholas Kesting showing up to Cain's office to get some bread Cain liked to bake and bring into the office . The picture used to explode his head in the credits was one taken of him talking to T.Ray Isaac. Final picture is actually not during Interplay, but a 2017 reunion of Leonard, Jason, and Cain for Fallout's anniversary.
A picture of Cain at his desk at Troika, Cain liked to carry a picture of Batboy around and say it was his second grade picture. Another picture of Cain with a bag head with Tiffany Chu, QA at Troika and now a producer. The first day Cain brought his dog to work on September 11th 2000, they were considering throwing a one year anniversary for the dog in 2001 but that did not happen for obvious reasons. A picture of some of them at an Arcanum press event in the Magic Castle in LA. A picture of probably the entire Temple of Elemental Evil team. Some pictures of him at his desk and the Wednesday D&D games, Cain tried to teach them 3.0/3.5.
Cain used to find almost dead crawfish, would try to keep them alive in a tank, one called Pinchy managed to get out and nobody could find them, Daniel Albert made a milk carton to shame Cain. They'd eventually find Pinchy dried up and dead.
At Carbine, a picture of his desk, he had a lot of decorations by this time, including a very large Arcanum cover poster. Went to Halloween as Charlie Brown, next year went as a pimp. Picture of being interviewed at his desk by G4, he's pretty sure this is what made it into the Fallout 3 making of DVD.
For his 2010 birthday they covered his desk in candy, and glued candy to the walls and the ceiling. Carbine was a fun time and had it's moment.
Obsidian Entertainment, B&W picture of him at his desk again, then a picture at his desk of Dan Spitzley and Robby Adatero labelled 'programmer meeting' of them playing with puppets. Reproduction of the wig photo at a 2018 Academy Awards party, same wig and statue, a bit older Cain. The Tim Cain thumbs up picture, which got turned into a meme at 'various places'. A plushie of a Cystpig Cain made for the first Outer Worlds game pitch, they had to go to New York in the winter of 2017 and he took it with him, didn't have room for it on his carry on, their art director had to bring it with his carry on and explain it to people. The Cysts are removeable, he ended up giving it to a Take Two producer. When Take Two had their ship party, they had kept it around and displayed it there. When they got the Moonman helmet, Cain snuck it home and took a few pictures in the day of Moonman's day to day life. Brought it back monday with no one the wiser.
Picture of almost everyone on the Outer Worlds team, he is not the fellow on the right.
A big chunk of the team came to the Game Awards when Outer Worlds was nominated.
Some pictures of the conferences he started doing, GDC 2012, Reboot in 2017 (T-Ray Isaac happened to also be doing a talk there.) They did a round table, but he can't remember what it was besides they laughed about it. Some pictures from his Australian conferences, including a round table with David Gaider. Went around to all the chocolate shops.
Final picture is Seattle when he was setting up Zoom/Teams for his contract work with Obsidian, his dog would come into the room and photobomb him.

 

[user]

Scenes from the life of a game developer, but for real:

 

[NecroLord]

Scenes from the life of a game developer, but for real:

Soon to be Avellone?
Nah, I think Chris is pretty loaded.
And Tim doesn't seem to do so bad for himself.

 

[ds]

Scenes from the life of a game developer, but for real:

If only - then they would actually be motivated to make something their audience likes.

 

[Alienman]

"Will code for frappuccino"

 

[StrongBelwas]

Very difficult to get narrative designers to stop thinking linearly. Most of them are used to writing stories in linear format like books/TVs/movies.
First red flag is when he is in a meeting with narrative designers and they say "The player goes here and the player does X." Cain questions them what happen if the player does something else there, or goes somewhere else entirely. Cain has had narrative designers respond to that with 'We will force the player to go there, and then we will lock them in that space. They cannot leave until we do the thing we want." Cain doesn't think that sounds fun to him, or most RPG players.
Cain tells them not to base the story on player actions, instead base it on player goals, and you don't care how the player does it. i.e, don't say 'The player goes to this goblin cave and gets this item', tell the player "I want this item." If it's not a unique item and the player can buy or craft it, all the better. If a merchant has been kidnapped, that's an even better hook. You can try and find out who kidnapped the merchant (i.e there were goblins arguing in his shop), when you figure out goblins did it, you can handle the cave in any number of ways, but what matters for the narrative is you come back to the quest giver with the merchant. Let the story unfold based on the goal, not the action of the player.
Cain hates the phrase Story Driven Game, which is why he kept calling Outer Worlds a Player Driven Story. Stories do not act, players act.
You have to have fallbacks if you go through with this. If you tell the player to get an item and it gets sold or lost, what do you do? Has to be considered from the moment you first start making the quest. Now, you can just implement mechanics to handle that. Quest important items can't be destroyed, but what happens if you apply that to the merchant quest? You get Essential NPCs, which Cain doesn't like, and he knows lots of the audience doesn't like. So, try and come up with an alternate way to the quest, with ending slides at the end maybe, if the merchant dies. Perhaps he was the only merchant who sold a certain item, or now the plot can't progress in a certain way. Players love this kind of impact and reaction to what happens, not a set thing that happens to every single player.
Don't tell the player what to do or where to go. Just tell them what the goal is and let them figure out what to do.
This can get very out of control very quickly though. Try and make it systemic wherever possible, not something you have to hand script. In the example of being asked to deliver an item, if it's something is craftable, you already have a system for getting it. If it can't be crafted, but you can put it as a loot drop, also a good and easy solution. If you really need a unique particular item you'll have to put it in the inventory of a guaranteed to spawn creature.
The thing that can be tricky is introducing a solution to the goblins and the merchants where the goblins ask you to do something or they are bribable. Now you have to write a dialogue, the one solution that requires hand scripting. Your time will be filled with making and debugging scripted solutions to your quests. Systemic ones can be tweaked and fixed easily. With hand scripted ones, each one is a unique situation, you fix one, you still have all the others to fix. Also, since dialogue gets localized early on (And thus locked in), you often can't change those if there is an issue. If you have unique scripted solutions that involve dialogue, you need to catch those bugs early, another reason to try and keep it systemic.
Stories that try too hard to force the player to do one thing or go one thing or solve it in a particular way make players feel railroaded. And at least in theory RPGs promise the player lots of options. If you force them to play a certain way, particularly a main quest, it's going to feel really bad. If that merchant rescue was a main quest and the only way through was to kill all of the goblins, that is not going to be good for a pacifist playthrough.
Setup situations with multiple ways through them, tell the player a goal, and then let events unfold. That alone can be good but enough, but keep a record of what the player does, and then have NPCs refer to those actions later on, and you have something very special.

 

[NecroLord]

Fallout reflects his mindset perfectly, I think.
Just got out of the Vault and you got 150 days left to find a Water Chip and save your Vault?
Nah, bro, you can go the fucking Glow if that's what you want.
Or do whatever else.
Also no "Essential NPC" either.
Arcanum followed some of the design principles of Fallout and had no Essential NPCs. You could exterminate entire cities just fine, though obviously there will be consequences.

 

[Quillon]

Very difficult to get narrative designers to stop thinking linearly. Most of them are used to writing stories in linear format like books/TVs/movies.
First red flag is when he is in a meeting with narrative designers and they say "The player goes here and the player does X." Cain questions them what happen if the player does something else there, or goes somewhere else entirely. Cain has had narrative designers respond to that with 'We will force the player to go there, and then we will lock them in that space. They cannot leave until we do the thing we want." Cain doesn't think that sounds fun to him, or most RPG players.
Cain tells them not to base the story on player actions, instead base it on player goals, and you don't care how the player does it. i.e, don't say 'The player goes to this goblin cave and gets this item', tell the player "I want this item." If it's not a unique item and the player can buy or craft it, all the better. If a merchant has been kidnapped, that's an even better hook. You can try and find out who kidnapped the merchant (i.e there were goblins arguing in his shop), when you figure out goblins did it, you can handle the cave in any number of ways, but what matters for the narrative is you come back to the quest giver with the merchant. Let the story unfold based on the goal, not the action of the player.
Cain hates the phrase Story Driven Game, which is why he kept calling Outer Worlds a Player Driven Story. Stories do not act, players act.
You have to have fallbacks if you go through with this. If you tell the player to get an item and it gets sold or lost, what do you do? Has to be considered from the moment you first start making the quest. Now, you can just implement mechanics to handle that. Quest important items can't be destroyed, but what happens if you apply that to the merchant quest? You get Essential NPCs, which Cain doesn't like, and he knows lots of the audience doesn't like. So, try and come up with an alternate way to the quest, with ending slides at the end maybe, if the merchant dies. Perhaps he was the only merchant who sold a certain item, or now the plot can't progress in a certain way. Players love this kind of impact and reaction to what happens, not a set thing that happens to every single player.
Don't tell the player what to do or where to go. Just tell them what the goal is and let them figure out what to do.
This can get very out of control very quickly though. Try and make it systemic wherever possible, not something you have to hand script. In the example of being asked to deliver an item, if it's something is craftable, you already have a system for getting it. If it can't be crafted, but you can put it as a loot drop, also a good and easy solution. If you really need a unique particular item you'll have to put it in the inventory of a guaranteed to spawn creature.
The thing that can be tricky is introducing a solution to the goblins and the merchants where the goblins ask you to do something or they are bribable. Now you have to write a dialogue, the one solution that requires hand scripting. Your time will be filled with making and debugging scripted solutions to your quests. Systemic ones can be tweaked and fixed easily. With hand scripted ones, each one is a unique situation, you fix one, you still have all the others to fix. Also, since dialogue gets localized early on (And thus locked in), you often can't change those if there is an issue. If you have unique scripted solutions that involve dialogue, you need to catch those bugs early, another reason to try and keep it systemic.
Stories that try too hard to force the player to do one thing or go one thing or solve it in a particular way make players feel railroaded. And at least in theory RPGs promise the player lots of options. If you force them to play a certain way, particularly a main quest, it's going to feel really bad. If that merchant rescue was a main quest and the only way through was to kill all of the goblins, that is not going to be good for a pacifist playthrough.
Setup situations with multiple ways through them, tell the player a goal, and then let events unfold. That alone can be good but enough, but keep a record of what the player does, and then have NPCs refer to those actions later on, and you have something very special.

how unlike TOW is to the examples he is giving...

it's sometimes difficult to take him seriously given how fucking crap TOW is