Pre-announcement Discord

2020

— 02/19/2020 5:04 PM

Do extrinsic rewards hurt intrinsic motivation to play? https://www.gamasutra.com/blogs/FabianFischer/20200218/357192/Reward_Fatigue.php

Raph Koster: The science is pretty clear in giving a resounding "yes" to that question.

— 03/03/2020 1:28 AM

But they were always the top Raiders

It's fine by me if only those ambitious guilds get to be top raiders. I just wish that cozy guilds wouldn't just get stuck completely with no chance of ever progressing further.

Raph Koster: Raids require high trust teams and therefore high commitment, because of the nature of the game challenge offered: a very large coordination problem with lots of "blind passes."

This by definition makes it a less accessible way to play, and limits the audience.

I very much dislike the way in which the actual game has become prologue, and favor elder games which are driven by lower trust dynamics.

(This reply written in the context of https://www.raphkoster.com/2018/03/16/the-trust-spectrum/)

— 05/16/2020 7:41 AM

Imho the question is, how much you can split the community up, without stopping to be an MMO. What makes Ryzom Ring stand out to me is that you could travel from world to world with a single character, that character progression would carry over, and that players from all those different worlds still participated in one shared economy.

Raph Koster: The risk there has always been that character transfers and the ability to mod would imply that one server can destroy the balance of another. One server with a mod where the admins enjoyed Monty Haul gameplay could create dupe bugs or infinite gold or whatever, and then it spreads to the others. One where they enjoy different gameplay could add an item which makes other worlds easy... and so on. And that's the benign version.

In designing more metaversy stuff, this was always at the forefront. Second Life had lots of problems with "gray goo" attacks from sim to sim, early on, and of course, the whole flying penises thing made possible by robust creative tools.

Blockchain worlds have had similar challenges, where they bump up against the fact that their distributed trust models means you can basically inject code into someone else's server... which might be anything. Saying "it's up to the destination to police things" is a failed model.

In Metaplace, we locked character stuff and currency stuff to central control, but allowed modding everything else. This allowed you to move freely between worlds. But it also meant you couldn't move objects from place to place unless they were part of the master, authoritative set.

— 06/06/2020 2:16 AM

i dunno if you will ever compel "machines" (i call them that beause they blow through games like machines and then get bored and move on) to slow down and "smell the immersion." i live with one, so ill ask

Raph Koster: I don't think it is about slowing down the goal-oriented players. It's about them not always dragging the "smell the roses" types along through peer pressure, level separation, and so on.

— 06/10/2020 5:43 PM

Doesn't Camelot unchained have a system sort of like what you're suggesting? I'm pretty sure crafters can sell their wears to their respective factions to help their side. FF14 also has the Ishgardian restoration event where crafters and gatherers were rewarded for contributing their items.

I think both are pretty neat idea's. I don't think i'd want to go anything beyond that though, I think player run economies like Albion Online has are much more fun personally. If I know a new dungeon is coming out and there's fire in it, I can mass create fire potions. It kind of fills that PvE crafting experience you're talking about lethality, only it's people not robots.

Raph Koster: I do want to say that AI buying goods is where MUDs started.

It basically kills a free market economy, because it is effectively government subsidy -- it means we mint money into the system on demand when players craft and sell to the NPC, and the NPC price for purchase is effectively a price floor for the good, which means that there is a guaranteed buyer with a minimum price. This results in prices never crashing from oversupply, and leads to more grinding not less

We tried things like having floating prices based on supply in UO, but then players griped about betrayed expectations -- when the shopkeepers started to refuse to buy the grinded useless items, the players got very upset and filed it as bugs.

[...]

So... yes,if we managed to build AI economic actions that acted just like players, it would work. But the other choice (and what, say, EVE and SWG picked) is instead to REMOVE the NPC actors.

Since making AI that acts human is a huge unsolved problem, but getting humans to act like humans is easy... it also means that you give more ways to play to the humans. Back when there were NPC shopkeepers, there were no human shopkeepers, crafters, delivery people, businesspeople, etc. The AI took all that Gameplay.

— 07/17/2020 3:56 AM

Raph Koster: You will all be glad to know we just had a major breakthrough with a crucial significant feature that I am not going to tell you about. But we are all signing off from work all psyched up today.

Well, we managed to get [redacted] to export [redacted] in a format that the [redacted] could read, and now we can [redacted] in [redacted], and it, like, totally validates a crucial game pillar and proves a bunch of experimental technology.

— 07/29/2020 2:15 AM

I have been looking for something as daring as UO and SWG ever since and most game developers of today seem to just follow a boring risk free pattern. I am rooting for you to break out of this.

Raph Koster: I think it is safe to say that you will see many elements from UO and SWG in the new game.

— 07/29/2020 10:50 PM

What do you think about including professions that need real life hard skills? Let's say you have a world that has a lot of ruins, ruins from an ancient civilization (think dwemer in elder scrolls or whaatever). You might be able to learn something from those ruins. Let's say there are some kind of stone tablets with notes on them, like hieroglyphs or whatever. Now to read them, you'd have to translate them. You could do that by "clicking on it and analyizing it with a specific skill" or you could just let the player figure it out on his own. Like we did with the real hieroglyphs.(Maybe even some kind of "long time hardcore puzzle like cicada 3301 - but ingame) Another example might be astronomy. Let's say there are planets and you could craft some equipment to observe them. Maybe you know from ancient text that if the planets align, that you get a buff. You could now try to observe them and calculate their distance (think of Aristarchus of Samos) - so you would actually write stuff down in real life and try to calculate stuff. Kind of like you'd do with theory crafting for stats but this time you hunt knwoledge about the world. I think you get the idea - did you ever experiment with something along this line?

Raph Koster: Yes, we did some of that as event stuff in SWG. I think the answer is a mix. Any time you have real world skill as a requirement for something, you are trading away audience. Of course, almost everything requires SOME real world skill -- it's a question of how big a barrier. Tab targeting has a lower barrier than action targeting, in terms of twitch skills... but it's actually less accessible to the average player (since the average player these days uses a phone or a console)... those are both kinds of skill, kinds of prior knowledge

with a sandboxy world, you can have some activities call for one sort of skill, and some call for another

[...]

Yeah, the kind of thing I was getting at was about real world skill versus RPG skill. RPG skill lets you do things that you have no ability to do in the real world, it's all dice rolls and stats. Systems that call for real world skill gate out those who don't have it.

Even RPG systems have SOME skill needed -- at least the better ones -- but there are always real questions of balancing real skill versus fulfilling a player's fantasy. [7:19 PM] A classic example of this is crafting systems, players have often disliked very active crafting because the type of player who likes crafting uses those systems as meditative breaks, peaceful low stress play. Active crafting works against that some.

— 09/26/2020 2:24 AM

Its old news but there is a weird connection between Playable Worlds Singularity 6 and Build A Rocket Boy Companies. They are all creating this MMO environment but have something in common which is allowing broad number of players to choose what they want to be or who they want to affiliate themselves with. @Raph Koster What’s goin on Raph ?

Raph Koster: Basically: the whole industry is moving towards games as a service, and lots of companies are taking notice. A lot of the established companies don't know how to do it. Retention is king in that world, and devs who have only made consumable content don't know how to design for that. MMOs are "hot" again for investment because investors know they are the original GaaS.

As part of that, "Metaverse" talk has bubbled up again, so there are investors who are interested in things that point towards that too... that's part of why you see people putting money into Manticore, for example.

Sandboxes in particular drive interest because people see that the casual and mobile market has hugely expanded games. So lots of folks are thinking that maybe sandboxes had the right idea, 20 years ago, to have more ways to play.

So, not as much about datamining, per se. But very much about long term industry trends.

— 10/24/2020 4:14 AM

Also what's the benefit of cloud when local computing may do something like this if I'm not mistaken? [link to UE5 demo]

Raph Koster: Cloud computing architecture is built around a few ideas that are a little different from just "host a server"

The biggest is launching and shutting down servers on demand, enabling what is called "elastic compute." Basically, it lets you add or remove computing power on the fly.

Devs currently mostly use this to add or remove individual game servers

But it gets more interesting when you pair it with other cloud features

One of the biggest is "microservices" -- the idea that instead of a monolithic server, you instead have smaller, more dedicated servers that do different thigns

Say, an AI microservice which handles the AI, can spin up more AI processes when needed, shut them down when not\

A related thing is that there are cloud hosts that are better at memory, others better at raw CPU, others which are basically GPUs in the cloud

Historically, game servers are built as one blob

If you architect your game to be cloud native, that means trying to take advantage of all this

That does not mean better graphics. The VAST majoreity of that UE5 demo is about graphics

Instead, it means better server capability

AI. Proc gen. Dynamic worlds. Persistence so that what players do matters.

Basically, doing things that were the daydreams of games 20 years ago but which have mostly not shown up

[...]

Scale of people in one place is ONE way to try to use the cloud. It's frankly the most boring and pointless way

[...]

Well... think of it this way. Back in 1997, with UO, I tried to make a parallel world. Every creature was going to get hungry, seek out shelter, have loves and hates. We couldn't afford it Players could build homes tile by tile, and a single house had -- HAS -- more interactive objects than the entirety of a Bttlefield or Fortnite match. But they couldn't chop a tree down, because we couldn't track that much data. Players could craft, but they couldn't customize what they made that much, because of the memory load. It had basic RPG combat instead of something more complex and active, because latency was bad. You couldn't dig in the ground, and mch of the world was static because it shipped on CD.

Today, the average online game is 100% static. Nothing saves. You can't build. You can't drop things on the ground, even. The creatures are just fodder. It's not an alternate world.

— 11/05/2020 2:09 AM

Having a formal, regular point of contact with the community like DE's devstreams would be amazing

Raph Koster: Our plan is to be very open, including posting up design docs and stuff for you to pick apart

I mean, we are making our idea, but we want it to be something you actually want, so we want you to vet it all

[...]

It is how I have always run my games

In UO we had weekly IRC meetings with the player base, we had every new feature go up in advance in "In Design" for discussion

In SWG we had weekly article updates on features, dev interviews, and players picked as reps for every class/role in the game

Nobody remembers these days, but UO basically invented "community management"

- 11/20/2020 2:22 AM

at some point you’re just recreating real life, then. wheres the game?

Raph Koster: The fact that people can play in different ways doesn’t mean each way needs to be tedious.

[...]

The key part that we see as crucially important is that people come to value the importance of the different ways to play. The warrior ought to value that crafter who makes their great weapons. The crafter should hold in high esteem the explorer who finds rare materials etc.

— 11/22/2020 12:36 AM

i'm not familiar with the backstory of the changes that made death less catastrophic in UO (maybe @Raph Koster was in favor of them?)

Raph Koster: UO actually had very NON-catastrophic death. Because items were relatively easy-come-easy-go, loss of gear wasn't crippling. Because UO allowed near instant travel all over the map, it wasn't that big a deal to be far away, either, not once you had a decent rune library.

What UO had was emotionally abusive murder which is not the same thing. Death itself was mild. It was the emotional attachment of having a griefer dance on your corpse or turn your pets into pies while you watched that caused so many of the problems. Dark Age of Camelot got pretty far simply by blocking chat between killer and target in RvR combat.

I didn't favor the way in which playerkilling was removed in UO -- I thought that a flagging system could continue to evolve and handle it. And eventually, SWG's and then WoW's (which is nearly identical to SWG's) did in fact do OK with the issue. By the standards of MMOs during the EQ era, WoW IS a "full PK" game.

— 12/11/2020 8:09 PM

That probably depends on what you consider a great MMO in the first place

Raph Koster: You don't ask easy questions eh?

A virtual world is a place. We put games in it. So first, a great place, I think.

Next, the games people like depend a lot on who they are. Very subjective.

But to my mind, that means offering a spread of things for people who play different ways.

These people may not realize it, but the world is better with different types of players in it.

The world needs to bring people together,. not drive them apart --not forced, but naturally, as people realize that we are better together, just like in the real world.

So it needs to not get in the way of that, basically.

Lastly... it's an alternate world, not our reality. People go there to be people they can't be, somewhere they cannot actually go, to do things they cannot actually do, with other people.

So it needs to find ways to enable people to reach for those other selves and other experiences

I think there are a lot of ways to accomplish the above... but it's funny how many games lose sight of some of those principles, I think

Excessive grind, no focus on fun, too much focus on numbers, too much emphasis on separating people, too narrow a play range, too boring a world... it easy to fall into some of those traps

And some player desires contradict others, of course

Bottom line, for me, though, it that it needs to feel like another world you fall into. Not just a game.

— 12/11/2020 8:23 PM

how do you feel about NPCs? What should their role be in a virtual world that's primarily intended to be player run?

Raph Koster: On LegendMUD, we lovingly handcrafted every NPC.

On UO, we built as robust a broad AI as we could to make them generic.

On SWG, we avoided having them at all if we could.

I think they historically serve two needs: one, to make the world come alive; and two, to give quests and tasks for players.

If the world is alive enough in other ways, SWG showed you can go really far without needing them as much. But I think the quest terminals replacing the quests were a fail.

Too dry, basically.

These days, it's been shown plausible to have players giving each other missions, so that opens new areas to explore

We also have much more capability to do large scale AI a la UO, so that opens other, different doors

The one thing that has clearly gotten less and less tenable is lovingly handcrafting them. Some pull it off, but it's obvious it costs a mountainous fortune