Original lead designer of the game here.

Richard's recollection is flawed on this (he wasn't close enough to the tech to know why it failed).

Reason #1: the closed loop fell victim to hoarding, particularly because the advancement system encouraged the production of goods with no value, as explicated in great detail in Zach Booth Simpson's detailed paper on "The In-Game Economics of Ultima Online."

Slides: https://studylib.net/doc/9778818/the-in-game-economics-of-ul...

Paper: https://web.archive.org/web/20090207141813/https://www.mine-...

Reason #2: the cost of pathfinding for the searches was too great. The Sims later solved this by having the agents in a very similar system use simple hillclimbing, with resources broadcast into a map.

wow, this is why I love HN: An incidental thread on a subject sometimes ends up having a key figure on that subject joining the discussion. :)

I never really got to try Ultima Online during its heyday (I was averse to MMOs and saw them as a step backwards from the narrative qualities of single-player games.)

Re: the ecology, other commenters here and on YouTube have also noted that it was too easy to do things what you shouldn't be able to do, like chasing down rabbits and deer with a sword. :)

Thanks and looking forward to your next game!

>because the players were human

It's a very compelling story, but that wasn't actually the case. The real reason was the flawed design. People just had to kill everything for the materials because of the way equipment was implemented.

I think one YouTube comment named another good reason: The animals were too easy to kill.

"The problem is obvious. The animals were too easy to kill. Imagine trying to chase down a real rabbit or deer with a sword. You will never catch either one and if by some chance you corner them the deer would actually stand a chance of beating you."

Mmo designers are very constrained in their game mechanics because they always have to deal with players that will exploit every possibility to grief. This unfortunately leads to things like extremely unchallanging encounters for random matchmade groups as any group will have players trying not just to get a free ride but actively trying to get the party to fail the objective.

As best I recall (it seems it's not easy to find sources now) the biggest problem was with player-killing and griefing. The devs left player-character interactions relatively unrestricted in the expectation that social order would be (largely) maintained by the in-character actions of the PCs. Similar things had worked in text-based MUDs of the time. When they naturally got a Hobbesian bloodbath instead (classic MUDs have many more hours of active games-mastering per player than any MMO) they began the retreat that led to the fake-roleplaying, theme-park experience familiar from World of Warcraft etc.

https://gametyrant.com/news/ultima-online-spent-3-years-deve... has some quotes from the creators.

"Similar things had worked in text-based MUDs of the time."

Not really, in my experience. In MUDs with unrestricted player killing, in-game policing did very little to stop it (unless you count intervention from immortals/wizards, I.e. admins)

Sounds like the game was perfectly realistic then.

That was in 1997, a time when nearly all players were using dialup modems and server hardware was pitifully slow compared to today's. The attempted ecology was no doubt very simple.

UO's original designer, Raph Koster, recently announced a new MMO project where he intends to revisit many of those old simulationist ideas.