Good puzzles.

[Coleco_work]

The same basic methods that can take an arbitrary list of words and fit them together to make a crossword puzzle could take a set of monsters and fit them into a world location in a way that would make a good encounter (puzzle).Hmmm ... it isn't the words that make up the puzzle that make it "good," it's the editor that edits the clues. The same arrangement of words can result in many very different puzzles. You could not replace Will Shortz with a computer and maintain the NYT's quality.

Coleco

[MahrinSkel]

True, but with no offense intended to the people that do it, most of the encounters that are created are not (and do not need to be) of NYT quality or complexity.

--Dave

[Coleco_work]

What is "it"? Crosswords or encounters? And if "it" is crosswords, your dismissal that they need not be of NYT complexity and quality equally applies to encounters. Perhaps even moreso: the average MMORPG player is likely much less discriminating than the average crossword puzzle enthusiast (at least with respect to what constitutes fun). But that may just be the NYT crossword snob in me.

Anyway, that is tangental. My point was that just because you have better MOB AI, it does not follow that randomly-assembled groups of MOBs will make for a better encounter. It may be better than current encounters, but the better encounters will still be those assembled by humans. They will still be trying to construct a larger set of puzzles from a smaller set of tools.

But I do not so readily accept that better AI will necessarily lead to a "better" game. For example, will better MOB AI result merely in players taking a larger group to the encounter, or waiting to tackle the encounter until they have levelled. That is, if players simply "brute force" your improved AI, is the gameplay really better? It seems to implicate extensively a game's fundamental design.

Coleco

[Bear]

Originally posted by Coleco_work
...Anyway, that is tangental. My point was that just because you have better MOB AI, it does not follow that randomly-assembled groups of MOBs will make for a better encounter.....For example, will better MOB AI result merely in players taking a larger group to the encounter, or waiting to tackle the encounter until they have levelled. That is, if players simply "brute force" your improved AI, is the gameplay really better? It seems to implicate extensively a game's fundamental design.

Coleco

You are correct about some players using a brute force approach. However, I have often noticed that the more boring and repetitive an encounter is, the more likely players are to run shortcuts like brute force. In the unlikely event that a game could be designed so that each enounter was unique, there would be fewer people trying shortcuts or exploits.

Of course it would not eliminate this behavior completely, some people always care more about "winning" than they do about having fun. But when you have already bought from/killed/quizzed the same mob untold times in precisely the same way, the temptation to just squash the thing and go on with your life can get overpowering.

Bear

[mikedsc]

some people always care more about "winning" than they do about having fun

I have nothing to say about this, except that it caught my eye and threw me for a loop. I was very surprised. :)

Somewhere in these forums (I just found it today, but I forgot again. It's in the SOC archives, I believe), there was a mention of an AI NPC called Galatea. I remember playing with it back then and enjoying it.

Your article reminded me of it and made me go and look for it. :)

Here are a pair of documents the creator wrote:
http://home.mindspring.com/~emshort/whatsif.html
http://home.mindspring.com/~emshort/NPC3.htm

It's about NPC design via Interactive Fiction. Take a look.