Monsters' movement patterns

I thought about this while commenting the EQ2's video here below.

Have you noticed how in ALL mmorpgs ALL the monsters ALWAYS move just in straight lines? They aggro and run to you, or they flee, more or less randomly. In between there's not much.

One of the things that caught my attention while I was playing God of War is how all the monsters had rather complex movement patterns that I would find hard even to describe technically. Complex rotations, retreats, fast dodges. They all look rather "fuzzy". Not so easy to recognize and predict, in particular when you fight more than one at the same time.

That's another element that has significant role in that game and one that completely misses in mmorpgs: the movement.

And another that I would really like being developed more, both aesthetically and for gameplay (different movement patterns during combat).

Add it to the "realistic aggro behaviours", and mobs attacking in organized groups (unfinished post).

Think how much it would be cool to assault a goblin camp and have all those goblins start to fight in groups, parse the environment to take cover behind trees/tents as they fire arrows at you and while another small squad of three or four are running toward you to engage in melee.

And then you can work to "branch up" from a typical goblin mob to create a number of different variations, depending on the weapons and armor they use, their rank and so on. Instead of one mob type cloned everywhere, you would obtain a more organic environment that could offer much more interesting and deep gameplay.

This is again what the genre has still to offer. You just need to not stop at a very superficial level and "dig the myth".

Then again, there are technical hurdles to overcome. This goes along with the lack of "physicalness". The sense of contact, weight, solidity. In mmorpgs everything that moves is immaterial. You cannot reach out and "touch". You just move through. Phantom-like. This isn't just a limit for the emotes (cannot really "hug", for example), but also for the combat, where you never really feel an impact. Stuns and roots are as far you can get. The monster cannot, for example, grab your arm and toss you away, or jump on you and keep you blocked under his weight. And if you are disarmed you are only losing the use of your weapon for a certain amount of time, you don't see your weapon bouncing away and you don't have to jump after it to use it again.

I think next-gen games will have to start to delve more on those patterns, see what's doable and push some more the technology.

That's innovation too. Without the need to look at other genres or fancy business models to experiment.

Unrelated to topic

Unrelated to topic:

Could you stop using relative URLs for internal links? I prefer using RSS aggregators to read blogs, and when the links inside a post refer to... the aggregator itself... not so good. I'll understand, of course, if you can't or won't. Just requesting it.

Re: Monsters' movement patterns

I could but it's pretty pointless at this point.

Never thought about the RSS, I used them relative so that I could eventually move the site without breaking the internal structure. I guess that's pointless as well now.

Re: Monsters' movement patterns

I think your spot on with the more lifelike movement ideas. They would just take immersiveness one level further. And 'fun' too. Learning an enemies movement patterns in order to defeat them.

The big thing is bandwidth and cpu time. With an MMO everything is kept as tight as possible bandwidth wise, as is cpu processing time. (calculating 10,000 swings of a sword being more difficult than 10,000 dice rolls.)

I've seen some people looking at alternative server models for the MMO-RTS space and i'm sure the MMO-FPS' have to take advantage of a lot of trajectory caclulations, so maybe we'll see some crossover from there.

It would be interesting if people could take advantage of end user processing power, but while there's no way of preventing client side hacks, it's a bit tricky. I wonder if it's possible to get end users pc's to calculate for somebody else.... that way hacking it would only result in someone else being boosted.

Re: Monsters' movement patterns

I think the bigger problem isn't about the processing power but more the "responsiveness". The need to reduce the margin of the approximation.

Meaningful movement would mean having a very reliable connection. The movement itself shouldn't be too hard to simulate as you could already achieve a good result with pre-scripted behaviours, maybe reacting to some variables, like the position of the player, his actions and so on.

After so many years I think the first steps in that direction could be possible.

User processing power is already used on some occasions, but always for the "superfluous". For example the newer CounterStrike has the ragdolls and a physics system, but they are all completely client-side. The server tells the client when a character dies, it's then the client to simulate the "death", even if different clients will show slightly different situations.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.