Page 1 of 2

Monster spawns

Posted: 24 Mar 2007, 19:04
by Crush
This topic is for discussing the monster spawning on the new server.

At the moment the implementation is the following: A few maggots are spawned in the center of one map at server startup and when they are killed they stay dead.

This isn't of course the way it is supposed to be. To keep the game going monsters have to respawn somehow.

What do you think would be the best way to handle monster respawn?

Whenever a monster dies a new one spawns at a random map location?
Every monster has a fixed spawn location and when it dies it reappears there after a fixed amount of time?
Constant monster spawn? (would make maps overcrowded after a while)
Maybe even a complex biosphere simulation where monsters eat, mate with each other and create offsprings that grow to maturity?
Monster generators?
Or do you have an even better idea?

Posted: 24 Mar 2007, 20:31
by Quiche_on_a_leash
How about having the population of the monster dependant of how much that type is hunted.
The population of other monsters on the map could be another factor.

Maybe there could be a food chain between the monsters. (?)

The monsters could spawn next to a random old monster.

NICE IDEA!

Posted: 25 Mar 2007, 02:34
by rtoot3
DUDE! that real life thing where they have babies and eat sounds AWSOME! that would be really cool! and while your at it, maybe, just after a monster is born they're blind for a while. and if you kill the parents in time, it wont see, and when it does see, it will see you and think your its parent. then it would be like a pet and follow you around, and help you in battle. but, you would have to give some of the food you find from killing monsters to your pet if it got hungry.

Posted: 25 Mar 2007, 03:10
by Crush
You shoudn't only think about what sounds "cool" at first glance but also what the consequences for the gameplay are.

Let's say we go for the realistic procreation system. The only source of new monsters are existing monsters that produce offsprings. What will the players do? They will hunt the most rewarding monsters and decimate them. In the end they will eventually exterminate the monsters. At first glance it sounds cool to have so much realism that the players can influence the game world by eradicating entire species. But the result will be that there would be less and less different monsters to hunt and the game would become boring.

Posted: 25 Mar 2007, 09:04
by EJlol
I like to see that the same kind of monsters have a little bit other stats from each other. So a maggot maxhp could have a random number between 10~20 etc. Also I like to see that monsters can hunt each other. About the respawning I think it's better if you can create area's on map.

Posted: 25 Mar 2007, 11:28
by Pajarico
It was suggested somewhere to have fixed spawn points that in the maps appear phisically as nests for species such as spiders, maggots, rats... which tend to procreate inside holes (?)(yes, I'm not zoologist :wink:). I would add to the list those white fluffies thingies which are like bunnies so they could jump from a lair when its their time to spawn.

So, in brief, my point is that i would go for fixed spawn points for those creatures and think about other ways for others monsters. There shouldn't be a spawn method for every monster, though (don't think is neccessary).

For example, flowers could spawn on different places provided that, when you kill one, its pollen flies and falls in a terrain where there is water and sun, etc (don't need to be too realistic, just zones where they can grow like the prairies) Maybe map tiles could have propierties about what monster can spawn?

I hope this is useful to you, Crush. Maybe I'm thinking to much on the phisicallity of it and you where asking for something closer to a code implementation or algorithm. If you wanted the second I would go for something simple, at least at first, such as:

* A fixed number + % of killed monsters of that kind during the last minutes..
* The amount of species is mantained spawning more monsters.
* There should be a maximum, though.

This would have the problem that if there are many players killing maggots and suddenly (or in a short amount of time) they stop killing maggots, there will be a lot of maggots spawned and noone to kill them.

Also, and I know that this will complicate things, I would like to see an implementation of limited natural resources so if there are to many maggots they will start to eat each other. this could solve the above problem...

Regards.

PS: I suggested somwhere on the wiki a mating crossbreeding feature.

Posted: 25 Mar 2007, 13:26
by Avaniel
I don't think that the perfect solution exists, but here are my views on a couple of them:

Spawn locations

Random location
The mobs pop into existence, on a random location.
- Not plausible, that's not how it works irl.
- Easy to program, no extra work for the mappers.

Spawn points/zones
- Medium plausibility, especially when used like Pajarico's rabbit-hole example.
- When there are many players and few mobs, players start camping the spawn sites.


Mob generating systems

Fixed number of mobs
As one is killed, one is spawned.
- Easy to program.
- Hard (or impossible) to break.

Biosphere simulation
Monster 'birth rate is proportional' to the number of mobs on the map (and available food).
- Possibility for mobs to become extinct. Be prepared to tweak this _a lot_.

Proportional 'birth rate'
A maximum number of mobs is set, the birth rate is proportional to the difference between the maximum and the current number of mobs.
- Gradual curve, compared to the 'Fixed number' method.

Additional thoughts
If a variation is preferred to the 'fixed number' method, make sure that it can't break by player actions. There are other biological models then just the predator-prey system.

nests

Posted: 25 Mar 2007, 14:56
by rtoot3
Crush wrote:You shoudn't only think about what sounds "cool" at first glance but also what the consequences for the gameplay are.

Let's say we go for the realistic procreation system. The only source of new monsters are existing monsters that produce offsprings. What will the players do? They will hunt the most rewarding monsters and decimate them. In the end they will eventually exterminate the monsters. At first glance it sounds cool to have so much realism that the players can influence the game world by eradicating entire species. But the result will be that there would be less and less different monsters to hunt and the game would become boring.
oh yeah. sorry :oops: ok so i gues i dont want that idea, but...
Pajarico wrote:It was suggested somewhere to have fixed spawn points that in the maps appear phisically as nests for species such as spiders, maggots, rats... which tend to procreate inside holes (?)(yes, I'm not zoologist ).
if there were little nests or holes, that would work. and the players wouldn't be able to hurt or destroy them.

Posted: 26 Mar 2007, 09:12
by Peacemaker
to avoid camping & cheating, I'm for random spawns like at the old server. (because its easy to implement and works)

But the server should be as flexible as possible which means fixed spawns for certain quest or boss monsters also would be nice. (but thats not important at the moment)

A dynamic system would be nice but it will slow down the server. - And I believe only a few people would recognize it as "dynamic". So its not worth the work.
here is a interesting link for a dyn. system which could be adapted for TMW: http://de.wikipedia.org/wiki/Wator (not available in english :( google for wator or wa-tor)

Thats a cellular automat which simulates fishes and sharks in an ocean environment (for TMW: fishes=monsters / sharks=players). There are 3 end scenarios:

1. the system runs forever (not possible in TMW because the sharks/players sometimes will not be online -> that implies scenario 2.)

2. all sharks dies (or are not in the dungeon) and the fishes rule the ocean/dungeon (in TMW words that means that the players were not able to enter a dungeon because its overcrowded by fishes/monsters).

3. all fishes die or where eaten up (killed) by the sharks (players). This implies that the sharks dies too because they are starving. (TMW: the player have killed every monster. The Monster cannot reproduce themselves any longer -> the dungeon is empty.)

Conclusion: Its too difficult to develop such a system for a server which should run as fast as possible. And specially scenario 2+3 is very bad and needs extra work and new ideas.

Posted: 26 Mar 2007, 11:14
by Crush
As a non-programmer you shouldn't worry about the performance. You can't estimate that anyway. Just say what you are thinking and let the programmers worry about how it can be realized.

Posted: 26 Mar 2007, 11:28
by Rotonen
I have had a simplified monster spawning and population control (as in population size and migration control, simplified ecosystem and ecobox simulation) system in mind for quite some time. Once we get the scripting backend to work, we'll be able to start to work on it (alongside with all the other dynamical features of the world we've planned so far).

Posted: 26 Mar 2007, 11:32
by Peacemaker
Crush wrote:As a non-programmer you shouldn't worry about the performance. You can't estimate that anyway. Just say what you are thinking and let the programmers worry about how it can be realized.
ok, from now on I'll just post ideas...

btw. I'm a programmer (and programming cellular automats is one of my hobbies) ...

Posted: 26 Mar 2007, 17:59
by EJlol
What about moving monsters? Like they spawn 2 days at map A and then they spawn 2 days at map B. after that they go back to map A.

Posted: 27 Mar 2007, 01:20
by yosuhara
EJlol wrote:What about moving monsters? Like they spawn 2 days at map A and then they spawn 2 days at map B. after that they go back to map A.
i don't see a point... definitely the dynamic system is the most attractive but it also has its flaws (like peacemaker pointed out), still, i like the current random spawn system... (thinking it over, a few fixed spawn-point-monsters would do no harm ;) )

Posted: 27 Mar 2007, 07:46
by ElvenProgrammer
I like the idea of the spawn rate being influenced by the whereabouts, especially I think a simple life game, i guess is a cellular automaton, shouldn't be that hard to add, so if you kill many monsters or a specific type in an area it will need more time for them to grow to a certain level again. This should force people to move to other areas. The drawback could be that since 99% of the players will be leveling the monsters population will be extremely low.