Creature stats table proposal

[Danimal]

Creatures needed data just keep on increasing as the game becomes more complex, and the previous creature.def is pretty lacking.

This creates issues like being able to declare super-creatures on level file that totally breaks the game. All creatures should obey their stats table and super creatures should be declared on the creature.def and given a proper skin/color distinction over its common version.

This means as well that the game should read from the map file the creature class to add and calculate its stats (add base equipment as well) before spawning it.

So this is what i think a creature stats table should have:

Creature class

***combat stats***
Base HP
HP gain per level

Base damage (physical/magical or both)
Damage gain per level

Base defense (physical/magical or both)
Defense gain per level

Walk speed (on different terrains, 0 speed on one means it cant pass that terrain)
Walk speed gain per level (% over base, should never go over 1.3 of base speed)

Skills name (and a level requesite before being able to use it, skills will have their own .def for things like effects and cooldown)

XP table (XP needed to level up, we should have 3 tables, weak creature, normal creature, strong creature; so weak monsters level up faster and stronger ones slower)

***Work stats***
Zero means it cant work on related room/task, All have Base value and Gain per level.

Dig
Claim
Blacksmith can work on forge
Researcher can work on library
Mysticism can pray and summon on related rooms

Favourite room/s (if a creature is idle and the room exists, it will go there to work )
It can be casino, library, temple, farm, training hall...

***Equipment***
If the creature can have equipment it will spawn with the one declared here, creature will only upgrade weapon/shield to one of the same kind.
This will allow to specialize some creatures to be axe/hammer/sword wielders (so they stick to weapon type) and include two-handers which will have their own animation (Ex: Heavy axe and Claymores for orcs and Barbarians) and pre-set animations depending on weapon (a hammer doesnt hit the same as an sword, i can prepare characters animations on exporting if they always stick to a weapon, this would make combat animation a lot better, and unrelated to code ).

***Others***
Rooms and size needed for spawn

Mood conditions:
Money/pay day wage
liked/hated creatures near

Max stat limitation? we could avoid this variable just by calculating properly the final stats, as easy as creating an excel table and adding the base and growth values.

Practical example:
Spider
20 Hp +5 Hp lvl
5,0 Atack +1,+0 lvl (physical magical)
2,1 Def +1,+1 Lvl (physical magical)
1,0.5,0 Speed on land,water,lava; x1.05, x1.05, x1.05 speed gain per level
Poison skill - lv3, Trap skill - lv6
weak creature XP table
Dig 0 + 0 lvl
Claim 0 + 0 lvl
Blacksmith 0 + 0 lvl
Researcher 0 + 0 lvl
Mysticism 0.3 + 0.1 lvl
Temple, Prison, Graveyard - Favourite rooms
No Equipment
Dormitory 4, Farm 4 (room requesites)
Wage: 10 gold per lvl
Cave hornet +1 mood
Swamp pup -1 mood

Warlock
40 Hp +4 Hp lvl
0,6 Atack +0,+2 lvl (physical magical)
0,4 Def +1,+3 Lvl (physical magical)
0.7, 0.4, 0 Speed on land,water,lava; x1.05, x1.05, x1.05 speed gain per level
Fireball skill - lv2, Meteor skill - lv5, Teleport skill lv10
medium creature XP table
Dig 0 + 0 lvl
Claim 0 + 0 lvl
Blacksmith 0 + 0 lvl
Researcher 10 + 4 lvl
Mysticism 5 + 1 lvl
Library, Casino, Temple, - Favourite rooms
StaffLv1 (base equipment)
Dormitory 9, Farm 9, library 9 (room requesites)
Wage: 50 gold per lvl
Troll +1 mood
Dragon -1 mood

***SOMETHING TO REMEMBER***
Sounds use by a classification depending on creature.def value
viewtopic.php?f=39&t=5695&start=25#p59177

[Bertram]

Hi Danimal,


Thanks for starting this discussion. With all the bugfixes OD has received recently, the game is becoming more and more in need of a good balance.

@Everyone:
For the record, the topic was spawned after some discussion here: https://github.com/OpenDungeons/OpenDungeons/issues/7

Creatures needed data just keep on increasing as the game becomes more complex, and the previous creature.def is pretty lacking.

Please, if some stats are lacking, tell us what you think is so we can take those in account as soon as possible.

There are several things intertwined in what you said, so I'd like to unknit them before we move on.

This means as well that the game should read from the map file the creature class to add and calculate its stats (add base equipment as well) before spawinig it.

Well, this is already done by the engine. Spawn pool and the possibility to define a creature definition file is already there.

All creatures should obey their stats table

From experience, defining stats bonus/level or a stat table is more or less the same thing. The thing is, to define a proper stat bonus per level value, you have to create scales and balance them between creatures before turning it back to a bonus per level.
The main advantage of defining stats per level is that you have much less values to type, store and parse when spawning a creature. And also, it is already in place.
Thus, IMHO, yes, values balanced between the overall available creatures should be done in a reference sheet somewhere, but:
- We should define what are the lacking stats as soon as possible to avoid breaking the balance later on.
- We should be able to turned back those values into stats bonus per level afterwards so I don't see the point in dropping the current system.
EDIT:
Values would be needed only if you want to add non-linear growth of certain stats, but atm, balancing will be hard enough without that, IMHO.
Maybe you can explain a few use cases?

This creates issues like being able to declare super-creatures on level file that totally breaks the game.
and super creatures should be declared on the creature.def and given a proper skin/color distinction over its common version.

The super-creature thing is a false problem IMHO: It is just a copy paste of another creature class with better stats.
The special skinning problem will also be handled by the promised-since-too-long entities catalog file.

Feel free to prove me wrong of course.

Best regards,

[Danimal]

Hi Bertram, i was barely starting writing, please read again. Also, i wouldnt touch non-linear growth, please lets not complicate this even more.

[hwoarangmy]

I definitely think that we need to start balancing. First, it will be the occasion to use the tests we all do to see if the balance is correct. Secondly, it will be more fun than seeing a troll fight a kobold for ages before dying, killed by its pickaxe.

[Danimal]

The troll vs kobold case is just the result of creatures not having defined a base attack (nor attack growth per level) and base HP, plus their tendency to cat stare contests and air attacking.
Once we set it like proposed, troll will start with something like 50 HP and 10 damage and kobold 10 HP and 1 damage. Then figths will be more realists, balancing them will be something we do after creatures properly engage and try to kill each other, so a spider cant kill a dragon.

[Akien]

The troll vs kobold case is just the result of creatures not having defined a base attack (nor attack growth per level) and base HP, plus their tendency to cat stare contests and air attacking.
Once we set it like proposed, troll will start with something like 50 HP and 10 damage and kobold 10 HP and 1 damage. Then figths will be more realists, balancing them will be something we do after creatures properly engage and try to kill each other, so a spider cant kill a dragon.

That makes me think that we might want to have a "threat" characteristic for each creature, that can have an influence on whether or not a weak creature will engage a stronger one. This would probably come into play in the algorithm for the fighting behaviour, so that a single kobold won't charge a dragon, but 12 kobolds might try their luck.

But maybe we don't want to hardcode it in the creatures definition file, but have it computed depending on a given creature's stats (level, hp, attack, def, etc.).

[Bertram]

Hi,

@Akien: As you said, I'd rather addition the physical+magical attack and use that as a threat value for a start. This way the "threat" of a creature would dynamically evolve.
All the rest of the proposal sounds fine to me.

What's left to define here is how creatures should decide to attack. This was discussed a bit in issues but not exactly stated here.

Once we all agree on something, I'll update issue #7 accordingly.

Best regards,

[Danimal]

Im fine as long as that doesnt make creatures stare at each other for a while before starting combat, but isnt it like what we already have?

[Bertram]

Im fine as long as that doesnt make creatures stare at each other for a while before starting combat, but isnt it like what we already have?

Even if everyone agrees there a need of action, Danimal raised a valid point here.

I think that if we make creatures care too much for their self-preservation, they will end up doing nothing more than fleeing battles and not going to work because too lazy, tired, in a bad mood, ... Reminds you some other type of creatures?

Don't misunderstand me, i think the current system concerning hunger and drowsiness is working quite fine already, but if we are to add support for a threat system (which is not far from the current one after all), we should make the fighting creatures very likely to go to combat, and workers likely to flee, whatever the threat, unless both are dropped next to an enemy.
This means the threat should make the creature flee if it is sure to die on first blow only, again, unless dropped next to the enemy.
This also means the threat is directly linked to damage done by the enemy creature in this case, IMO.

What do you think of my proposal?

Best regards,

[hwoarangmy]

Don't misunderstand me, i think the current system concerning hunger and drowsiness is working quite fine already, but if we are to add support for a threat system (which is not far from the current one after all), we should make the fighting creatures very likely to go to combat, and workers likely to flee, whatever the threat, unless both are dropped next to an enemy.
This means the threat should make the creature flee if it is sure to die on first blow only, again, unless dropped next to the enemy.
This also means the threat is directly linked to damage done by the enemy creature in this case, IMO.

What do you think of my proposal?

The good point is that it is simple. The problem is that you cannot attack one strong creature with loads of weak ones. With this logic, ants would not survive

[Bertram]

The good point is that it is simple. The problem is that you cannot attack one strong creature with loads of weak ones. With this logic, ants would not survive

Eh eh, beware of the ants.
I do think it's not a problem since I wouldn't want all my weaklings to go get crushed by themselves unless I decide to drop them on the big boss myself.

[Danimal]

I do think it's not a problem since I wouldn't want all my weaklings to go get crushed by themselves unless I decide to drop them on the big boss myself.

Sending in the cannonfodder so the ranged do its job its a common tactic, we are evil masters after all. Im not fond of any treath system and think a creature should only run away when low on health or morale (kobolds being the exception, they should run from anything else except other kobolds). Just think on common military men, they are whipped into submision and made totally non thinking and brass order dependant, the same should be for our creatures.
The "1 hit kill" fear is not a bad condition, but a weak creature attacking depends on how many allies are around, but those allies migth just be doing something else instead of engaging, that takes us to the same problem we have with the current system.

Im still defending to just have a few simple conditions a creature decides to figth or not, Hp and morale. Werent they supposed to be evil bloodthirsty monsters?

[Bertram]

Hi,

Im still defending to just have a few simple conditions a creature decides to figth or not, Hp and morale. Werent they supposed to be evil bloodthirsty monsters?

I must say you have a point. Having conditions that makes a creature consider other creatures make the the thing much more complicated and not easily reliable.
It's a choice we have to take now:

@everyone:
--> Shall we make a creature consider the threat and allies along, or should we make it go when healthy enough?

Danimal's proposal to check only one's own health sounds pragmatic, simple and easier to maintain. I'd rather go that way.

What do you think, guys?

Regards,

[Akien]

@everyone:
--> Shall we make a creature consider the threat and allies along, or should we make it go when healthy enough?

Danimal's proposal to check only one's own health sounds pragmatic, simple and easier to maintain. I'd rather go that way.

What do you think, guys?

I think we should try to keep most gameplay mechanics simple until they are implemented well enough to be fine-tuned. What I mean is that the focus should be on getting simple and understandable gameplay mechanics for the user and rule out all "behavioural" bugs (e.g. two enemy creatures staring at each other). So I'm in favour of using a simple check such as Danimal's for now.

When we have well-defined and balanced creatures, maybe with skills and spells too, we can have a look at making their behaviour more interesting. If there's still a need :-)

[hwoarangmy]

As I've already said in other occasions, IMHO, OD is a strategy game and I do not consider that my creature's default behaviour has to be perfect. If they do not teleport instantly when the dungeon heart is attacked, for example, I will not think it is a bug. That's the player's job to optimize creature's behaviours

That's why I think we should keep it simple as everybody said for now

I also think it is a good thing to have creatures that flee. But we could relate it to some creature definition value so that when they see a dangerous creature, they attack or flee depending on that probability (I am not talking about weak units, that's another subject).

[Bertram]

Ok, so we have a consensus about making the creatures only check their "health" before attacking.
By health, I mean HP, drowsiness, hunger and later mood.
I do think several of those stats should be low to make a creature flee. And by low, I think it would be < 1/5 of the total.
I also think that a drowsy or hungry creature should no longer naturally heal each turn (I'm not speaking of when sleeping, of course :]).

I also think it is a good thing to have creatures that flee. But we could relate it to some creature definition value so that when they see a dangerous creature, they attack or flee depending on that probability (I am not talking about weak units, that's another subject).

Then, you're thinking about a fixed value depending on the creature specy? Would then a "courage" percentage value be what you think about?

Best regards,

[Danimal]

drowsiness and hunger would be redundant if we check (and implement) a mood system: both of those would put a creature in a bad mood if not fulfilled and thus make it run. A happy creature should figth anyways if dropped over an enemy no matter how drowsy or hungry it is.

Then, you're thinking about a fixed value depending on the creature specy? Would then a "courage" percentage value be what you think about?

This takes us back to the same problem, What if i have a legion of high level bugs and set it against a demon? would all of them run because individually they are no match? we check for nearby allies? -> back to the same problem

I think the best option is the most simple one, HP and mood check. All attack at first sigth but if any of those fails, the creatures runs away

[Bertram]

All attack at first sigth but if any of those fails, the creatures runs away

Could you elaborate on "if any of those fails"?

[hwoarangmy]

Then, you're thinking about a fixed value depending on the creature specy? Would then a "courage" percentage value be what you think about?

Yes, something like that. Natural fighters (even weaks) will almost always go to fight. But some like wizards would flee because they are not natural fighters. IMHO, that's not related to the surrounding creatures. Note that increasing the "courage" as the creature is leveling up makes sense. This way, a level 1 wizard would much more flee than a level 30, which sounds like normal.

drowsiness and hunger would be redundant if we check (and implement) a mood system: both of those would put a creature in a bad mood if not fulfilled and thus make it run. A happy creature should figth anyways if dropped over an enemy no matter how drowsy or hungry it is.

I agree, I think is is better to separate things. But I think Bertram's proposal of influencing heals regenration depending on hunger/drowsiness is interesting. Note that actually, hunger/drowsiness affects how the creature works. That means that if a player do not make enough dormitory/hatchery, his creatures will not level up. Which is already a big handicap.

And by low, I think it would be < 1/5 of the total.

Considering that they only consider going eating/sleeping when under 50%, I would say that by low, I would consider 10% or less

drowsiness and hunger would be redundant if we check (and implement) a mood system: both of those would put a creature in a bad mood if not fulfilled and thus make it run. A happy creature should figth anyways if dropped over an enemy no matter how drowsy or hungry it is.

What if i have a legion of high level bugs and set it against a demon? would all of them run because individually they are no match? we check for nearby allies?

No, we don't look at nearby allies. I was thinking that, for example, a level 1 wizard would have up to 50% chances to flee (depending on his ennemies stats) when choosing if he engages. IMHO, flee would be something like panic rather than deciding to not fight because of some logic reason.

[Bertram]

No, we don't look at nearby allies. I was thinking that, for example, a level 1 wizard would have up to 50% chances to flee (depending on his ennemies stats) when choosing if he engages. IMHO, flee would be something like panic rather than deciding to not fight because of some logic reason.

As far as I can see, I think the panic button option might be not trivial to re-add right away when redoing the whole system.

IMHO, I would not add it in the first implementation cycle, and then make it up to 30% max to flee once implemented, when we are fine-tuning that part. Would you be ok with this guys?

As I can see, there is already some work to do before we are to this point, anyway.

Regards,

[hwoarangmy]

IMHO, I would not add it in the first implementation cycle, and then make it up to 30% max to flee once implemented, when we are fine-tuning that part.

Yes, as Akien said, there is no need to rush in there. That can be done later if needed. For now, we can keep flee for wounded creatures.

[Bertram]

Sounds fine to me.

I'll try to sum it all up, since a lot has been said. Once everyone is okay, I'll close issue #7 and recreate one with what is agreed.

[Bertram]

Summary on what has been said upon creatures stats and overall configurability:

(Please review all this and tell me where it seems wrong out of what has already been agreed. )

First Issue/thing to do:
The creatures.def file should be adapted so that creatures are defined within [Creature] [/Creature] tags, so we can use several lines and make this much more readable.
The stats may be within [Stats] [/Stats] tags if it's helping for readability and code.
The file should be auto-documented using #comments.
See Github issue #11 for an example.

Second Issue:
Creature stats lacking:

Base HP (at level 1)
HP gain per level

Base physical attack
Physical attack gain / level
Base magical attack
Magical attack gain / level

Attack cooldown. (The time before a creature can trigger another attack.)

Base physical defense
Physical defense gain / level
Base magical defense
Magical defense gain / level

Base attack range (in tile)
Attack range gain / level

Walk speed (on different terrains, 0 speed on one means it cant pass that terrain) -> Done
Walk speed gain per level (% over base, should never go over 1.3 of base speed)

XP table (XP needed to level up, we should have 3 tables, weak creature, normal creature, strong creature; so weak monsters level up faster and stronger ones slower)
--> 30 values per creatures. Note that the creature's XP should go back to 0 at each level up as it isn't the case ATM. The XP to reach for next level should be displayed in the info window.
We should create an excel table and add the base and growth values for every creatures, helping to balance the game.

HP Heal rate per turn (if not drowsy, hungry)

Worker's stats:

Zero (0) means it can't work on related room/task, All have Base value and Gain per level.

Dig -> Done.
Claim -> Done.
Dig gain/level
Claim gain/level

Github Issue #11
Rooms and size needed for spawn
Mood conditions
Money/pay day wage
liked/hated creatures near

Github Issue #34:
DropOnDeath Item dropped on death

Github issue #126 - Creature sounds configurable per class:
Creature sounds will have to be set per specie, like reptile-walk, insect-attack and the engine will have to play corresponding sounds.

To do/keep in mind for later:
Blacksmith can work on forge
Researcher can work on library
Mysticism can pray and summon on related rooms

Favourite rooms (if a creature is idle and the room exists, it will go there to work in the given order.)
It can be casino, library, temple, farm, training hall...

Skills (and a level requesite before being able to use it, skills will have their own .def for things like effects and cooldown)
Linked to Angel Script support.

Equipment
If the creature can have equipment it will spawn with the one declared here. -> Done.

Creature will only upgrade weapon/shield to one of the same kind.
This will allow to specialize some creatures to be axe/hammer/sword wielders (so they stick to weapon type) and include two-handers which will have their own animation (Ex: Heavy axe and Claymores for orcs and Barbarians) and pre-set animations depending on weapon (a hammer doesnt hit the same as an sword, i can prepare characters animations on exporting if they always stick to a weapon, this would make combat animation a lot better, and unrelated to code ).

Combat:

- The fighter creatures should check its "health" + "mood" before going or keep going to combat or fleeing.
- "Health": If HP < 10%, flee to bed.
- "Mood": If mood < 10%, flee to dungeon randomly.
- Note that mood is decreased by hunger, drowsiness, and later, other creatures, lack of pay wage, and other things I didn't see.

- Github issue #90: The worker creatures will flee a few tiles away when receiving their first blow, and they won't take on fighter creatures, only other workers.
When seeing enemy creatures, they will keep claiming, digging, until first blow.

- All creatures can be forced to fight by dropping them next to an enemy creature.

Later:
We can add a % of "courage". This is a dice roll based on a fixed configurable value -/+ enemy total attack against own total attack, that can trigger a flee back to dungeon.
This "panic test" should be done only once before falling into battle mode, not each turn. The base value shouldn't be higher than 25-30%

Have I missed something?

Best regards,

[Bertram]

I forgot to add attack cooldown. (The time before a creature can trigger another attack.) Added now.

[Danimal]

That seems about rigth; just dont forget the place to declare which .mesh to load

The "panic test" doesnt sounds bad actually, what i want to avoid the most is having creatures motionless while they should be doing something, be it figthing or running away.

By the way, i think the need to sleep or eat should only be trigered when under 30% (or slowing their drop rate), im having creatures going between dormitory and farm endlesly and ignoring training because when one need is high the other is low. Also, chikens run like hell, my poor minions cant get a hold of them, and the troll is thinking on munching on kobolds instead