Turning monster pairs (v0.15)

[Gambit37]

This is both a difference from DM and also a personal preference. Monsters that can travel in pairs and that take up two corners of a tile (worms, hellhounds, rats), do not turn together. When they do turn, it's possible to get really weird effects depending on how they are overlayed and which direction they are facing.

I know this behaviour is because each monster is now treated independently. But it looks really, really weird and let's RTC down in the believability stakes. I feel that RTC should emulate DM in this regard.

[beowuuf]

This was covered ages ago, I can't rememeber wehat I said at the time! There are only a few creatures that do it...with worms it's not too bad.

I've gotten used to this, so while I agree it's not 100% realistic, its better than George having to work out how two monsters can move at the same time if they meet up on the same tile...programming for simultaneous turning will be painful I'm guessing.

[Gambit37]

Perhaps you've got used to it, fair enough. But there are some situations where it looks really silly:

<a href="http://website.lineone.net/~matthill/rt ... irs.gif</a>

I wouldn't think it would be too hard to fix this. This is all that needs to be covered:

* If two monsters are on the same tile, turn them together
* If one monster is on a tile, and another wants to share it, only allow it if they are facing the same direction.

I guess the trick is writing a bunch of code that removes the independent behaviour when two creatures are on the same tile, then managing them together. If one is then killed, the single worm should revert to independent behaviour. Surely this is just a few extra if-then's in the monster AI routines?

I also think that in situations where the worms turn to face away from the player, it's both too high on screen and the scaling factor is wrong. Because of the forced perspective in DM, worms facing away should appear lower down the screen - if you see it on screen with one facing towards the player, it's head looks too big in the distance. This is also true of hellhounds and painrats.

[Gambit37]

Incidentally, does anyone know what DM Java does regarding monster pairs turning?

[Gambit37]

This doesn't seem any different in v0.16. I wonder, what are your feelings on this issue George?

[Lubor Kolar]

In .20 I met Purple Worms and both were on the left side of the square. One was facing my party and another was with its tail to the party. This is really bad behaviour.

[George Gilbert]

The last point is valid and is a bug, I'll look into that.

As for the general case (e.g. in the above screenshot), then I'm happy with it as it is and monsters should retain their independence. If you force monsters to turn together / only allow two on a tile when they're facing the same direction then it massively limits their attacking capabilities.

As it happens it's actually easier to program the dependent AI than the independent one, but I like the idea of each monster being free to roam as they wish.

Anyway, aren't worms able to slither over each other

[Gambit37]

Yes, I guess worms can slither over each other, but have you seen what it looks like with hell-hounds or pain rats? REally, really, silly.... <!--EZCODE EMOTICON START --><img src=http://www.ezboard.com/intl/aenglish/im ... ohwell.gif ALT=""><!--EZCODE EMOTICON END-->

[beowuuf]

You could always force the other monsters to be drawn a fraction higher so it looks like a swarm

But it would look stupid with hell hounds, and probably not too sharp with pain rats, but the worms would look cool

[Gambit37]

I spent some time watching this in detail this evening using PC DM.

1) If two monsters share the same tile, they are treated as a pair and move together.

2) If one of the pair is killed on the tile, the remaining monster moves to the centre of the tile after a short delay. Other creatures can no longer enter this tile as it's considered 'full'.

3) Extra monsters never enter this 'full' tile, even if the creature hasn't moved to the centre. The other creature will seek out a tile where it can face the party and attack unobstructed.

4) Sometimes the monster won't move to the centre if it's 'fearful'. There seems to be a case when the second monster of a pair enters this state after its partner is killed. It won't attack you, and won't move to the centre of the tile.

Not sure what you want to do with these observations. I guess it's just me who thinks that the monster turning in RTC looks ridiculous....

[Lubor Kolar]

I thing monster AI should be improved.
As some wrote, independent monsters are more tough to fight, because they can split from their groups and spread in the chamber, which reduces party maneuver tactics possibilities (less free tiles). This could be a problem, but I like if game is a bit harder.
But I thing monster AI should take care about monster groups. There should be detected which one from this two monsters has a bigger priority (master) when deciding what this pair will do. The other monster from this pair (slave) will accept what "master" decides.
This should be implemented only for movement and only for monsters which 2 of them fit on one tile. Smaller monsters (mummys etc.) or big monsters (scorpion, dragon...) could act totally independent.

Let's see my example:
There are two worms (w1 and w2) in the chamber, each on its own tile. When they detect party is approaching and stays in corridor, both worms will move towards the party.
W1 is now facing party and w2 is one tile behing w1. W2 of course want to advance to fight the party. It should be allowed *only* when w1 and w2 will face the same direction. If that condition occurs, w2 advances and they will make a pair. If that condition does not occurs, it would be nice that w1 will move/turn to create this condition (it's logical - w1 will be happy if w2 could help him with the fight).

In the moment w1 and w2 are on the same tile (they are a pair), RTC should detect, which one is master and which one is slave (in my exaple, w1 is master and w2 is slave).
They could attack as they want, they could move as they want (each monster independetly of the another), but they must synchronize turning. Let's continue with my sample:
Now w2 gets hurt and has to perform some action. AI detects that it wants to turn back and run away. At this moment AI should detect, if master monster "confirms" this move and if so, the w2 waits for the moment w1 should make and action and then both worms turn. Result is that w1 and w2 are turned clockwise and next action will be another clockwise turn. Next action could be w2 runs away, but w1 wants to fight and it will turn back. But at this moment pair is divided, there is no master and slave and w1 and w2 now act independetly.

Making master and slave is IMHO needed because if it wasn't, the following situatuion could happen:
W1 is healthy and wants to fight, W2 is hurt and want to flee. One round W2 would like to turn clockwise (to be prepared to flee) so pair is turned clockwise. Another round W1 would like to attack so the pair turns back. But W2 would like to flee ..... And we could have "dancing pairs" performation.

I think it is not so difficult and it would solve problems we have now.

[Gambit37]

The simple way around this is to do what FTL did: move and turn a monster at the same time.

If you watch worms for any length of time, you'll notice that sometimes they will move from one tile to another but also change direction at the same time.

I.e., if they are facing you (lets say north-south facing) and attacking, and you wound them enough that they flee, sometimes they won't turn around and will simply shift to the next available tile and face, in this case, east-west.

This may look a little odd, but I believe this could be implemented into RTC simply enough and would solve some of the current problems.

[beowuuf]

Firstly, I assume in RTC the tile is actually like this:

a b c
d e f
g h i

and that two square monsters occupy b, d, f and h, normal creatures occupy a c g and i, and all creatures can occupy e, which blocks access by anyone else under normal conditions.

To keep independance, how about

a) Having 'two square' monsters blocking more of the tile than they are just now (a straight line on the square in RTC, like a b c, g h i, etc)

b) introduce a 'panic move' - creatures that cannot turn will immeditely walk forwards a step.

This second part will only affect 2 square monsters, as large and normal creatrues can always turn on the spot
A two square monster trapped on a tile with another that wants to turn will immeditely try to advance forward.

Imagine two worms facing 180 degrees from each other, one facing the party
The one facing the party cannot move, tries to move forward, can't so attacks the party
The one not facing the party tries to turn, cannot, panics and walks forward where it can then turn, and re-enter the tile facing the same way as the first

You could further refine this by having a random chance that if a two sqaure monster is blocked, the whole tile will be rotated by 90 degrees for monsters - this would affect single creatures too, as if it moves aside so the two square creature could turn, or as if a pair turned in time as DM

Just a thought...

[Lubor Kolar]

This is not so easy because it could work only when following conditions are true:
* both monsters from the pair should have the same speed (which could be true for two worms, but not for pain rat & worm)
* there must not be any delay between monster 1 action and monster 2 action. Let's imagine a worm pair - this condition means that both worms should attack and the same time, I will try to draw a time graph:

monster 1 (+ is action, - is action recovery time)
---+---+---+---+---+---+---+---+---+---+ time->

monster 2 (+ is action, - is action recovery time)
-+---+---+---+---+---+---+---+---+---+-- time->

[cowsmanaut]

in DM both screamers and theifs seemed to exhibit pack behaviour. if there were a lot of them (or at least more than one) then they would attack relentlessly. However when that number was brought to 1.. they often wouldn't be so macho. If you are thinking "theives? they didn't move in groups" you are right but they did seem to hang out with the ghosts if there was a ghost nearby they seemed to get more brave. Perhaps it's more distractive behaviour since they also tended to attack from behind or the side a lot too.

I think that it is always good to have simple instincts coded into various different enemies as it gives variety and allows for building different strategies for the various types since reactions from each differ.

First thing is to decide on the attitude.

If it's hit, does it run, or does it get pissed and chase you?
If it has friends is it more likley to attack you, or are the whole lot afraid of you?
Do they fight amongst eachother as wolves and rats sometimes would do in the wild if food is scarce?
Do they preffer to just attack or would they rather split and flank if they can?
what's their eyesight or motivation factor.. how far can they see you from or how long will they be willing to chase you?
can they recognise dammage to the party? (this is a good one actually as you can figure on a beast like the worms perhaps they can smell blood and so a wounded party member could perhaps draw more to you) does a weakened party look more apealing to them and make them more likley to attack.

anyway.. just throwing ideas out