Summoning Speculations

[jens]

I'll add mana to monster.txt, as soon as 3.3 is out :-)

Need some info though:
What should the syntax for the directive be?
What scale should we use for mana? I'd like to have a scale similar to @, but it depends on how monster spells are coded.
How fast will monsters regain mana? Should this be included in the directive?
Can you point me to the branch where this is implemented?

[jens]

Another summoning idea I'd like to throw out there, and it's on topic if you go by the topic title ;-)

I'd like all items to be useful in some context, and I just thought of how to make scrolls of summoning useful: If you summon monsters next to other monsters there is a chance that the summoned monsters will start attacking the other monsters. Guessing it's lots of coding for this, but I think it's a neat idea :-) And I like the flavour of reciting a scroll of say summon demon to fight against your enemies, then loosing control over the demon and having to fight that as well :-)

[fizzix]

Question: When a monster is summoned, does the player have a free turn before the monster can move (as in when the player enters a level) ? If not, changing that would improve the survivability of fighting summoners in the open.

No, summoned monsters come with a delay of 0. That means that they can get a turn before @ if they are fast enough.

[fizzix]

I disagree. The code for this is already in a branch on github, waiting to be merged with a monster.txt that assigns monster mana (which is a non-coding job that sadly nobody has signed up to yet). I can add the mana cost of monster spells in half an hour (adding a field to list-mon-spells.h). In fact, I'll make it the first thing I do for 3.4, in order to allow plenty of time for consequent rebalancing and ironing out of bugs.

I was unaware of this! I can try to do this along with jens, but I think we should first make a list of spell costs for all monster spells.

I'm guessing arrows and boulders count as spells for now, right?

[Magnate]

I was unaware of this! I can try to do this along with jens, but I think we should first make a list of spell costs for all monster spells.

I'm guessing arrows and boulders count as spells for now, right?

Yes, they do - but there's no reason they can't have 0 mana cost. In fact, that's exactly what they should have, as that will enable us to get rid of the terrible MIN_NONINNATE_SPELL hack, and just use if (!rs_ptr->cost).

For mana costs and monster mana values, please see NPP in the first instance - takkaria has said we'd like to start from there.

For the existing branch, please see http://github.com/elly/angband.

[ChodTheWacko]

you know what would be interesting - if tunneling always made a tunnel of width two. It still allows tunneling to be used for stuff like phase door and short cuts but I think it cripples ASC ability.


- Frank

[Derakon]

For what it's worth, I generally feel that NPP monsters have way too much mana. Or their spells are way too cheap. It's been awhile since I played last, but I took a paladin all the way to Sauron and I never once saw a monster run out of mana. I've seen other people say that basically the only way to kill Sauron/Morgoth in NPP is to drain their mana (ideally by forcing them to cast healing spells all the time, as opposed to letting them blow you up over and over)...it'll be tricky to get the balance on that right. Mana effectively acts to extend the health bar on monsters with healing spells, so how much longer does that health bar need to be?

Of course once the mana is gone the monster is also much less dangerous, so that's something to consider too. Especially if you introduce some mechanism for the player to drain monster mana. You don't want the Sauron fight to turn into the player coating Sauron in antimagic slime and then going to town on his defenseless hide.

[fizzix]

For what it's worth, I generally feel that NPP monsters have way too much mana. Or their spells are way too cheap. It's been awhile since I played last, but I took a paladin all the way to Sauron and I never once saw a monster run out of mana. I've seen other people say that basically the only way to kill Sauron/Morgoth in NPP is to drain their mana (ideally by forcing them to cast healing spells all the time, as opposed to letting them blow you up over and over)...it'll be tricky to get the balance on that right. Mana effectively acts to extend the health bar on monsters with healing spells, so how much longer does that health bar need to be?

Of course once the mana is gone the monster is also much less dangerous, so that's something to consider too. Especially if you introduce some mechanism for the player to drain monster mana. You don't want the Sauron fight to turn into the player coating Sauron in antimagic slime and then going to town on his defenseless hide.

I was looking at the values today and I agree that it seems that most monsters have too much mana. Of course NPP has the ability to drain mana and V does not, so we should obviously adjust for that.

I think we need some more information. So the next time you fight a tough unique, pay attention to:

1) How many game turns did the battle take. (needed because of regeneration)

2) How many spells did it cast

If we had that information for several endgame uniques, we could probably make a reasonable guess for mana allotment. If we get the framework up and working, we probably will need a lot of discussion and input to get things right.

[bio_hazard]

Just going to throw this out there- what if monsters could summon as is *one time*, but then they would be prevented from summoning again *as long as any of the previous summons were still around*.

This would keep the challenge of fighting summoners, but would also make the task more manageable. If you want to farm the summons for XP or drops, you can kill them all. If you want to take out the unique easily, you have to leave at least one of the summoned creatures alive.

Some questions on implementation would be
- What exactly would "still around"? Would a TO-ed monster still count as "around"? What about a slower monster that the original summoner outpaces? Basically alive on level or within sight of the original summoner?
- If a summoned monster can itself summon, whether that would need to be adjusted or not.

[fizzix]

Just going to throw this out there- what if monsters could summon as is *one time*, but then they would be prevented from summoning again *as long as any of the previous summons were still around*.

In a way, something like this already happens. Summoned monsters that aren't brought along via FRIENDS or ESCORT(S) are constrained to appear within 2 squares of the summoner. If all those spaces are filled, no more summoned monsters can appear.

[buzzkill]

Some time ago someone suggested creating 'limbo' where spawned and/or summoned creatures are drawn from. Supposing that limbo was populated at the time of dungeon creation based upon the dungeon level (spawns) and denizens (summons). Monsters that are 'removed from the level' (destruction) can be said to have a chance to be returned to limbo, so that they may have a chance to reappear.

IMO limbo should be large, probably having a population greater that that of the dungeon level itself. Cleaning it shouldn't ever really happen, though it shouldn't be impossible either. If limbo is empty, the level can be considered truly cleared.

[fizzix]

I just played a game with two changes. The first were more sensible alertness values and that is likely to make it into V. The second change was more controversial as it had to do with summoning.

The change is as follows. When a monster summons, it's summoning power is computed as (monster level) * (current dungeon level). Every summoned monster contributes as its (monster level)^2. Monsters that have the FRIENDS are computed as (monster level/5)^2 to account for their considerably weaker strength. When the sum of the summoned monsters is equivalent to the summon power, no more monsters can be summoned.

Summoned monsters do not come with FRIENDS and ESCORTS. Summoned monsters can now appear up to 4 squares away from the summoner.

Here are the results:
Summoning is generally weaker but not all that much so. Certain summons are a lot more interesting from a tactical standpoint. Your choice is forced when 20 time/gravity/chaos hounds are pulled in by a hound summon, but that's not true when you have 1 of them. It's now an interesting decision whether to continue fighting or not.

Not allowing FRIENDS makes summoning stronger overall. The reasoning is that you can't clog up the room with wimps from a weak summon. The monster can always pull in more. With more variety, there's more chance of getting a nasty monster or a tough unique. At the higher levels, Morgoth and Sauron had just as difficult summons as I'm used to. In fact, probably more difficult. There were no "weak" summons from Morgoth where I got a bunch of barrow wights or Hezrou. Every one of his summons brought in at least 1 tough monster. I think I burned more ?banishment and ?mass banishment on the last battle than I'm used to. Sauron barely attempted to summon...

Removing ESCORTS makes summoning weaker in a few cases but makes it stronger overall. You often can get a bunch of free hits off of Morgoth or Sauron if they make the mistake of summoning a troll or orc unique that fills the room with its buddies and prevents Morgoth from adding more monsters. Now if they come, they're alone. This makes gothmog/lungorthin/kronos/gabriel individually less powerful when summoned. But often the summon that brings them in is more powerful because it lacks the easy counter of ?banish. It's harder to deal with Gothmog + Gabriel + Tarrasque than it is Gothmog + 30 pit fiends. Not that much harder, but it is harder.

Anyway, I was pleasantly surprised by how the changes played. And I think it's a good step towards revamping summoning.

[half]

Monsters that have the FRIENDS are computed as (monster level/5)^2 to account for their considerably weaker strength.

Did you really mean to count them at a 25th the value? That seems a little extreme -- like counting a Time Hound as a level 10 monster.

[fizzix]

Did you really mean to count them at a 25th the value? That seems a little extreme -- like counting a Time Hound as a level 10 monster.

25 is roughly how many would come in if they were in groups, so yes. I want monsters like barrow wights, or hezrous to count as virtually nothing when summoned. Enough hounds should come to fill up all available spaces. That's what they would do previously.

[half]

25 is roughly how many would come in if they were in groups, so yes. I want monsters like barrow wights, or hezrous to count as virtually nothing when summoned. Enough hounds should come to fill up all available spaces. That's what they would do previously.

Fair enough, I suppose it has been a long time since I played Vanilla and 25 in a group seemed like a lot to me. It sounds like a good idea overall to me, but bear in mind my lack of recent Vanilla experience.