Summoning

[XuaXua]

I'm against the LOS thing simply because managing summons when they're slower than you (fast races, running, swiftness, earth elementals) is already painful enough.

Summon cap notwithstanding, my opinion is the increased rate (or chance; not sure how summons work, is it a % chance to vanish per turn, or is it a flat decreasing rate of existence?) of duration expiration is the best choice.
- I don't care for the "change of mindset to non-fighting" when out of LOS
-- it creates additional coding and checks
-- they're vanishing anyway soon if you have an increased rate, so the conceptual quantity of damage will reduce and if monsters flee when damaged enough, even moreso.

- I don't care for the "vanish entirely" when out of LOS
-- makes cutting corners a problem (As pointed out)
-- makes speed difference even more of a problem, even considering galehar's implementable
-- reduces flee distance potential
-- makes lantern of shadows even less useful

[galehar]

I'm against the LOS thing simply because managing summons when they're slower than you (fast races, running, swiftness, earth elementals) is already painful enough.

Summon cap notwithstanding, my opinion is the increased rate (or chance; not sure how summons work, is it a % chance to vanish per turn, or is it a flat decreasing rate of existence?) of duration expiration is the best choice.

Summons have a duration.

Faster duration reduction is a good fix, but no attack outside of LOS is simply a better one. It is more efficient, has none of the downsides of the duration reduction and is even thematic (you need LOS to control your summons).

[XuaXua]

I'm against the LOS thing simply because managing summons when they're slower than you (fast races, running, swiftness, earth elementals) is already painful enough.

Summon cap notwithstanding, my opinion is the increased rate (or chance; not sure how summons work, is it a % chance to vanish per turn, or is it a flat decreasing rate of existence?) of duration expiration is the best choice.

Summons have a duration.

Faster duration reduction is a good fix, but no attack outside of LOS is simply a better one. It is more efficient, has none of the downsides of the duration reduction and is even thematic (you need LOS to control your summons).

If the summon is in LOS, will it still attack a target that is outside of LOS (it should)?

[galehar]

If the summon is in LOS, will it still attack a target that is outside of LOS (it should)?

Why should it? Doesn't that encourage dancing around corners to try to have your summons in LOS attack monsters which cannot attack you? What's the upside?

[XuaXua]

If the summon is in LOS, will it still attack a target that is outside of LOS (it should)?

Why should it? Doesn't that encourage dancing around corners to try to have your summons in LOS attack monsters which cannot attack you? What's the upside?

Your summon cannot defend itself from attack from a monster at edge of line of sight.

[galehar]

If the summon is in LOS, will it still attack a target that is outside of LOS (it should)?

Why should it? Doesn't that encourage dancing around corners to try to have your summons in LOS attack monsters which cannot attack you? What's the upside?

Your summon cannot defend itself from attack from a monster at edge of line of sight.

Then take a step toward it. I think it's better than the other way around, summons being vulnerable is less important than the player being invulnerable. But it's a minor issue anyway. The important thing is that the summon should be in LOS to be able to attack.

[sardonica]

If the summon is in LOS, will it still attack a target that is outside of LOS (it should)?

Why should it? Doesn't that encourage dancing around corners to try to have your summons in LOS attack monsters which cannot attack you? What's the upside?

You're really pounding the "nerf Summoning" drum on every thread.

The upside is that summoning is about sending monsters to do your bidding, and if that involves dancing around corners, in exchange for less XP and a big hassle, that's cool?

[galehar]

You're really pounding the "nerf Summoning" drum on every thread.

Have you read the thread's title? And I'm not pounding anything. Nerfing summoning has been planned for ages and is agreed by every devs and most players. Especially the out of LOS abuses.
I'm just discussing how to do it.

[varsovie]

The important thing is that the summon should be in LOS to be able to attack.

If it mean that a summon at the edge of your LOS can attack everything (even at distance, outside of your LOS) I'm OK. It still allow to *corner fight* dangerous enemies like Orc priest.

[galehar]

The important thing is that the summon should be in LOS to be able to attack.

If it mean that a summon at the edge of your LOS can attack everything (even at distance, outside of your LOS) I'm OK. It still allow to *corner fight* dangerous enemies like Orc priest.

Seems like the discussion has gone into an infinite loop. So let's try to sum up. Being able to kill monsters with 0 risk isn't interesting. So your post has convinced me. And since it seems I'll be the one implementing it, here is how summons will be in 0.12 (unless someone convince me otherwise with real aguments):
A summon can attack a monster only if it and its target are both visible to the player. This only applies to free summons, not permanent allies or god gifted allies.

Regarding caps, I'm not so sure how to do it, and maybe MarvinPA will help with formula and implementation. For now, here is my opinion:
There will be a per spell (depends on power) and a total limit (depends on summoning skill) to the number of summons. When at the limit, casting is disabled.

[Stormfox]

A summon can attack a monster only if it and its target are both visible to the player.

As in literally visible, or within the player's LOS? If the former, then what's the solution for a summoner who can't see invisible against monsters with invisibility capability? Even if the summons themselves can see/sense invisible?

(I know, there are half a dozen ways to deal with invisible monsters that don't require see invisible.)

[sardonica]

A summon can attack a monster only if it and its target are both visible to the player. This only applies to free summons, not permanent allies or god gifted allies.

Do you mean "within the radius of LOS of the player" or literally "what the player can actually see?"

[palin]

And since it seems I'll be the one implementing it, here is how summons will be in 0.12 (unless someone convince me otherwise with real aguments):
A summon can attack a monster only if it and its target are both visible to the player. This only applies to free summons, not permanent allies or god gifted allies.

This is very nice, except a summoner is this way very penalized on one of his strength: early oklobs. But maybe this is not a problem.

[galehar]

Right, being in LOS should be enough, not being actually visible. Remember the reasoning: we want to limit the ways a player can kill monsters while taking no risk. Having an invisible monster in LOS is usually quite risky.

[sanka]

This is very nice, except a summoner is this way very penalized on one of his strength: early oklobs. But maybe this is not a problem.

I do not think oklobs can attack you trough your summons. Not that oklobs are that hard for non summoners.

[XuaXua]

Having an invisible monster in LOS is usually quite risky.

That's what Summon Canine Companion is for.

[pubby]

What if some summoning buff spells were added that only effected summons in LOS? The spells would give you a status which buffs nearby visible summons, so if they go out of LOS and come back they regain their buff. Summons would be nerfed a to make tough fights be much easier to use buffs in sight rather than fighting without buffs out of sight.

This would give you an obvious advantage and strategy to have your summons fight in LOS without introducing convoluted penalties.

Few ideas of spells:
Battle cry - boosts attack, leaves you out of breathe
Alarm call - boosts evasion, leaves you out of breathe
Group resistance - gives allies + resistances, gives you - resistances
Group levitate - levitates others
Heal other - mentioned this in OP

[pubby]

How is reducing their attack, evasion, resistances, and damage they take from airstrike less convoluted than making them go away???

I wasn't considering a nerf to be a penalty which is what I was calling convoluted.

But yes, it's more complex and requires more effort than just making them go away when out of LOS.

[Torious]

I recently won 3 DESu's with decent scores. I played them as pure summoners; no other spells except utilities such as blink, regen, rmsl, etc. My concern w.r.t. nerfing summons (which I agree is necessary) is how it will affect pure summoners.

The discussed LOS-based nerf seems fine.

The summoning caps are a difficult issue, because summons are super hard to balance. Imagine putting a cap of 4 dragons on Summon Dragon at decent spell power. In the current form of Summon Dragon, that would be enough to clear most of the game. But the Zot:5 vault, Cerebov, Antaeus; that's just not gonna happen, unless you take a billion turns (and you already are using more turns than conjurers and getting less XP). One of the reasons summoners create 15-20 dragons in those fights is that they simply lack the raw damage output you get from high-spell power conjurations.

Also, earlier in the game, you can currently clear Lair with just ice beasts and scorpions, although there are some long and tough fights vs. death yaks and hydras mostly. Those tough fights involve spamming scorpions, but you still get your ass kicked and gotta run, recover, and re-approach. Ofcourse you can argue that you should come back later with Summon Ugly Thing (still not easy fights), but fact is that putting a cap on the no. of scorpions will significantly weaken summoners in lair. Also, spell power won't be that high at that point, so a cap based on spell power is extra tough.

Btw, the above is based mostly on my DESu's, so summoners of other races will have an even harder time. I guess the large part of the imbalance comes from spamming dragons.

I don't have a clear-cut solution, but one approach could be the following:

Have some caps in place, much like suggested, but also drastically scale summon power with spell power, so that you can actually take on Cerebov with 5 dragons, if you've invested heavily in Summoning (lv. 24+ or so). This should also significantly weaken summons (dragons) for late game conjurers etc. that just train Summoning to the minimum to get dragons. In case I'm not clear: At high spell power, individual summons should be a lot stronger than they currently are imo. And then put a cap on the number of summons.

It's actually more fun to work with fewer, but more powerful summons. This is evident from mid-game use of Summon Greater Demon (that's one well-balanced spell imo btw). One greater demon can take out a few enemies, but often turns hostile after some time, and your abjuration may not always work immediately. That requires quite some tactical decisions...

Oh yeah, this whole problem stems from channeling abuse, which stems from undead (form) being way overpowered. Remove all undead (forms) in 0.12

EDIT: I wanna add that, looking at the top 1000, there are like 10 summoners (*Su) that are not MuSu's. So even with summon spells as they are, I think it's obvious pure summoners have a hard time (apart from mummies, which are totally broken). Please consider that and work on balancing summons to prevent abuse but still make pure summoners a viable class, even score-wise. The decent scoring summoners (like my DESu's) only do so by skipping a shitload of exploration, too, I might add.

[galehar]

I don't think ensuring that you can get the 15 runes with a single skill should be a design goal. If you have to diversify to kill Cerebov, it's fine. You don't even have to kill him, you can also distract him with your summons and try to snatch the rune and run. You can also have greater demons and ugly things to support your dragons if we go with a per spell cap as I've suggested.
I'm not sure what you mean by scaling the summons' strength with power. Most summoning spells already summon tougher monsters at high spell power. If you mean "scaled" monsters, with higher HD and damage, then I disagree. Spell power already affect number of summons per cast and type of summons. If it also affect the per-spell cap, then there's plenty of room to balance and scale individual spells.
Channeling is another issue. It's related sure, but I'd rather see it discussed in its own thread if anyone has suggestions about it.

[KoboldLord]

EDIT: I wanna add that, looking at the top 1000, there are like 10 summoners (*Su) that are not MuSu's. So even with summon spells as they are, I think it's obvious pure summoners have a hard time (apart from mummies, which are totally broken). Please consider that and work on balancing summons to prevent abuse but still make pure summoners a viable class, even score-wise. The decent scoring summoners (like my DESu's) only do so by skipping a shitload of exploration, too, I might add.

It is more likely that the lack of FooSu on the scoreboards comes from the fact that summoner-play is boring and nobody who wins with it will be in a hurry to experience the same style of play again. I don't care how easy MuSu and FeSu are; they are unfun and therefore terrible.

Forcing a FooSu to train a secondary skill with which to support their summons makes the playstyle more interesting and fun, and is therefore an improvement to the style even if it is strictly a nerf. If five dragons cannot kill Cerebov, this is not really a problem because I can hit him with a bar of holy wrath at the same time as my dragons go to town on him. Getting hit with Fire Storm builds character.

[pubby]

I don't think ensuring that you can get the 15 runes with a single skill should be a design goal. If you have to diversify to kill Cerebov, it's fine.

By 'diversify' do you mean pick up the handful of tools that are used by the majority of 15 rune wins (haste, holy/antimagic weapon, lich form, etc)? Diverse skills does not lead to diverse characters.

One of the strongest things about the 3 rune games is that it can be won with practically any skill set and play style. I'm not sure why the 15 rune game abandons this idea, although it's probably because the monsters aren't varied enough and skills reach 27 too early.
</offtopic rant>

It is more likely that the lack of FooSu on the scoreboards comes from the fact that summoner-play is boring and nobody who wins with it will be in a hurry to experience the same style of play again. I don't care how easy MuSu and FeSu are; they are unfun and therefore terrible.

The fact that summoning is not fun is the primary reason I made this thread and so I'm very interested in ideas on how to fix this. I would much rather play a fun but unbalanced summoner than vice versa.

Forcing a FooSu to train a secondary skill with which to support their summons makes the playstyle more interesting and fun, and is therefore an improvement to the style even if it is strictly a nerf.

Forcing is perhaps too strong of a word - you should be able to win without learning support spells or investing in melee or ranged, just like you can win a melee character without learning charms.

[Torious]

I don't think ensuring that you can get the 15 runes with a single skill should be a design goal. If you have to diversify to kill Cerebov, it's fine. You don't even have to kill him, you can also distract him with your summons and try to snatch the rune and run. You can also have greater demons and ugly things to support your dragons if we go with a per spell cap as I've suggested.
I'm not sure what you mean by scaling the summons' strength with power. Most summoning spells already summon tougher monsters at high spell power. If you mean "scaled" monsters, with higher HD and damage, then I disagree. Spell power already affect number of summons per cast and type of summons. If it also affect the per-spell cap, then there's plenty of room to balance and scale individual spells.
Channeling is another issue. It's related sure, but I'd rather see it discussed in its own thread if anyone has suggestions about it.

With a cap on the number of summons, individual summons would have to be stronger at high spell power, to compensate for the lack of damage output a pure summoner has. That's why I mentioned Cerebov as an example: Throwing 8 OODs in his face does a lot more damage than 5 golden/storm/quicksilver dragons, which he will quickly dispose of with firestorm, when you have to stay in LOS. I don't know how Crawl works internally, but give them more HP, buff damage, etc. Maybe "level them up". Maybe summon pearl dragons.

Anyway, it's not about Cerebov, but if summons are just nerfed and not re-imagined somehow, pure summoner is not very viable anymore. That may be OK with the devs, but that's the choice to make. If the devs would prefer summons to have only a supporting role, then nerf away. In that case, it might make sense to remove summoner as a class.

Regarding diversifying, there is always diversification in a build anyway; charms for Haste, necro/transmut for Necromutation, maybe transloc for CBlink, etc. A "pure" summoner is not that different from an FE who's firestorming it up all game. How's that diverse btw? Not that I'm complaining... As a final note on this, keep in mind that it'd be nice to still be able to get a highscore once in a while with a summoner. That means focusing skills and quickly getting your 1 endgame spell (firestorm, summon dragon, ... ). When going for a fast win there just isn't time to learn a second strong spell as a summoner (due to half XP), especially when not playing DE.

I support changing summons (change is def. needed), I'm just warning for destroying the summoner as a class/playstyle. I believe summoners are fun and can be even more fun!

===============

Some random ideas to improve on summoner gameplay, assuming caps etc.:

* A low cap on Summon Dragon, but long duration, so they can level up (or however that works).
* Summons killing things extends their duration, like how TSO blesses summoned angels.

* Make Regeneration targetable like Haste, so you can regen your dragons.
* A Charms/summoning spell that affects your summons (berserks them or so), as someone suggested earlier in the thread ("Battle Cry").
* A spell like "Creature Bond", which makes the summon permanent while this is in effect (shows a "Bond" status, like "Regen") and can be renewed. Part of the damage the summon receives is reflected to the player (or some other, similar drawback mechanism).

Ofcourse I'm not sure if this is any good in practice, but the line of thought is to work with lower numbers of summons, and get a little more creative with them. Perhaps this can also increase synergy with some other spell schools (mostly charms in these examples).

* Malign Gateway: Make it placable/targetable and useful in smaller spaces. Pretty much noone wastes spell slots on this. I used it current game for the heck of it, and it just confirmed this again.

* Give a +1 summoning aptitude to a octopodes in exchange for losing a tentacle . Demonspawn +1 summoning apt. might make sense as well, with the summon demon spells and all. That might make for interesting hybrid builds as well.

===============

Good luck figuring something out!

[crate]

"only 5 dragons" will get you a 3 rune win no problem

[XuaXua]

Better Ally Management.

[varsovie]

So let's try to sum up. Being able to kill monsters with 0 risk isn't interesting. So your post has convinced me. And since it seems I'll be the one implementing it, here is how summons will be in 0.12 (unless someone convince me otherwise with real aguments):
A summon can attack a monster only if it and its target are both visible to the player. This only applies to free summons, not permanent allies or god gifted allies.

Of course 0 risk is boring(and even with risk, Su is actually boring for me). But the whole point of the game is to keep the risk as near as possible of 0 to win (or as far as possible, and still win ).

For the implementation itself, your statement means that :
Canines won't be able to attack invis anymore unless you have see-inv.
Clouds will disrupt the allies, it can be a balance problem with summons/foo that can produce clouds (or fire over water) and some vaults.
The lantern of shadow (or shadow mutation for Ds) will reduce summon effectiveness.
Will transparent walls block the LOS regarding to summons?
Will the limitation also apply for Sticks to snakes, because yes it's non-permanent allies, but it's Tr instead of Su, and uses non-renewable resource to cast. Will it apply to summons gained through summon scrolls or evocable items.
An also, will the LOS limit also apply to enemy or ally summoners? I hope no, because having to flee from a bunch of useless summons wouldn't be fun.

Other than that, I'm looking forward to try Su again after your patch.

[galehar]

I don't think ensuring that you can get the 15 runes with a single skill should be a design goal. If you have to diversify to kill Cerebov, it's fine.

By 'diversify' do you mean pick up the handful of tools that are used by the majority of 15 rune wins (haste, holy/antimagic weapon, lich form, etc)? Diverse skills does not lead to diverse characters.

Diversifying means picking up a second source of damage, not learning some utility spells.

With a cap on the number of summons, individual summons would have to be stronger at high spell power, to compensate for the lack of damage output a pure summoner has.

Or have the cap depends on spell power. I keep repeating it, but you don't seem to grasp the concept. With high spell power, you can summon more dragons. Also, at high spell power you already have a better chance for a golden dragon or a quicksilver dragon.

Anyway, it's not about Cerebov, but if summons are just nerfed and not re-imagined somehow, pure summoner is not very viable anymore.

I think that's the point yes. Maybe pure summoner will still be viable for a 3 runes, but at some point, you'll have to get your hands dirty. It's the same for melee and poison, those 2 builds need also diversification. With them, it's even worse, winning 3 runes without some ranged attack or with only poison spells is a conduct game.

As a final note on this, keep in mind that it'd be nice to still be able to get a highscore once in a while with a summoner.

I don't think ensuring that all playstyle are able to reach a high score should be a design goal.

Canines won't be able to attack invis anymore unless you have see-inv.

I've already answered that before. Sorry for my poor choice of words, I meant the target needs to be in LOS, not visible (this is actually obvious if you think about the reason of the change).

Clouds will disrupt the allies, it can be a balance problem with summons/foo that can produce clouds (or fire over water) and some vaults.

Most clouds don't block LOS. And yes, if you or your summons spam fire spells in shoals and throw steam everywhere, they'll be less effective. Can't see how that's a problem. Just don't do it. You'll need to adapt your tactics to the change.

The lantern of shadow (or shadow mutation for Ds) will reduce summon effectiveness.

Yes. That's fine.

Will transparent walls block the LOS regarding to summons?

I think so, yes. Grates and trees might give some headache. Again, we need to think about the reasoning. Summons shouldn't be able to attack monsters that cannot attack the player. But with the various LOS rules and targetting types, it can get complicated. Let's start with the obvious cases, we can decide on the edge cases later.

Will the limitation also apply for Sticks to snakes, because yes it's non-permanent allies, but it's Tr instead of Su, and uses non-renewable resource to cast. Will it apply to summons gained through summon scrolls or evocable items.

I think it should apply to all temporary summons (s2s, evoked elementals, ...) but not permanent ones (unholy creation, efreet, ...). TSO summons might be an exception and be treated like permanent ones.

An also, will the LOS limit also apply to enemy or ally summoners? I hope no, because having to flee from a bunch of useless summons wouldn't be fun.

No, it would be too complicated to make the AI adapt to it. And there's no reason for it anyway.

[Torious]

With a cap on the number of summons, individual summons would have to be stronger at high spell power, to compensate for the lack of damage output a pure summoner has.

Or have the cap depends on spell power. I keep repeating it, but you don't seem to grasp the concept. With high spell power, you can summon more dragons. Also, at high spell power you already have a better chance for a golden dragon or a quicksilver dragon.

I can grasp the concept just fine, but apparently I wasn't clear: My point is that summoner damage output is a lot worse than a conjurer's, for example. That's why you "need" 15+ dragons for Cerebov. Let's say you cap at 5 dragons at high spell power (seems reasonable), then you've just made that fight impossible to win (yeah, sure you can steal, that's not the point). That's why I suggested basically beefing individual summons, to still get decent damage output.

If that's not gonna happen, that's fine, that just means you need to diversify as you say. It just means it will be even more rare to see pure summoners 15-runing. That's just a design choice and I can see how it makes sense. However, "pure" FEs, IEs, EEs, AEs and what not can 15-rune just fine. Perhaps the difference is the dual spell school (conj/fire f.e.) and "variety" in sometimes having to use OOD. Perhaps, in some far off future, summons can be reworked to revolve around managing and buffing a few very strong, semi-permanent summons instead of spammed hordes. That seems most interesting to me.

[KoboldLord]

I can grasp the concept just fine, but apparently I wasn't clear: My point is that summoner damage output is a lot worse than a conjurer's, for example. That's why you "need" 15+ dragons for Cerebov. Let's say you cap at 5 dragons at high spell power (seems reasonable), then you've just made that fight impossible to win (yeah, sure you can steal, that's not the point). That's why I suggested basically beefing individual summons, to still get decent damage output.

Your analysis is incorrect. Summoner damage output is higher than conjurors, but distributed over a longer period of time. Five dragons will easily take out any target that can't abjure them away, and a target that can abjure them away will still go down if you have a way to quickly restore mp when your summons are removed from play.

Have you ever actually tried taking out Cerebov with summons? It isn't actually all that hard, even with a normal build that doesn't have infinite channeling.

If that's not gonna happen, that's fine, that just means you need to diversify as you say. It just means it will be even more rare to see pure summoners 15-runing.

If a 'pure' character is 15-Runing, that means there's a problem with that build's main skill.

That's just a design choice and I can see how it makes sense. However, "pure" FEs, IEs, EEs, AEs and what not can 15-rune just fine.

What? No they can't. Not routinely, anyway. Conjurations-focused casters diversify into loads of magic skills for buffs and defenses, and in order to 15-Rune they usually want to diversify into melee, evocations, etc. too. You can't run to a safe room and mash 5 for more mp if you're in Hell.

[Torious]

Your analysis is incorrect. Summoner damage output is higher than conjurors, but distributed over a longer period of time. Five dragons will easily take out any target that can't abjure them away, and a target that can abjure them away will still go down if you have a way to quickly restore mp when your summons are removed from play.

Have you ever actually tried taking out Cerebov with summons? It isn't actually all that hard, even with a normal build that doesn't have infinite channeling.

Yes, you're right, and in fact I routinely take out Cerebov with dragons. Dragons are overpowered. Try taking him out with 5 dragons. That's not gonna happen. Partially because yes, Cerebov abjures. But also because per-round damage output (that's how I should have phrased it) is lower than OOD, for example. It's not that it should be equal by any means (summons have other benefits), but with a cap of say 5 dragons, that fight becomes a lot like fighting a death yak with a few war dogs. Thus, it would be nice to have a way to buff your fewer summons, or have them be stronger to begin with (at high spell power).

That's just a design choice and I can see how it makes sense. However, "pure" FEs, IEs, EEs, AEs and what not can 15-rune just fine.

What? No they can't. Not routinely, anyway. Conjurations-focused casters diversify into loads of magic skills for buffs and defenses, and in order to 15-Rune they usually want to diversify into melee, evocations, etc. too. You can't run to a safe room and mash 5 for more mp if you're in Hell.

Well aside from melee (unless you just mean Fighting skill) this is just for "support" or "utility", as mentioned above in this thread. If you mean iron shot, OOD, firestorm, necromut, haste, rmsl/dmsl, regen, high evo/invo for channeling and maybe tloc for cblink is diversification, a summoner does the same (with Greater Demon instead of OOD, f.e.). That's a no-brainer, and largely unrelated to your damage source...

[crate]

There is nothing about summoning that is forcing you to not avoid Cerebov and grab the rune.

I am pretty sure that 5 dragons will kill everything you will see in a 3 rune game with no problems at all.

I do not understand why being unable to kill the single hardest enemy in the game for summons with just summons is bad when against the vast majority of things in the game they will remain extremely powerful.

[galehar]

Perhaps, in some far off future, summons can be reworked to revolve around managing and buffing a few very strong, semi-permanent summons instead of spammed hordes. That seems most interesting to me.

I think this niche is better filled with permanent allies. They need improvement too, especially with the UI. Beogh could have some buffing invocations. Evilmike suggested giving him a touch range healing invocations for example. But I'm drifting off-topic.

[BlackSheep]

The lantern of shadow (or shadow mutation for Ds) will reduce summon effectiveness.

Yes. That's fine.

How would the shadows summoned by the Lantern be affected by a LoS change?

[sgrunt]

I've glanced through this thread, and the major thread of logic that sticks out to me is that summoning, by itself, is not inherently broken; what is broken is summoning with an effectively infinite mana supply, as with the case of MuSu of Sif Muna.

If we wish to fix that, we should be looking for a simple solution that targets that specific case - we need a tweak that tones down the effectiveness of summons as the mana spent on them increases.

I've had an idea floating around for a while of a "soft summoning cap" - given summons of a specific type (demons would be one broad category; other categories could be established around other spells), the chance of a successful cast of the spell actually summoning a create falls off with the number of creatures of that type you currently have summoned (and, possibly, increases with power). To be appropriate as a solution for the stated problem, the soft cap would need to be chosen such that there's, say, a 50% chance of a successful summon when you have enough summons to have burned through an average mana bar.

Let's crunch some numbers. An XL 14 human summoner with 14 in the appropriate skills has 34 maximum MP, and can cast Summon Demon, a level 5 Summoning spell, just short of seven times; empirically, they regenerate one MP while casting so as to make it a full seven times and have zero MP left over. Given the aforementioned skill setup, they cast the spell at power 51 (of 200) outside of any enhancers.

Let's say that we want the chance of a successful summon to be 1 in (1 + a * n), where n is the number of spell levels of summons of that type currently present and a is a constant we want to find. (You might recognise this as a low-pass function, which has the desired property of being high for low values of n and low for high values of n while always remaining between 0 and 1 for non-negative values.) At the limit of the mana bar in the above case, we have n = 35; thus, if at this point we want the chance to be 1 in 2, then a = 1/35.

If we want to factor spell power into it, and we want to keep the above case working as intended, then we could, for example, set a = 50 / 35 * (spell power) - observe that as spell power increases, so does the chance of success for a given number of spell points. Of course, if we want spell power to be more or less of a factor - say, we want it to be half the factor in the above - just replace (spell power) with, e.g., ((spell power / 2) + 50 * 1/2).

The aim in taking spell power into consideration would be to keep the half point roughly consistent over the range of spell power. Returning to our human summoner above, at XL 27 and 27 of the appropriate skills, he has 49 MP and can cast a pure summoning spell at 171 power; thus, we'd aim for 171 spell power to result in a half chance at 49 summon points. If we fit the two data points above, we can obtain a ~= 50 / (35 * (spell power / 6 + 50 * 5/6)).

We could, of course, refine things further by taking some more data points into consideration, but on the surface this appears to be adequate for most summoning cases; I'd have to see how it performed in real game conditions to be able to tune it further.

Thoughts?

[crate]

There are several things that make summoning really powerful:

1) Because you make other things, it is inherently safer than any other method of killing enemies (especially since the vast majority of attacks in the game require line-of-fire). It also lets you retreat from nearly every fight.

2) You are allowed to spend all of your mp before you even see an enemy, or if you do see an enemy but it doesn't wake up. You cannot spend 50 mp on conjurations without an enemy getting an action in (unless you paralyse/petrify the enemy or something). In addition most summons last for a super long time so you can just walk around with a few effectively-permanent allies. Also you can combine this with 1) to, for instance, retreat to a staircase and then return to the fight with up to 8 new allies instantly.

3) Summoning power doesn't grow linearly with the number of times you cast the spell. Two dragons are significantly more than twice as powerful as one against most enemies. This can be somewhat compensated for by making a single cast of a summoning spell generally less powerful than a single cast of a same-level conjuration, though I'm not sure how much this is actually the case. But because of 1) and 2) generally you will never have to fight with just a single summon anyway.

Also it is convenient that spell power matters little for most summons. So with wizardry/Vehumet/Sif you can cast summoning spells at nearly full effectiveness with less skill than conjurations of the same level.

I am not going to offer suggestions for making summoning less powerful but pointing out the reasons it is powerful seems like a good thing to do.

[varsovie]

The lantern of shadow (or shadow mutation for Ds) will reduce summon effectiveness.

Yes. That's fine.

How would the shadows summoned by the Lantern be affected by a LoS change?

Because they are non-permanent and non-god allies bound to the player (or the lantern). Since the LOS goes from 8 to 7 with shadow, it means that the summons will fight closer.

[sanka]

I've had an idea floating around for a while of a "soft summoning cap" - given summons of a specific type (demons would be one broad category; other categories could be established around other spells), the chance of a successful cast of the spell actually summoning a create falls off with the number of creatures of that type you currently have summoned (and, possibly, increases with power). To be appropriate as a solution for the stated problem, the soft cap would need to be chosen such that there's, say, a 50% chance of a successful summon when you have enough summons to have burned through an average mana bar.
...
Thoughts?

I must admit I'd like the idea of a soft summoning cap, however, I would like to see something much more transparent. It sounds complicated to follow the actual number of your summons and guess which one is good to summon.

Of course we could display the actual chance for casting a particular summon, but since you mentioned a per spell chance this may be a lot of info. You may need to switch to the spell screen too much for my tastes.

Summoning is tedious enough - I'd rather give a fixed cap for each spell and if you reach it you cannot cast it anymore - so you need not to check anything, the game will tell you if you reach it.

But, as crate has pointed out, summing is very strong because it's very safe. I think the first thing to nerf should be the LOS requirement one. This does not makes summoning very dangerous, but at least a little bit more dangerous than now.

[BlackSheep]

How would the shadows summoned by the Lantern be affected by a LoS change?

Because they are non-permanent and non-god allies bound to the player (or the lantern). Since the LOS goes from 8 to 7 with shadow, it means that the summons will fight closer.

I didn't mean why would they be affected, I meant how, but my question was born out of a misunderstanding of the effects of the lantern on LOS. Lantern sets stealth to zero, but I got confused and thought it sets LOS to zero (which would be funny).


Regarding a soft cap, I think such a system would just encourage more of the same behavior. (i.e. spamming summons and channeling) Maybe if there were miscast effects for failure?

I think dpeg's general guidance to make changes gradually over time is sound. Start with a LOS change, see how things work out, then if more work is needed add a cap of some sort.

[njvack]

I've had an idea floating around for a while of a "soft summoning cap" - given summons of a specific type (demons would be one broad category; other categories could be established around other spells), the chance of a successful cast of the spell actually summoning a create falls off with the number of creatures of that type you currently have summoned

By 'successful cast' do you mean failures would result in miscasts?

Getting an increasing chance of miscasts (not blocked by Sif) as some function of number of active summons and spell power would also limit summon spam. Even minor miscasts would glow you enough to make you decide whether you really need that sixth dragon.

[pubby]

I've glanced through this thread, and the major thread of logic that sticks out to me is that summoning, by itself, is not inherently broken; what is broken is summoning with an effectively infinite mana supply, as with the case of MuSu of Sif Muna.

This is only the case because summoning spells (in this case, anyway) are cast before fights instead of during. Unlimited channeling is very cool (and balanced) with conjurations and such - remember it still takes a whole turn - I'd hate to see it go just because of the two worst designed spell schools in the game.

I hate to sound like a broken record but having summons cost MP til they die as I mentioned in the OP should solve this. It prevents you from channeling to get extra summons, but allows you to quickly replace them. Summoning before fights will leave you with just as much MP as summoning during.

[XuaXua]

I hate to sound like a broken record but having summons cost MP til they die as I mentioned in the OP should solve this.

This could work, conceptually. Every cast of a summoning spell reduces your MP maximum by the MP cost of the summoning spell until the summons disappear. Can't recover or access that MP until that time.

Summon 4 dragons (MP = 9?), reduces your total by 36 MP. It's sort of like that maintainable Hexes/Charms concept that's being worked on.

It's also simple and easy to track. Of course, can it be abused with the lower MP monsters? Yeah; 50 MP isn't that tough to obtain.

[galehar]

I've had an idea floating around for a while of a "soft summoning cap" - given summons of a specific type (demons would be one broad category; other categories could be established around other spells), the chance of a successful cast of the spell actually summoning a create falls off with the number of creatures of that type you currently have summoned (and, possibly, increases with power).

I think it would increase even more the summoning/channeling synergy. With channeling, you can keep spamming and get more summons than without. I don't see how it's better than a hard cap. It's less efficient and more complicated. You also have to track the fail chance somehow, whereas with a hard cap, the spell is just disabled at the limit.

1) Because you make other things, it is inherently safer than any other method of killing enemies (especially since the vast majority of attacks in the game require line-of-fire). It also lets you retreat from nearly every fight.

Yes, although it's kind of the point, so maybe it's a feature . The easy escape is a bit worrying, but I'm not sure what could be done about it.

2) You are allowed to spend all of your mp before you even see an enemy, or if you do see an enemy but it doesn't wake up. You cannot spend 50 mp on conjurations without an enemy getting an action in (unless you paralyse/petrify the enemy or something). In addition most summons last for a super long time so you can just walk around with a few effectively-permanent allies. Also you can combine this with 1) to, for instance, retreat to a staircase and then return to the fight with up to 8 new allies instantly.

We could reduce some durations, but that wouldn't really address the issue. Maybe if they don't take stairs that would help a bit too?

3) Summoning power doesn't grow linearly with the number of times you cast the spell. Two dragons are significantly more than twice as powerful as one against most enemies. This can be somewhat compensated for by making a single cast of a summoning spell generally less powerful than a single cast of a same-level conjuration, though I'm not sure how much this is actually the case. But because of 1) and 2) generally you will never have to fight with just a single summon anyway.

Caps should help here.

Also it is convenient that spell power matters little for most summons. So with wizardry/Vehumet/Sif you can cast summoning spells at nearly full effectiveness with less skill than conjurations of the same level.

Well, if caps depend on power, that would increase the importance of spell power. Maybe some high level spells could be nerfed a bit when cast at low power.

I am not going to offer suggestions for making summoning less powerful but pointing out the reasons it is powerful seems like a good thing to do.

Thanks for your analysis.

I hate to sound like a broken record but having summons cost MP til they die as I mentioned in the OP should solve this.

This isn't really different than my hard cap proposal. The upside is that it is very simple and easy to understand. But I think a per spell cap might be more interesting and easy to balance. A global cap should depend on the summoning skill, not the number of MP. A staff of power allowing you to control more summons doesn't make much sense. With your system, the best summoner has a lot of spellcasting, rings and staff of power or wizardry. A staff of summoning wouldn't be much useful and summoning skill maybe less important than spellcasting.

[varsovie]

I've had an idea floating around for a while of a "soft summoning cap" - given summons of a specific type (demons would be one broad category; other categories could be established around other spells), the chance of a successful cast of the spell actually summoning a create falls off with the number of creatures of that type you currently have summoned (and, possibly, increases with power).

I think it would increase even more the summoning/channeling synergy. With channeling, you can keep spamming and get more summons than without. I don't see how it's better than a hard cap. It's less efficient and more complicated. You also have to track the fail chance somehow, whereas with a hard cap, the spell is just disabled at the limit.

Why not instead of a miscast, make the summon over the soft limit have a chance to being hostile. So super-summoning before or during a combat is non-effective and dangerous.
It still have the disadvantage of being more complicated, and possibly abusable : eg, if after 5 dragons, they have 90% chance of being hostile, with enough patience/chanelling/Mu/boredom, it might still be possible to always have 8 dragons before going downstairs.

[crate]

The thing about maybe making the next dragon you summon hostile when you already have five dragons ... is you already have five dragons, so they can just kill the one dragon. You make casting summoning in-combat bad by doing this, but in-combat summoning is not problematic so you are not really fixing anything.

[dd]

Sorry if this was suggested already.

How about if summons expire faster when they're further away from you? Each summons has X amount of time before they vanish, and each turn X = X - (distance between player/summons). The distance could be lower-capped to 5 so you couldn't abuse it by staying adjacent to your summons to keep it from disappearing.

[nicolae]

The thing about maybe making the next dragon you summon hostile when you already have five dragons ... is you already have five dragons, so they can just kill the one dragon. You make casting summoning in-combat bad by doing this, but in-combat summoning is not problematic so you are not really fixing anything.

What if, if you summon too many, the ones you already have turn hostile, but still count as your own summons for caps and such?

[SmilePic]

I dont' really get the restriction of summon being able to attack monster only in summoner's LOS.
I can imagen raging dragon being tickled to death by goblin that the dragon can see but summoner can't. =)
It's efective nerf but it somewhat ruins common sense.

[BlackSheep]

It's a case of gameplay trumping realism. While the goblin vs. dragon image is colorful, the 12 dragons stomping on Pan lords while the player is over by the stairs channeling is the bigger problem.