Ranged Combat overhaul, 0.6 target

Trunk has the Arcane Marksman class and we plan to keep it. However, it comes with new spells which brand ammunition permanently. As discussed, this has to be repaired, for 0.6. The resulting proposal targets some more related changes. Stuff that is still missing:

  • How to change the two AM starting books and the third book.
Hopefully done to a suitiable degree. — due 2010-01-01 14:14
Yes, absolutetely. — dpeg 2010-01-03 18:49
  • How to differentiate bows from crossbows.
    • Could be okay for now: different availability, different mulching rates, different brands and aptitudes. — dpeg 2010-01-03 18:49
  • Generation and mulching numbers; these should be decided when implementing this.
    • Not sure what this point refers to. — dpeg 2010-01-03 18:49

List of ammunition brands

  • Flame
  • Frost
  • Poison
  • Returning
  • Chaos
  • Penetration
    • For me, it goes too far into Conjuration territory, and makes the two combat techniques too similar. — b0rsuk 2009-12-12 10:47
    • Yes, not used as spell. Can only be found on some missiles, so is rare.
  • Reaping
  • Dispersal
  • Exploding
    • I propose to replace this with a new type of “area” ammo, thrown or fired farther with a sling. — b0rsuk 2009-12-12 11:26
  • Steel
  • Silver
  • Pain

(Removed electric, added pain.) It might be an option to turn the materials (steel, silver) into attributes like elven and orcish. But I would wait with that until the basic system works. I could imagine that holy wrath could be an interesting brand, since there is no spell equivalent. \\Note: Pain not used. — dpeg 2010-01-08 02:04

Brands per missile type

  • Needles: poison, curare (as now).
  • Stones: Never generated with brands, but slings can be branded.
  • Large Rocks: No brands whatsoever.
  • Javelins: Returning, Penetration, Steel, Silver, Chaos.
  • Nets: Steel, Silver, Chaos.
  • Hand Axes etc: Can be branded using spells.
  • Darts: All brands except for Returning, Penetration.
  • Bullets: Flame, Frost, Poison, Chaos, Steel, Silver, Exploding.
  • Arrows: Flame, Frost, Poison, Chaos, Reaping, Dispersal.
  • Bolts: Flame, Frost, Poison, Chaos, Steel, Silver, Penetration.
Done. — due 2010-01-01 09:01

This list is not the final word — changes may come up. The point is to illustrate how restricting brands for missile types can make ranged combat more interesting.

Ammunition can be generated randomly with these brands. Chance depends on type (e.g. silver could come up more often on bullets than on bolts) and need not be flat within a type (details not here). Each brand (temporary from branding the launcher or not) increases mulch by a fixed, across-all-types factor

Branding spells

  • Spell-branded melee weapons and launchers allow rebranding.
  • Unwielding a temporarily branded item cancels the brand.
  • The new branding spells work on the launchers and temporarily brand them. No effect on permanently branded launchers (as for weapons).
  • Branding spells for: Flame, Frost, Poison, Pain. These work on weapons and on launchers — no new spells are needed. Apart from pain, temporarily branded launchers can be fixed using a scroll of vorpalisation. Done, except for Pain.
  • Not all ammunition brands should have spells, see above.
  • Launchers can only carry some brands, see below.
  • Launcher with egos have no effect on branded ammunition.
All done, apart from Pain; requires implementing a Pain missile brand and discussion should be had as to how this is to function. — due 2010-01-01 14:14
Let's not do pain brand branding on missiles. (On launchers may be okay.) — dpeg 2010-01-03 18:49

Combo brandings (e.g. a flamed poison arrow) are possible and not overpowered. However, they are a minor detail, so ignored here. The quality of the new Ranged Combat system does not depend on them.

All combo brandings except for flaming poison arrows (in order to not nerf Nessos) have been disabled. — due 2010-01-01 14:14
That's okay but the fact that flaming + poison is now possible is spoiler information. Would it be feasible if players couldn't use this, and Nessos's arrows were hardcoded to deviate from the norm? — dpeg 2010-01-03 18:49

Launcher brands

  • Remove speed brand, except for randart launchers (as with melee).
Done — due 2010-01-01 09:01
  • Keep the vorpal brand but make it rarer. Done. — due 2010-01-01 14:14
  • Launchers can carry brands their ammunition is eligible to, with the exceptions of Returning, Dispersal, Exploding, Steel.
  • Allow the Penetration brand but only on randart launchers whose missiles are eligible for the brand (eronarn). Done.
  • While stones never come with brands, slings can be found with the appropriate brands. Done.
  • Replace Protection ego on launchers with Evasion. Done.
  • Chances for brands depend on launcher type (no details here).
    • What are the current chances? — dpeg 2010-01-03 18:49


Here is what everybody could agree with:

  • Remove hand crossbows.
  • Subsume Darts skill under Throwing.
Both done. — due 2010-01-01 09:01
  • Items under Throwing skills get a Str/Dex weight. The effect of that weight can be specified later — we want a Troll with Throwing 24, Str 32, Dex 8 to have problems using darts; a Halfling with Throwing 24, Str 8, Dex 24 should have problems using javelins.

More may happen with needles, see below. If we keep the skills Throwing, Slings, Bows, Crossbows, then Throwing and Slings should crosstrain since they both use stones.

That's not really a sufficient reason. Throwing hand axes doesn't use Axes skill, after all, so why should throwing vs. slinging stones? — eronarn 2009-11-26 11:58
I think it would be possible to allow that cross-training without breaking consistency. It's not necessary, of course. Just one more of those idea to boost Slings a bit more. Should probably be kept for later. — dpeg 2010-01-03 18:49


Needles and Blowguns

A fickle issue. Everybody agrees that a good niche for needles is this:

  • Silent. Done, and already existent.
  • No direct damage; if damage, then over time (a la poison). Status effects are also okay. Currently, this is how they work.
  • Apart from Poison, the needle brands are exlusive. We could use a few more in addition to Curare, for example Paralysis, Rotting, Blindness (see 2901420). Note that since all needles are branded, the branding spells are useless on blowguns.

Another suggested brand: Sleeping or Tranquiliser. Should cause Ensorcelled Hibernation-esque sleep effect that doesn't check cold resistance. Could be as rare (or rarer) than curare, and provide a non-spell like assistant to stealth-based fighters. — Jude 2009-11-26 08:52
Another one: Anti-magic. Player hit loses all MP, monsters, lacking MP, will be unable to cast for a while. — kilobyte 2009-11-26 09:59
Several of these are in. But not all — please move the missing ones to Ranged Combatdpeg 2010-01-08 02:04

  • Both skill and blowgun enchantment should matter more when it comes to duration/potency of the effects (again see FR).
Done, new brands have been added for needles. Please see the list at the end of the page. — due 2010-01-03 14:55

There were voices that needles and darts would be too similar, but I am not sure as the effects are completely different. Nevertheless, here are some ideas that came up to differentiate better:

  • Reduce range of darts to 4, decrease damage variance.
Not currently done, but I don't think reducing the range will help here. — due 2010-01-03 14:55
Decreasing the damage variance is a good thing. On D:1 I had a character find a suit of crystal plate mail. On D:2 a kobold hit me for 8 damage with an unbranded dart… but too often, darts don't feel like they're doing anything at all. On the other hand, it makes no sense to give darts smaller range than blowgun needles. — eronarn 2009-11-26 11:44
Counter-opinion: disagree with “it makes no sense”. Darts don't need a launcher; needles should be compensated. Measured against reality, this nerf makes sense. — wtanksleyjr 2009-11-26

Okay, I propose to leave in the specific needles skill for now. There are some interesting ideas here for how to make needles better, we should collect them and apply at a later stage: see Ranged Combat. (Feedback and testing for the new needle brands should come first. And “later” means post-0.6.) — dpeg

Random thoughts

  • Large rocks mulch into stacks of stones (mostly flavour). done
  • Penetration mulching is not applied for each creature hit. Otherwise we just encourage players to micromanage aiming at fewer monsters than are in the corridor.

Targets left to complete

  • Blowgun mechanics (silence) No change required: blowgun shooting is silent, as are needles hitting targets.
  • Blowgun brands (disease, tranquiliser, etc) Done.
  • Penetration mulch mechanics
Still not done. Will probably have to wait. — due 2010-01-03 14:55
  • Protection → Evasion Done

Throwing mechanics

Dpeg wants to differentiate throwables by statting in a way that matters. Discussion on IRC (sorear, Eronarn, doy, gunofdis) has produced the following plan:

  • Every item has a base delay, a base to-hit, a base damage, a strength to-hit bonus, and a basic range
  • Delay = base delay, possibly modified by body size/handedness; no stat influence; modified by skill exactly like a weapon
  • To-hit = base to-hit + perstr*strngth + X*dexterity + W*throw; X is around 0.5 or 1
  • Damage = base damage * (100% + Y*strength) * term for skill; Y is around 10%
  • Range = base range + Z*str; Z is around 0.2
  • Optional: Strength to-hit modifier and range are derived from item mass

This will be experimentally implemented and the numbers will be played with to get relative balance. It may be extended to launched ranged combat if successful.

Thoughts from — eronarn 2010-01-02 08:27
Blowguns, and possibly dart, str/dex utilization could work like so:
Needles do damage which is increased by Str but always quite low. First, worn armor is checked. Dex and skill both provide a chance to bypass it, multiplicatively. If it's hit, the impact makes noise at some chance based on strength and projectile weight, and damage is decreased appropriately. Any leftover damage (usually none through armor, but e.g., large races might be able to get through leather, or damage-enchanted blowguns could be used) reaches the monster's “racial” AC. For most “fleshy” monsters this is reduced dramatically and then applied to protect against the needle's damage - thick skin doesn't help against blowguns much. Obviously, this doesn't apply to monsters like iron trolls. If 1 or more damage make it through, the poisoning effect applies. This poisoning effect should be skill-dependent (delivering the poison to a useful location such as the neck), but default to the no-skill effect if the Dex*Skill-boosted check wasn't able to bypass armor (the projectile was poorly placed).
Note that Stabbing skill for blowguns could consist of adding your skill in Stabbing to your skill in Throwing for delivering poisoning effect. Or averaged in, or they could have weightings, etc.
This outcome check would be distilled into four simple messages: “The blowgun dart clatters off the foo's armor!” “The blowgun dart fails to penetrate the foo's tough hide!” “The blowgun dart pierces the foo!” “The blowgun dart pieces the foo's neck/eye/whatever!”

Some comments on this by dpeg:

  • The system seems okay. I will not go into the numbers right now, but rather describe some effects I'd like to see come out from the mechanics.
  • Throwing large rocks should be noticeably slower than throwing stones (for example). (Ideally also as slow for monsters.)
  • Needles should never do direct damage on impact.
  • Range should never go beyond LOS. That means 8 — I know about the diagonal abuse but we won't address that now.
  • I agree that Strength should influence range and speed (i.e. damage) for heavy thrown items. I am thinking of a Str 6 Ogre/Troll only reaching 3 squares or something.
  • Strength should have probably no (or very little) impact on range for stones. But it should have an effect on damage for stones.
  • To-hit should be affected by Dexterity, but inversely proportional to item mass: Dex wouldn't matter for large rocks, would matter a little for javelins, and a lot for darts and most for needles.
  • For Needles, we should make clear how Dex vs skill works. Here's one idea: Dex affects to-hit (strongly), skill doesn't. On the other hand, skill (alone) affects chance for the needles's brand to apply, and if so, how long.
  • Bolt damage and range should explicitly not depend on Str or Dex. To-hit should depend on Dex. Base delay for bolts should be slower than for arrows and be mitigated by skill and Str, perhaps using a term skill*Str? (I know that the above formula only deals with throwables; just saying.)
Are you sure you want to use Dex and Str in this way for crossbows? Seems to me there should clearly be one stat that is most important for a given launcher – otherwise the mechanics feel muddy. With the system proposed here, when someone asks “I'm a dedicated crossbow-user; should I raise Dex or Str,” the answer is going to either depend on some fiddly calculations, or else be that it doesn't matter. Neither of which seems desirable. Since the Crossbow-Str link seems more natural, I recommend using the speed-Str link as here (tho added to skill rather than multiplied) and making accuracy depend almost entirely on skill, with Dex playing at most a minor role. — lemuel 2010-01-08 17:48
  • Returning to the formulas:
    • For Range, we could also use a system that penalises low Str (depending on item mass), e.g.
      range = min(3, base - mass/(15*Str))
    • damage = base * (1 + sqrt(mass)*Str/20/60) * f(skill)
      This means double damage with large rocks at Str 60. It means +10% for javelins with Str 20. And no noticeable effect for the other throwables (in contrast to what I said above for stones, but that's because stones are too light, see below.)
    • I don't like the given to-hit formula as it stands. What about this:
      to-hit = base * (1 + Dex/sqrt(mass)/30) * f(skill)
      This would leave large rock to-hit essentially independent of Dex; javelins would get +12% at Dex 30; the light items would profit a lot from high Dex (up to +100% for needles at Dex 30).

For comparison, here are the numbers (item / base damage / mass):

needle          0       1
stone           4       2 
dart            5       3
arrow           7       5
bolt            9       5
large rock     20     400
sling bullet    6       4
javelin        10      80
throwing net    0      30

> Why are arrows heavier than stones? Even darts are heavier than stones? I suggest to gives stones a weight of 6 or so.

New needle egos

  • Sleep
  • Paralysis
  • Rage
  • Confusion
  • Sickness
  • Slowing

Names and numbers will have to be changed. Currently, each brand has a base power, with an additional power based on your throwing skill (or HD, for monsters), modified by the enchantment (positively or negatively) of the blow gun. Also needing to be looked at here is resistances to these.

Good list. I feel that Rage will be useful only if it (temporarily) also turns the monster neutral (i.e. will attack anyone nearby) — we could use B0rsuk's insane once we have that. (No additional effect on the player, of course.) Is Sickness useful in the hands of the player (surely is in the hands of a monster.) What are the generation chances for these brands? — dpeg 2010-01-03 18:49
If regular poison and curare are going to stick around, sickness would seemingly only be useful vs trolls and maybe hydra. — porkchop 2010-01-08 19:13

Per due's request - Slings really do not need to train Throwing skill at the same time any more. Their Shield + Sling buff, along with all of the other launcher buffs makes them stand alone quite well now. Crosstraining with Throwing is still a great idea, but if you're using a sling you're also probably trying to skill up Shields as well now, and the xp diverted into Throwing is really starting to hurt you. Additionally it messes with Acquirements and Gifts. Imagine how funny it would be actually wanting a Sling randart for the first time since you started playing Crawl, and getting a stack of Darts instead. — vandal 2010-01-04 01:01

Testing Feedback

purge: Needles of sleeping are too powerful currently. I think an appropriate starting nerf would be be make them 100% mulch on use. Confusion and slow are not as bad since there is still some danger, but perhaps their mulch rate should be increased as well.

Logged in as: Anonymous (VIEWER)
graveyard/ranged_combat_overhaul_0_6.txt · Last modified: 2011-05-02 16:15 by galehar
Recent changes RSS feed Donate Powered by PHP Valid XHTML 1.0 Valid CSS Driven by DokuWiki