Combat Maneuvers & Weapon Reform

New header by dpeg 2010-04-27 13:36: This wiki page is about special stuff (“moves”, “combat maneuvers”) you want to do with melee weapons. Here are some guiding principles:

• No active abilities. Whatever weapons will do, it should be triggered passively. Of course, circumstances (nearby monsters, short-term history, stats) should play a role.
• The new features should be positive. Because there's no ability or toggle, we don't want players to avoid melee combat for fear of a negative effect. One example: pushing back a monster is cool, but it could turn out that players quite often don't want this behaviour.

Note that these rules should guarantee that melee stays very light on the interface. Simply standing there and attacking adjacent monsters should be as good as it is now; for a better performance, positioning etc. might play a somewhat larger role.

Also note that many attacks will be negative in corner cases, imagine cleaving with an axe when you target is surrounded by jellies. This is rare, and we can make Ctrl-dir (* dir) attack safely (i.e. no move) and/or add logic to prevent negative moves.

We don't want these moves (only) because they're nifty. Much rather:

• Weapon differentiation. Currently, it doesn't really matter if you're wielding a mace or an axe or a sword. The distinction of weapon types currently employs skill, rarity (hard to find a very good mace etc.), brands (the egos a weapon can have depend on the type), numbers (speed etc.), handedness and, to a lesser extent stats, god and whether its a butcher tool. All of this is good and should stay, but the fact remains that for many fighters, the weapon they use is not a very relevant choice.
• More tactical combat. Currently, it is almost always optimal to entrench yourself in a corridor and beat the opposition one on one. One way to challenge this is using monsters (interesting attacks, or AI). But another one is to make fighting in the open more attractive.

Special moves should generally belong to a weapon type (e.g. axes) but it is conceivable that some of them apply to only special weapons (e.g. whips) or to other sets (e.g. two-handers).

Below we collect ideas for such moves. There is the question of when and how often the effects should trigger. While it seems natural to link those chances to weapon skill, I think it's much better to use stats instead (and remove the effect of stats on damage). In this way, we can hope that two late-game axe fighters with Str 10 and Str 30 would still play differently.

kilobyte 2010-04-27 14:40 (reply to the old version): I find most of such “special moves” to be really contrived and artificial. Let's limit these to only weapons where such a property comes naturally. This means, Polearms should all have reaching, and dpeg's idea to remove the Stabbing skill outright and make it use the Short Blades skill (if you're wielding one) or dex for a minor bonus otherwise sounds really tempting. Cleaving for axes would for example be either silly or an usability nightmare. The two weapon types I mentioned really could use a buff – Polearms were strongly hit by the 1.5-handed and AC nerfs so they suck for anyone non-merfolk, and Short Blades… well, you know the story.

squashmonster 2010-04-29 19:50: If enough good special move ideas can be found, you could further differentiate characters by making different stats apply to different special moves. For example, with axes, cleave could be based on dex and charge on strength. This would make a Str 30/Dex 10 and a Str 10/Dex 30 character much more different in play style. And if int gets used for some, that fixes the “int is not useful for fighters” problem quite handily.

gaff 2010-07-03: I was playing princess, a 7drl, and it makes heavy use of this sort of move. It actually adds a lot of depth to the game and makes it much more interesting than the usual retreat into a corner and mash a button. If these moves add depth why not make them available to skilled fighters wielding any weapon? I've added some of the interesting moves to the list.

b0rsuk suggested elsewhere in the wiki that different weapons could have different Stabbing effects (like clubs have now).

blinkfrog 2010-08-10 There was some special attacks in Dungeon Crawl: Vampire Patch

Those don't look very passive to me. — mrmistermonkey 2010-08-10 11:59

I looked at everything on this page, and it looks like a nightmare for new players. I don't want to see people have to look up the advantages of their chosen weapon before they can effectively play a melee character; that'd drive people away. And if you don't have to know about it to play an effective melee character, then that means this amounts to little more than overcomplicated flavor and there's just about no payoff for all this work. It's not like mages where they learn spells one at a time and can refer to the description of each; these are a package deal you get from square one when playing what most would assume should be the game's “simplest” characters. If the goal is, as I believed before I started reading the move descriptions, to differentiate the weapon types and ideally give incentive for a person to train more than one melee skill, a much simpler solution might be better: Just give each weapon class advantages against some opponents. Axes and maces get an advantage against zombies and skeletons, maces get a bonus equal to the EV penalty of a humanoid enemy's armor, spears get some advantage against animals and drakes/dragons, etc. Stabbing for short blades and universal reaching on polearms would make those two pretty differentiated, and if you can come up with something else that significant for other weapons that'd be cool, but if it's not zero or one significant special features per weapon type (with whips counting as their own type), plus some intuitive advantages or disadvantages against certain broad enemy types or situations, you're asking for trouble. — brickman 2010-11-07 23:00

Yet another thing to consider: are monsters going to use these? When players see that they have these perks, they'll expect monsters to get them too. This isn't a problem for stuff like reaching or staves blocking, but for things like axe cleave they will be hard to balance. — minmay 2011-04-17 15:32

(Excuse the snazzy subtitle, but I wanted this subdiscussion to get into a subsection of its own.)

I don't like tying any of this stuff to stats, as making a weapon class's unique feature require dex is effectively saying that you shouldn't use such weapons for an AC or casting build, which is basically unbelievable. “Two late-game axe fighters with Str 10 and Str 30” wouldn't exist, as the one with 10 strength would have no reason to use axes at all, and would instead have made the obvious decision to choose a weapon type aligned with that character's favored stat. The notion that this would increase variety is very misguided - it'd instead result in each stat prescribing specific weapon classes (and Squashmonster's proposal is just as bad, as it's still dictating stat-specific playstyles). For the same reason, dex shouldn't be much of a factor in stabbing (if at all), as low-dex races would be largely locked out of the stabbing playstyle. All weapon effects should be based on weapon skill, and possibly fighting skill. — og17 2010-05-19 02:54

I don't have the time to refute this profoundly negative comment in full details, but here are some hints: (a) assumption that melee users have to keep looking like they do (or used to do), e.g. all Str; (b) confusion between “weapon move” and “weapon class's unique feature”; © strong assumptions (wrong of course) about how large the effects will be; (d) presenting a conclusion as the only one, as so often, when there is a plethora of options. — dpeg 2010-09-26 23:29

For (A), tying stats to weapons gives players yet another reason to pump a single stat - how could this possibly result in the opposite? (B) makes little sense, as without these moves, the only “unique feature” of weapon classes is if they can butcher or chop off hydra heads (and currently, short blades and stabbing). (C), too, as power is irrelevant; unless the effects are negligible, players will want to make use of them. (D) is empty, as if my position is “don't use stats,” it's a given that I'm not going to support options that use stats.

Are you misunderstanding my objection of how this would limit player options? Your response seems pretty off the mark. — og17 2010-09-27 20:28

I agree with dpeg that using stats for weapon moves would gives players a reason to diversify their stats instead of focusing on one. The idea is that instead of putting everything in str and ending up with a dex:10, str:30 fighter, you would choose to put a few points in dex because that would enable you to perform a certain move which is tactically interesting for your char. Of course, you need to be able to perform it in heavy armour, or else it doesn't make sense. You would end with a dex:13, str:27 fighter that plays a bit differently. The difference between a str of 27 and a str of 30 isn't huge anyway, so players only need a little incentive to choose to diversify. — galehar 2010-09-27 21:39

You're handicapping yourself if you take a 13 dex 27 str fighter and fuel a dex ability with only 13 dex. It'd be more effective to put that high str to work with a str ability (and a str-pumped character). If your fighter wanted to use a dex ability, you'd go evasion and pump dex so you'd have a stronger ability and be making full use of those stat points' other effects. And it's not the stats that make things “a bit different” in your example, as your player could make identical decisions under a statless system; burdening the player with additional armor-like stat reqs doesn't result in meaningful diversity to begin with (as you say), and it actively discourages actual diversity to boot. This is why this doesn't work. — og17 2010-09-27 23:21

Why not give each weapon type (at least) two abilities, perhaps useful in different situations, and tie one to each stat? It might require some extra creativity to give maces a dex ability, though. — danei 2010-09-28 15:34

Okay, here we go once more. You try to give a proof that using stats cannot work. This proof relies on certain assumptions which are not valid.

• A melee fighter will want to use Str and nothing else. There may be EV fighters and AC/EV hybrid fighters in the future.
• There is much more to stats and melee than just the moves we are discussing here.
• The intention of the weapon moves never was to dominate melee combat. Rather, we plan to make melee combat tactically more interesting, and in ways that characters biased towards Dex will occasionally prefer a different tactical solution (run vs stand, say) than a Str-biased character.
• Diminishing effects: even (and especially!) if the moves were as crucial as your replies suggest, adding another point of Str to a high-Str character may be suboptimal. You get more special moves if you boost Dex instead. (The chance for the Str-move increases only very little; the chance for the Dex-move increases much more.)

I believe the underlaying problem in your logic is that you assume a “fighers choose one stat and that's it” system. This is probably the case now, but there's absolutely no reason why it has to stay like this. And that's why comments like ”…this doesn't work” puzzle me to no end.

To end on a positive note, here is how I envision future melee fighters choice processes: (a) Which armour type aimed at? (b) Shield or not? © Which weapon type to use? (Within this weapon type, I am fine with players ultimately, i.e. at high enough skill, going for the heaviest weapon given the handedness.) Note that moves don't play a role. They are secondary: a character may use an axe for a number of reasons — and I want the moves to be useful for an full-timer basher as well for a hybrid. They will have different stats and therefore see the various moves with different probabilities and thus occasionally make different choices. — dpeg 2010-09-27 23:58

Attack monsters adjacent to the target:

.gKO.   The axe-wielding @ attacks the ogre and may also get a hit on the kobold;
..@..   if lucky, even on the goblin.

Every axe attack has a chance to also attack an adjacent (to the original target and the player) monster. This chance increases with Strength and the weight of axe and decreases with the number of attacks. (I am thinking of Str 12 to have a chance for one additional attack, Str 18 for sometimes attacking a third target etc.) You get the chance for subsequent attacks even if the original attack misses.

Open problem: lack of orientation! Should additional axe attacks go clockwise or counterclockwise (assuming there's a choice)? Might be best to make that random.

This should simply hit every creature adjacent to both you and the targetted enemy simultaneously, with the effect being weaker with more enemies. Skill should determine the percentage of base damage that's carried over, not how many creatures are hit. This may not be true to a real-life swing, but it'd work simply and well. — og17 2010-05-19 02:15

Even if it's obvious, it's probably worth stating out loud that even if it activated 90% of the time this ability would still be inferior to tunneling for all opponents except those with smite-targeted attacks or who are going to cast buff spells if you wait too long. The difference is that you have an advantage if you get caught in a situation where tunneling wasn't an option, such as if you're swarmed by enemies faster than you (like bees) or find yourself in a bad spot after a teleport or chute, or when dealing with ranged attackers or open levels. I think you're going to have to make it activate somewhat reliably (like, once ever two or three attacks, or even every time if you decrease the damage to each target) if you want it to make any difference at all, since it requires putting yourself in a situation which would be tactically inviable if you don't cleave. Activating reliably but doing reduced damage per enemy would do a lot to make this ability carry its weight in code, at least in the player's mind. — brickman 2010-11-07 22:30

og17: Cleaving is disabled if an ally would be hit. Other problematic targets, such as corrosive enemies are included, as it wouldn't be hard to position yourself so they're not a concern, and you'd want to attack them if you had a corrosion-proof axe anyway.

I don't think you should worry too much about problematic targets. It's thematic for a crazy axe-wielder to hack away without worrying about precision. If you want precise striking don't take axes! It's a nice tradeoff that Axes give you free attacks at the expense of hitting things you might not want to hit. -Wes

Above seconded. — evktalo 2010-09-13 18:59

pepperfez: I know how popular argumentum ad D&D is around here, but I think they nailed the implementation of cleave. Whenever an axe wielder kills a monster, they should get a free attack on an adjacent target. The player's turn would just continue until they fail to strike and kill something. This would allow a sufficiently strong character to carve through summon spam easily, but not overpower strong opponents - maybe making the endgame a little more accessible to non-magical, non-TSO-worshiping characters. It would also avoid making axes feel like area of effect weapons because each target would be struck individually. Thematically, cleaving represents cutting all the way through a target, so it really only makes sense if the target is killed.

Not a bad idea. Conjures an image of an axe-wielding maniac cutting through the hordes quite well. — evktalo 2011-09-25 08:57

evktalo: “Semi-active cleaving” requires you to strike each target manually. On a successful roll that triggers the weapon move, you get a low-delay opportunity to strike the square clockwise or counter-clockwise to the original target. If you make the roll again, you can keep going to the same direction. You can strike empty squares with ctrl-direction.

Pros:

• Being able choose to not cleave targets that you don't want to hit removes the “problematic targets” problem.

Cons:

• Many keypresses (9) for a full cleaving swing.
• Possibly quite difficult to communicate the opportunity (and method) of cleaving to the player.

This implementation would be kind of semi-active, but not activated from a menu, only with movement/melee keypresses when the opportunity comes. This would be quite unique, but possibly taxing for the player (have to pay attention to notice opportunity & press correct key combos). pepperfez's idea above is similar in this manner, although probably easier to grasp and more natural regarding controls.

All polearms and whips can attack 2 squares away from the player (like the current Reaching brand). There's no chance involved here. (Yes, this is more like an ability, but it's worth it.)

This would remove the current Reaching brand (as all applicable weapons would have it). Instead, polearms (except for tridents) would get an Impale brand.

Don't they already have Piercing etc? What would Impale do? — evktalo 2010-06-03 14:09

Attack two in a row. (This is ideal for corridor combat, I know. But it fits — we are free not to use this, of course.) — dpeg 2010-09-26 23:29

See reaching_weapons for suggestions how to improve the reaching interface.

This has been in trunk for a long time, and it's good; however, there's an interface problem: it's not easy to tell if a monster is wielding a polearm without examining it in console. It's not cool having to examine every gnoll every few turns. This information needs to be in the console monster list somehow - I'd suggest using colour to indicate what weapon class a monster is wielding, except that's already been stolen by the danger level indicator. — minmay 2012-01-09 17:21

What about increasing the danger level of the monster if the monster has reach? — XuaXua 2012-01-09 17:38

A proper move for polearms would be impede: depending on Strength, a monster may lose (half a) turn when approaching you. (Idea by someone-else.) If doing this, we might want to use facing.

How about a free attack on a monster that moves adjacent to the player (perhaps limited to one direction per turn or one attack per turn?) -Wes

Again, both ideas support fighting in tunnels (which may or may not be okay). — dpeg 2010-09-26 23:29

Stabbing is possible against aware monsters. (So it works best against sleeping or paralysed monsters, works still good against confused monsters, and is hard to set up against aware monsters.) Chances increase with Dex and severely decrease with weapon weight (or speed). The damage boni for stabs are drastic; so stabs against aware monsters should be toned down or made rare.

The good thing about this is that short blade users could rely more on stabbing (if they employ Dex).

Also see Stabbing Revision. I still think that it's possible to reduce Short Blades, Long Blades, Stabbing to two skills.

I'm aware of the current shortcomings of short blades; however, it's very difficult to feel sympathy for the weaknesses of weapons capable of literally one-shotting most anything in the game. Branching into another weapon type (such as the conveniently cross-training long blades) or taking a supplementary chunk of magic or divinity is a perfectly good solution as far as I'm concerned - short blades can't have their powerful stabs along with general innate usefulness in a stand-up fight, it'd be ridiculous. Current stabbing mechanics (using Short Blades instead of the now-discarded Stabbing skill) would be more than enough of a special feature for the class.

I'd think long blades should have an effect of their own, rather than what would amount to a dull “critical hit” chance, though I've no ideas past some kind of damage-over-time bleeding effect (with blood trails, naturally). — og17 2010-05-20 03:58

I also like separating long and short blades' special abilities. For long blades how about 'parry': chance of blocking an attack thereby reinforcing the idea of long blades as defensive-oriented weapons that synergize with Shields/The Shining One. Or perhaps 'riposte' which gives a chance to counterattack after blocking an attack (also synergizing with Shields). Consider limiting to one parry/riposte per turn. -Wes

Parry could be too weak to be noticeable: it only comes up in a few melee fights (it is too passive, so to speak: the player cannot trigger it). — dpeg 2010-09-26 23:29

Parrying. I think I've figured how we can have the cookie and eat it ! The way I see it, there are two things to worry about, in no particular order a) Too weak - not noticeable, yet another passive/boring protection mechanism b) too strong - power spiral, players keep adding defensive measures: Armour, Shields, Charms, now passive parrying !
Active parrying. Active parrying would provide a big defensive bonus at the cost of dealing no damage at all (or just enough for windshield kills). You may balk at the idea of an active weapon special move, but it makes sense ! Because it's purely defensive, players are not going to spam it. They will use it thoughfully, tactically - when needed. You don't win fights by standing still, and you would still take enough damage so it doesn't become a new method of pillar dancing. No, active (but 0 damage to monsters) parrying could be used when waiting for Teleportation to kick in, when a monster is in noxious fumes but not yet confused, when enemies are poisoned. Summoners, players with allies might like parrying.
There's no need for an extra key - wait (also search) would be fine, no direction needs to be specified anyway.
Parrying efficiency could depend on Str, weapon speed and weight. This would make weapons of Speed especially good (they could be renamed to “of parrying”)
Obviously, parrying would only work on melee attacks. Trampling would override parrying. No spells and no missiles. It would use the Shields skill (which could subsequently be renamed to Parrying, or Blocking). This would make switching gears from shields to 2H swords a bit easier. — b0rsuk 2011-07-13 20:46

I don't see why parrying would have anything to do with Str or Shields; surely more related to Dex, Dodging and skill with current weapon.

Parrying does seem the most realistic and actual differentiation of the long blade from the other extant weapons classes. One conjures to mind the various arts of fencing and general swordsmanship found in differing incarnations the world over. It's difficult to picture a well-trained swordsman who couldn't use his weapon to block or otherwise deflect incoming strikes.

Perhaps parrying could only be available sans shield, for two-handed options? Or perhaps it might be tied to the 'long swords' skill itself, to keep it more relevant after one maximizes weapon speed? Perhaps the player might have an active selection of passive 'guards' (Ox guard, Plough guard, Fool's guard, guard of the day”, to borrow from Liechtenauer).

On an aside, and apeaking of Liechtenauer, perhaps a increasing swords skill could open up one or more passive master attacks. Few things would seem as cool for your long blade as discovering you had the ability to deliver a Mordhau, or “Murder-blow”, right? — LordKristopf 2013-09-10 23:16

Attacking with a staff has a chance to also attack a monster adjacent to the player but not to the original target (b0rsuk):

...g.  Attacking the kobold with a staff may result in
.K@O.  also attacking either the goblin or the ogre.

Chance should depend on Dex (only).

This is overly similar to axes, and cleaving seems natural there. Staves could instead be a defensive-geared choice and offer an additional chance to block an attack, as below. — og17 2010-05-19 02:15

While staves could be fast enough for a double-attack, that can be easily done by increasing the weapon speed. I much prefer Staves as a blocking / parrying concept, possibly using Shields skill to enhance use. — XuaXua 2012-07-20 19:23

It shouldn't use Shields skill, that just makes it even closer functionally to using a one-handed weapon - and even if it doesn't use Shields skill it's already dangerously close. — minmay 2012-07-21 00:51

You can parry/block some melee attacks (turnerjer).

Chance increases with Dex and decreases with base weapon delay.

Executioner's axe: delay 200%
Battleaxe:               180% 
Halberd:                 160%
Lajatang:                140%    
Quarterstaff:            120%

Of course we sacrifice some realism here, but it seems good that staves users would get many blocks and other two-handers rather not. While it seems unfair to not allow blocks for longswords (say), users of those weapons could at least wear a shield, if they wanted to.

I only like this for staves, though it's very good there. Generally speaking, taking a two-handed weapon should mean that you're trading defense for offense - a battleaxe shouldn't come with a ghostly shield that helps make up for the lack of a physical one, as doing so greatly weakens the choice of using a two-handed weapon in the first place. — og17 2010-05-19 02:15

I agree with og17. Very good for staves. — evktalo 2010-06-03 14:00

I also agree with og17. Quarterstaves have always been optimum weapons for offence and defence, quoted by several as the best 'close quarters weapon' (in Medieval times). Blocking should be reliant on strength, though. Parrying on dexterity, yes. When you block, you're trying to absorb the energy of the blow with your strength. If you're too weak, your arms will let the blow through… but even if you're strong, you might break a bone or two trying to block a heavy weapon. With parrying you just expend a tiny amount of energy to redirect the attack in a different direction… and then you dodge the direction you directed it towards. — studiomk 2010-08-24 23:02

I don't see why a huge unweildy 2-handed weapon would be the best thing for parrying attacks. To me it seems like it'd make a lot more intuitive sense to give this to staves and then give one class of one-handed weapons a chance of parrying, equivalent like SQRT(weapon skill) shield points. Specifically I'm thinking one-handed long blades get a bonus that works only against opponents with their own one-handed melee weapons (and thus only humanoids), since that's pretty much what sword fighting is all about anyways; a sword fighter with a shield thus finds himself with slightly more defense than any other fighter with a shield, since he can block with both the shield and his weapon. Two-handed swords get no change. Staves, meanwhile get this against all melee attacks. — brickman 2010-11-07 22:43

While realism clearly dictates only applying parrying to melee attacks, it would be nicely economical and transparent to just make it use the shield statistic. We've all seen action movie heroes bat arrows out of the air, so the mechanics of it are fairly intuitive. pepperfez 2011-05-18 11:49

If you have moved straight toward your target prior to an attack, damage is multiplied by (1+(Str-10)/5)*(S-1)/S with S being the number of spaces moved straight (capped at four). “Straight” shouldn't restrict to the eight main directions, the following should be okay:

.....4O.
....34..
...23... 
.@1.....

By lemuel in 2803259.

I really like the idea of the charge, however I think it should be restricted to one space of movement, and possibly not tied to a certain weapon type. – Thann 2012-05-21

Chance of (Dex-14)/Dex that attacks are targeted at the square you occupied last turn, instead of this turn.

Again by lemuel in 2803259.

This is of use when advancing towards a ranger attacker, or fleeing from one. It is also useful in melee, if there is more than one enemy around.

This is an idea that could go in right now, in my opinion. It doesn't require much (like in-game explanations, or interface-wise) and would provide a good testing ground for the whole concept. The only reason currently speaking against Sidestep is that EV is overpowered. — dpeg 2010-09-11 22:51

After attacking, if you travel around the target you get a free attack:

........
....@...
...3g1.. 
....2...
........

E.g. Here if you attack g then move to 1 you get a free attack on g. Moving to 2 then 3 then back to @ also grants attacks each turn.

I absolutely love this idea. Not sure if this should restricted to some type of weapons and/or linked to some stat. But it has a number of very obvious advantages: (1) no (interface) choices required: we can assume that if the player attacked a monster last round, he will be happy with another, free attack next round. (There will be exceptions, but players will learn to live with the rule.) (2) It encourages combat in somewhat open space: you won't get any use out of this move from standing in corridors.
This is another one that could be used right away, in my opinion. (Well, once there is a specific proposal for the mechanics.) — dpeg 2010-09-11 22:51

Some considerations: Movement speed (should spriggans get all of these attacks? A delay cap might be appropriate), any other situation (if there are any, I'm not sure; sidestep above would present one) where it would be optimal to run in circles around something instead of attacking it directly–Do we want players to do that?
And what happens if g is located where 1 is now? Is it possible to move around it? If not, is such a disparity between cardinal directions and diagonals desirable? — danei 2010-09-23 16:14

I don't think these are problems. Note that there is only a chance to get that attack while running. So if you move, you may or may not get an attack on the monster, but the monster will surely hit you.
As for proper phrasing of the rule, it could be stated like this: If you attack an opponent and in the next turn move but stay next to it, then you get a chance for another attack. (There is the alternative of getting a chance for an attack when moving next to some monster regardless of short-term history, but this would go into stabbing so much and also be so strong that I cannot see this being used.) — dpeg 2010-09-26 23:29

Moving towards a wall causes you to vault off the wall and switch places with the target. For example you can do this transition by moving up into the wall:

..###..    ..###..
...@...    ...g...
...g... -> ...@...
.......    .......

Optionally this could also expose the target to being backstabbed next turn.

Rest for two turns then move to grant a temporary speed increase for a few turns.

This is too meta-gamey for my taste. (Players will count turns etc.) — dpeg 2010-09-26 23:29

Then maybe just do rest a turn, speed bonus next turn? — Luc 2011-12-03 22:55

This is an active ability. Just because you don't use the 'a' key to use it doesn't mean it's passive. — minmay 2011-12-04 21:36

Suggested at least by psyshvl, modified by evktalo.

A strong blow from a heavy mace sends the target staggering backwards. This is not enough to move them to another square, but if the target is thrown against a wall, extra damage occurs.

Knockback shouldn't move opponents, because it would often be undesirable. Also, extra damage shouldn't occur from monsters bumping against each other, because that would occur often in a standard melee situation, i.e. where you are in a corridor with a mass of enemies lining up against you. Instead, the aim is to make it desirable to manouver your character a bit to get the benefit of knockback against your enemies.

This would be more interesting with an additional freeze-like ministun, no matter how weak. — og17 2010-05-19 02:15

Yeah, that could be more interesting than plain damage. — evktalo 2010-06-03 14:03

I like the idea of a small stun effect here, perhaps wherein the recipient is either unable to move for a turn, and/or has a penalty to chance to hit/cast a spell. The idea being that the one hit has been 'dazed' or otherwise incapacitated to some degree — veliq 2011-01-20 16:04

Flails and whips are hard to block. Also to parry, if the above staves idea goes through. This is more interesting against the player, as few monsters use shields. Tentacle slap attacks could also get this.

Skilled use of short blades or whips may disarm enemies. Should be balanced so that this is an actual advantage against early orc warriors and ogres. Mostly this is likely to be more interesting against the player. ADOM has ratling fencers, if you want to check out how this plays. (Actually, it's kind of annoying.)

Nagas start constricting (appropriate) enemies they are attacking with a successful roll.

b0rsuk: Alternatively, it could just be made “Stabbing” attack for unarmed characters of all races.

But we want to make nagas (both player and monster) more different. — dpeg 2010-09-26 23:29

Stabbing could be an implicit move: it works on sleeping monsters and it also works (and should do so better) one others, e.g. distracted monsters. The latter part is ideally suited for the ideas on this page. See Stabbing Revision.

An opportunity attack of large guys vs small ones. Related to the ideas here. See Trampling.

Content below this line untouched by dpeg:

• Blunt weapons: It makes sense they could stun opponents. The implementation I like: if a hit causes a monster to drop to 25% health or less, it is paralysed for a short time (2-5 turns). Implications: 1) less problems with finishing enemies who run away 2) minor advantage in close fights, but you still have to GET to that 25% 3) sometimes it may be safer to the opportunity to run away 4) while one enemy is stunned, it may be better to take care of another one.
• Move and Attack: Currently, crawl rewards defender perhaps a little too much. Attacker has to step forward and take a hit. So here's an idea - if a creature/player enters a square adjacent to an enemy, he gets a one-time discount to attack (the attack is faster). Say, a move burns 100 units of time (I have no clue). Traditional attacks take 100 units of time. Move and attack would take 100 for move, but the first attack would take only 20. So 120 total. Each subsequent attack would take 100. Weapon types would differ in their suitability to move and attack:
  • Maces&Flails would get little or no discount, they're heavy and momentum-based. You are not going to be running while swinging a heavy mace. Advance slowly and hold your ground.
  • Axes - similar to maces, they're heavy, little or no discount.
  • Short Blades - big discount, they're quite light.
  • Polearms - biggest discount, because it's enough to point the weapon towards an enemy and move forward. No swing necessary, so there's little momentum. Only heaviest polearms like bardiche could be problematic.
  • Long blades, Staves - moderate discount

— b0rsuk 2010-02-15 16:26

I'm not sure I agree that there is something wrong with fighting defensively; Crawl is all about minimizing risk, and fighting defensively will always be safer even in your example. Why spend 120 time units attacking when you can hit . with the monster 1 square away, spend 100, and (most importantly) get the first hit? Unless your system somehow let players get the first hit while moving next to monsters in the same turn, it wouldn't really change anything imo. — vandal 2010-02-15 17:03

Yes, the optimal move would still be to stand still and attack. But in situations where players needs to adjust his position slightly before attacking, it would allow greater flexibility. Imagine Orcish Mines - player ducks behind the corner to attack an orc while taking cover from a priest or a wizard. It is not about transforming a penalty into an advantage, but about reducing the penalty for some weapon types where it makes sense. b0rsuk

The actual effect of move and attack could even be different for different weapon types. A speed bonus for one, a damage bonus for another, maybe some sort of extra effect for another, etc. — danei 2010-09-23 22:42

Unstoppable Force, every swing crushes the opponent's defenses, resulting in them losing 1-2 AC per attack. The debuff duration would be refreshed with each additional attack and recover fairly quickly if you were no longer hitting the monster (say 1 point per 2 rounds without being hit).

Another idea is Knockback, which attempts to send the opponent flying backwards with the force of your attack. The higher the opponents hitdie and weight versus your strength and M & F skill, the less likely it is send them flying as far or even work. If they hit a wall or another opponent they take additional damage, and if it was an opponent it could possibly knock them back a little bit too. It would be useful for getting space between yourself and your opponent or to temporary take an opponent out of combat. Whips could have a Disarming attack which causes the opponent to drop their weapon or any items they are holding. — psyshvl 2010-02-15 15:02

Just my two cents, but I really am not in favor of any weapon causing knockback, or any of these weapons having a % chance of their effect actually working. The weapon passive should be “always on” or else it is too much like just another proc based ego such as electric. Knockback specifically is not a good idea because of how it would practically effect combat. It sounds neat at first, but think about what it would actually do during fights, and an ability that doesn't work on end game monsters / monsters with high hit die would relegate Maces to being a less popular weapon type much like Short Blades are now. — vandal 2010-02-15 16:57

I agree with this completely, offensive abilities should trigger on every attack. — og17 2010-05-19 03:09

Armour “piercing”. Maces (and flails) pretty much bypass armour. This could have reduced effect on “soft” enemies (only hard materials conduct force well enough) - orc warlord in his plate mail is going to feel most of the blow on his skin, lich has no soft parts, which could slow down the hit, but a cacodemon, with its rubbery flesh will just shrug it off. Alternatively, stun. Or maybe minimal amount of damage set somewhere in the middle, so that you are sure that if you hit, that kobold will go *SPLAT*. This would go well with axes' cleave - axes chop off arms and legs, but those orcs still have teeth. Mace kills one orc, but you can bet it will die. This would make maces and flails really reliable weapons - if you hit, your enemy takes damage, regardless of his carapace or whatnot. — someone-else 2010-02-15 19:02

Pierce Used by maces or all piercing weapons. Always reduces x of monster AC (x being a fixed number, maybe different for different maces depending on weight or basedamage). With stronger weapons having a higher monster AC reduction it will be able to scale. Tho there is the chance of finding a eveningstar or demon trident on d:1 which could be overpowered unless there is some other modifier like XL/strength that only allows x AC reduction no matter what weapon you got.

Stun Used by maces or all crushing weapons. Well it has allready been mention but with different ideas of how it should be activated. Stun activated when you roll 90% or above of your maximum basedamage roll. This makes more sense imo than having a 10% chance of stun, IF you would want a chance based stun that is. In other words you need to get in a good hit to be able to stun.

Increase monster walk/attack delay Not sure which weaponclass would get this but the idea is simple, when you land a hit the monster gets +x to delay in movement- and/or attackdelay. If increased attackdelay I guess staves/swords would be most suited flavorwise, perhaps staves more than swords. — rkd 2010-09-22 22:38

I'd like to see weapons used to block melee attacks. If different weapons classes are able to block different sets of weapons, then it starts to matter what kind of weapon you wield and what kind of weapon your opponent wields. Here's my proposal for how well each class blocks each other class:

A formula something like this would determine how much your shield rating is increased by:

BlockingFactor * BaseDamage * (WeaponSkill + OppWeaponSkill + ShieldsSkill)/ (27*3 + WeaponSpeed * WeaponSpeed * BaseDamage)

BlockingFactor is from the table. OppWeaponSkill is your skill level in the class of weapon that your opponent is using. WeaponSpeed and BaseDamage are existing properties of the weapon you are wielding.

The polearms are a special case. Before you are hit by a given opponent, use the blocking factor from the table. After being hit once, use a 0 blocking factor until you or the opponent moves. This represents using the length of the weapon to keep your opponent at bay.

I'd also advocate a few changes to skill training to help develop the skills to block all manner of weapons:

• When you block a weapon hit, you have a chance to train the skill of the weapon the opponent is using.
• When you are hit by a weapon, you have a chance to train the skill of the weapon the opponent is using, but only if your current skill in that weapon is zero.

I don't know whether this does more to help melee fighter's defense or whether it would end up as a stealth buff to conjurations and ranged weapons, but I think it would make the weapon types matter more. — jejorda2 2011-01-20 17:36

That would be a micromanagement hell. Even now the need to swap melee and ranged weapons is a major frustration for the player for no extra fun and is an avid argument in favor of playing a caster. With this, if you have a decent skill in a couple weapons (or don't have one at all) you'd want to always select the most fitting one for the current enemy. —sinsi 2011-03-29 17:01

Same page, but totally new approach. Here are the goals:

• To a great degree, differentiate weapon skills from each other.
• To a lesser degree, differentiate weapons within the same skill from each other.
• Provide some incentive for training multiple weapon skills.
• Make this system easy to grasp by limiting how many exceptions there are:
  • Ideally, all weapons can be differentiated by just their Skill, Damage Type, Size, and Handedness.
• Make this system intuitive to use by:
  • Limiting the triggering circumstances to be relatively passive
  • Limiting the triggered effects to be almost always desirable

Let's start by trying to categorize the weapons:

We run into these exceptions so far:

• Flails. Their “chaininess” doesn't show up in their stats anywhere - they are, in terms of stats, identical to maces.
• Falchions. They're smaller than other long blades, and more like shortswords, but also supposed to be choppy (there's even a note wondering if they should be chop rather than slice).
• Handaxes. They're smaller than other axes.
• Scythes. They do their own thing.
• Lajatangs. Likewise!

So, let's try it with the addition of some new weapons to aid in categorization, and some tweaks to existing stats:

Let's draw out their family trees, too. These show hypothetical progressions from a 'starter' weapon (at left) to an 'endgame' one (at right), with 'families' connected by weapons that have similar effects.

Maces & Flails

                   
("simple" maces) club --> hammer --> mace ---------------------> great mace                   
                                \                                          \
                        (flails) --> flail --> kusari                       --> giant club --> giant spiked club ("giant" weapons)
                                       \                                                  /    
                        (spiked flails) -----> spiked flail --> dire flail               /    
                                       /                                                /     
  (morningstars) ankus ------------> morningstar --------------------> eveningstar -----

         (whips) whip -----------------------------------------------------> demon whip, sacred scourge

Short Blades
                                  --> tonfa (parrying mace)
                                 /
    (parrying knives) --> main-gauche --> sai
                     /
(stabby knives) knife --> dagger -----------> quick blade
                                \
                  (stabby blade) ----> katar 
                                /
                      (blades) short sword --> sabre
                                  \
                   (choppy blades) --> kukri --> machete

Long Blades

(exotic)  estoc, hook sword, urumi

(fencing) rapier --------------------> jian

 (choppy) falchion --> scimitar -----> khopesh 
                  \
      (one-handed) --> long sword -------> demon blade
                          \
             (half-handed) --> bastard sword --> double sword --> katana
                            \
                (two-handed) --------> great sword --> triple sword      

Axes

 (sickles) sickle

   (picks) pick ------------------------> heavy pick

(hatchets) hand axe --> war axe
                           \
                (battleaxes) --> broad axe --> battleaxe --> exe axe

Polearms

                                      (pollaxe) --> glaive --> bardiche
                                              /  
                          (axe+spear) --> halberd
                                     /
                      (sword+spear) --> naginata --> war scythe
                                   /
(sharp sticks) shortspear --> spear ---------------> longspear
                                   \
                        (tridents)  --> trident ---> harpoon --> demon trident

                               (exotic) scythe

Staves

(balanced sticks) halfstaff --> quarterstaff --> longstaff --> demon sceptre
                                            \
                                             --> sectioned staff
                                (exotics)     \ 
                                               --> sodegarami
                                                \
                                                 --> lajatang
(exotics)

Note that this isn't completely precise because I've tried to avoid directly addressing the damage stats (dam/acc/speed) of weapons.

Next we'll go into the stat and size changes a bit.

Right now the effects of stats on combat are pretty minuscule. They should be bigger - dex should matter to hitting and str should matter to dealing damage similar to the degree that int matters for spell power.

That being said, all melee weapons benefit from you being strong, and all melee weapons benefit from you being dextrous. You should never become less effective with a weapon simply from increasing dex or strength.

Stat weighting should instead be abstracted: “strength-weighted” weapons should simply do more weapon effects that check your strength, and fewer that check your dex. This doesn't need to be an actual variable, though.

This is a bit murky for most players. We can do better here. A simple proposal mostly identical to the current system:

• Handedness refers to the hands required for a weapon's intended use. It is only loosely correlated with size. That's because some tiny weapons are designed to be used with two hands, and some physically large ones are usable one-handed.
  • One: This weapon isn't any better for you, even if you do have a free hand.
  • Half: This weapon has a one-handed mode and a two-handed mode.
  • Two: This weapon isn't usable in one hand due to size/weight.
  • Double: This weapon isn't usable in one hand because it requires two hands to operate it.
• Weapon sizes parallel monster sizes, and are described in terms of their size compared to the user they're designed for.
  • little (spriggan) - only very short weapons (like knives)
  • small (kobold) - shorter, one-handed weapons
  • medium (human) - most other weapons
  • large (ogre) - only very long weapons (like pollaxes)
  • big (giants) - there are no human weapons of this size
• There are penalties for using a mis-sized weapon:
  • Weapon is..
    • four sizes smaller: you cannot use it at all.
    • three sizes smaller: -1 hands, but awkward to use.
    • two sizes smaller: -1 hands.
    • one size smaller: -.5 hands.
    • your size: no penalties.
    • one size larger: +.5 hands.
    • two sizes larger: +1 hand.
    • three sizes larger: cannot wield it.

Some examples (noting that some existing weapons would need size or handedness tweaks):

• An ogre:
  • can wield a 'little' weapon as a 1H weapon, but with a penalty
  • can wield any 'small' weapon, or any 1/1.5H 'medium' weapon, or any 1H 'large' weapon as a 1H weapon
  • can wield a 2H 'medium' weapon, a 1.5H 'large' weapon, or a 1H 'big' weapon as a 1.5H weapon.
  • can wield a 2H 'large' weapon or a 1.5H 'big' weapon as a 2H weapon.
  • cannot wield a 'big' 2H weapon.
• A kobold:
  • can wield a 'little' or 'small' 1H weapon in 1H.
  • can wield a 'medium' 1H weapon in 1.5H
  • can wield a 'medium' 1.5H weapon, or a 'large' 1H weapon, in 2H
  • cannot wield a 'medium' 2H weapon, or a 'large' 1.5H weapon
• A spriggan:
  • can wield a 'little' 1H weapon in 1H.
  • can wield a 'little' 1.5H weapon or 'small' 1H weapon in 1.5H.
  • can wield a 'little' 2H weapon or 'small' 1.5H weapon or 'medium' 1H weapon in 2H.
  • cannot wield a 'small' 2H weapon, a 'medium' 1.5H weapon, or any 'large' weapons.

With the stat changes understood we can move on to the groupings:

Here's the available palette of weapon effects to work with:

Note: some of these are identical, yet have different names. Consider this a “placeholder”: for example, Hook/Trip/Sweep could be unified under one name if that one name doesn't look weird for any of the weapon types it appears for.

Here's the available palette of triggers to work with:

And here's the unification of the above - the effects weapon groups have, and the triggers required to use those effects. In other words, Trigger X checks skill Y to decide whether to do Effect Z.

Weapons have either 1 or 2 groups, so far, and get the effects of each group. Each group has 1 or 2 effects. So weapons have from 1 to 4 weapon effects.

A polearm can have “balanced”, which uses staff skill. This means that some weapons will benefit from learning multiple skills. The idea is that you can use these weapons fine without such a bonus skill learned, but that some of a weapon's effects may depend on having one or more. These other skills will only train when weapon effects kick in - so quite rarely, like aux unarmed training UC.

This will make races with versatile weapon apts better: they can spend less XP on gaining weapon effects. One-sided races, like merfolk, will have a harder time using certain specialty weapons.

Tis brings up the question of crosstraining. Here's how it currently works:

Staves <-> Maces <-> Axes <-> Polearms <-> Staves
Short Blades <-> Long Blades

One modification might be to make staves crosstrain with longblades instead of maces. Historically, staves were used in a manner closer to longblades than to maces, and it results in this much smoother progression:

Short Blades <-> Long Blades <-> Staves <-> Polearms <-> Axes <-> Maces

One issue is that shortblades and maces are both left with only one thing to crosstrain to.

Here's another alternative approach:

 
             Polearms <-> Staves <-> Long Blades
                  ^         ^             ^
                  |         |             |
                  v         v             |
Short Blades <-> Axes <-> Maces           |
    ^                                     |
    |                                     |
    ---------------------------------------

Here, everything crosstrains to at least two skills, but axes and staves crosstrain to three skills. Looking at it in terms of distance, this has a nice effect: there's a maximum of one intermediate step from one weapon skill to another.

This seems weird at first. However, a lot of the reason for that is that there are simply no two-handed shortblades. Consider that short blades/axes/maces will gain more variety in this system (via picks, cleavers, etc.) Axes then contain an actually quite wide variety of weapons, and it's not as much of a stretch to say that a small axe crosstrains well with a kukri or a machete, and staves being versatile is clearly fine.

Any exe axe ⇒ rapier weirdness should be taken care of at the effect level - big choppy axes should make use of strength-affected effects.

Crosstraining is one mechanic that may encourage diversity of weapon skills. I have an idea for another. You have to know how firearms work to make a good bulletproof (usually kevlar) vest.

In practice, monsters attacking you would get a accuracy -3 malus when attacking you with a weapon you know. If my Maces skill is 5 and an orc warrior attacks me with a morningstar, he gets -3 to hit. The bonus is static (doesn't increase with levels), this is key because it's supposed to promote skill diversity. Unarmed Combat would provide a benefit against all monsters who don't use weapons, including various beasts.

For the record, I think perks for knowing more than one weapon skills would be good for these reasons: Currently you marry a weapon skill, and it's a zero-sum game. It's best to have xp in one skill and fight with one weapon. If you switch to another skill later, you may feel bad about “wasting” xp in previous skill. This change would make players like me feel a bit better about it. And auxiliary unarmed attacks are quite unimpressive for most races, you're better off using a shield or a 2H weapon. Another reason - interesting choices are good. Polearms and maces are commonly used by monsters, so players who choose this class would benefit more than long blade users. Short Blade users would also benefit a lot. Actually the bonus could differ between weapon classes, you could have Short Blades give -6 penalty if it's deemed necessary. The users of staves would get the short end of the stick.

— b0rsuk 2011-06-29 21:03

I'll just quickly say I like this a lot. Thanks for the thorough write-up. — evktalo 2011-03-28 17:05

Everything in this section should happen. — electricalbatross 2011-07-08 22:45