Monsters you are not allowed to melee

[duvessa]

There seems to be a particularly dramatic push lately to force characters to get ranged attacks if they want to be able to actually kill monsters. Satyrs, new sirens, spriggans actually generating, etc. The one that pushed me to finally make this thread is octopode crusher, which seems to exist for no purpose other than being impossible to melee.

I think this is highly undesirable. It's often advertised that Crawl presents problems with many solutions and lots of different characters are viable. When you have monsters that basically block one of the two major classes of attacks, you're weakening that, and also just forcing players into one option, removing any interesting tactical decision (or character difference) there might have been between them. It's particularly bad when melee attacks are the ones being blocked since they are overwhelmingly more interesting than ranged attacks (because of positioning). So right now if you want to play a melee character you basically have to take attack wands everywhere if you want to be able to kill spriggans and the like, and I guess you just have to avoid octopode crushers altogether unless you are a large species or unusually strong. I find this very annoying, particularly in a game where melee attacks are already almost universally acknowledged to be less powerful than ranged ones.

And to be clear, I don't think the solution to this is to introduce "octopode deflectors" that pull your crossbow bolts and poison arrows out of the air and throw them back at you. My complaint is that there are monsters designed to shut down half the options in the game, not that they are shutting down the wrong half.

(Note that this is a trend I have noticed since 0.6, it just seems to be getting particularly bad lately, perhaps because of the incredible speed at which monsters are being added now.)

[dck]

After a small exchange with dpeg I meant to start a thread addressing some of the monsters that are been added lately and how they seem to go against crawl's strongest element (positioning and generally all melee combat). My intention was to take a more generic angle at what is causing the creation of these monsters to begin with (given that adding content, particularly content that is meant to be experienced in most if not all games, requires high standards for said content not to be detrimental to the game) and why they're been added in such gratuitous amounts lately. I think the "melee-proof" enemies are a good place to start since preserving the ability to use the one damage tool that is intrinsically linked to positioning is very important, if not vital.

[gammafunk]

Monsters that impose some extra kind of danger while you melee them don't qualify as "impossible to melee," and your melee attacks aren't "being blocked." You can still use your positioning and situate yourself so that the chance of being thrown into a bad situation is relatively small. This may mean changing the order in which you deal with a group of monsters or fighting the crusher 1v1. It only throws you to a location that you can see, so you there's always the possibility of constraining LOS appropriately while fighting it to prevent the throwing attack from sending you well and truly out of safety. There also fact that it simply may not throw you before you kill it, and yes, you can use ranged attacks to soften it up first.

We have other monsters in 3-rune games with constriction (also affects your ability to position) or spines or even simply dangerous weapon brands that the player may not have the ability to resist at present (or ever, in the case of distortion). They all present special challenges to melee in different ways, and the throwing mechanic just increases that variety. I think melee will continue to live its long and happy life just fine.

[crate]

Constriction mostly does nothing. I honestly think it is a failure as a monster mechanic (player constriction is ok but kind of boring). So using that as an example is um, lacking.

[duvessa]

(stuff about throwing)

The point is not that using melee in these situations is literally impossible, but that it is obviously inferior to ranged. Perhaps a clearer example is monsters with cause fear.

We have other monsters in 3-rune games with constriction (also affects your ability to position) or spines or even simply dangerous weapon brands that the player may not have the ability to resist at present (or ever, in the case of distortion).

I don't like any of constriction, spines, or monster distortion, for these same reasons (and constriction and monster spines were introduced relatively recently in DCSS...).

[n1000]

This post is severely weakened by the presentation of several controversial points as being self-evident.

It's particularly bad when melee attacks are the ones being blocked since they are overwhelmingly more interesting than ranged attacks (because of positioning).

This is not at all obvious to me. MP is a limitation that affects conjurations but not melee, surely this is an interesting feature?

When you have monsters that basically block one of the two major classes of attacks

If such monsters exist the case needs to be made here. I am familiar with the argument against spriggans but not the ones against satyr or octopode crusher. You say the latter, "seems to exist for no purpose other than being impossible to melee." To who? Why?

forcing players into one option, removing any interesting tactical decision (or character difference) there might have been between them.

A counterpoint is suggested. By weakening a character's preferred method of dealing with monsters, the player is taken out of his or her comfort zone and must adapt to the threat in a novel way. Recognizing the limitations of one's character is now a non-trivial component in identifying which situations are threatening (threat recognition being a key skill in crawl). There ought to be threats which are easier to melee and ones which are easier to fight using ranged options.

you basically have to take attack wands everywhere if you want to be able to kill spriggans and the like

I don't think this is the case. I also don't think melee is categorically a weaker option than ranged, though I do not like how ranged weapons work, that's an issue for another day.

[TeshiAlair]

I like the idea that certain branches punish/reward certain builds: TSO followers wreck Tomb, Undead are better at Pan, rPois helps for Swamp/Snake, ranged is better at shoals, axes and air mages are better at Orc, etc. As long as there are choices as to where to go, I don't think its an issues that certain areas are more punishing for types of builds than others. That being said, more branch diversity I feel should be a goal of Crawl in general, though time-consuming and incredibly difficult one.

[duvessa]

It's particularly bad when melee attacks are the ones being blocked since they are overwhelmingly more interesting than ranged attacks (because of positioning).

This is not at all obvious to me. MP is a limitation that affects conjurations but not melee, surely this is an interesting feature?

This one really did seem self-evident to me...with ranged attacks you have to care about positioning much less than you do with melee attacks, since there are far more spaces that you can attack, and it is generally easier to escape monsters that are further away from you.

When you have monsters that basically block one of the two major classes of attacks

If such monsters exist the case needs to be made here. I am familiar with the argument against spriggans but not the ones against satyr or octopode crusher. You say the latter, "seems to exist for no purpose other than being impossible to melee." To who? Why?

It's a monster that does really bad things to you if you melee it (unless you are large) so it seems like its intended design is that it prevents you from meleeing it. Satyrs have cause fear, which not only outright prevents you moving towards them, it outright deflects melee attacks (and has no effect that I know of on ranged ones).

A counterpoint is suggested. By weakening a character's preferred method of dealing with monsters, the player is taken out of his or her comfort zone and must adapt to the threat in a novel way. Recognizing the limitations of one's character is now a non-trivial component in identifying which situations are threatening (threat recognition being a key skill in crawl). There ought to be threats which are easier to melee and ones which are easier to fight using ranged options.

There can and should be situations where ranged attacks are better. But recognizing that a monster is the monster it is, and not another monster, is not an interesting case of threat recognition. I think it's fine that sometimes it is better to shoot at a centaur than it is to run up and melee it. I don't think it's fine that "wand of fireball" is pretty much always better than melee for killing a spriggan druid (at the depth where spriggan druids are dangerous, of course).

[crate]

since there is apparently a possibility that this topic is not completely pointless I guess I will write up a somewhat more constructive post.

I will begin by talking about crawl's uniques and an overarching problem with said uniques. As a whole, crawl's uniques have a huge flaw: they are overwhelmingly not actually worth fighting. The risk-reward is completely out-of-whack. Killing Frederick is by no means impossible. In fact it's not even terribly hard. For doing this, you get a whopping total of a bit less than 3800 xp, and maybe if you are lucky a good weapon (let's be optimistic and say you get a good weapon for your character 20% of the time.)

For comparison, frost giants have noticeably less powerful spells and give about half as much xp; they also have a pretty decent weapon (battleaxe of freezing). They do not come in packs; Frederick does not come with a band. I believe frost giants are typically located deeper in the dungeon than Frederick but this might be not entirely correct. I will assert without proof that it is significantly less dangerous to kill two frost giants than to kill one Frederick. Alternatively, you could compare Frederick to stone giants. It takes about 2.5 stone giants to make one Frederick of xp. This is also signficantly easier than killing one Frederick.

Therefore, why would you ever fight Frederick? This is not a case of making the player make a decision; this is a case of the risk-reward balance being far too skewed toward "risk" and thus Frederick is too dangerous to be interesting (or, alternatively, he should give much more xp, but then you run into the problem of crawl probably still having more xp than players need ... I'm not going to get too deep into fixing crawl's uniques in this post). Frederick is not a special example; I picked him because he is simple so it is easier for me to compare him to frost giants and stone giants than to compare a unique like Mennas or Nikola or Jorgrun to other monsters. Why do you think the advice given on this forum for dealing with Mennas or Mara is uniformly "don't fight"?

I cannot say for sure that octopode crushers are problematic in a similar way; I have no experience with them. However, I will assert the following two things: melee and ranged attacks are, in general, of similar power levels in crawl. In general, monsters in crawl are not particularly susceptible to melee attacks or to ranged attacks. Now I will suggest this third thing: that octopode crushers are significantly less susceptible to melee attacks than they are to ranged attacks, since they have an ability that specifically throws you away if you get into melee with them. If you believe all three of these are true, then we are in the same situation I just discussed re: Frederick and fight-or-don't-fight. The correct choice is to never melee the monster, since the risk-reward is too skewed. They are too strong against melee to be interesting.

It is possible that my suggestion that octopode crushers are significantly more dangerous in melee than at range is wrong, and of course my conclusion here relies on a certain degree of "significance". I don't actually care much about this particular monster; this is just to illustrate a point that I find many people (including the devs, based on my conversations with them) sometimes miss when talking about crawl monsters.

---

On a different note I would suggest that monster abilities that reposition the player are bad gameplay, and this is my more specific problem with crushers.

I find the most interesting part of crawl tactics to be positioning. Positioning is twofold: it is both the location of your character in relation to the other monsters on the floor (both those in-los and those out-of-los and even the monsters that do not exist but you as the player fear might exist), and the location of your character in relation to the dungeon features on the floor. For the majority of enemies in crawl, the player has very nearly absolute control over both of these two factors (yes, enemies move around, but they do so in an entirely predictable manner for the majority of monsters in crawl). Some enemies play with the first factor of positioning: they move monsters around in a non-predictable manner (most commonly a monster moving itself via blink, but also things like blink allies encircling or convoke). I don't have a problem with this, as long as it is not the norm; there is still skill in positioning yourself in a good location with respect to the static dungeon features. Monster dig plays with the second factor, by actively altering the dungeon around the player. I do not like this much personally, but I do not have a game design argument for eliminating it, so I will just say that it is similar in effect to moving monsters around. In either case there are still interesting aspects of positioning that matter for the player.

Forcibly moving the player around destroys both elements of positioning. The player is moved around in relation to the monsters on the level, and additionally the player is moved around with respect to the dungeon features. If we take this to an extreme--monster teleport other, which places you on a completely random space on the floor--then positioning before the attack is literally useless. You have taken what I find the single most interesting tactical element of crawl and completely eliminated it from the game. Monsters blinking the player or octopode crushers moving the player around are obviously not so extreme, but it should be evident how much less control over positioning the player has, and thus how much less interesting positioning becomes. I do not think this is a good thing for monsters to do.

---

An off-topic exposition on constriction below, though I do not expect changes here:

Spoiler: show

Constriction is bad primarily because it is a melee effect. There are three important cases to consider:
1) The player is faster than the monster
2) The player is the same speed as the monster
3) The player is slower than the monster

I will begin with 3)
In this situation the player cannot outrun the monster without using consumables (if the player has haste spell and using said spell makes the player faster then you can either consider the MP used as the consumable, or you can consider this case 1), whichever you prefer), so constriction limiting player movement has little effect. It may block a blink, but mainly this only matters if the player tries to blink after getting constricted, which is easily avoidable. By and large the effect is just increased damage for the monster and slightly increased HP for the monster (since the player will want to spend turns moving to break constriction instead of attacking, most likely).

Now situation 1)
In this situation the player can outrun the monster without consumables. Constriction is only ever important here in concert with other monsters. However, since the constricting monster itself is slow, the player can, tediously, separate it from any monsters of different speed with little danger. So let's just assume that all the monsters the player is dealing with are slow. Since this is a melee effect, one possibility is you actually end up in a situation like the octopode crusher one above, where you just aren't supposed to melee constrictors. I have the same problems with it here that I did above. If meleeing the monster is still fine, then in practice constriction just ends up being extra damage. You can reposition into hallways if you like (digging exists) so being stuck next to other, non-constricting monsters is not really a thing.

Situation 2) is actually the most troublesome, because of quirks about how constriction works.
In this case, if you just stand and fight the monster, then constriction mostly just acts as more damage for the contrictor (and more HP if the player spends moves trying to break it). The problem here is that, paradoxically, constrictors are actually easier to escape once they move next to you than monsters that do not constrict. You see, when the player breaks constriction, the constrictor is stunned for a short time. This means that if you somehow end up next to a constrictor but want to escape, the best plan is to purposely get constricted, break constriction, and then run away. Even if the end result of this is not undesirable (I would say it is, but will not prove it) the fact that it is counterintuitive is problematic.

So in the end my conclusion is that all constriction does is 1) add damage/hp to the constrictor and very occasionally 2) block blinks (note it does nothing to teleports). If 2) is a thing worth having, why not have that be its own thing completely? If you want to affect player movement, then having it done via a ranged ability (nets, barbs) is substantially better than doing so via a melee ability, since cases 1) and 2) are very different.

[XuaXua]

So right now if you want to play a melee character you basically have to take attack wands everywhere

Oh no, 2 inventory slots, maybe slightly more if you throw a net or some javelins. It's terrible that Fighting Skill can somehow slightly compensate for lack of Throwing Skill.

[and into]

The OP overstated the case somewhat, in my view, but I think duvessa is overall correct in raising an alarm about this. A very large number of monsters have been added recently that make melee a poor option, or just make it more frustrating; maybe a lot of these will be removed before 0.14, but at the moment the pendulum has indeed swung far in this direction. crate addressed a lot of this in detail, but another part of why this is bad that hasn't (I don't think) been addressed is that smacking stuff is for nearly all characters your go-to method for killing things that do not provide a threat to you in the current situation, as this streamlines dealing with those non-threats in the fastest and most convenient way possible.

[WalkerBoh]

I agree with most of your points, crate. But with regards to constriction, I'd like to add that it's also very relevant for blocking going up/down stairs. This is actually the situation that I find it affects me the most. Not that this negates your points, I just wanted to mention it.

[crate]

I agree with most of your points, crate. But with regards to constriction, I'd like to add that it's also very relevant for blocking going up/down stairs. This is actually the situation that I find it affects me the most. Not that this negates your points, I just wanted to mention it.

Yes, true, I forgot about this. Again, if this is a good thing, I would suggest it should just be its own thing! (And it is, with wardens.)

[gammafunk]

Ok, one last (and long) response, and then I'll have to sit on the sidelines for a while.

It's a monster that does really bad things to you if you melee it (unless you are large) so it seems like its intended design is that it prevents you from meleeing it. Satyrs have cause fear, which not only outright prevents you moving towards them, it outright deflects melee attacks (and has no effect that I know of on ranged ones).

In the case of throw, it does add a complication to the use of melee, but it's one that the melee-oriented player can deal with (yes, with melee). In general for monsters with melee-only effects, if the player can maintain distance and perform enough ranged attacks to kill this monster, then depending on the cost these attacks (MP, wand charges, and evocable charges no longer available), and depending on the severity of this melee-only effect (affected by what else is happening in LOS or on the level), the player can be in a "ranged is clearly better" situation for these kinds of monsters. But "ranged is a lot better than melee" won't be true in every or even most situations if the mechanic is reasonably designed. In general for crawl, the player can elect to not have access to many ranged attacks will do just fine by relying on good positioning. The basic idea behind these melee-only effects is to add variation in the kind of danger one experiences with melee, not to prevent melee. I agree that no or at least very few monsters should fall into the "ranged is *always* better no matter what" category, but at least for constrict and throw, these don't fit that description at all. If we want to think about whether the danger added to melee by each of these mechanics is sufficiently interesting, we can do so, but not starting from some premise that adding some difficulty to melee means we have removed melee as a good option.

About the number of monster mechanics and the number of monsters in general (which influences the number of so-called "don't ever melee" monsters), one item we have for post 0.14 is to decide how much we'd like to reduce both unique monster mechanics and the monster list. New monsters tend to get a new mechanic these days, because developers like making new and interesting things (wide range of views on what is interesting). But we also can and do take the time to remove and reduce.

Regarding crate's post, the issue of risk-reward for uniques is not one I want to address now aside to say that we also have plans to go over the uniques and make adjustments/removals.

I think it's false that risk-reward is all out of whack for melee vs. crushers. They aren't less susceptible to melee compared to range in some way that makes a melee player always decide to never fight them; melee works pretty well for reducing their hp to 0 or less, in fact, as they don't have considerable defenses. As a player, I would only invoke this "do not fight with melee" rule only when the amount of danger due to being thrown is clearly to high given the *current situation*. Can I constrain my LOS so wherever it throws me is very likely to be sufficiently as safe, and if not, given the additional risk would I be exposed to, how likely am I able to handle the situation given my repositioning/escape options. I might indeed to the conclusion: "I'm not going to melee this thing, I'll escape or use ranged", but I might not, and in general the decision-making and risk-control opportunities are interesting to me.

The idea that mechanics that reposition the player are inherently bad is also wrong. It doesn't "destroy" or "completely eliminate" positioning from the game; in fact you'll likely use positioning again to regain control over the situation. Just like you can miss when attacking with melee or ranged, you can be forced to reposition, and none of these three things become uninteresting because of having to use them multiple times. I agree that it can reach levels of frustration if too many things mess up your positioning too frequently, but we're talking about relatively uncommon mechanics that are not triggered 100% of the time even when a monster has them. These reposition effects shouldn't be present in most fights, I agree, but removing them completely is bad design in my view.

[tasonir]

One that always bothered me was red devils being able to walk backwards and attack at the same time. The ability itself isn't bad, but when they spam it 3-4 turns in a row, it's frustrating. Give them an exhaustion timer, or a lower chance to use it, imho. It isn't that bad because they're generally weak and so it's rarely deadly, but it is annoying.

I haven't heard too many complaints about spriggans, and I've melee'd the forest several times, what's rough there? Air mages have roughly infinity EV, but you can eventually land a hit...They can also dodge ranged of course but there are workarounds through spells like fireball, OOD, etc.

[crate]

but removing them [things that reposition the player] completely is bad design in my view.

In case this isn't totally clear, I cannot agree with this.

The idea that mechanics that reposition the player are inherently bad is also wrong. It doesn't "destroy" or "completely eliminate" positioning from the game

The specific example I gave actually does. This is why I gave that particular example. There is absolutely no correlation between your position pre-teleport and post-teleport. If an enemy action completely undoes any positioning you have taken before that action happens, why bother with positioning in the first place? There is a cost to moving around, after all: it takes turns (and effort from the player).

It is true that you can control things like "blink player to a random place in player's los" to some degree. One example: you can dig a corridor, and make sure you lure said monster to that corridor. I think that the ways of influencing player-repositioning effects are generally things like that example; the best I can say for them is they are tedious. The idea of allowing player-created traps has been shot down in the past specifically because it encourages repeatedly luring monsters to the location of the trap, for comparison (and this is really very similar).

I feel like all of the potential benefits from moving the player are just things that you can also achieve by moving monsters around instead (compare blink allies encircling to dispersal ammo), except moving the player has additional drawbacks so it's worse.

edit:

I haven't heard too many complaints about spriggans, and I've melee'd the forest several times, what's rough there?

Well for one thing, they make freeze (a level 1 spell) approximately as effective as a 14-skill, 14-fighting, +5 vorpal lajatang (a very good melee weapon). There's no reason to waste time using attacks that can miss against them. I don't personally think they're interesting because of this, but I'm not going to argue about it (especially not here).

[dck]

@tasonir: The problem with high EV monsters and melee is that crawl supports them poorly and the player doesn't have a lot of feedback to what he is doing; particularly in the case of spriggans (hilariously in the case of riders when one connecting hit can kill them twice) it becomes a game of whack-a-mole that you only play when the monster is harmless, if it isn't harmless a wand of fireball, clouds or even the freeze spell or conjurations+ battlesphere in general are the appropriate way of dealing with the threat and that isn't very fun because having such a narrow way of dealing with enemies isn't something players are used to in crawl.

@gammafunk: Re: Mechanics that reposition the player -I think most mechanics that forcibly reposition the player are at best tolerable because the monsters that have them don't normally exist. Right now there are two enemies that do however use them in excess and it really shows when you're fighting said enemies: Draconian shifters and more prominently chaos champion Ds.
At least with shifters you can pretend things are not all that awful by blocking their LoF with other draconians if there are no walls nearby, but chaos champions reposition you and the enemies nearby constantly (irresistibly so and with no LoF need) and the only thing you can do to have them not reposition you constantly is killing them or running into them at a point where you can break LoS instantly.
I don't know what your experience with those repositioners is, I don't know what you think about positioning being a core element of crawl -I know for me it's frustrating and tedious to kill shifters since they got blink other and that chaos champions destroy all semblance of tactics whenever they are present.
Also, given than positioning is an active action players do it's frankly just frustrating that those monsters casting a single spell can destroy everything the player has achieved with those actions, it just encourages not actually using tactics or using terrible ones like luring everything into a LoS corner (or digging your own).

[Baldu3]

Just dropping by to add support to the motion, as I also find that crawl is the most interesting puzzle game to try to solve, especially when trying to play as melee. Playing a caster, learning fireball/bolts/clouds tend to solve almost all puzzles without effort or though, playing ranged tabbing solves almost all puzzles without effort or though, but playing melee there are still large numbers of situations that require short periods of though and effort and reward me with the satisfaction of success.

Consumables are a way to deal with harder puzzles and unexpected challenges. While I havent fought crushers yet, the reason I highly dislike sirens and druids is that I see them a lot like item destruction, if I meet them as a melee they *will* cost me finite resources to solve an uninteresting puzzle which wont be available later for an interesting puzzle.

[Laraso]

I'm just going to put in my two cents and say that I hope this discussion doesn't lead to the removal of octopode crushers, because I just ran into one the other day and I have to say, it's pretty badass. I think a monster that grabs you and throws you around with pure brute force is interesting and fun.

[dpeg]

I don't agree with everything that's been said here but many, many thanks for this thread, especially to duvessa and crate for the long and detailed postings. This topic is absolutely crucial to Crawl design, and discussion is long overdue. Some general comments: there is no universal best solution, this is very much a matter of taste. Some questions are: how much player attention should a single, standard monster get? How much a unique? Are we in favour of unique monsters, or do we rather go for (fewer) mechanics which we then distribute in interesting patterns across monsters?
Answers to these questions are subjective, but it is imperative that we develop some form of design goals in this area. Otherwise we'll be meandering helplessly in the design space, the Brownian molecules of monster making. I will do my best to help establish some form of design direction for this (I have very little investment in monsters, which may help).

This thead being rich as it is, there is also too much going on. crate's observation about uniques is relevant, but I am grateful that a thread about this has been sourced out.

I believe that we have to make an acquisition of all the mechanics and special tricks used currently employed by monsters. Followed by an assessment whether these live up to expectations (this can only come once we have agreed on these expectations!). In other words, it's a long and slow process. A typical thought process could be (I am using constriction, because it's been brought up before, and because it's one of the only two bits I contributed): How does constriction affect gameplay? For various builds? Do we see a way to make constriction affect gameplay in a better way, i.e. should numbers or the mechanic be changed? Would constriction work better on other monsters? Are we best off scrapping it completely?

It's true that there have been many new monsters with decidely original mechanics, and a number of old monsters have been cut or given new mechanics. There were good reasons for this: Crawl started out with very many very boring monsters (only differentiated through numbers). I keep citing Jeff Lait of POWDER who once said something along the lines of "If a new feature only requires copying code and changing numbers, then it's not good enough." This is an exaggeration, but he said this on purpose. Of course, in Crawl we should actually talk at some point whether we see it like this. It's not clear that more, and more individual monsters improves the game; neither is the opposite clear.

Let me give an example of where something similar happened in Crawl before: vault making (the maps, not the branch).
When we started out, basically every new vault (by anyone) was taken. Along the way, the syntax and thus the possibilties improved drastically (this started with adding more and more des file keywords and culminated in elaborate lua-driven vaults). Needless to say, vault makers proceeded to explore the possibilities, in all directions: vaults that are nastier than the ones before, or vaults that make use of the fancy new technical tools. After a while, there was a phase of reconsideration: in short, our standards of what makes a good vault have risen, and old vaults were systematically examined, updated and sometimes scrapped. To be sure, the absolute number of vaults is (and always was) increasing. Some vaults are as contentious as ever. However, we now have developers with a much better idea of what vault making should be about (time to write these up, actually!) and in my opinion, the quality of Crawl vaults has gone up noticeably.

I am saying this because I expect a similar trend with monsters. We have been experiencing a lot of creativity going into the possibilities of monster design. It is probably time to stand back and look at the whole situation, and to establish some general principles. There will be weeding out, and lots of changing, but the fact that new monsters are being designed, implemented and tested is a great feature in itself. It makes Crawl a living game.

One final word (I apologise for not even getting close to the details in this posting; this is really a huge subject!): it is extremely helpful to get negative feedback. Especially if that feedback is not just "I really dislike new monster X" but comes with some form of explanation. (The OP gives a good example!) However, it also helps to sometimes say which new or changed monsters are good in your opinion. Not (only) for stroking developer egos, but rather to see which ideas seem to work already!

For example, my recent MfFi challenge game met the new deep earth bands under real game conditions a first for me. I was very impressed with how they worked. They caused problems I didn't expect, and which slowed my (real time) game speed -- but I wasn't annoyed. Granted, this could've been the cool of the new, but I think they have lots of potential.

[ackack]

I think it's worthwhile for people to list some of the recent monsters they view as being anti-melee, for concreteness. Here are some of my examples with short explanations:

Spriggan druids (see "the problem with spriggan druids" thread for much more detail);
Lost souls (fast, run away if you try to engage in melee, discourages fighting other things when they are around);
New curse skulls (always try to get to edge of LOS (requires much gimmickry to get around this), summon big packs that obstruct motion toward them);
Chaos champions (dck explained these nicely; they've been toned down somewhat since first introduced, which is good because they were bonkers back then);
Demonspawn corrupters (these are less egregious, but in general summoning on this scale rewards some sort of penetrating or AoE attack to get through the wall of beef)

[WalkerBoh]

Other examples:

Water nymphs and thorn lotuses: Require flight to stay in melee with.

Spatial maelstroms and apocalypse crabs: Chaos effects make them extremely nasty to get hit by/be near.

Death drakes: Freaking miasma.

Ghouls: Freaking rot.

Raiju: To a lesser extent than above, but it's very hard to get in melee. Wisps have the same problem but it's less annoying because they can't attack from range.

Lurking horrors: Obviously.

I only play melee-centric characters, but that list along with demonspawn chaos champions, curse skulls, and sirens make up my "I wish I had a ranged attack right now" list. I've never had issues with hitting spriggans and the octopode crushers are flimsy enough that I haven't had issues with them either.

[Lasty]

Therefore, why would you ever fight Frederick? This is not a case of making the player make a decision; this is a case of the risk-reward balance being far too skewed toward "risk" and thus Frederick is too dangerous to be interesting

Overall, I think you're making very interesting points that warrant consideration. When I read this, my first thought was, "is that bad?". I think there's something interesting about having monsters that are too dangerous to fight. Sometimes it will mean taking unexpected steps to avoid fighting them. Other times it will mean being forced to fight them even though you would prefer not to -- and may need to preemptively burn resources to defang the encounter. I think a lot of people would say that they've had memorable experiences with current uniques -- I know I have.

[crate]

Therefore, why would you ever fight Frederick? This is not a case of making the player make a decision; this is a case of the risk-reward balance being far too skewed toward "risk" and thus Frederick is too dangerous to be interesting

Overall, I think you're making very interesting points that warrant consideration. When I read this, my first thought was, "is that bad?". I think there's something interesting about having monsters that are too dangerous to fight. Sometimes it will mean taking unexpected steps to avoid fighting them. Other times it will mean being forced to fight them even though you would prefer not to -- and may need to preemptively burn resources to defang the encounter. I think a lot of people would say that they've had memorable experiences with current uniques -- I know I have.

So what's wrong with giving them actual rewards? I did not say that crawl uniques are too strong to be worth fighting in any situation. If Frederick gave 100000 xp, he would absolutely be worth fighting. At 3500, he is never worth fighting. Clearly there is some middle ground where he would be worth fighting sometimes, but not always. Is making the reward commensurate with the difficulty of achieving that reward a bad thing? I would suggest that the current situation is far worse design, and additionally it is bad design in a way that is not fun for players. Killing off a unique should be an accomplishment, so why not make it actually be one (by giving the reward you would expect)?

Grinder and especially Pikel are examples of uniques whose xp values are actually reasonably in-line with their difficulty. Pikel is a scary dude, but by killing him you get as much xp as you would find on the entire rest of his dungeon floor combined (at least if you get an early Pikel) ... which is good since he's the most dangerous thing there. He's not just "oh that's another monster" of xp.

I just don't see any situation in which making xp rewards for uniques meaningful is worse than the current state.

(this is of course exactly the same as making Frederick easier without adjusting his xp, from the standpoint of risk-reward assessment, which is of course the point of my original discussion on this)

[and into]

In terms of crate's points about uniques, I'd just add that "riskier, but with good reward" is more interesting than "only slightly riskier than average monsters, with slightly better reward." At least so long as you avoid "so risky it is suicide" type stuff.

There are more ways for uniques to matter beyond XP too, obviously. Pikel is also a good example here—early game dude with good chance to spawn with whip of electrocution, which is hugely useful, but also makes him much more dangerous when he does have that. The slaves going friendly is neat but usually not a major consideration in whether I try to kill Pikel, however it does gesture toward other ways that killing a unique could be made interesting or give rewards. A unique who has a prisoner that is freed if you kill his captor and offers to become a permanent ally could make it more tempting to engage in a somewhat risky battle; you could easily add this to someone like Frederick or any generic bounty hunter unique. I've always thought Aizul should be guarding a mini-treasure vault that can only be opened if/when you take Aizul out. Maybe Erica should get a bit of a buff and her weapon should have a 50% chance to be an artefact scimitar generated with flaming and some other randomly chosen properties. A unique that gave better rewards the faster you killed it (though it may be hard to relate this information to player in a clear but also thematic manner). Etc.

Some uniques are good without necessarily having special rewards, too, but the longer you go in the game, the fewer of those there are, because late game you usually have more than enough tricks up your sleeve to escape one dude, and diving without exploring all of a level is proportionally much less of a punishment.

[Jeremiah]

I will begin by talking about crawl's uniques and an overarching problem with said uniques. As a whole, crawl's uniques have a huge flaw: they are overwhelmingly not actually worth fighting. The risk-reward is completely out-of-whack. Killing Frederick is by no means impossible. In fact it's not even terribly hard. For doing this, you get a whopping total of a bit less than 3800 xp, and maybe if you are lucky a good weapon (let's be optimistic and say you get a good weapon for your character 20% of the time.)

I always thought the point of killing uniques was to eliminate the threat of meeting them again at some later time when they would be even more dangerous (ie when you are already facing some other threat.) Since monsters do sometimes wake up and most of them can move.

(obviously this doesn't apply to Boris.)

[Wahaha]

Other examples:
Water nymphs and thorn lotuses: Require flight to stay in melee with.
Spatial maelstroms and apocalypse crabs: Chaos effects make them extremely nasty to get hit by/be near.
Death drakes: Freaking miasma.
Ghouls: Freaking rot.
Raiju: To a lesser extent than above, but it's very hard to get in melee. Wisps have the same problem but it's less annoying because they can't attack from range.
Lurking horrors: Obviously.

Apocalypse crabs, death drakes, ghouls, raiju, lurking horrors are not anti-melee.

The monsters that are the most anti-melee in the game for me are jellies and brown ugly things. Adding back jelly splitting is just insane.

[Baldu3]

I was just reminded that plagued shamblers melee attacks make potions unusable and they emit a cloud of miasma on death. When fought in the open it's possible to outrun the miasma but in corridors it doesnt seem possible to outrun the miasma.

[dck]

They are mostly anti-naga bombs.

[damiac]

It's funny, but this looks like precisely the sort of thing the experienced crawl players love to jump all over when new people complain about.

"I can't just kill every single thing in the game with 1 method of attacking"

That's interesting. When a guy who likes to play mages complains "I can't get by just blasting conjurations", the response is always "Well, who said you're not supposed to pick up a melee weapon. There's no such things as mages, bla bla bla..."

If a guy who wants to play his CeHu as a ranged character complains about ammo, or about ranged weapon power falling off later in the game, the answer is "Who told you hunters are supposed to just use ranged weapons? You have to use melee".

So it's very funny to me to see someone who loves saying that kind of thing to other players, complaining now that it's too hard to win with just melee weapons. My question to you is, what's a melee character? That's not a thing in crawl, this isn't a game design problem, you're just playing wrong. I'm glad to see things are going in that shake up the game enough where the super experienced players have to actually change up their strategy a bit.

[duvessa]

??? Maybe there are some people who do that but I have always maintained that it is very easy to win using only ranged weapons or only conjurations. In fact (especially in the case of the former) it is probably easier than using only melee. And I would consider it highly undesirable if using melee was required, as I said in the first post.

[crate]

@damiac: This is actually a completely different situation.

Ranged combat can kill any specific enemy in crawl just fine (except there are a few uniques/spriggans who have both high EV and permanent dmsl, but you should notice that I already said this is not a good state of affairs!). You don't use a melee weapon because you are incapable of killing specific enemies with ranged combat. Instead you do so because you cannot kill every single enemy you will encounter in the game with ranged combat (except you can because there is too much ammo in the game, but that's a separate topic entirely). It is the same situation with conjurations (I'm lumping conjuration-like spells like LRD in here): any individual enemy in the game is perfectly killable via conjurations (and even the dmsl thing doesn't apply here, since there are spells that ignore EV!). However having a way to kill enemies that does not use MP is helpful for not running out of MP while fighting groups of enemies.

This case is different: melee does not run out the way ammo does or the way MP does. You could argue that perhaps having individual enemies who are absolutely not killable with melee is fine, but I personally don't agree and think it is less interesting, like I described above.

[damiac]

Melee is by far the best killing tool throughout the game. Ranged is good, but weaker, and ammo dependant. Conjurations are more situational, and subject to MP. Evocations are typically very limited, but powerful. Rods are something you might never even find in a 3 rune game.

So, yeah, melee doesn't run out of ammo or MP, allows more than 1 attack per turn, and is capable of incredible damage. Its one weakness is that you have to be next to an enemy (or 1 square away for spears). It makes sense to me that some enemies would exploit that one weakness, and force the primarily melee character to use a secondary means to dealing damage.

You use melee over ranged because ranged damage gets a bit lackluster in zot, and I'm assuming in some of extended. Ammo is also a concern if you're not using an ammo gifting god, depending on which part of the game you're in.

You use melee over conjurations to save MP, because you're out of MP, or because you don't have the proper damage type available in conjuration form.

You use ranged or conjurations over melee because you want a little bonus damage before the enemy closes to melee range, where you switch to said melee weapon.

At least, this is how it seems to my experience. Yes, there are certain enemies like sirens, orb spiders, etc, where having a ranged option makes it easier, but it's far easier to win with "pure melee" than with "no melee". So new enemies that make it a little harder to go "pure melee" actually make a lot of sense to me. There's no god, race, or background that forces you to just use melee, so why should pure melee be viable?

Where is it that "melee attacks are already almost universally acknowledged to be less powerful than ranged ones"? That's the opposite of my experience. I've seen a couple people say that on the forums, yet I never see anyone recommend to people that they should play a ranged character for an easier win. In my experience, ranged is overly powerful only very early in the game. In the hands of a centaur, you can extend that to dungeon and lair branches since you can kite so easily, but that's more of an issue that kiting is OP, or that being faster than 1.0 is OP.

Also, what enemies are "absolutely not killable with melee"?

I do agree that plague shamblers seem designed to just screw over nagas.

[crate]

what enemies are "absolutely not killable with melee"?

sirens in shoals and orb spiders try pretty hard

(I would probably actually like sirens if they were in a place without water)

[duvessa]

Yes, there are certain enemies like sirens, orb spiders, etc, where having a ranged option makes it easier, but it's far easier to win with "pure melee" than with "no melee".

Yeah I just really strongly disagree with this. I think pure ranged combat is just about the easiest type of character in crawl, especially after entering Lair, and the only reason I don't give it the title of "most effortless win" is because deep dwarf exists (and is better off using Makhleb).
(It's also not really relevant to the point of the thread.)

[damiac]

Well, I've done spiders as pure melee, and yeah, orb spiders are a HUGE pain in the ass, but not at all unkillable. I think things like that are good, where pure melee is quite sub-optimal, but possible. Electric eels probably fill this same category, at least early on. Of course, those are typically optional enemies.

I don't think high EV creatures are a bad thing. I've melee'd my way through the forest, it didn't really seem that bad.

I haven't experienced octopode crushers yet, but those don't sound impossible, they just seem like you need to use different tactics than normal enemies. Or the same, boring tactic that usually works, see monster, walk back toward a cleared area, kill monster.

With sirens, I suppose you have a good point. They're not just hard to kill with melee, they are literally not killable, as you said. I don't like the idea that one could find themselves mesmerized at full health, with no ranged weapons, and just have to sit there and die while the siren very slowly kills them with boredom.

So, perhaps I took a bit too much joy in what seemed like fair turnabout, because I do agree situations like that shouldn't be possible. I do like the idea of forcing all characters to consider secondary forms of attack, although I dislike the fact that I will suddenly stop getting (possibly) good gifts with trog and oki if I decide to train a ranged skill past a certain threshold, because most of my gifts will then be ammo, which I don't need.

EDIT: NINJAD

Yes, there are certain enemies like sirens, orb spiders, etc, where having a ranged option makes it easier, but it's far easier to win with "pure melee" than with "no melee".

Yeah I just really strongly disagree with this. I think pure ranged combat is just about the easiest type of character in crawl, especially after entering Lair, and the only reason I don't give it the title of "most effortless win" is because deep dwarf exists (and is better off using Makhleb).
(It's also not really relevant to the point of the thread.)

Yeah, looking at the YAVP board I see two CeHU wins, back to back, from what looks like fairly inexperienced players, with 11 or so weapon skill. So maybe I was just expecting too much from ranged, or I got the wrong stats on level up or something on my CeHu. It really seemed like once I hit zot, my longbow became useless, so I switched to a triple sword of freezing, and saw much better results. Of course, the fact that my first win was a CeHu, even if I did switch to long blades, probably helps to prove your point. My first character to make it to Zot was a DDBe, so you can see I was playing off Duvessa's list of easy wins.

I hate it when Duvessa's right...

[crate]

I never said there were any monsters in crawl that were literally not killable in melee (I was talking in hypotheticals above, as crushers themselves are a hypothetical to me). If you read my posts here you should have noticed that that's not even the problem I have with crushers!

[damiac]

I never said there were any monsters in crawl that were literally not killable in melee (I was talking in hypotheticals above, as crushers themselves are a hypothetical to me). If you read my posts here you should have noticed that that's not even the problem I have with crushers!

What? I was agreeing with you in regards to sirens. If you don't have a way to fly, or a ranged attack, and get mesmerized by a siren, you're kinda screwed, right?

As far as the problem with crushers, I don't really see as a "problem", but they are game changers. Of course, like I said, I haven't experienced them yet, I'm only going off what people have said about them in this thread.

I still disagree with the main premise of the thread, that monsters against whom melee is very sub-optimal are a bad thing. I also very strongly disagree with the notion that melee attacks are weaker than ranged ones, unless you're talking about high level conjurations here, and not just bows and such. I will say I'm against anything that is completely immune to melee, ranged, or conjurations, and sirens are practically immune to melee in their current state.

[gammafunk]

I never said there were any monsters in crawl that were literally not killable in melee (I was talking in hypotheticals above, as crushers themselves are a hypothetical to me). If you read my posts here you should have noticed that that's not even the problem I have with crushers!

This illustrates why threads titles (at least in GDD) shouldn't contain hyperbolic statements: it can derail the discussion and/or confuse people at some point.

Regarding the overall issue of anti-positioning/reposition abiltiies monsters can have, crate and others have pointed out how we should recognize that positioning is central to crawl, which I think is a conclusion no one can escape if they hope to win at crawl; certainly the developers know this as well. My list of anti-positioning mechanics that are more egregious would also include teleport other and fear. Tele other is a problem because it's mostly an interface screw ("oh, I got teleported? right, the hellwing."). Fear is a problem because it truly does prevent all use of melee (in contrast to mesmerize), and the enemies that currently use it aren't particularly effective at exploiting it.

However not all repositioning effects are the same, and Dig or fog are not even the first choice for dealing with the group of "move player somewhere near" mechanics. Most often just repositioning to exclude some LOS will be enough. Forcing the player to adapt LOS away from "best kill point for this group of (or potential group of) monsters" to "best point to decrease LOS of this monster since it can reposition me" isn't the same as a player trap, either. The answer is not always to dig, and the player generally has no requirement of a fixed cell to which enemies must be navigated. Furthermore the player is doing a lot of this "find a kill spot" positioning already, and that's part of the puzzle-like aspect of melee that people like. We shouldn't treat the player's current position as some kind of unassailable thing, but unlike HP, which is threatened by monsters at every opportunity, the player's position shouldn't often be changed.

Repositioning monsters relative to the player can also do some interesting things, but changing some of LOS is a fundamentally distinct thing. We might need to reduce the number of these mechanics and be a bit more careful with how we reposition the player for those remaining, but it's better for crawl that it's in the gameplay toolbox.

[Utis]

With regard to the role of positioning: I also find this one of the most interesting aspects of crawl. And the relative predictability of monster movement is a most important part of it. Long before I played crawl, I played some DROD:

http://drod.caravelgames.com/

Now, crawl isn't a puzzle game, but back then I always thought that a blend of some DROD mechanics with tactical fight and RPG strategic character development would be a wonderful game. Crawl comes closest to this.

So, I would like to see monsters with something special in the way they move with regard to player movement. Maybe DROD could be harvested for some ideas.

[dpeg]

Sirens retreating into deep water was mentioned as a problem. Here are some recent commits (a lot more has been going on, please read the changelog.

Spoiler: show

  Code:

commit 90c8bdd2ab93d0ffddaed583e2b4238c93a3a166
Author: DracoOmega <draco_omega@live.com>
Date:   Sun Mar 9 11:40:59 2014 -0230

    A couple mesmerisation tweaks
   
    Reduce mermaid/siren mesmerisation power, increase base duration,
    make breaking mesmerisation grant brief immunity to being
    remesmerised.

commit e1b2bb2750fdc6f07279ba58f8099cf7734926f0
Author: DracoOmega <draco_omega@live.com>
Date:   Fri Mar 7 13:29:53 2014 -0330

    Reduce siren pulling frequency
   
    Sirens were unintentionally able to pull more than one space a turn
    if they were standing in water at the time, and this reduces the
    chance still further (to at-most, once per 10 aut).

commit bb6088fec99f460e3e42ac86610d8cb552bdf78b
Author: DracoOmega <draco_omega@live.com>
Date:   Fri Mar 7 13:22:32 2014 -0330

    Adjust siren movement behavior
   
    This changes the metric for how sirens select their movement destination
    in a couple ways. They no longer try to surround themselves completely
    with deep water, but instead prefer at least one neighbouring tile
    be shallow water or land. The presence of a single tile of deep water is
    valued more highly, but multiples more lowly. Moreover, they will no
    longer always reposition if a more optimal spot is visible to them, so
    long as their current location is 'okay enough', which should reduce the
    distances they move while still trying to give them sufficient terrain
    advantage.
   
    In general, this should cause them to make themselves inaccessible far
    less often, as well as move away less often, period.

[Amnesiac]

I haven't read everything, but I think spines should be removed, because it's troll that you HAVE to fight a monster since it's faster, but you hurt yourself doing so if you don't use wands, but it's not strong enought to encourage using wands, so you end up almost inevitably hurting yourself and puttiong yourself in danger for some reasons known only to the dev of monster spines. This is also imbalanced between AC and EV forcused characters, afaik, and the latter sometimes doesn't have other choice, especially Fe and Op.

While I'm at it, monster kiting is unnecessary annoying, imo. Wasn't monster low hp escape trigger removed from most monsters for the similar reason at some point?..(not that I know the exact reason, but I remember that even popcorn monsters used to try running away)

[and into]

The lack of symmetry between AC and EV isn't a problem per se (actually it is good to make those forms of defense more distinct), but I agree nonetheless that spines aren't interesting on monsters. Keep Ds player mutation, give porcupines a chance to give you a weaker-than-manticore version of barbs every time they hit you in melee, if want to make them special, otherwise remove them.

[Tedronai]

[...]give porcupines a chance to give you a weaker-than-manticore version of barbs every time they hit you in melee

Don't forget to give Hell Sentinels the uber-version of barbs (this is not a joke, I support that solution)

[Siegurt]

I haven't read everything, but I think spines should be removed, because it's troll that you HAVE to fight a monster since it's faster, but you hurt yourself doing so if you don't use wands, but it's not strong enought to encourage using wands, so you end up almost inevitably hurting yourself and puttiong yourself in danger for some reasons known only to the dev of monster spines. This is also imbalanced between AC and EV forcused characters, afaik, and the latter sometimes doesn't have other choice, especially Fe and Op.

While I'm at it, monster kiting is unnecessary annoying, imo. Wasn't monster low hp escape trigger removed from most monsters for the similar reason at some point?..(not that I know the exact reason, but I remember that even popcorn monsters used to try running away)

Wands aren't required (throw/fire stuff, or wait until it runs out of shots and approaches you, you did know they have a limited number right?) I'm not really sure what you mean by "EV focused chars don't have a choice"

At very least in this case you do have a choice (between taking the movement damage to hit them, or waiting it out, or using ranged) vs. with the old siren behavior you didn't have a choice (if you can't attack at range 2, you're stuck with no options for killing it) The new behavior sounds like it should eliminate the most terrible aspect of that though.

[Baldu3]

...with the old siren behavior you didn't have a choice (if you can't attack at range 2, you're stuck with no options for killing it)...

I think the small misunderstanding in this thread about last week's sirens isnt that melee characters would actually not have any way to get rid of them, but rather that they would guarantee item loss even in uninteresting situations. Whether wands/ammo/scrolls, something like 75% of the time sirens would be irrelevant popcorn that would each cost 2-3 curares/wand charges/javelins that would end up at the bottom of deep water. The other 25% of the time they could lead to often frustrating but at least somewhat interesting situations. I havent met the newest version of sirens but the note about them made them seem much better.

They are still a testament to the thread's subject as the challenge they can bring is almost completely irrelevant for characters that are firstly shooters/blasters.

[Amnesiac]

Wands aren't required (throw/fire stuff, or wait until it runs out of shots and approaches you, you did know they have a limited number right?)

I said that the monster is faster, so I was talking about porcuperines, did you read my post carefully?

I'm not really sure what you mean by "EV focused chars don't have a choice"

I mean that EV focused chars don't have a choice but to be EV focused chars and that's especially true for Fe and Op, so they hurt themselves even more, when fighting porcuperines.

Their spines are an absolutely uniteresting feature (as well as comodo sickness) that doesn't actually encourage any tactics, their spines are not really dangerous enough to burn consummables, so they inevitably make you trade your health for getting rid of them, which doesn't actually encourage you to do any tactical decisions. And forcing you to use ranged combat just for annoying monsters is quite unreasonable, like duvessa said.

[crate]

EV is actually better against spines than AC I believe, not that that makes them any more interesting.

edit: nevermind see below

[dck]

It used to be, then this commit changed it and now EV is completely useless against them.

[and into]

It seems to be an unstated assumption that magic and melee should have complete parity or equality or something, but why should we assume that? It doesn't actually measure up to game play. Melee is something that absolutely all characters engage in at various points in the game. It has the best and most straightforward interface, and even on a very casting-centric character, when something non-threatening (relative to level and power of your guy) is in the way, you are going to plow through it in melee rather than cast. So having a lot of enemies who screw with your melee too much detracts from game play in a way that, e.g., silence does not.

Speaking of which, silence doesn't just stop spells, you know? I've avoided Mennas much more often when I am playing non-casting guys. I've killed Mennas several times on the strength of Shadow Creatures and a single application of fast (from potion, wand, or spell) alone, so I actually don't see Mennas strictly as "anti-mage." However the fact that I cannot use ?blinking or ?tele if the battle goes south means I am much less inclined to fight Mennas unless I am completely confident and don't mind burning the !might and !speed or whatever.

To take another example: MP draining attacks are, again, something that punish you more if all you know how to do is cast spells, but it punishes all characters to some degree since lots of things, from god abilities to evocations, use MP. But it isn't an "either-or" the way that super high EV monsters *greatly* devalue a whole range of tools, but *all* forms of melee in particular, while other tools (EV-ignoring attacks) become essentially necessary.

So aside from the question of the difference between magic and melee in game play, there's also a question of degree. A lot of monsters in Crawl are well designed such that certain kinds of attacks or tactics are weaker, while others are stronger, but the swing is not too extreme in either direction. Blink frogs are a good example of an exceptionally well designed monster in Crawl, for instance. They devalue but do not "block" or completely trivialize certain kinds of attacks, still have certain weaknesses (no rPois, e.g.), and they make you rethink your positioning but are dangerous enough for their depth that this is fun rather than annoying. (Also they aren't too common and nothing else except uniques, including player ghosts, and I guess jumping spiders, use blink in Lair. That distinctiveness is important or else you'd get tired of blink frogs much more quickly.)

The problem with a lot of the enemies that were raised earlier can be summed up as
1.) however important and well-supported magic is, melee is *the* central mechanic in Crawl;
2.) these enemies devalue melee (or just render it obnoxious) to an extreme extent, far beyond the effect silence/MP drain have on spell casting;
3.) several enemies guilty, to varying degrees, of 2. were added at once and thus are/were fairly common (and thus way too common) in certain parts of the dungeon in Trunk.