Fighting reform

[galehar]

Well if you make C different for different weapons you are making the new formula more complex and spoilery, just like the current one. I don't really see the point in changing at all if that is the case.

Huh, not really. Speed is still linear, it just has a different slope. Doesn't seem very complex to me. On the other hand, the current delay formula seems simple, but how it affects the damage output isn't really obvious. Aren't you guys the ones who keep having to explain it over and over again to newbies?
It seems to me that you're feeling attached to the current formula just because you're used to it. Go back to the wiki page, and take a good look at the graphs, then come back and tell me that this formula is just fine, and if you were designing a brand new game, you would use the same one.

[crate]

I think the current formula is not particularly good from a design standpoint, but I don't design games: I play them. It is very good from a gameplay standpoint imo.

If you want to make the current formula less spoilery you can just, you know, tell people how it works in-game. Anyone who doesn't immediately realise that going from 0.75 to 0.7 attack delay is a bigger damage increase than going from 1.05 to 1.0 attack delay is not trying to do the math.

[KoboldLord]

The current system typically involves checking a spreadsheet for the skill required for minimum weapon delay, or memorizing those values for the best weapons in each type. The damage boost to a double sword from getting to 10 long blades skill is roughly equivalent to Haste just from the weapon delay improvement alone, and the damage boost from getting it up to 16 long blades skill is equivalent to another Haste multiplied into that Haste. And then you use a real Haste so you have three nested Hastes multiplying into each other.

Haste is stupid overpowered, as I'm sure we all agree. Weapon delay is an even bigger damage multiplier, though, and that means that weapon delay is a big bleating alarm klaxon that you dare not let alone for even a moment until you've hit the cap for your weapon type, because there are simply not that many skill investment opportunities in the early and mid-game that are going to be more valuable than Haste riding another Haste piggyback.

Damage++ is a boring investment, even if it is obviously the most powerful by a huge margin. As a conjuror, I usually have the liberty to turn off my conjurations when I finish earning a new toy. I could up the damage a little, and often I should, but that's boring and I have books and books full of new toys to play with on that character. My current damage output will do, especially if I have the opportunity to invest in something that approaches my problems from a different direction, like necromancy for Sublimation of Blood or charms for Swiftness, rather than just spending it all on damage++. With a melee character, I don't have that sort of choice. If I'm going to put my early-game xp in anything that isn't damage++, it had darn well better be something absolutely game-changing because it's going to cost me a Haste multiplier.

[seth]

My thoughts as someone who has been pretty much exclusively playing melee characters for a few years:

Maces: Not sure how the AC bypass really is different from increased damage. Feels spoilery that that is the benefit. I don't even know how AC helps me, except that i know when I have more of it, I take less damage. So, Maces have more damage than the number on the weapon states, but the last bit is a hidden variable? I'm not sure why knockback has been eliminated from discussion. I think some people were saying it could lead to unwanted situations. I doubt that, though. I think this would be a different way of encouraging people to leave corridors, since if you have your enemies all lined up you won't get a benefit from it. I think the main use of it would be to juggle enemies when they surround you--when one enemy is knocked back you can then hit the other enemy and hoped it is knocked back while the other one steps toward you.

In addition, as a melee player it gets boring that the decision you make the most in tricky situations is to run into a corridor or use an item. I'd like a few more ways to control a situation just by virtue of being a melee character.

Axes: If you only have a chance at hitting three enemies with cleave, it seems a little underpowered--not much benefit to leaving a safe corridor. But hitting all enemies seems maybe a little silly. What about the five in front of you and to the sides? Mostly, I like this.

Long Blades: Higher damage seems boring, and almost as though it'd become the no brainer choice. I mean, at least with long blades you'd know what you were getting, unlike Maces (for AC bypass). But I guess parry and riposte are ruled out, so Idk what else they could do. Critical hits? eh...

i like Staves and short blades and polearms

[FalconNL]

I'll add my two cents to this discussion as well. I really like twelwe's idea of differentiating the different weapon types, but as others have mentioned the proposed idea makes things too complicated. Instead, I propose each weapon class just have a single inherent property, which also serves to differentiate their playstyles.

Short blades: Very high speed, low damage. Doesn't need to be changed too much from its current incarnation: gives a major damage bonus when the victim is impaired (sleeping, confused, etc.) Since stabbing will pretty much solely be done with short blades the stabbing skill is redundant, so get rid of it and just tie stabbing bonuses to the short blades skill. It doesn't make much sense to stab someone with a mace anyway. Play style: sneaking up and stabbing.

Long blades: High speed, medium damage. Special ability: has a chance of blocking melee weapon attacks. Play style: flexible, both corridor and open combat work fine.

Staves: Medium speed, low damage. Special ability: has a chance of blocking unarmed melee attacks. Play style: fairly bad to attack with, but they will only be wielded by casters, who won't be attacking physically anyway.

Maces & Flails: Low speed, high damage. Special ability: Big and heavy and thus they have a higher mindelay than the other weapons, but when they hit, they hit hard. Great against the harder targets in extended, like Pan lords and such. Play style: stick to corridors as much as possible.

Axes: Medium speed, medium damage. Special ability: Axes are swung around, and since they are heavier than long swords they have enough momentum to hit multiple targets without stopping. Deal 50% damage to all adjacent enemies other than the one you attacked. Play style: out in the open.

Polearms: medium speed, medium damage. Also doesn't need to be changed much. Special ability: Reaching. Play style: mostly corridors.

Bows/Crossbows/Slings: Fine as they are. Play style: sniping at a distance.

[galehar]

Maces: Not sure how the AC bypass really is different from increased damage. Feels spoilery that that is the benefit.

It's only good if the monster has any AC. It's very good if it has a lot of it. It's spoilery until we make the AC/EV properties of monsters more visible.

Axes: If you only have a chance at hitting three enemies with cleave, it seems a little underpowered--not much benefit to leaving a safe corridor. But hitting all enemies seems maybe a little silly. What about the five in front of you and to the sides? Mostly, I like this.

Yeah, hitting 5 enemies is what I've suggested in a previous post. Will update the wiki.

Long Blades: Higher damage seems boring, and almost as though it'd become the no brainer choice.

I haven't said higher damage, I said better stats. There's also speed and accuracy you know. Why would it be a no brainer? They would usually be a bit more effective than another weapon, but less so in specific cases (axes be better against multiple opponents, maces against high AC, staves against high EV,...). It's similar to comparing vorpal with the other brands. And vorpal isn't a no-brainer.

[Bloax]

It doesn't make much sense to stab someone with a mace anyway.

What do you think would be more effective against an ogre;

Stabbing it in the neck with a dagger, or bashing it's skull in with a huge hammer?

The former would be a fatal, but not instantly deadly wound (mace to the face incoming); while the latter would definitely kill it. (Because surprise - you just smashed the insides of it's head.)

[FalconNL]

What do you think would be more effective against an ogre;

Stabbing it in the neck with a dagger, or bashing it's skull in with a huge hammer?

The former would be a fatal, but not instantly deadly wound (mace to the face incoming); while the latter would definitely kill it. (Because surprise - you just smashed the insides of it's head.)

Good point. I think the problem here is in the term used. I was going by the literal meaning of stabbing, i.e. to use your weapon in a thrusting manner, which does very little with a blunt weapon, whereas you're using the figurative meaning of attacking a helpless target. Perhaps Stabbing should be renamed to something like Sneak Attack, which is more weapon-neutral.

Oh, and the dagger would be more effective, provided you can aim it accurately enough to sever his spinal cord, since at that point he would be unable to use his arms, preventing the club to the face scenario

[Sar]

What do you think would be more effective against an ogre;

Stabbing it in the neck with a dagger, or bashing it's skull in with a huge hammer?

Well if you, say, sever its carotid artery, it would lose consciousness very fast.
Regardless (stabbing works on undead as well) it's more of a gameplay concern to keep the playstyle viable. If you could both stab effectively and fight effectively with maces or long blades etc. there wouldn't be much reason for short blades to exist.

[ebarrett]

It doesn't make much sense to stab someone with a mace anyway.

What do you think would be more effective against an ogre;

Stabbing it in the neck with a dagger, or bashing it's skull in with a huge hammer?

The former would be a fatal, but not instantly deadly wound (mace to the face incoming); while the latter would definitely kill it. (Because surprise - you just smashed the insides of it's head.)

really, some people in here act like they never stabbed someone irl

[seth]

It's only good if the monster has any AC. It's very good if it has a lot of it. It's spoilery until we make the AC/EV properties of monsters more visible.

I would love it if the EV/AC properties became more visible--I can start to see why this would be an appealing choice.

I haven't said higher damage, I said better stats. There's also speed and accuracy you know. Why would it be a no brainer? They would usually be a bit more effective than another weapon, but less so in specific cases (axes be better against multiple opponents, maces against high AC, staves against high EV,...). It's similar to comparing vorpal with the other brands. And vorpal isn't a no-brainer.

It's hard for me to treat accuracy as an important stat. Maybe that's because when I first started playing, either the game's manual or the wiki said "you don't typically have to worry about this stat". I would like it if accuracy seemed more important, and if when we equipped a weapon we saw an expected hit rate or something. It's hard to choose accuracy over damage or something else when it's unclear how much it's really affecting you.

I can certainly see how if long blades are faster than other weapons (aside from short blades) that would be enough have people choose it.


I hope I didn't come across too negatively in my other post--ultimately I'm very excited for these changes

[Eji1700]

Just wanted to ask-
Whats the reason that no character can start with a sword and instead must go short blades cross training?

[dpeg]

Well dpeg the thing is that right now all weapons are decently accurate once you pass a (pretty low) threshold of fighting/weapon skill. So if you don't include a large rewrite of how accuracy works then it is mostly negligible that different weapons are more accurate; the bigger one does more damage in very nearly all realistic situations.

Exactly. And as far as I understand, it would really help to increase that threshold. It is fine if all weapons are eventually very accurate. But arriving there later than now will make the decision of weapon transition more interesting. The point is not have players give up their executioner axes but to make them use those later (recall that this point was about "the heaviest weapon is best at all skills").


I snipped the section about the interface because missing/hitting is one of the most readily observed features, in my experience.


Regarding weapon type diversification: if we cannot make the distinction between long blades and maces (say) meaningful, there is no problem with removing one type.

[inkydood]

Fi/Gl can start with falchions.

Which still begs the question, why are Long Blades so exclusive starting out compared to Maces, Axes, and Polearms?...Also, why do Mace/Flail users start with a Mace and not the weaker whip?

[eeviac]

I've never understood the heavy long blades restriction myself - hand axes and maces are virtually identical to falchions.

[galehar]

falchions have 1 more damage than hand axes, and 1 less delay than maces. They are objectively better than both. If they were available, it would be a no-brainer for species with equal aptitudes. We have stopped adjusting stats, because it became obvious that formulae needed to be changed first.

[eeviac]

And tridents have reach and better base damage than all of the above. Following that logic every Fi/Gl would start with one, but they don't. People just use whatever they want to use. I'm not convinced that falchions are so much better than other low tier weapons that they should be restricted.

[Hannoskaj]

Well if you make C different for different weapons you are making the new formula more complex and spoilery, just like the current one. I don't really see the point in changing at all if that is the case.

Huh, not really. Speed is still linear, it just has a different slope. Doesn't seem very complex to me. On the other hand, the current delay formula seems simple, but how it affects the damage output isn't really obvious. Aren't you guys the ones who keep having to explain it over and over again to newbies?
It seems to me that you're feeling attached to the current formula just because you're used to it. Go back to the wiki page, and take a good look at the graphs, then come back and tell me that this formula is just fine, and if you were designing a brand new game, you would use the same one.

Just a remark on this point: this system could on the contrary be displayed in an intuitively understandable fashion.

Just display these two values in weapon description:
Speed at skill 0: 1.5
Speed at skill 27: 3
for a quickblade, or
Speed at skill 0: 0.5
Speed at skill 27: 1.5
for an exec axe, for example.

The non-spoiled assumption would anyhow be linear progression.
Moreover, the player might not then need to know that two-handers always go to delay 7, it would be shown on each weapon.

And it does allow more freedom for balancing. Finally, I think it also makes things more explicit for balancing itself: it is really done on these two values, C is only a representation in formula.

(Maybe the scale could be changed by multiplying by ten, so as to avoid decimal numbers)

[XuaXua]

There is absolutely no reason anyone who doesn't have penalized polearms skill shouldn't pick Trident.

[acvar]

Let me start by saying that I have in the past been a professional game designer. From that here are a few insights.

Game mechanics need to be overly simple. Complex equations are just a way of hiding bad game design. They confuse and frustrate players, and are against the crawl philosophy of no spoilers. If you need to use exponets, squares, and other complex equations then your system is poorly designed at its base. Do you really want professional mathamaticians as your only target audience?

You need to keep things relevent throught the entire game. The fact that to hit probability becomes irrelevent after a short time is a bug. Giving weapons minimum str requirements would be a bug. Having the effects on speed have obvious break points is a bug.

Base anything is bad. Humans being a basic race that all others are compared against just makes humans a challenge race that is good at nothing and does not get played much. Maces being a basic weapon that all others are compared against will have the same effect.

Now for a more on topic suggestion.

To hit bonuses/penalties need to matter. They currently are irrelevent after the mid game. This can easily be fixed by either scaling monster EV to HD or by increasing the bonuses/penalties of the weapons themselves. If a hand axe is +3 to hit and an executioners axe is a -30 to hit then there is a very good reason for low level players not to use the exe. axe. Only high level characters can use the exe. axe and consistently hit with it. This change would be easier to implement then changing all the monsters EV, and does not require complex formula.

The proposals that a certain weapon be the base and have no special abilities will not work. If the weapon class recieves no bonuses for its vanillaness then it will be weaker in some situations and unused. If you give it better stats to compensate then it will be better in 90% of all situations and will be the obvious choice. You are just setting yourself up for failure. You might if you work really hard balance it all out, but why choose a path that forces you to work hard? Just pick a bonus for all weapons and be done with it. Make sure that those percs are fairly uniform (no shield breaking) and the task of balancing things is trivial.

[roctavian]

I would respectfully disagree with a lot of what you've said. Complex equations can be used to support simple inputs with clear outputs, and the player doesn't need to know how to do each step; sports players are rarely professional mathematicians, despite the relevance of geometry to their given sports. Much of what happens in Crawl happens "under the hood" so to speak, and this information is deliberately withheld to allow players to focus on strategy and gameplay rather than numbers. At present, the way minimum delay works in the game forces players to look at the numbers (Okay, I hit minimum delay with this weapon at X skill, so I need to train up to...) and renders many skills irrelevant after a certain point. I'm not involved in Crawl development at all, but I can easily say that the purpose of this reform is to remedy that issue.

As far as the rest -- that to-hit bonuses and penalties need to matter throughout the whole game -- is more in a gray area. I would argue that not every concept in a game necessarily needs to be relevant at every point, particularly when the game is as complex as Crawl. I think there is a good case to improve the EV of appropriate monsters later in the game, perhaps. But the realities of combat in the late-game are different than in the early game, as many threats take on different forms (hellfire, torment, etc) and have different countermeasures (line-of-sight, teleportation, and so on). Part of what makes Crawl interesting is that in any given game, a player's strategy will have to change and adapt depending on the situation. Accuracy and evasion have binary consequences (did I hit? or did I not hit?) and I think making them a more important part of later-game play wouldn't really make the game more interesting.

[acvar]

I would respectfully disagree with a lot of what you've said. Complex equations can be used to support simple inputs with clear outputs, and the player doesn't need to know how to do each step; sports players are rarely professional mathematicians, despite the relevance of geometry to their given sports. Much of what happens in Crawl happens "under the hood" so to speak, and this information is deliberately withheld to allow players to focus on strategy and gameplay rather than numbers. At present, the way minimum delay works in the game forces players to look at the numbers (Okay, I hit minimum delay with this weapon at X skill, so I need to train up to...) and renders many skills irrelevant after a certain point. I'm not involved in Crawl development at all, but I can easily say that the purpose of this reform is to remedy that issue.

Sorry but this is just wrong. The point of any game is to create positive feelings in the one playing it. In a game it is good to challenge the player. It is unacceptable to frustrate a player. Frustration occures when a person feels a lack of control. A person can not feel in control when they do not understand what is going on. Having all the displays right in front of you does you no good if you do not know how to adgust those displays. If it was all good we would not have people asking for/searching for spoilers instead of just playing the game. Why do people waste time asking for spoilers instead of playing the game? The search for spoilers is a symptom of a bug in the game. We have all seen it. People ask how things work. Somebody gives them the equations for it, and they invariably ask "so whats all that mean?" The answer is always by neccessity unsatisfactory since it is less then the truth that the equation represents. In other words if the equation basically does "X" then why don't you just simplify the equation and make it actually do "X".

As far as the rest -- that to-hit bonuses and penalties need to matter throughout the whole game -- is more in a gray area. I would argue that not every concept in a game necessarily needs to be relevant at every point, particularly when the game is as complex as Crawl. I think there is a good case to improve the EV of appropriate monsters later in the game, perhaps. But the realities of combat in the late-game are different than in the early game, as many threats take on different forms (hellfire, torment, etc) and have different countermeasures (line-of-sight, teleportation, and so on). Part of what makes Crawl interesting is that in any given game, a player's strategy will have to change and adapt depending on the situation. Accuracy and evasion have binary consequences (did I hit? or did I not hit?) and I think making them a more important part of later-game play wouldn't really make the game more interesting.

You are agreeing with me whether you realize it or not. There is nothing wrong with introducing new obstacles that require new methods to overcome, but it is bad design to discard obstacles. You don't want to add a new ball to be juggled just to remove an old one. You want to keep increasing the number of balls to be juggled.

[pratamawirya]

In other words if the equation basically does "X" then why don't you just simplify the equation and make it actually do "X".

I'm not really sure what you're trying to imply here, but I believe nobody in their right mind would ever make complex equations just for the hell of it.

[BlackSheep]

The simplification you're referring to is visible on the graphs that are often used for illustrative purposes on the dev wiki. The equations may be complex, but that's because the desired output isn't a straight line.

[acvar]

The simplification you're referring to is visible on the graphs that are often used for illustrative purposes on the dev wiki. The equations may be complex, but that's because the desired output isn't a straight line.

I understand this, but why is it desired that it not be a staight line? The fact that it is not a straight line is only useful if the players understand it is not a straight line, understand exactly how the curve is created, and as such can informatively respond. Otherwise it is just a bunch of coding masterbation.

[pratamawirya]

In other words if the equation basically does "X" then why don't you just simplify the equation and make it actually do "X".

I'm not really sure what you're trying to imply here, but I believe nobody in their right mind would ever make complex equations just for the hell of it.

you've never taken a look at the crawl source have you

i've seen commanderc's posts but i'd like to keep my mind open anyway

[galehar]

I understand this, but why is it desired that it not be a staight line?

Well, the speed graph is a straight line. That's the point of the change. People won't need to understand the delay formula if they know that increasing their weapon skill improve the speed and damage in a linear fashion. You don't need to be a mechanic to drive a car do you?

Sorry but this is just wrong. The point of any game is to create positive feelings in the one playing it. In a game it is good to challenge the player. It is unacceptable to frustrate a player.

I think game design philosophies diverge strongly on this point between commercial games and roguelikes.

Sorry but this is just wrong.

The fact that you used to be a professional doesn't mean you're always right. Your arrogance doesn't encourage discussion.

[snow]

I completely agree with everything Acvar has said. He knows what he's talking about.

To be fair though I think crawl is too old and too complex for some problems to be fixed. Plus there's a lot of "if it's not broken don't fix it" stigma when it comes to changing things that are very archaic and make no sense at all. For example a deep dwarfs damage shaving: it has this overly complex formula with nested dice rolls and the vast majority of the time it just shaves off 1 damage. It would make sense to simply change it to "reduce all damage by 1" but I don't see that ever happening.

[acvar]

I think game design philosophies diverge strongly on this point between commercial games and roguelikes.

Well the goal of a commercial game designer is to make a game that people enjoy playing so I guess the goal of a roguelike game designer is to build a game that people don't enjoy playing? Would you care to defend that?

To be fair though I think crawl is too old and too complex for some problems to be fixed.

The linux kernel is older and more complex but still gets improved.

Plus there's a lot of "if it's not broken don't fix it" stigma when it comes to changing things that are very archaic and make no sense at all.

The problem is as you have pointed out we all know it is broken. That is why people are entertaining ways to fix it in the first place. If you are going to take the time to fix it why not do it right?

[galehar]

Well the goal of a commercial game designer is to make a game that people enjoy playing so I guess the goal of a roguelike game designer is to build a game that people don't enjoy playing? Would you care to defend that?

You said: "It is unacceptable to frustrate a player." That really sounds like how modern commercial games are designed. However, roguelikes don't have such a goal. Permadeath can be very very frustrating. It's only because of all this frustration that the first win is so gratifying.
Most commercial games are designed to "entertain". The difficulty curve is smooth, some even adapt to the player skill so as to not frustrate him. Roguelikes are design to challenge the player. They are merciless, but the joy of overcoming the challenge is much higher.
Really, the game design goals are completely different.

[BlackSheep]

For example a deep dwarfs damage shaving: it has this overly complex formula with nested dice rolls and the vast majority of the time it just shaves off 1 damage. It would make sense to simply change it to "reduce all damage by 1" but I don't see that ever happening.

The formula referenced:
int shave = 1 + random2(2 + random2(1 + you.experience_level / 3));

At level 1, you're getting 1 plus a coinflip. Every 3 levels you get a chance to shave more. Shaving exactly 1 HP all the time would make sense if you wanted to nerf DD severely.

[dpeg]

It will come as no surprise, but I cannot stand the "we professionals know how to do it" attitude. This may go for masonry, but in all forms of art, I have doubts. (I will always admit that production values of professionally generated art is higher, but as I see it, production is all too often confused with quality.)

The objective of a professionally developed game is to sell, either copies or server time etc. This does in no way imply that those are better games. As it happens, Crawl used to have a professional among the devs for a while as well (he did the Health and Magic bars, among other things, and he left for time reasons, a real pity). I talked to him quite a bit, and one of the things he shared was this: a modern game cannot afford to offer actual decisions ("non-linear gameplay") for long, think the first half of the game. Effectively, having the first half of the game be like tutorial/story would be often advisable. I'm not saying that's bad, but it's clearly something completely different from where we are with roguelikes.

Some more differences:

• Fun is an extremely subjective notion. Permadeath may not sell anymore (it used to), but that does not mean that there wouldn't be enough fans of permadeath, or a genre around it (not just roguelikes, also shmups, for example).
• Certain MMORPGs are heavy on grinding. By the above, there will be players who love that. I despise it. Also, might there be a connection between "commercial game" and "grinding is acceptable"? How does acvar's dictum of "It is unacceptable to frustrate the player" mesh with grinding in WoW and that game's success?

All in all, I see a lot of rubbish talk around here, that will not improve Crawl one bit. By contrast, dtsund's Crawl Light fork helped tremendously (and is supported by the devs).

[snow]

I haven't looked at the formula in ages. The Wiki says it averages 1.5 at level 1 and 3.5 at level 27... reducing it to 1 across the board isn't that big of a nerf. Anyway my main point was how obscure and overly complex the formula is. Even something like 1+xl/n would give a similar curve and be much simpler.

[danharaj]

On the topic of professional game design, I can pick any game in my modest library and pick its flaws apart. Design flaws are abundant in games, even ones that are critically acclaimed.

I understand this, but why is it desired that it not be a staight line?

Straight lines are boring.

[XuaXua]

Let me start by saying that I have in the past been a professional game designer. From that here are a few insights.

Unless your other alias is "Notch", you opened up a can of worms, buddy.

[galehar]

I haven't looked at the formula in ages. The Wiki says it averages 1.5 at level 1 and 3.5 at level 27... reducing it to 1 across the board isn't that big of a nerf. Anyway my main point was how obscure and overly complex the formula is. Even something like 1+xl/n would give a similar curve and be much simpler.

You seem to assume that you need to know and understand every formula in the game to play it well. This isn't true. Nobody cares what's the expected value and variance of the DD damage shaving, what's important is that it stays relevant throughout the game. Good players don't play with spreadsheets or lua scripts calculating statistics for them. They just develop a feeling of the game mechanics.

What's important is that the mechanics are easy to understand, not the formulae. For a game mechanic to be easy to grasp, it must not be affected by too many variables. 2 or 3 variables is usually enough. That's important, not the shape of the function. There's nothing wrong with using polynomial, exponential and logarithmic functions. They can be very useful. There's no reason to limit ourselves to linear functions.
Increasing your weapon skill increases your attack speed, whatever your current skill level. That's simple and all you need to know. There's no need to calculate the delay yourself.

Of course, there will always be obsessive players who are convinced that they need to run the maths to play the game optimally. Crawl isn't designed for those players. If you want to play optimally, write a bot. If you want to be a good player, just keep playing, watch others and try to understand the game mechanics, not the formulae.

[glexus]

As a novice with an investment in mechanics being easier to understand, can a change be made such that it's more fun and gives more feedback when you're missing or not doing damage in combat?

I've recently understood the need to raise armour/shields when the message log tells me I miss due to that. However, when I miss or don't do damage to monsters at other times, it's very boring and frustrating and there's not a lot of feedback. He's right there, I'm jabbing away, why isn't he dying? Can it say something like:

"You slash at the orc, but he easily dodges out of the way"
"The orc laughs as you hopelessly swing your blade in his general direction"
"The orc growls angrily as you scratch his armour with your pathetic strike"
"Your novice weapon skills mean you fumble the attack on the orc"
"Your substantial blow fails to even scuff the orc's armour"
"Your blade rings out as it crashes to no effect against the orc's chestplate"

I suppose what I really want to give the player is a better sense of why the battle is progressing in the direction it is. Sometimes, I get a bunch of hits in and nearly kill someone without him touching me, and then suddenly fail to be effective with the next ten hits, with no apparent change between the two states.

The wand resist messages are better in this regard, but not very flavoursome.

[galehar]

Huh, it's already the case. Maybe you should pay closer attention to the message log. Dodge messages have an adjective, and damage strength is expressed by punctuation. However, I'm not sure they are really useful, maybe there is too much variance. Also, the thresholds have been set arbitrarily and I'm not sure anyone has looked closely at them.

[glexus]

Huh, it's already the case. Maybe you should pay closer attention to the message log. Dodge messages have an adjective, and damage strength is expressed by punctuation. However, I'm not sure they are really useful, maybe there is too much variance. Also, the thresholds have been set arbitrarily and I'm not sure anyone has looked closely at them.

I will look again when I next play, I'm going from recent memory right now. I know there's notifications like "you miss" and "you closely miss", but that doesn't tell me whether the orc got a good dodge roll or I got a bad attack roll (i.e. is it useful for me to try hitting again or will the orc always dodge?).

I don't recall any messages at all regarding opponent armour strength though?

[galehar]

that doesn't tell me whether the orc got a good dodge roll or I got a bad attack roll

How would the distinction be of any use? Whether you have bad accuracy or it has good EV, the result is the same, hard to hit.

I don't recall any messages at all regarding opponent armour strength though?

Same here, I don't think making the distinction between low damage and good AC roll would be useful. It would be better to give the monster's AC and EV and some info about your accuracy and damage.

[Galefury]

A few months ago I suggested adding some estimation of monster AC and EV to the list of monster resists in the monster description, and people seemed to like the idea. The hardest part for me would be coming up with quantifiers that don't sound like crap (best I could do is "It is somewhat evasive", which is probably not good enough). And of course someone has to come up with breakpoints, which is not trivial because one would need good knowledge of monster stats to assign a good message to a stat range. Might make a good implementable, or maybe the design could be finished in another thread.

[galehar]

Well, putting numerical values might be ok too. I am not against numbers as long as they are meaningful. AC and EV are quite important, and they can easily be compared and understood as numbers since that's what we show for the player. Player's accuracy and damage however would be better represented with adjectives IMO.

[Galefury]

Monster EV and AC are different from player EV and AC, so I'm not sure that showing a number for them would be good. For example 20 EV/AC on a monster is huge, while it is pretty standard on a player. Damaging a 20 AC monster can be very hard with some weapons or spells, while as a player you still take lots of damage from lots of stuff. Probably the to hit values of monsters and players are very different? The damage definitely is. No matter the reason, monster and player defensive stats cant really be compared based on their own number alone, so I think adjectives might be better here too. Showing a number will inevitably lead to people relating those defensive stats to their own, which doesn't work at all.

Just showing the numbers would make design a lot simpler of course. Maybe numbers could be used as a temporary solution, released, and one version later the numbers get replaced by adjectives? Maximum confusion!

It might be hard to come up with a good textual representation of player accuracy and damage, because whether they feel high or low depends so much on what you are fighting. This also applies to numbers of course. I'm not sure they are worth listing because of this. The same applies to MR by the way. The only information I get out of the MR listing is "I have more now", and only if I can remember the previous adjective and put them in the right order. It works a little better for AC and EV because they are almost always relevant, and you see them constantly, but even there I don't usually know if the current value is good or bad for where I currently am, just that I probably want more.

For monster defensive stats, a simple 5-step scale might be good enough. Something like not evasive, slightly evasive, evasive, very evasive, uncannily evasive. Except with better words.

[glexus]

For monster defensive stats, a simple 5-step scale might be good enough. Something like not evasive, slightly evasive, evasive, very evasive, uncannily evasive. Except with better words.

What about a scale relative to other things you've fought? E.g. if you've fought an orc knight it could say that an orc mage has about 1/2 the armour of the orc knight, or something. This at least links it to played experience.

[Galefury]

No, that's terrible. Less useful and harder to code.

[snow]

The AC and EV are already listed on the bots so they're not exactly hidden. It might act as a give away for heavily disenchanted armor if anything but the base AC and EV are shown.

I think the system is fine as is to be honest.

[galehar]

I have pushed a new branch to test the linear melee weapon speed. Feedback appreciated.