Combat and Weapon Skill reform (SECOND VERSION)

[Steel Neuron]

EDIT: After all the awesome feedback received, I have rewritten the proposal with a much simpler, more conservative version. I would really appreciate some feedback in this second iteration, especially if you contributing by discussing the former. I'll keep the entire original thread in spoiler form, for reference:

Spoiler: show

Hey there!

First things first: I'm a noob at crawl. Probably many of the things I propose here are tainted by lack of game knowledge, and I should know better than to open my face hole about something I'm inexperienced on. However, I feel like I want to contribute to Crawl in the long run (maybe once I get as far as to beat it, I'll try and make a patch!) so I guess getting into game design discussions is going to be fun =)

So, take these suggestions with a grain of salt. Anyway, sometimes opinions from an outsider are useful as they are more likely to show a different perspective. I am writing this with full knowledge that it will be hard or maybe impossible to implement (or simply people won't like it) but i'd enjoy it if it brings some debate. Maybe, just maybe, you will like it, in which case that will be a great motivation to look into the source code and start working on it!

Getting on topic... I LOVE this game. However, my major gripe with it is that melee combat is comparatively boring, both in terms of execution (tab tab tab), flavour (lack of weapon differentiation) and restrictive skilling. I have read in the Wiki the initiatives for a weapon and melee skill reform, and I agree with most of it. I understand weapon skills are intended to provide your character with an interesting choice. However, in its current implementation, it feels more like an arbitrary restriction. Unless you're interested in stabbing, whether or not your weapon skill was the "right" choice entirely depends on the loot you get. Delaying your choice of weapon skill until loot unravels doesn't seem to be an option, and even if it was, it would just delay the issue.

To make melee a bit more interesting, I'd like to propose an alternative scheme that makes it possible to mix and match passive abilities and exploit them even after switching weapons. In fact, the key point of this proposal is enforcing the idea that an adventurer may want to keep more than one weapon type handy and switch depending on what challenge he's facing. Complementary weapon types should be the objective. Here is what I'd change, step by step:

Changes:

• Remove all weapon specific skills: Just do away with all of them. It will make sense in a moment.


• Rework the Fighting skill; split it into Endurance, Short Range Combat and Long Range Combat : Here is one of the times where my newbie perspective might come handy. "Fighting" is a misleading name for a skill that increases your HP. its effects, along with those from the specific weapon skills would be redistributed as follows:
  
  • Endurance: Increases your max HP. Maybe some secondary effect for flavour?
  • Short Range Combat: Improves Damage in Melee, accuracy, and attack speed (essentially, it combines the offensive part of Fighting with the Weapon skills)
  • Long Range Combat: Improves Damage at range, accuracy and attack speed (similar)
  So far, a character with level 27 on Endurance and level 27 on Short Range Combat would be equivalent to a character with level 27 fighting and level 27 in all current weapon skills. This may be seen as a raw increase in power for a given amount of points, but can be tweaked if it poses a balance issue. Anyway, this is just to set the framework for the skills that would be introduced. Also, as you will see, with this system it will be more viable to become a true melee specialist, perhaps making things like a Demigod without any magic a viable endgame build, by using combinations of weapons in a smart, tactical way. As another good side effect, decoupling Fighting into its offensive and defensive components makes it possible for a magic user that just wants to dabble into melee to focus his efforts, getting Short Range Combat to max level rather than both Fighting and a weapon spec. This also applies to glass cannon assassins and other similar builds, who might want to maximize damage output without harming the flavour of their role by increasing HP.
  
  
• Make Flails and Whips a separate cathegory from Maces: Any attempt to give those proc abilities should differentiate between these two cathegories... I think we can all agree a mace is mechanically different to a whip!



• Introduce specific "Combat Maneuver" skills: This is slightly different than the approach of giving every weapon type a different proc. Instead of weapon skills, these new "move" skills would make it in, which provide passive procs and bonuses for several weapons, being specially effective (+30%) for one weapon type. In fact, one such skill is already in the game in a format very similar to what I'd love to see, and that skill is Stabbing. So here is a possible list for melee:
  • Stabbing: Enhances the damage of attacks made against monsters unaware of your presence. Works best with Short Blades. Works for Long Blades and Polearms.
  • Disarming: Chance to disarm an armed monster, throwing his weapon to a random adjacent tile. Works best with Whips. Works for Short Blades, Staves and Unarmed combat.
  • Arcing Blows: Increases the likelihood of performing an arcing blow, that will hit monsters in adjacent tiles to your target. Works best with Axes. Works for Long Blades, Staves and Maces.
  • Dazing Strikes: Increases the likelihood of performing a dazing blow, and the time the monster will spend dazed. Works best with Maces. Works for Staves, Unarmed Combat and Whips.
  • Piercing Thrusts: Increases the likelihood of performing a piercing strike, reaching an additional tile in the direction of the attack, ignoring the AC of the first target. Works best with Polearms. Works for Long Blades.
  • Rending Strikes: Increases the likelihood of peforming a rending strike, making the enemy bleed for X% of the damage over time. Works best with Long Blades. Works with Short Blades, Axes and Polearms.
  • Sweeping Attacks: Increases the likelihood of peforming a sweeping attack, throwing the enemy to the ground. Works best with Staves. Works for Whips and Polearms (Essentially it forces the monster to perform an action, "XXX monster stands up" that will take longer for heavier and slower monsters).
  • Opportunity: Increases the likelihood to react quickly to enemy movement, attacking as the monster tries to move into, or away from an adjacent square. Works best with Unarmed Combat. Works for Short Blades, Maces and Whips.
    
    

That would be all for melee. It may look like a lot of skills, but considering the weapon specific skills are out, it adds up to a similar amount. There should be a similar list for ranged maneuvers. Please note that I did my best to respect the philosophy of weapons in DCSS, which means: All effects are passive, and all procs are positive, with no such things as knockback or other effects that would be undesirable (with the exception of cleaving attacks, which could bring up a prompt in the relevant scenarios).

The fun thing, in my opinion, of this skillset is that it's possible to mix and match. Obviously, multiple proc effects could only happen where it's applicable and makes sense (it should be possible to make an arcing + dazing blow with a mace, but not an arcing and piercing strike with a sword). Anyway, there are many reasons why a player might want to train many of these at once (unlike the current system). For instance, an assassin really devoted to daggers could train all three of Stabbing, Opportunity and Rending Strikes, therefore becoming deadly in melee, where as an assassin hybrid would prefer to stick to Stabbing. For the aforementioned assassin, this would naturally bring up many options to the table. Aside from being specially good at daggers, he would naturally be able to switch to unarmed combat in order to deal with mobile enemies, or take a long blade for a more hit and run style.

I tried to make every skill so it appeals to a different situation, and they're often opposed in nature. For instance, Sweeping Strikes and Opportunity are respectively good at dealing with slow and fast monsters. The first one will make huge monsters trip and fall, and it will take them long to stand up, giving you the time to finish them off unharmed. This wouldn't be too useful against a bat. However, Opportunity will make the bat's life miserable, by attempting to strike every time it moves in and out. Let's expand this example:

Arbus the Monk has decided to level up Opportunity and Sweeping strikes. Upon entering a room, he finds two dangerous zombie bats and a huge ogre. Quickly, he thinks of all his possibilities. He is confronted with the following choice:
a) Equipping his Staff and going for the ogre. By making it trip, it will be slowed down and easily dealt with. However, the bats might hurt him bad in that time.
b) Going bare handed, using his lightning fast reflexes to hit the bats as they move around, while focusing on the ogre.
c) (and this is the cool part of the system) Realize he's carrying a whip, which isn't a weapon he particularly expected to use, but that uses both abilities to a medium degree. Using the whip, he may be able to make the ogre trip while striking the bats. Perfect combination!

This should always remember a viable tradeoff, which implies the "preferred" weapon for a specific maneuver shouldn't be more than 30% more effective than the rest at doing that particular thing.

Other examples of maneuvers that help in different scenarios are as follows:
Disarming- Only useful against armed opponents (obviously).
Rending- Good for hit and run strategies. Use when you can kite the monster, and not against swarms of fast creatures.
Piercing- Use in small enclosed areas, and against heavy armored creatures. Not to use in the open.
Arcing- Opposed to piercing. Use against surrounding enemies.
Etc


I'll suggest some more combos to give you food for thought, as they highlight the emergent properties of this system:

Runaway combo: Mace with Dazing Strike, Arcing Blows and Opportunity. By equipping a mace with these skills trained, it is possible to strike monsters that are chasing you (both reflexively and actively) potentially hitting more than one at once and dazing them, allowing you a clean escape.

Ogre Killer: Whips with Disarming and Sweeping Attacks: Everyone knows the power of Ogres comes from their weight and terrible spiked clubs. Tackle both at once by disarming them with your whip and keeping them unbalanced.

The Hallway Grinder: Polearms with Piercing and Rending Strikes: Line all your enemies up in a hallway and sew through them. By piercing and applying rending strikes, the second monster in line will already be hurt when you get to him, so less work to do!

As you can see, there are many possibilities, that offer several degrees of specialization in weapons, and much, much more flexibility in building melee classes. As a character levels up, he might want to start with vanilla Short Range Combat and then specialize along the way depending on loot, or build a specific strategy based on mixing and matching certain procs, changing from weapon to another, etcetera. In this system, I imagine players like true medieval knighs were: Fighters competent with all weapons, that switch from one another when the situation requires for it. While weapons by themselves don't change, they indirectly gain a lot of charisma, since each weapon type is the ONLY capable of certain combinations. For instance, you will find no weapon capable of stabbing, arcing and rending other than the Long Blade. So if you want an assassin that deals a lot of DoT damage and can also defend against multiple opponents in a pinch, that's your weapon of choice.

I also think that this system is not too obscure for new players. All these are clear in what they do and there's less punishment for using a "wrong" weapon that you have acquired than currently. Also, beginner, more simple races could have a straight boost to learning "Short Range Combat" and a penalty to specific maneuvers so it is even simpler to decide how to build, while not penalizing the understandable "wow, awesome loot, I want to try it!" feeling that all of us newbs have. All of these effects are fed to the player through the combat log. Stuff like "Your sweeping attack makes the orc fall over!" "The orc stands up" are not too dissimilar from the current "You slice the Rat like a ripe Choko!" and don't get in the way while learning a game, even if you're not completely sure how they work.

That was it. Sorry for the long read! I hope you guys enjoyed it, and be sure I'll keep updating it with your suggestions and ideas if you did.

Okay, here goes round two, trying to address all the feedback given in the thread, while still aiming for a vaguely similar idea.

Changes:

• Remove Stabbing: Would be redundant with the changes.


• Slightly rework Fighting : It should read something like "Increases your HP, and your damage, speed and accuracy with all weapons, more effectively in those you don't already have training with".
  Let's explain the rationale behind this with some more detail, as it's the pivotal point of this second proposal. Despite the fact I haven't cleared the game yet, I have studied the mechanics because my interest in DCSS is mostly from a development perspective. One thing that rubs me slightly the wrong way is that Fighting and the weapon specs are additive, which means someone who is taking the latter is always better off picking the former, which is reflected clearly in all melee-oriented YAVP posts. I would rather see them complementing each other somehow. So this change would make fighting give you the benefits of roughly 70% of what weapon specs do, but for all weapons, at the cost of slightly overlapping with those. Let me illustrate with an example: The effectiveness of a character with certain weapon is shown as an arbitrary percentual number, which reflects attack speed, accuracy and damage:
  
  
  • A character with 0 Fighting and 0 Daggers would be 10% effective with daggers.
  • A character with max Fighting and 0 Daggers would be 70% effective with daggers.
  • A character with max Daggers and 0 Fighting would be 90% effective with daggers.
  • A character with both Fighting and Daggers at max level would be 100% effective with daggers.

As you can see there is a certain overlap, that in my humble opinion reflects the idea of specialization with more accuracy. A player that has no interest in anything other than Daggers may just train that skill and obtain almost full profficiency with them. However, someone that wants a more generalistic approach may take Fighting, and be reasonably good with all weapons. Not in the level of a specialist, but unlike the current implementation, it makes it an actual choice rather than rending all non-specialized weapons useless. Lastly, a character that wants to stay flexible yet exceed at one particular weapon may train both, achieving perfection in said weapon while still being versatile. This would also have benefits that stem from the last change I suggest (coincidently, while it may appear than training fighting and neglecting the specs is a better tradeoff, the new clause I propose for weapon skills would counter that).


• Keep on adding passive procs to weapons (not necessarily tied to type). Also make Stabbing a weapon proc: This isn't really a suggestion, but rather "keep up the good job". More procs such as Axe Cleave and similar would progressively be added. They should be reflected in the intrinsic text of the weapon and also imply how good the weapon is at it. For instance a dagger could read "this weapon is good at stabbing" an axe "this weapon is good at cleaving" and a long sword "this weapon is decent at stabbing and cleaving".



• Weapon skills also make you better at certain procs: Weapon skills would stay, right as they are now (aside from the overlap with Fighting), but in addition to what they do they should include a clause "Learning XXX makes you better at YYY". For instance, "Learning Daggers makes you better at Stabbing" or "Learning Axes makes you better at Cleaving". Note this doesn't explicitly state that this exact weapon type is required, only that you become better at that kind of action. This is both realistic and intuitive; someone trained in using Axes will tend to use all weapons like an axe. Also, it doesn't complicate the learning process or make it less simple, since an inexperienced player may just make the connection "I'm training axes- I must prioritize axes" and the benefits will be reaped.

So, there goes round two. I tried to address everything that has been said in this thread (thank you for the feedback). How do you feel about this iteration? It should be easy to implement (no huge reforms, just progressive weapon proc implementation as it was being done already, plus the clauses on weapon training). The Fighting tweak is optional and could be kept out, but I think it would help with versatility and improvisation, by enforcing dilemmas such an axe wearer thinking of using a particularly powerful sword, because he's able to cleave with it effectively. The corollary is that cross training and dual speccing would become useful (with or without Fighting) as a character would not only be good at both weapons trained, but also better than a strict specialist at utilizing those weapons that can do more than one proc. A specialist in daggers and swords, for instance, would be better at the stabbing aspect of a sword than a character that has trained Swords only.

[XuaXua]

What would need to be done is not limit the skills to weapons but limit the weapons to your proposed skills. Give each weapon a ranking representing the potential to perform a particular act (where the potential is nonexistant, ranking is 0). Using that, weapons can be differentiated.

[rebthor]

I'm really confused about your case

Arbus the Monk is trained at using both staves and his bare hands, as can be expected from a Monk. He has decided to level up Opportunity and Sweeping strikes. Upon entering a room, he finds two dangerous zombie bats and a huge ogre. Quickly, he thinks of all his possibilities. He is confronted with the following choice:
a) Equipping his Staff and going for the ogre. By making it trip, it will be slowed down and easily dealt with. However, the bats might hurt him bad in that time.
b) Going bare handed, using his lightning fast reflexes to hit the bats as they move around, while focusing on the ogre.
c) (and this is the cool part of the system) Realize he's carrying a whip, which isn't a weapon he particularly trained for, but that uses both abilities to a medium degree. Using the whip, he may be able to make the ogre trip while striking the bats. Perfect combination!

Emphasis added

When combined with

Remove all weapon specific skills: Just do away with all of them. It will make sense in a moment.

[byrel]

Honestly, it sounds like a bear to balance. I don't really have a problem with the current system. You'll find a workable end-game weapon for any of the skills by Vaults, and they're fixing the flavorless bit. As far as execution goes? I think the tactics required for melee are more intricate than anything except maybe a stabber-blaster combo. Tab is often the wrong key to press in melee; usually running away is a much better choice at first.

I don't mind your idea; It's quite well-written, and if crawl were that way I'd enjoy it too I think. But I don't mind it the way it is.

However, you did give me a chuckle in the middle of it:

Upon entering a room, he finds two dangerous zombie bats and a huge ogre.

I think this is the first time I heard zombie bats called dangerous! Now if it were an ogre and two killer bees...

[Steel Neuron]

What would need to be done is not limit the skills to weapons but limit the weapons to your proposed skills. Give each weapon a ranking representing the potential to perform a particular act (where the potential is nonexistant, ranking is 0). Using that, weapons can be differentiated.

That can work as well. In fact, it could be used to relax the boundaries between weapon types.

Rather than a falchion reading "This weapon belongs to the Long Blades cathegory" It could read "This weapon may stab and strike in an arc in trained hands. It's specially appropiate for rending strikes.".

I guess it depends on how easily the message gets across, but mechanically it's the same.

[Psieye]

You do realise that what you're proposing isn't so much a 'patch' as an entirely separate branch of development on git for an almost entirely different game?

I can see where the desire for this proposal comes from: you don't want to be invalidated from most weapons you get just because it's the wrong type. You feel there is not much to melee besides hitting things at range 1 (or 2 if reaching). My impression is: that feeling will disappear if you get into the lategame. "Melee" means determining the best position to be in for a fight (after determining which target has highest priority), when to train up non-melee skills, decisions on evocables/invocables/buff spells. Yes, "melee" doesn't mean "never casts spells", it just means "primarily uses a range 1 physical weapon to kill things". In other words: there's already enough to think about for a melee, no need to add even more for the sake of having more to think of in the early-midgame.

not penalizing the understandable "wow, awesome loot, I want to try it!" feeling

Arguably, once you've mastered the tactical decision making to get by in crawl, the game becomes a story of how you spent your limited EXP budget and limited available loot. That is: the game has meaning BECAUSE there are hard choices to make. That pain of not being able to use whatever you come across serves to amplify the satisfaction when you do find things that fit your requirements - in addition to having made you think about whether it's worth changing your game plans to accommodate the unexpected awesomeness.


Not that your proposal is a bad idea (if we completely ignore the cost/benefit analysis), just that it's trying to fix a problem crawl doesn't really have. Or perhaps it does exist but we just don't care once we're consistently hitting the later stages of the game.

[Steel Neuron]

I'm really confused about your case

Arbus the Monk is trained at using both staves and his bare hands, as can be expected from a Monk. He has decided to level up Opportunity and Sweeping strikes. Upon entering a room, he finds two dangerous zombie bats and a huge ogre. Quickly, he thinks of all his possibilities. He is confronted with the following choice:
a) Equipping his Staff and going for the ogre. By making it trip, it will be slowed down and easily dealt with. However, the bats might hurt him bad in that time.
b) Going bare handed, using his lightning fast reflexes to hit the bats as they move around, while focusing on the ogre.
c) (and this is the cool part of the system) Realize he's carrying a whip, which isn't a weapon he particularly trained for, but that uses both abilities to a medium degree. Using the whip, he may be able to make the ogre trip while striking the bats. Perfect combination!

Emphasis added

When combined with

Remove all weapon specific skills: Just do away with all of them. It will make sense in a moment.

True, I tried to add some flavour to the narrative and I muddled the message. I'll rewrite it so it makes more sense. I meant that he's training at staves and bare hands from a flavour perspective.

[Steel Neuron]

You do realise that what you're proposing isn't so much a 'patch' as an entirely separate branch of development on git for an almost entirely different game?

I do. It's a big change, but wouldn't necessarily differ from Crawl philosophy. So I agree it's a big branch but not sure what you mean about a different game.

Not that your proposal is a bad idea (if we completely ignore the cost/benefit analysis), just that it's trying to fix a problem crawl doesn't really have. Or perhaps it does exist but we just don't care once we're consistently hitting the later stages of the game.

This is a bit surprising. I researched a bit to see if people were unhappy with the melee system and this is what I found: https://crawl.develz.org/wiki/doku.php? ... pon_reform . Also, Axe cleave is on trunk and other initiatives are being weighted. I'd say that this post is, at least, in the general direction of what other people want.

I decided to write this post because it appeared to me I wasn't the only one who disliked the melee system. Rather than dislike, I think the appropiate feeling is that it could be improved. Maybe I did wrong in phrasing the proposal as a way to "fix a problem". I guess the objective is enhancing tactical combat without harming usability. This might be my newbieness talking but it looks to me like Crawl is a deeply strategical game, but not so tactical. By comparision, ToME would be tactical but not strategical. More emphasis seems to be put in management of resources (as you said, a tale of how you spent your EXP) and less on actual fight maneuvers, and that's what I tried to address.

Keep in mind that this proposal comes from the perspective of a beginner. You see that as something negative; "you haven't experienced the endgame". However, I think appealing to new players is very interesting if the genre is to become a little more known. Most newcomers are going to experience the road to D:10 over and over again, better give them some tools to make it more fun.

[byrel]

I guess the objective is enhancing tactical combat without harming usability. This might be my newbieness talking but it looks to me like Crawl is a deeply strategical game, but not so tactical. By comparision, ToME would be tactical but not strategical. More emphasis seems to be put in management of resources (as you said, a tale of how you spent your EXP) and less on actual fight maneuvers, and that's what I tried to address.

Keep in mind that this proposal comes from the perspective of a beginner. You see that as something negative; "you haven't experienced the endgame". However, I think appealing to new players is very interesting if the genre is to become a little more known. Most newcomers are going to experience the road to D:10 over and over again, better give them some tools to make it more fun.

Generally, when I die at least, I made 1-5 tactical errors that killed me. That includes quite a handful of dead Merfolk Gladiators before Lair. Bad tactics is what keeps people from clearing lair; up til then, strategy is either trivial or its consequences won't show up yet. (Trivial strategy is skilling. For a melee character, this is nearly always a no-brainer. Late-game consequences come from things like wasting cMut and enchant armour, which won't hurt, and might help slightly early game) Simply put, people generally don't die from bad strategy till after Lair.

Tactics is the biggest challenge to a melee character early game already. I don't think that's really a problem.

[Steel Neuron]

I guess the objective is enhancing tactical combat without harming usability. This might be my newbieness talking but it looks to me like Crawl is a deeply strategical game, but not so tactical. By comparision, ToME would be tactical but not strategical. More emphasis seems to be put in management of resources (as you said, a tale of how you spent your EXP) and less on actual fight maneuvers, and that's what I tried to address.

Keep in mind that this proposal comes from the perspective of a beginner. You see that as something negative; "you haven't experienced the endgame". However, I think appealing to new players is very interesting if the genre is to become a little more known. Most newcomers are going to experience the road to D:10 over and over again, better give them some tools to make it more fun.

Generally, when I die at least, I made 1-5 tactical errors that killed me. That includes quite a handful of dead Merfolk Gladiators before Lair. Bad tactics is what keeps people from clearing lair; up til then, strategy is either trivial or its consequences won't show up yet. (Trivial strategy is skilling. For a melee character, this is nearly always a no-brainer. Late-game consequences come from things like wasting cMut and enchant armour, which won't hurt, and might help slightly early game) Simply put, people generally don't die from bad strategy till after Lair.

Tactics is the biggest challenge to a melee character early game already. I don't think that's really a problem.

I understand.

So I'm convinced that weapons aren't an issue in terms of balance or tactics. However, what about flavour, diversity, options? While you're making a good case about how weapons work in the endgame, there seems to be one single way to do things. You said it yourself "Trivial strategy is skilling. For a melee character, this is nearly always a no-brainer". DCSS has a crusade against no-brainers. Magic is so beautifully complex when compared to melee that, while functional, it leaves me wishing for more, and apparently people in the article on the wiki have a similar opinion.

Keep in mind that my proposal is a skill change reform, not a weapon reform. From the player's perspective, things would work the same: press the direction of the monster and damage happens. In fact, you could skill as to mimic the former gameplay spending the same amount of skill points. However, these skills try to offer different approaches to the same problem.

Would, assuming similar balance, prefer to play with the system I propose or something similar? Or stick to the true and tested formula? (it's not a rhetorical question, I really want to know how you'd feel about it)

[byrel]

I understand.

So I'm convinced that weapons aren't an issue in terms of balance or tactics. However, what about flavour, diversity, options? While you're making a good case about how weapons work in the endgame, there seems to be one single way to do things. You said it yourself "Trivial strategy is skilling. For a melee character, this is nearly always a no-brainer". DCSS has a crusade against no-brainers. Magic is so beautifully complex when compared to melee that, while functional, it leaves me wishing for more, and apparently people in the article on the wiki have a similar opinion.

This is actually why I like playing casters (particularly complex support-oriented casters) generally more than melee. But that's very much a personal preference of course. I do like the new polearm distinctive, reaching; it makes the weapon decision more interesting. And I really do hope that cleave and such end up spicing this up. Melee variety, I agree, is an issue. What you suggest would, in fact, reduce this problem some, but it's a major overhaul that will be plagued with balance issues for years to come. I think a more incremental approach would be more practical.

I also note that yours is (probably) slightly more realistic. While it ignores the different muscle memory required for different weapons, it's not like there isn't a lot of common knowledge. Though I understand that realism isn't really a goal, it's better to have than not, all else being equal.

[Psieye]

This is a bit surprising. I researched a bit to see if people were unhappy with the melee system and this is what I found: https://crawl.develz.org/wiki/doku.php? ... pon_reform . Also, Axe cleave is on trunk and other initiatives are being weighted. I'd say that this post is, at least, in the general direction of what other people want.

Hmm ok, my articulation wasn't precise. So, I can agree that it's in the general direction of what people want: there is more that could be done for the tactical side of the early-midgame. However, the way you've written your proposal you've shot yourself in the foot by starting with "let's do away with all these weapon skills". That brainstorm wiki page you linked says:

The distinction of weapon types currently employs skill, rarity, brands, numbers, handedness and, to a lesser extent stats, god and whether its a butcher tool. All of this is good and should stay, but the fact remains that for many fighters, the weapon they use is not a very relevant choice.

Further down the page, a 'totally new approach' is proposed but again it states:

Here are the goals:

- To a great degree, differentiate weapon skills from each other.

Neither of the proposals on there are suggesting what you propose: get rid of the weapon skills altogether and replace them with completely different things. This combined with the fact enemies should get to use the same weapon effects as you proposed would mean the game will be 'almost entirely new' - in terms of how tactics (for non-melee playstyles as well given monsters using these effects) and EXP budgets would have to change.

However, I think appealing to new players is very interesting if the genre is to become a little more known. Most newcomers are going to experience the road to D:10 over and over again, better give them some tools to make it more fun.

Such tools do exist: magic, gods and evocables. It's my view that a new player will get the most fun out of exploration of the dungeon and figure out what hostiles are bad news and how to deal with them. Further in my view, melee is the simpler playstyle choice so new players can focus on getting the fun described above and not have to worry about things like EXP budgeting, weapon choice, gods (because Oka/Trog are nice, easy choices for a newbie) and spells while they iron out their tactics. There already is a hella lot they still have to think of: consumable resource management. I'd think throwing in a more intricate tactical system where you have to pay even more attention to what weapons to use in which situation would be... overwhelming to new players.

This is really a question on user demographics: are most new players to crawl going to love being bombarded with even more delicate details to balance out? We don't want to dumb the game down for the sake of attracting new players but I'm not sure going the opposite direction would be helpful. I appreciate you've designed your proposal around trying to make it intuitive and it is intuitive to pick up the concepts. But it adds to the burden of learning (and balancing) proper tactics and further befuddles newbies on what the best decision should be in a given tactical situation.

DCSS has a crusade against no-brainers. Magic is so beautifully complex when compared to melee that, while functional, it leaves me wishing for more, and apparently people in the article on the wiki have a similar opinion.

Main point is: the brainstorm wiki page is trying to make weapon choice more meaningful but still keep it a strategic choice. You want to bring that choice down to the tactical level. What weapon type you pick will affect your tactics, but you can't just juggle between the weapon types at will to suit the situation.

[rebthor]

So I'm convinced that weapons aren't an issue in terms of balance or tactics. However, what about flavour, diversity, options? While you're making a good case about how weapons work in the endgame, there seems to be one single way to do things. You said it yourself "Trivial strategy is skilling. For a melee character, this is nearly always a no-brainer". DCSS has a crusade against no-brainers. Magic is so beautifully complex when compared to melee that, while functional, it leaves me wishing for more, and apparently people in the article on the wiki have a similar opinion.

In the early game, magic is also a no-brainer. Raise your spell skills to get to acceptable fail chances and progressively upgrade your spells in your starting book until you reach a point where you can branch out. It's no different than raising weapon skill to min delay and progressively upgrading your weapon to the next one in the class that you find.

It's only once you've built a character to the archetype that most people start to branch out.

[Steel Neuron]

The distinction of weapon types currently employs skill, rarity, brands, numbers, handedness and, to a lesser extent stats, god and whether its a butcher tool. All of this is good and should stay, but the fact remains that for many fighters, the weapon they use is not a very relevant choice.

Here are the goals:

- To a great degree, differentiate weapon skills from each other.

Hmm. I see what you mean.

It might be a semantics problem on its core, though. My proposed skills could still be called "daggers", "polearms" and the like, and work similarly, perhaps with a boost to damage stats on that specific weapon. The core principle behind this is portability between skills, so that you can't be easily pigeonholed, and a bit more flavour to the weapon skilling choices, but if the consensus is that the current weapon-specific skills are fine, I can get behind that.

Neither of the proposals on there are suggesting what you propose: get rid of the weapon skills altogether and replace them with completely different things. This combined with the fact enemies should get to use the same weapon effects as you proposed would mean the game will be 'almost entirely new' - in terms of how tactics (for non-melee playstyles as well given monsters using these effects) and EXP budgets would have to change.

I don't see why enemies "should" use these skills. I think it's better to see it as "could". A straight forward implementation would be to have monsters not use any of these maneuvers, and implement them slowly as the devs see fit on whatever monsters it makes sense. I also disagree about EXP budgets. Every former build could be perfectly realized with the same EXP budget; rather than going for 27 levels in fighting and a weapon skill, you need 27 levels in Endurance and SRC/LRC. It is true that new possibilities would arise, but that doesn't imply the former EXP budgets are invalid. If these changes made the game easier, that'd be a balance concern, but as long as they're equal in power to former choices or more of a challenge, nothing would be altered in a meaningful bad way.

This is really a question on user demographics: are most new players to crawl going to love being bombarded with even more delicate details to balance out? We don't want to dumb the game down for the sake of attracting new players but I'm not sure going the opposite direction would be helpful. I appreciate you've designed your proposal around trying to make it intuitive and it is intuitive to pick up the concepts. But it adds to the burden of learning (and balancing) proper tactics and further befuddles newbies on what the best decision should be in a given tactical situation.

That's an interesting question indeed. I think anyone who picks a Roguelike is in for some mentally challenging gameplay, but I thoroughly agree with you about intuition and the burden of learning. Melee is significantly easier to understand, and less profound, than other game mechanics. This should stay like that, as what any new player will do is melee his way out of problems. That's why I think keeping the mechanics to "walk towards where you want to strike" is a fantastic idea, and weapon abilities should stay passive. I see them being a problem if different instances of the polearm evoke started popping up (which took me a while to realize). However, purely passive, always beneficial procs, in my format or any other such as what axes are getting in trunk, can't do any harm.

It appears that the consensus is against the principles proposal. It's fine, I just wanted to try and have an interesting chat! I'll keep learning about the game, and this does help =).

Anyway, I hope more weapon changes are pushed and that we see more stuff like axe Cleave in the future. Despite the great arguments I've been reading in this thread, I still find melee somewhat lacking, in my humble opinion, but I guess there are many angles to approach that.

[Lasty]

I like your proposals, and I think they're largely well thought-out. However, they do radically change Crawl and they would take a lot of work to implement (game balance would be shifted throughout the entire game; monsters would need to be adjusted to handle new player powers; loot drops would get better overall for people with melee/ranged skills; god gifts would improve significantly and might need to be toned down; etc). While I think they would add something interesting, it isn't something that's really needed.

Still, I think your proposal has interesting elements that I could imagine would spark some ideas that get implemented.

Of the skills listed, several of them effectively give free attacks. Opportunity, Dazing Strikes, Sweeping Attacks give free attacks when they trigger (as I understand it); Arcing Blows and Piercing Thrusts give extra attacks when multiple opponents are positioned correctly. The only ones that do something else are Disarming, which reduces enemy damage and Rending Strikes, which increases your damage.

The first group of free attacks seems like the most problematic, since they're free attacks you get without having to do anything you don't already want to do. Of them, Opportunity seems like the most problematic by far, because it allows you to strike first against all enemies and also kill enemies by running away from them and having them follow you.

[Pereza0]

I think many of these ideas are good, and pretty sure some of them would be fun if implemented partially or gradually. But they are too radical to implement as a whole, changing the core of the game so much would mean a step back in the sense that everything would have to be rebalanced and that takes time (as pretty much everybody is saying).

I think that the skills should be left untouched, (except stabbing, that should be gone) and that this weapon specific abilities should correspond to particular weapons. Having a single skill for all weapons makes having the variety that currently exists in crawl pointless (trainable abilities for a few is not good enough) and removes all the specialist/generalist decision about weapons. It also makes artifact weapons a no brainer, it would be better just to pour all your XP in the weapon skill and just use the best artifacts you find since their advantages will almost certainly be a lot better than those the weapon specific abilities can give you (and you can still train those if the artifact you get can't be topped).

These things would work nicely on a fork or an entirely different game balanced to suit it

[Steel Neuron]

Thanks everybody for the input. I now clearly see the big holes in my idea!

I'm working on a second version of the proposal. Much simpler, progressive, and immensely more respectful with the status quo. I'd love you to comment once I have it done. Again, thanks for the lively debate!

[varsovie]

I think I kinda miss the point of the suggestion. It's suppose to be mostly for newbies, diversifying the melee at a tactical level early on, but :
---With a Skill and % based system, any meaningful effects will appear in mid game rather that early game
---It removes a little bit of clarity (train maces for mace) and add some complexity

If we where to implement this, I would propose that we first keep the current skill system and add the effects (aka special move) on some weapon/Unrandarts to try and balance them (% of success depending on current weapon skill). Then maybe thinking renaming the skills, don't forget current weapon skills are involved in the to hit chance, increases the damage and more importantly reduce the weapon delay, all those effects must find a place in the new system and it's may takes a while to balance.

Also the fighting skill should remain has it is, since you're only separating it in three component, one of them always way better than the others.

Also I don't like OPPORTUNITY that much, it's already "present" in some spells/abilities/mutation, and there is some situation where you DON'T want to attack. (sword +hydra or UC +hydra - 1 turn...)

Anyway, good job Steel Neuron, for a newcomer it's hell of a well written post.

[Davion Fuxa]

One thing I like about the current system is that is mostly has depth without complexity. Complexity isn't a bad thing but in DCSS it is nice to have at least one concept that isn't complex; while also retaining a lot of depth in how you experience melee combat. This idea isn't bad but taking into considerations everything outside of Weapons and Melee, it may be something that would be better suited for DCSS2 so to speak.

[rebthor]

I also disagree about EXP budgets. Every former build could be perfectly realized with the same EXP budget; rather than going for 27 levels in fighting and a weapon skill, you need 27 levels in Endurance and SRC/LRC. It is true that new possibilities would arise, but that doesn't imply the former EXP budgets are invalid.

I think this may be because you don't have a lot of experience with the game 27 in a weapon skill is not that common. I mean, I personally won an MiBE - not a class starved for XP by any means - with 26.2 axes, and an MfGl with ~20 polearms. Once you hit the min-delay breakpoint on the weapon you're going to use, the gain from training weapon skill is not that great.

And since I'm not that great, I checked cerebovsquire, crate and minmay's recent wins too and found similar things on all their melee characters. They are not getting full levels of weapon skill until well into the hells if they even get them at all.

[Steel Neuron]

Okay, here goes round two, trying to address all the feedback given in the thread, while still aiming for a vaguely similar idea.

Changes:

• Remove Stabbing: Would be redundant with the changes.


• Slightly rework Fighting : It should read something like "Increases your HP, and your damage, speed and accuracy with all weapons, more effectively in those you don't already have training with".
  Let's explain the rationale behind this with some more detail, as it's the pivotal point of this second proposal. Despite the fact I haven't cleared the game yet, I have studied the mechanics because my interest in DCSS is mostly from a development perspective. One thing that rubs me slightly the wrong way is that Fighting and the weapon specs are additive, which means someone who is taking the latter is always better off picking the former, which is reflected clearly in all melee-oriented YAVP posts. I would rather see them complementing each other somehow. So this change would make fighting give you the benefits of roughly 70% of what weapon specs do, but for all weapons, at the cost of slightly overlapping with those. Let me illustrate with an example: The effectiveness of a character with certain weapon is shown as an arbitrary percentual number, which reflects attack speed, accuracy and damage:
  
  
  • A character with 0 Fighting and 0 Daggers would be 10% effective with daggers.
  • A character with max Fighting and 0 Daggers would be 70% effective with daggers.
  • A character with max Daggers and 0 Fighting would be 90% effective with daggers.
  • A character with both Fighting and Daggers at max level would be 100% effective with daggers.

As you can see there is a certain overlap, that in my humble opinion reflects the idea of specialization with more accuracy. A player that has no interest in anything other than Daggers may just train that skill and obtain almost full profficiency with them. However, someone that wants a more generalistic approach may take Fighting, and be reasonably good with all weapons. Not in the level of a specialist, but unlike the current implementation, it makes it an actual choice rather than rending all non-specialized weapons useless. Lastly, a character that wants to stay flexible yet exceed at one particular weapon may train both, achieving perfection in said weapon while still being versatile. This would also have benefits that stem from the last change I suggest (coincidently, while it may appear than training fighting and neglecting the specs is a better tradeoff, the new clause I propose for weapon skills would counter that).


• Keep on adding passive procs to weapons (not necessarily tied to type). Also make Stabbing a weapon proc: This isn't really a suggestion, but rather "keep up the good job". More procs such as Axe Cleave and similar would progressively be added. They should be reflected in the intrinsic text of the weapon and also imply how good the weapon is at it. For instance a dagger could read "this weapon is good at stabbing" an axe "this weapon is good at cleaving" and a long sword "this weapon is decent at stabbing and cleaving".



• Weapon skills also make you better at certain procs: Weapon skills would stay, right as they are now (aside from the overlap with Fighting), but in addition to what they do they should include a clause "Learning XXX makes you better at YYY". For instance, "Learning Daggers makes you better at Stabbing" or "Learning Axes makes you better at Cleaving". Note this doesn't explicitly state that this exact weapon type is required, only that you become better at that kind of action. This is both realistic and intuitive; someone trained in using Axes will tend to use all weapons like an axe. Also, it doesn't complicate the learning process or make it less simple, since an inexperienced player may just make the connection "I'm training axes- I must prioritize axes" and the benefits will be reaped.

So, there goes round two. I tried to address everything that has been said in this thread (thank you for the feedback). How do you feel about this iteration? It should be easy to implement (no huge reforms, just progressive weapon proc implementation as it was being done already, plus the clauses on weapon training). The Fighting tweak is optional and could be kept out, but I think it would help with versatility and improvisation, by enforcing dilemmas such an axe wearer thinking of using a particularly powerful sword, because he's able to cleave with it effectively. The corollary is that cross training and dual speccing would become useful (with or without Fighting) as a character would not only be good at both weapons trained, but also better than a strict specialist at utilizing those weapons that can do more than one proc. A specialist in daggers and swords, for instance, would be better at the stabbing aspect of a sword than a character that has trained Swords only.

[Pereza0]

Hmm I agree on the part that special abilities such a stabbing and cleaving should tie to the corresponding weapon skills.

I don't like the idea that having fighting fully trained allows you to use every weapon in the game with 70% capacity, the approach would be the following

Train Fighting and be mildly proficient with all weapons -> use branded and artifact weapons you find around the dungeon to compensate -> get an outstandingly good artifact weapon -> stop training fighting and specialize in that weapon getting the weapon ability bonus along with whatever properties it has.
It also makes fighting pointless if you are playing as a specialist, 27 leves arent worth the 10%.

Im not really a big fan of fighting as of 0.11 (gotta play trunk) since reading its description in the skill interface does not give you a good idea of what it does, and it seems most people train it for the health anyway. If it was up to me I would rename it resilience(maybe make the player slightly more resistant to poison and elements?) and put it with the dodging&co or remove it entirely.

[rebthor]

@SteelNeuron - it's not that your ideas are bad at all. In fact they're pretty interesting. The issue still is (as others have mentioned before) that they are creating a completely different game that what crawl has right now in melee.

What happens to min delay? How do you swing your axe faster than you do now in your system? You do know that fighting already increases your accuracy and damage in melee and is additive with the benefit from current weapon skill?

So for example, but for weapon delay, training fighting is always better than weapon skill because the benefit you get from weapon skill to accuracy and damage is similar to the benefit you get from fighting. And fighting gives you more HP.

[Steel Neuron]

Hmm I agree on the part that special abilities such a stabbing and cleaving should tie to the corresponding weapon skills.

I don't like the idea that having fighting fully trained allows you to use every weapon in the game with 70% capacity, the approach would be the following

Train Fighting and be mildly proficient with all weapons -> use branded and artifact weapons you find around the dungeon to compensate -> get an outstandingly good artifact weapon -> stop training fighting and specialize in that weapon getting the weapon ability bonus along with whatever properties it has.

That would definitely be an option. However, you're presenting it as optimal and I assure you it is not.

If you inted to specialize in a single weapon and never switch again, Fighting will be underused. If you had trained that specific weapon type exclusively from the beginning, by the endgame you would be similarly strong (90% compared to 100% in relative terms) with roughly half the exp budget. I'm not saying your approach is bad (sacrifice of exp in exchange for early game flexibility) but it definitely isn't the only one.

Compare your approach with this other I propose: Do not train fighting. Start cross training two weapon skills; Axes and Swords for instance. Get through the game by alternating between good axes and swords. Towards the endgame, try to find a good artifact that includes the procs from both, and stick to it (like a Sword that can also cleave). This character would be limited to swords and axes, but would arguably be better than the Fighting based one you proposed at using those two weapons.

Bottom line is: There are varying degrees of specialization. The generalist, the specialist, and a middle ground.

It also makes fighting pointless if you are playing as a specialist, 27 leves arent worth the 10%.

If you are playing as a strict specialist, I would agree Fighting may not be the best option. That is completely intended. However, the versatility offered is a great asset, specially considering how the procs you are trained for may appear in weapons of different types. Also, remember Fighting boosts your HP.

Im not really a big fan of fighting as of 0.11 (gotta play trunk) since reading its description in the skill interface does not give you a good idea of what it does, and it seems most people train it for the health anyway. If it was up to me I would rename it resilience(maybe make the player slightly more resistant to poison and elements?) and put it with the dodging&co or remove it entirely.

I could agree with that. Still, I believe the game would benefit from a "generalist" skil that provides similar choices to what I described. I wouldn't mourn the death of Fighting, however.

[Steel Neuron]

@SteelNeuron - it's not that your ideas are bad at all. In fact they're pretty interesting. The issue still is (as others have mentioned before) that they are creating a completely different game that what crawl has right now in melee.

Thanks for the compliment! About the changes being too radical, I would've agreed more based on the first instance of my proposal, but now, I think they're conservative enough. Besides, Trunk is already experimenting with different scaling formulas, which I'll get into in a moment.

What happens to min delay? How do you swing your axe faster than you do now in your system? You do know that fighting already increases your accuracy and damage in melee and is additive with the benefit from current weapon skill?

Now this is something I do have a strong opinion on. The arbitrarily long road to min delay is a fundamentally flawed mechanic, in my opinion. It's obscure to inexperienced players, unintuitive, and doesn't comply with the philosophy aspect of DCSS that pursues clarity in mechanics and learning without need for spoilers. Thankfully, I'm not the only one that thinks this and fundamental changes to the attack speed scaling formula are already being tested, as you can see here:

https://crawl.develz.org/wiki/doku.php? ... g_reform&s

It discusses a linear growth of attack speed through skill levels, which is a much cleaner implementation. Granted, balance is broken, but the consensus in the wiki seems to be that a sacrifice must be made in this case.

My proposition would fit these changes without an issue. The linear attack speed increase (as well as damage and acc increase) is tackled by both Fighting and the weapon specs, weighted as I described with the percentages. Therefore, with full fighting exclusively you'd be 70% closer to max attack speed in everything, with full weapon spec and no fighting 90%, and 100% with both. Which would translate to the hyperbolic growth for attack delay described in the Wiki.

I'm even going to go a bit further (hope it doesn't sound pretentious) but I think this proposal could perhaps offer a tool to combat the inevitable imbalance that will appear when the linear attack speed formula makes it into the game (which seems to be scheduled regardless). As the wiki reads: "Addressing [the attack speed formula] has a big effect on balance, maybe we can use [the other fighting reforms] to restore it?". This proposal could offer a way to leverage the increase or decrease in power of certain melee weapons by different stages of the game, by compensating with the procs system, either by making them stronger or more abundant where it fits.

So for example, but for weapon delay, training fighting is always better than weapon skill because the benefit you get from weapon skill to accuracy and damage is similar to the benefit you get from fighting. And fighting gives you more HP.

This is a description of the status quo, but not a justification of why it should be like that. In fact, it doesn't sound intuitive to me at all. Why should someone trained in brawling be better at using a rapier than someone who has dedicated his training to that weapon? This is one of those cases in which mechanics go against the inexperienced player's intuition and would be better the other way around.

[Lasty]

I think the round 2 proposal is much much implementable; it strikes me as being largely in keeping with existing plans except for two things: 1) making fighting a more generalist weapon skill, and 2) allowing for multiple weapon procs per weapon by tying procs to skill. At this point, I quite like the proposal, though I'd dicker about the numbers somewhat, and I'm not sure I like how HP is being increased: is it tied only to fighting, or is it tied to both fighting and specialized weapon skills?

If I understand the system right, then training Fighting improves HP and damage/to-hit/speed (or "skill" for short), and training a specific weapon improves skill slightly faster (9/7ths as fast) and also one specific weapon proc. Given that, most characters will be better off training fighting until they know what weapon(s) they're likely to be using. Also, some specific weapon skills will really suffer: Cleave, for example, is not particularly good even when it procs every time, so players would rarely train Axes.

Also, reaching is an excellent ability, and works well on polearms, but it gets a lot worse if it doesn't proc consistently. If it becomes a no-training, always-proc ability for polearms only, polearms become more powerful because they have a unique proc in addition to all the new procs, but if it becomes a sometimes-proc ability, polearms will be much worse (especially when used at > 10 speed).

[some12fat2move]

Also, reaching is an excellent ability, and works well on polearms, but it gets a lot worse if it doesn't proc consistently. If it becomes a no-training, always-proc ability for polearms only, polearms become more powerful because they have a unique proc in addition to all the new procs, but if it becomes a sometimes-proc ability, polearms will be much worse (especially when used at > 10 speed).

There could just be a damage penalty for reaching that decreases with polearms skill.

PS: I think this is a discussion that should happen once all the weapon effects have been established. The proposal wants to make weapon effects scale according to weapon skill and have multiple weapon effects per weapon type, which is an interesting idea, but we don't know exactly what the ramifications would be since we don't have all the weapon effects on the table yet!

[Steel Neuron]

Also, reaching is an excellent ability, and works well on polearms, but it gets a lot worse if it doesn't proc consistently. If it becomes a no-training, always-proc ability for polearms only, polearms become more powerful because they have a unique proc in addition to all the new procs, but if it becomes a sometimes-proc ability, polearms will be much worse (especially when used at > 10 speed).

There could just be a damage penalty for reaching that decreases with polearms skill.

PS: I think this is a discussion that should happen once all the weapon effects have been established. The proposal wants to make weapon effects scale according to weapon skill and have multiple weapon effects per weapon type, which is an interesting idea, but we don't know exactly what the ramifications would be since we don't have all the weapon effects on the table yet!

Unfortunately, this is a problem of chicken and egg. In one hand, we lack the information to discuss this proposal untill all procs are known. On the other, it would be better to implement it progressively, step by step, and that would be easier with a smaller amount of procs.

I think the right moment would not necessarily be when all procs are defined. True, some weapons will be more exciting than others, but if you ask me, that's already the case with polearms.

[Steel Neuron]

I think the round 2 proposal is much much implementable; it strikes me as being largely in keeping with existing plans except for two things: 1) making fighting a more generalist weapon skill, and 2) allowing for multiple weapon procs per weapon by tying procs to skill. At this point, I quite like the proposal, though I'd dicker about the numbers somewhat, and I'm not sure I like how HP is being increased: is it tied only to fighting, or is it tied to both fighting and specialized weapon skills?

If I understand the system right, then training Fighting improves HP and damage/to-hit/speed (or "skill" for short), and training a specific weapon improves skill slightly faster (9/7ths as fast) and also one specific weapon proc. Given that, most characters will be better off training fighting until they know what weapon(s) they're likely to be using. Also, some specific weapon skills will really suffer: Cleave, for example, is not particularly good even when it procs every time, so players would rarely train Axes.

Also, reaching is an excellent ability, and works well on polearms, but it gets a lot worse if it doesn't proc consistently. If it becomes a no-training, always-proc ability for polearms only, polearms become more powerful because they have a unique proc in addition to all the new procs, but if it becomes a sometimes-proc ability, polearms will be much worse (especially when used at > 10 speed).

That's a good analysis. Regarding polearms, I didn't consider "reaching" as a proc. It's an oddball and it's likely to stay like that, specially since people at the Wiki seem to be limiting themselves to passive procs. I foresee the Polearm intrinsic proc (the one tied to the skill) as some sort of "pierce", that would reach the attacked monster and the tile behind it. Reaching could be kept as an entirely separate mechanic, like it is now, lacking a better alternative.

About HP, HP increase would be linked to fighting, not to weapon skills (like it is now). I'd personally take HP out of fighting altogether, but that's a battle for another day: I wanted to keep the proposal simple.

About your point that characters are better off training fighting first: I don't think that's necessarily the case. If they do that and decide to stick to a single weapon, once they specialize, there will be an overlap. There would have been double the EXP spent for only a 10% increase (and the HP), as compared to a true specialist. I see the "train fighting first, specialize later" as a generalist approach, but not the optimal in terms of exp budget. Your fears would be true if there wasn't an overlap; if Fighting and weapon skills were additive, it would definitely favour that route.

About Cleave, I think it's one of these mechanics that sucks when you're doing well, but comes in handy when things go south. Or at least it should be, it balanced properly. If you're doing your tactics right, isolating enemies, fighting in corridors... Then true, Cleave is useless. But in those cases you aren't in trouble anyway. Get surrounded by wasps however...

[rebthor]

What happens to min delay? How do you swing your axe faster than you do now in your system? You do know that fighting already increases your accuracy and damage in melee and is additive with the benefit from current weapon skill?

My proposition would fit these changes without an issue. The linear attack speed increase (as well as damage and acc increase) is tackled by both Fighting and the weapon specs, weighted as I described with the percentages. Therefore, with full fighting exclusively you'd be 70% closer to max attack speed in everything, with full weapon spec and no fighting 90%, and 100% with both. Which would translate to the hyperbolic growth for attack delay described in the Wiki.

Please see my earlier comment on how few characters train to anywhere near 27 skill level in a weapon - even primary melee fighters. Unless you are going to completely change the weapon speed and damage formulas, this is going to cause most characters to be much weaker in melee.

[Steel Neuron]

What happens to min delay? How do you swing your axe faster than you do now in your system? You do know that fighting already increases your accuracy and damage in melee and is additive with the benefit from current weapon skill?

My proposition would fit these changes without an issue. The linear attack speed increase (as well as damage and acc increase) is tackled by both Fighting and the weapon specs, weighted as I described with the percentages. Therefore, with full fighting exclusively you'd be 70% closer to max attack speed in everything, with full weapon spec and no fighting 90%, and 100% with both. Which would translate to the hyperbolic growth for attack delay described in the Wiki.

Please see my earlier comment on how few characters train to anywhere near 27 skill level in a weapon - even primary melee fighters. Unless you are going to completely change the weapon speed and damage formulas, this is going to cause most characters to be much weaker in melee.

Not necessarily, depends on the slope of the line. Provided the linear formula reaches the same value at the current point of saturation (where the progression flattens) it wouldn't translate to a loss in power. In fact, it would translate into an incentive to keep on leveling weapon skills as the arbitrary limit on min delay is released. In other words, by the level current weapon skills stop being practical, the new weapon skills would be equally powerful, but there would still be further progression.

Anyway, that particular change (adoption of a linear formula that grows all the way to level 27) seems scheduled to make it in regardless of my proposal, and it's reportedly working well. So yes, all attack speed formulas are going to change completely (assuming the tests go well), I'm just trying to propose a system that would fit the upcoming changes.

[byrel]

What happens to min delay? How do you swing your axe faster than you do now in your system? You do know that fighting already increases your accuracy and damage in melee and is additive with the benefit from current weapon skill?

My proposition would fit these changes without an issue. The linear attack speed increase (as well as damage and acc increase) is tackled by both Fighting and the weapon specs, weighted as I described with the percentages. Therefore, with full fighting exclusively you'd be 70% closer to max attack speed in everything, with full weapon spec and no fighting 90%, and 100% with both. Which would translate to the hyperbolic growth for attack delay described in the Wiki.

Please see my earlier comment on how few characters train to anywhere near 27 skill level in a weapon - even primary melee fighters. Unless you are going to completely change the weapon speed and damage formulas, this is going to cause most characters to be much weaker in melee.

I don't know about this. I'm always looking for the bardiches and executioner's axes, and usually find them eventually on pure melee charcters. And then lvl 26 is minimum delay.

Naturally, if I find a demon trident, etc. first, I'll never raise it high. But that's rare.

[Lasty]

Please see my earlier comment on how few characters train to anywhere near 27 skill level in a weapon - even primary melee fighters. Unless you are going to completely change the weapon speed and damage formulas, this is going to cause most characters to be much weaker in melee.

My understanding is that there's already a plan under way to eliminate the current incentives to level weapon skills to a specific threshold by making speed reduction percentage-based rather than a flat decrease per level. Steel Neuron's proposal is meant to overlap with that change, not be an introduction of that change.

About your point that characters are better off training fighting first: I don't think that's necessarily the case. If they do that and decide to stick to a single weapon, once they specialize, there will be an overlap. There would have been double the EXP spent for only a 10% increase (and the HP), as compared to a true specialist. I see the "train fighting first, specialize later" as a generalist approach, but not the optimal in terms of exp budget. Your fears would be true if there wasn't an overlap; if Fighting and weapon skills were additive, it would definitely favour that route.

I'm not sure I understand what you mean by overlap. You've stated that if you train Fighting to 27 / and (for example) Axes to 0, then you're 70% as good as it is possible to be with axes, and if you train Axes to 27 and Fighting to 0, then you're 90% as good with axes as it is possible to be, and that if you train both Axes and Fighting to 27, then you're 100% as good as it is possible to be with Axes. I understood that to mean that the your total ability to use axes would be a factor of your Fighting and Axe skills combined according to some formula that weights the two. It sounds like for low values of Fighting / Axes, there's very little benefit to training both -- is that right? If so, that seems less than ideal, since it's just a new type of newbie trap.

Could you give an example of what you see as being the correct total skill levels (expressed as a %, for consistency) for a character with 5 Axes/0 Fighting, 5 Fighting/0 Axes, 5 Fighting/5 Axes, 15 Axes/5 Fighting, 5 Axes/15 Fighting, and 15 Axes/15 Fighting?

I know this is just dickering about the numbers, but I think it's important that someone who invests in both skills to get a reasonable return from both skills.

[Steel Neuron]

I'm not sure I understand what you mean by overlap. You've stated that if you train Fighting to 27 / and (for example) Axes to 0, then you're 70% as good as it is possible to be with axes, and if you train Axes to 27 and Fighting to 0, then you're 90% as good with axes as it is possible to be, and that if you train both Axes and Fighting to 27, then you're 100% as good as it is possible to be with Axes. I understood that to mean that the your total ability to use axes would be a factor of your Fighting and Axe skills combined according to some formula that weights the two. It sounds like for low values of Fighting / Axes, there's very little benefit to training both -- is that right? If so, that seems less than ideal, since it's just a new type of newbie trap.

Could you give an example of what you see as being the correct total skill levels (expressed as a %, for consistency) for a character with 5 Axes/0 Fighting, 5 Fighting/0 Axes, 5 Fighting/5 Axes, 15 Axes/5 Fighting, 5 Axes/15 Fighting, and 15 Axes/15 Fighting?

I know this is just dickering about the numbers, but I think it's important that someone who invests in both skills to get a reasonable return from both skills.

There's no newbie trap as long as the skill descriptions are well presented. A newbie that trains both would be doing very well, for the following reasons:

1- He would be 30% stronger than someone who only trains fighting, at his or her weapon of choice, plus would also have the benefits of the procs.
2- He would be 10% stronger than the pure weapon specialist, plus would have extra HP, plus the flexibility of being able to pick other weapons and abuse the main proc he trains for in different weapons.

As you can see, the "train both at the same time" scenario has desirable qualities above sticking to either Fighting and Daggers. It would be a viable tradeoff.

A preliminary implementation (in a scale from 0 to 27) could be something like:

Effective Axe Level = (0.7*fighting + 0.9*axes*(1-fighting*2/81).


Using that formula times 100/27 to get to our abstract percentual value, your scenarios would yield the following results:

• 5 Axes, 0 Fighting => Effective axe level 4.5. Percentual effectiveness 16.666%
• 0 Axes, 5 Fighting => Effective axe level 3.5. Percentual effectiveness 12.9%
• 5 Axes, 5 Fighting => Effective axe level 5.2. Percentual effectiveness 19.34%
  
  
• 15 Axes, 0 Fighting =>Effective axe level 13.555. Percentual effectiveness 50%
• 0 Axes, 15 Fighting =>Effective axe level 10.5 Percentual effectiveness 38.9%
• 15 Axes, 5 Fighting =>Effective axe level 15.333. Percentual effectiveness 56.79%
• 5 Axes, 15 Fighting =>Effective axe level 13.333. Percentual effectiveness 49.38%
• 15 Axes, 15 Fighting =>Effective axe level 19. Percentual effectiveness 70.37%
  
  
• 27 Axes, 0 Fighting =>Effective axe level 24.3. Percentual effectiveness 90%
  
• 0 Axes, 27 Fighting =>Effective axe level 18.9. Percentual effectiveness 70%
  
• 27 Axes, 27 Fighting =>Effective axe level 27. Percentual effectiveness 100%
  

I added some more values to make it more clear.

With the concrete data, you can see that all paths are viable. Going for strict specialization will offer the highest effectiveness-to-exp ratio, at the cost of versatility. Going for Fighting will be economical and offer decent aptitude for all weapons. Going for both will make you achieve excellence with one weapon (10% more than with pure specialziation, 30% more than with pure fighting), while retaining the flexibility, at the cost of twice the exp. The overlap is positive in that it offers choices.

Also, if you still think that leveling both skills is suboptimal, considering the EXP investment, also think that apart from the extra 30% effectiveness on your main weapon, there is also the added benefit of procs being shared between weapon types.

[crate]

Right now no one uses multiple types of weapons and I think that is a good thing. I don't see a reason to make melee combat more strategically complex (especially since, as minmay said earlier, it is already the most tactically rich part of crawl) and additionally to give it inventory management problems. It seems like this proposal is based on making carrying multiple types of weapons desirable, and I think that's flat out bad.

[Steel Neuron]

Right now no one uses multiple types of weapons and I think that is a good thing. I don't see a reason to make melee combat more strategically complex (especially since, as minmay said earlier, it is already the most tactically rich part of crawl) and additionally to give it inventory management problems. It seems like this proposal is based on making carrying multiple types of weapons desirable, and I think that's flat out bad.

It's about making it an option, not about making it the option. It should still be perfectly fine to be a true specialist. However, it does propose a tool to circumvent the current restriction, by training in a different way.

In other words: I think Fighting as a skill is redundant. I'd like it gone. However, rather than taking it out, it can be reworked to give players who have a different play style a different option, and being generalistic and mixing weapon procs sounds exciting to me. But of course, I don't expect everyone to have the same taste

[crate]

Right now fighting is actually more useful than it is in your proposal (for a single weapon type), I don't know why you think that making it worse is making it better.

[Steel Neuron]

Because by being additive with the specs, regardless of how good or bad it is, it is always optimal to specialize in one weapon, as the sum of both skills will put the chosen weapon well above the rest.

The changes to fighting aren't necessarily intended to make it better ot worse. Rather, it's about making it overlap with weapon specs so the latter don't become mandatory in all cases. Also the new fighting wouldn't necessarily be worse, as it would also help increasing attack speed under the new linear formula.

Anyway, what I can't fight against is the preference of other players for the game to be kept focused on single-weapon development. I personally find that aspect bland, but if the majority does enjoy it because of inventory management issues, simplicity or otherwise, I'm not who to complain!

[Psieye]

The critical question then, to which all other details in this proposal are secondary, is: "does crawl want generalists in melee?" Where 'generalist' means "I'm kinda ok with more than 1 melee weapon type".

I'm with crate on my answer to that: do not want such generalists.

[Steel Neuron]

The critical question then, to which all other details in this proposal are secondary, is: "does crawl want generalists in melee?" Where 'generalist' means "I'm kinda ok with more than 1 melee weapon type".

I'm with crate on my answer to that: do not want such generalists.

That is indeed critical. I can understand that, thanks for pointing it out.

If it is true that most people prefer this, I will withdraw the proposal (not that I was too serious about it), as I said in the beginning it's just a thought exercise and a way to get acquainted with how people feel about game mechanics. I aspire to do some code for DCSS some day and this is a more direct way to get in touch with its innards.

In that case, the only thing I desire is that upcoming weapon procs aren't clean cut to a single weapon type. I'd like to see a bit of variety in that, such as weapons that can cleave and stab, polearms that can reach and rend, and so forth. That on itself would be enough to make melee more exciting for me. I can see that balancing for multiple weapon use could be difficult and not necessarily add anything in value. Making weapons exciting would make me enjoy them more, so I hope the development effort keeps going in that direction.

[byrel]

The critical question then, to which all other details in this proposal are secondary, is: "does crawl want generalists in melee?" Where 'generalist' means "I'm kinda ok with more than 1 melee weapon type".

I'm with crate on my answer to that: do not want such generalists.

We do already have features that encourage some generalism; hydras->blunt, bailey->reaching, etc. I think specialism will always be the more powerful choice against generic monsters in any reasonable skill arrangement. But the game already rewards generalism slightly, so 'making generalism easier' is hard to rule out pro forma.

For that matter, the simple fact that pain brand is additive, while holy wrath is multiplicative, tends to encourage toting around a quickblade of pain and an executioner's axe of holy wrath. Not to mention stabbing... The more you tactically differentiate weapon types, the more you reward generalism. The skill system clearly needs to penalize generalism, to make specializing worthwhile. Exactly how much generalism needs penalized is a worthy question.

[Lasty]

Right now fighting is actually more useful than it is in your proposal (for a single weapon type), I don't know why you think that making it worse is making it better.

I don't think that's true. As I understand his proposal, Fighting still gives HP, it just also gives a larger bonus to hit/dam/speed than it does right now. Given the exact numbers he gave, if you have 15 in one skill and 5 in another, Fighting at 15 / Axes at 5 gives only slightly worse hit/damage/speed then Axes at 15 / Fighting at 5, but with an extra 10 levels of HP from Fighting (and 10 less levels of Cleaving proc).

I maintain that with the numbers given, Fighting would be the correct choice of skill almost every time.

[Steel Neuron]

Right now fighting is actually more useful than it is in your proposal (for a single weapon type), I don't know why you think that making it worse is making it better.

I don't think that's true. As I understand his proposal, Fighting still gives HP, it just also gives a larger bonus to hit/dam/speed than it does right now. Given the exact numbers he gave, if you have 15 in one skill and 5 in another, Fighting at 15 / Axes at 5 gives only slightly worse hit/damage/speed then Axes at 15 / Fighting at 5, but with an extra 10 levels of HP from Fighting (and 10 less levels of Cleaving proc).

I maintain that with the numbers given, Fighting would be the correct choice of skill almost every time.

Your analysis is correct, except perhaps the "slightly worse" appreciation. I think a fairer comparison is the 15 axes 0 fighting versus 15 fighting 0 axes case, and it's 50 against 38.9, a relative increase of 128%.

Still, you might be right about fighting being better. It's just a number tweaking issue though. 0.6 instead of 0.7 of Fighting's relative weight, or perhaps less HP bonus.

[crate]

Well I was doing the comparison between e.g. axes 20/fighting 0 and axes 20/fighting 20. I didnt do any math but you will really feel the difference if you compare the two in crawl as-is (in fact I believe that fighting skill does just as much for your damage/hit as weapon skill currently). Going from axes 20 (proposed axes 20/fighting 0) to to axes 23 or whatever (proposed axes 20/fighting 20) damage-wise is not very noticeable though.

I didn't bother to think about whether training fighting or weapon skill in the proposed system here is better ... imo having both skills in crawl is fine and it works well as-is. Unless you also want to eliminate spellcasting skill....

[galehar]

(in fact I believe that fighting skill does just as much for your damage/hit as weapon skill currently).

Almost. Damage is mutliplied by (1 + weapon_skill / 25) and also by (1 + fighting_skill / 30). For accuracy, they contribute exactly the same way. They both add random2(skill) to your to_hit roll.

[Psieye]

I can get behind "weapon skill dictates proc chance of certain passive melee effects, which are near independent of which base type your weapon is" on the condition the proc chance benefit plateaus out by around Rank 10~12.

[Pereza0]

(in fact I believe that fighting skill does just as much for your damage/hit as weapon skill currently).

Almost. Damage is mutliplied by (1 + weapon_skill / 25) and also by (1 + fighting_skill / 30). For accuracy, they contribute exactly the same way. They both add random2(skill) to your to_hit roll.

Oh wow, I would have never imagined that

[Lasty]

Almost. Damage is mutliplied by (1 + weapon_skill / 25) and also by (1 + fighting_skill / 30). For accuracy, they contribute exactly the same way. They both add random2(skill) to your to_hit roll.

Didn't know that -- thanks, galehar!

Given that, Steel Neuron's proposal is even more limited than I realized:

1: Make Fighting factor into weapon speed at a rate inferior to the specific weapon skill.
2: Tie procs to specific weapon skills and make each weapon eligible for 2+ procs.

[Steel Neuron]

Almost. Damage is mutliplied by (1 + weapon_skill / 25) and also by (1 + fighting_skill / 30). For accuracy, they contribute exactly the same way. They both add random2(skill) to your to_hit roll.

Didn't know that -- thanks, galehar!

Given that, Steel Neuron's proposal is even more limited than I realized:

1: Make Fighting factor into weapon speed at a rate inferior to the specific weapon skill.
2: Tie procs to specific weapon skills and make each weapon eligible for 2+ procs.

Plus

3: Give slight diminishing returns to Fighting and specific weapon skills, so that, even in the event of full specialization + fighting, other weapons are usable.

However if, as a couple people have confirmed, they don't want characters to become more generalistic, then I'll be more than happy to stick to point 2.