Race proposal: Homunculus

[Croases]

Small alchemical constructs that learn fast but cannot truly master any skill.

  Code:

Homunculi are constructs resembling small humanoids with disproportionally large (and deceptively strong) limbs, created through alchemical rituals and typically employed to guard the study of their master. They have a natural knack for slipping through the traps, curses and magical barriers that characterize these places, and their ability to assist in alchemical research is bolstered by their prodigious capacity to absorb new knowledge: A newly created homunculus learns to speak within days, and can grasp the basics of magical theory and fighting tactics soon afterwards. However, their artificial nature limits the amount of information they can process, preventing them from ever reaching the level of mastery shown by the warlords and archmages of other species.

Although they are typically created in service to a master, a smattering of free homunculi are known to exist, and their largest community even supports its population by building new "offspring".

Attributes:

  Code:

Starts with 10 STR/12 DEX/12 INT, +1 DEX/INT every 3 levels, +4 MR/level
+20% HP, +10% MP
Small size; can wield weapons and shields as a medium-sized character (effectively a fancy way of saying "extra EV")
Construct resistances (rTorment, rN+, rot immunity, unbreathing)
Fast learning 3: All experience towards your skills is increased by 50% (or some bigger number, enjoy your +8 Fighting aptitude...)
Lorekeeper 3: Your skill levels cannot exceed 12 (...while you can)
Unburdened: You are immune to curses, nets, webs, paralysis, petrification, roots and other effects that limit or prevent movement. You are also unaffected by effects caused by equipping and unequipping items (no amulet penalties, can ignore *Drain/*Contaminate/*Fragile)
Quickdraw: You can rapidly change your weapons and armor. Swapping shields or auxillary armor takes 0.5 auts. Swapping body armor takes 1 auts.

Homunculi cannot worship Ashenzari.

Aptitudes:

  Code:

Fighting +5
Short Blades +3
Long Blades +3
Axes +2
Maces & Flails +5
Polearms +4
Staves +4
Unarmed Combat +5
 
Throwing +4
Slings +3
Bows +2
Crossbows +2
 
Armour +4
Dodging +5
Stealth +5
Shields +3
 
Fire Magic +4
Ice Magic +4
Air Magic +4
Earth Magic +4
Poison Magic +4

Spellcasting +5
Conjurations +5
Hexes +3
Charms +5
Summoning +5
Necromancy +3
Translocations +4
Transmutations +5
 
Invocations +3
Evocations +5
 
Experience 0

Alternatively, all +5. This has the advantage of not tricking new players into thinking that homunculus aptitudes matter.

This is my take on a construct race. It is mainly intended to be what trolls are thought to be, but aren't: A race with a great early game that struggles later on. As such, Lorekeeper 3 is the core part of this proposal.

In addition to a strong early game (thanks to demigod-tier attributes, large HP pool and rapid early skill advancement), they also get considerable freedom to swap out items in and out of combat, which in conjunction with their broad skillset (you'll get 12 on everything by mid-game) allows them to switch roles on the fly. I don't think equipment swapping is a particularly undesirable mechanic, since it's neither unintuitive nor tedious, and the ability to switch from heavy to light armor in particular may be an interesting option. There are races that encourage much more undesirable behaviors (kiting and Ce/Fe/Sp, killholing and Fo) anyway, so I guess racial and divine mechanics have some leeway in this department (since they don't necessarily exist in every game).

I also like this particular tradeoff ("easy early game and flexible equipment setup in return for low skill levels for late game") because it's the exact opposite of Ashenzari's conduct ("massive bonus to skill levels in return for a sub-par early game and hard-to-change equipment"), and while symmetry is not a design element, it seems fun to have symmetry between the mechanics of a god and a species that it is inimical to. Lastly, people who hate paralysis and amulet penalties will probably enjoy playing this, and probably also die a lot since it doesn't seem like a particularly strong race. As making players die is the majorest of major design goals, I think this is a fairly solid proposal.

[duvessa]

  Code:

Fast learning 3: All experience towards your skills is increased by 50% (or some bigger number, enjoy your +8 Fighting aptitude...)

How is adding a new mutation better than just increasing all the aptitudes by ~2.34, which does the same thing?

  Code:

Homunculi cannot worship Ashenzari.

Why? Trading your god slot and quick swapping for higher skill levels seems fair. Do you want to ban all the other gods that give you the benefits of higher skills, too? Like Okawaru, Vehumet, etc...

This species doesn't sound weak or difficult to me, it gets amazing damage/defenses in exchange for giving up the ability to have super-high stealth, cast level 8 or 9 spells (almost all of which suck anyway) and use some high-end weapons to their full potential (notice that demon whips are basically unaffected). Sure, maybe you have worse ac/ev than a human once you've gotten your fifth rune or so, but you're unlikely to notice at that point.

I think you are overvaluing high skill levels, honestly. It's fun to train skills to high levels, which is why people do it, but if any player looks through the times they've trained a skill past 20, chances are that it was a suboptimal investment almost every time that they weren't with Trog. So being capped at 12 doesn't seem like an especially harsh handicap to me. I suspect this race would "struggle later on" less than quite a few existing races; Fo, DE/Gr/Te, and Mu come to mind.

[Croases]

How is adding a new mutation better than just increasing all the aptitudes by ~2.34, which does the same thing?

Can you increase aptitudes beyond +5? That was the idea, giving you Sprint-like skill advancement that cuts off completely once you transition into the midgame.

But yes, they may well stay +5 without that mutation, you'll hit the cap very quickly with or without a 50% bonus anyway.

Why? Trading your god slot and quick swapping for higher skill levels seems fair. Do you want to ban all the other gods that give you the benefits of higher skills, too? Like Okawaru, Vehumet, etc...

Ashenzari is forbidden because you're unaffected by his conduct, you can swap out of cursed items. He wouldn't give much of a benefit to your skills anyway, since your ceiling is 12 and he wouldn't increase it further, like how he doesn't raise skills beyond 27.

So Okawaru and Vehumet are fine, though Heroism wouldn't help you (or do you think it's better for Heroism/Fencer's Gloves/Ashenzari skill boosts to bypass the 12 cap?) If the "unaffected by curses" part of the unburdened quirk seems excessive, then worshipping Ashenzari would be fine too.

This species doesn't sound weak or difficult to me, it gets amazing damage/defenses in exchange for giving up the ability to have super-high stealth, cast level 8 or 9 spells (almost all of which suck anyway) and use some high-end weapons to their full potential (notice that demon whips are basically unaffected). Sure, maybe you have worse ac/ev than a human once you've gotten your fifth rune or so, but you're unlikely to notice at that point.

I think you are overvaluing high skill levels, honestly. It's fun to train skills to high levels, which is why people do it, but if any player looks through the times they've trained a skill past 20, chances are that it was a suboptimal investment almost every time that they weren't with Trog. So being capped at 12 doesn't seem like an especially harsh handicap to me. I suspect this race would "struggle later on" less than quite a few existing races; Fo, DE/Gr/Te, and Mu come to mind.

If that is the case, this is a big concern for the proposal - I'd have expected 12 to be very severe, given the scarcity of tournament wins with the lorekeeper 3 conduct (hence why I thought forcing it in a species could be interesting), but I guess lorekeeper 3 is hard partly because you can overskill. Do you think further lowering the skill ceiling would increase the difficulty to an acceptable level?

[archaeo]

  Code:

Fast learning 3: All experience towards your skills is increased by 50% (or some bigger number, enjoy your +8 Fighting aptitude...)
Lorekeeper 3: Your skill levels cannot exceed 12 (...while you can)

Everything else aside, it seems like it'd be a lot simpler to just divorce skills from XP entirely with this idea; you might as well hand out a skill level every even level instead. Maybe that'd be more broken than this already seems like it would be, but still.

[Croases]

Everything else aside, it seems like it'd be a lot simpler to just divorce skills from XP entirely with this idea; you might as well hand out a skill level every even level instead. Maybe that'd be more broken than this already seems like it would be, but still.

Good idea, sounds reasonable to start at 4/5 across the board and get +1 to all skill levels for every even level, capping out at XL14/16 (mid/end-Lair). I was concerned that the super-rapid skilling would create a lot of unwanted "you've mastered X, now enable something else" messages, and this would handily solve that issue while still giving you a lot of early game punch, especially since you'll be semi-competent at using every weapon you find.

Still not sure where a low-skill race would fall on the power scale, though. I should try to impose that limit on a human and see how it plays.

[Tiktacy]

Very, VERY interesting race idea. Flavorful, unique, and most likely quite enjoyable. Ashenzari should be permitted, being able to get around your innate mutations through god abilities is fun. I don't think fedhas would be too keen on following a machine though, but that's just bikeshedding.

I greatly disagree with archeo on on his idea for skill levels upon XP level.

I think this race might be a little overpowered though, I don't think fast learning is needed to make this race good, I would scrap that. As duvessa pointed out, many very strong weapons have a min delay reached around level 12, I often find myself stopping at around level 12 with most skills as well.

Overall, its a great idea, but I think as it exists its problematicly overpowered. Removing fast learning, and maybe putting the cap at 10 should fix that. You might want to also add a mutation that makes all skills available for training at all times, so that way the player doesn't have to carry extra junk with them to train stuff like books or a spear or something once they cap out in their main skills.

[archaeo]

I greatly disagree with archeo on on his idea for skill levels upon XP level.

I figure that if you're going to change the skill system so dramatically for a species, you might as well make it fairly simple and transparent instead of requiring some oddball mutations that just mess with the formula. If the problem is that the skills come too slow over 27 XL, you could speed that up.

Like lots of species proposals though, I think "skill cap god" makes more sense than "skill cap species." One could build a god around skills getting maxed out at 12 more easily than a species, imo.

[all before]

would you worship a skill cap god if its altar appeared on d:7?

[dynast]

I have a hard time not reading this idea as a catchy advertisement:
"Are yo tired of getting netted by hill giants?"
"Hate those pesky orc sorcerers paralyzing you?"
"Optimal play is preventing you from trying every gear in the dungeon?"
"Your hp,mp and apts suck?"

[HardboiledGargoyle]

Like lots of species proposals though, I think "skill cap god" makes more sense than "skill cap species." One could build a god around skills getting maxed out at 12 more easily than a species, imo.

not iMo, however. Consider converting to god late, or from the god at any point. And what would the god's benefits be? unclear, while on a species, it's easy to see how well it could interact with near any god in nice combinations, but if it were a god, comparative advantages to different species would vary too much based on their natural apts. Something like DD on the other hand would be better as a god since there's no reason no-heal-from-resting should be tied to good axe and earth aptitudes or other typically racial modifiers.

[duvessa]

Ashenzari is forbidden because you're unaffected by his conduct, you can swap out of cursed items. He wouldn't give much of a benefit to your skills anyway, since your ceiling is 12 and he wouldn't increase it further, like how he doesn't raise skills beyond 27.

Oh, I thought you meant they were immune to mummy curses, not that they could swap cursed items. (Why can they swap cursed items? If they can't worship ash then the ability to do so is completely irrelevant for everything except scythe of curses.)

I'd have expected 12 to be very severe, given the scarcity of tournament wins with the lorekeeper 3 conduct

Never use player stats as an indication of difficulty or good play. Most players don't even know what skills do, let alone know their costs, let alone bother to formulate anything close to an optimal strategy, let alone actually want to use that strategy. And very few players want to intentionally handicap their character in order to get a meaningless badge.

Do you think further lowering the skill ceiling would increase the difficulty to an acceptable level?

Well, I think your core premise of "easy early, hard later" is flawed. It sounds like your line of thought is "the post-early game is boring, so I'll suggest a species with a very artificial way of making the post-early game not-boring, but I don't want it to actually be weaker than other races, so I'll make the early game easy (and boring) instead." Wouldn't it be better to fix the post-early game in general, instead of leaving it broken on every other race?
But if you literally just want a race that is bad after the early game, yes, obviously lowering the skill cap would accomplish that. I don't know why that would even be in question, I think it's pretty obvious that a race with skills capped at 0 would be bad eventually.

[WingedEspeon]

(Why can they swa

That is 100% not true and you know it.

[duvessa]

I was going to edit "completely" to "essentially" because I knew someone would pick on the fact that yes, technically it means you aren't penalized for wear-IDing equipment on D:1 that you almost certainly have no reason to wear in the first place, but my connection crapped out and I didn't care enough to come back and do it

Also, are quote tags really that hard to use?

[WingedEspeon]

Preventing wear IDing on D:1 is the entire reason for curses to exist. Having a ring possibly be a +6 ring of slaying is a good enough reason to want to put it on.

[Tiktacy]

Ashenzari is forbidden because you're unaffected by his conduct, you can swap out of cursed items. He wouldn't give much of a benefit to your skills anyway, since your ceiling is 12 and he wouldn't increase it further, like how he doesn't raise skills beyond 27.

Oh, I thought you meant they were immune to mummy curses, not that they could swap cursed items. (Why can they swap cursed items? If they can't worship ash then the ability to do so is completely irrelevant for everything except scythe of curses.)

I'd have expected 12 to be very severe, given the scarcity of tournament wins with the lorekeeper 3 conduct

Never use player stats as an indication of difficulty or good play. Most players don't even know what skills do, let alone know their costs, let alone bother to formulate anything close to an optimal strategy, let alone actually want to use that strategy. And very few players want to intentionally handicap their character in order to get a meaningless badge.

Do you think further lowering the skill ceiling would increase the difficulty to an acceptable level?

Well, I think your core premise of "easy early, hard later" is flawed. It sounds like your line of thought is "the post-early game is boring, so I'll suggest a species with a very artificial way of making the post-early game not-boring, but I don't want it to actually be weaker than other races, so I'll make the early game easy (and boring) instead." Wouldn't it be better to fix the post-early game in general, instead of leaving it broken on every other race?
But if you literally just want a race that is bad after the early game, yes, obviously lowering the skill cap would accomplish that. I don't know why that would even be in question, I think it's pretty obvious that a race with skills capped at 0 would be bad eventually.

Very good feedback overall, but quite harsh. Nice to see duvessa being duvessa as always.

The interesting part about this species isn't that its good in the early game and bad in the late game. In my mind, what is interesting about this species is that he can potentially have very high levels in all skills and basically have access to a toolbox of spells and weapons, sort of a jack of all trades. I do think the max cap needs to be lower to make it less overpowered but still have the toolbox aspect to it.

[Lasty]

I think the low skill cap is an interesting proposal, but duvessa's first post did a good job of talking about why this species will be pretty easy as designed.

A few other points:
* I don't like the "immune to a bunch of status effects" aspect of the proposal, since being immune to a threat is the least interesting way to change a player's interaction with that threat.
* Quick-swapping all equipment feels silly to me. You still don't want to swap all your equipment in a fight (unless re-equipping is instant), so this just changes how far you want to run before re-equipping, and may change how much of your kit you want to re-equip at a time. In general, the equipment a player wears is the best equipment that player has available, but I suppose this encourages players to quick-swap a bunch of gear whenever they want to cast spells that are tricky to cast with all their skills as 12 -- deflect missiles, say, or monstrous menagerie. It potentially increases the amount of extra crap you want to haul around with you.

[WingedEspeon]

Personaly I think that this spieces would be rather intersting to play even if it is easy. 12 as a max skill level is pretty harsh on melee combat, so you might be a bit shoehoned into being primarily a caster though.

[Croases]

Lasty's point that immunities are a bad way to handle a threat is well-made, so before I figure out how to introduce a new species and do a hackneyed attempt at coding this, here are some alternative benefits to replace their immunity to movement-restricting status:

- Bad status effects are delayed for 2 turns, showing up as faded text while inactive (duration can be lowered to 1 turn if necessary). Everything works like petrification, which should create some interesting tactical situations (and not just for paralysis - you also get to blink out before Dimensional Anchor kicks in, get an extra hit in before sleep and !cancel off marks before everything on the floor knows where you are).

- Whenever you quaff a potion, you have a 33% chance not to use it up. This encourages them to use consumables liberally and possibly enhances their role as a species that gets by through item flexibility. Flavored as their bodies better handling alchemical contoctions due to their nature, so sometimes they can consume part of a potion and still get its effects.

- Deviceheal++, all potion effects last 50% longer (including ambrosia, lignification and such). I don't like this one as much as the two above, since it's a fairly generic benefit, but maybe the turn delay should work on good status too, so you get a lot of healing/might/whatever but lose out on emergency options.

Which of these do you think would be the most appropriate? Alternatively, the paralysis-and-other-things immunity could just be replaced by nothing, especially since the popular opinion is that they would be strong enough without extra perks.

I will, however, defend quickswapping - swapping weapons/rings takes 0.5/1 auts and is very commonly practiced, so swapping auxillary armor at 0.5 auts and body armor at 1 auts will probably have combat utility as well. I also don't think it will require the player to haul around more stuff than rings do, perhaps two or three armor pieces at best, but I guess that will require testing.

[Floodkiller]

Lasty's point that immunities are a bad way to handle a threat is well-made, so before I figure out how to introduce a new species and do a hackneyed attempt at coding this, here are some alternative benefits to replace their immunity to movement-restricting status:

- Bad status effects are delayed for 2 turns, showing up as faded text while inactive (duration can be lowered to 1 turn if necessary). Everything works like petrification, which should create some interesting tactical situations (and not just for paralysis - you also get to blink out before Dimensional Anchor kicks in, get an extra hit in before sleep and !cancel off marks before everything on the floor knows where you are).

- Whenever you quaff a potion, you have a 33% chance not to use it up. This encourages them to use consumables liberally and possibly enhances their role as a species that gets by through item flexibility. Flavored as their bodies better handling alchemical contoctions due to their nature, so sometimes they can consume part of a potion and still get its effects.

- Deviceheal++, all potion effects last 50% longer (including ambrosia, lignification and such). I don't like this one as much as the two above, since it's a fairly generic benefit, but maybe the turn delay should work on good status too, so you get a lot of healing/might/whatever but lose out on emergency options.

Which of these do you think would be the most appropriate? Alternatively, the paralysis-and-other-things immunity could just be replaced by nothing, especially since the popular opinion is that they would be strong enough without extra perks.

I will, however, defend quickswapping - swapping weapons/rings takes 0.5/1 auts and is very commonly practiced, so swapping auxillary armor at 0.5 auts and body armor at 1 auts will probably have combat utility as well. I also don't think it will require the player to haul around more stuff than rings do, perhaps two or three armor pieces at best, but I guess that will require testing.

I personally think the species would still be interesting and very strong if you took out pretty much all the other gimmicks. It's already basically a Demigod who can worship based on just its innate starting stats: 10 stats above the stat budget average with average stat growth, above average MR gain, great HP and MP (even with a limit on Fighting and Spellcasting), and no penalty to experience for leveling. If you just stuck with the super good aptitudes and the skill cap as the gimmick, it would be simpler to code up (only need to put in the skill cap limit and whether divine abilities/fencer's gloves can boost them), and much easier to tell how balanced it is/what adjustments would need to be made (such as lowering the skill ceiling, reducing starting stats, etc.)

[Floodkiller]

Bumping this because I've been interested in this for a while and too lazy to make it, but I pushed myself to get something done during the holidays:

http://www.github.com/Floodkiller/crawl/tree/cynocephalus

Cynii are humanoids with the head of a dog, with origins from the eastern deserts. Cynii are unusually attracted to the dungeon, with mentions of the Orb generating blank stares, tilted heads, and involuntary drooling.

Attributes:

  Code:

Starts with 8 STR/8 DEX/8 INT, +1 STR/DEX/INT every 5 levels, +3 MR/level
+0% HP, +0 MP, +0 XP
Normal size, no bodyslot restrictions
Limited potential (each skill can be raised to a maximum of 12, instead of 27).

All aptitudes are +5 across the board.
___________
I made some changes implementing the species proposed in the OP, as seen. I knocked the attributes down to human, and gave them undead stat growth. This was to make the attributes reflect the limited skill gimmick. I also currently have them at no other gimmicks; this may change to SInv and hats only, but I'll wait until I get more solid playtesting to find out (and if they get bikeshed into something else).

Current known bugs:
Oka's Heroism and weapon cross training (I'm not sure how???) are currently allowing them to boost their skills past 12, which I intend to prevent. Ashenzari does not boost the skills past 12, which is working as intended.

Yet to implement features:
Matching player species monster (currently using MONS_JACKAL).
Some other stuff I'm sure I forgot

If you have the time, please give it a playtest or two and provide feedback! I'll be able to work more on it and give it more solid playtesting myself once I'm done traveling home for the holidays.

[Croases]

Cynii are humanoids with the head of a dog, with origins from the eastern deserts. Cynii are unusually attracted to the dungeon, with mentions of the Orb generating blank stares, tilted heads, and involuntary drooling.

I hope that part doesn't get bikeshed away because it's the best flavor.

I'm going to assume that they all have either corgi or shiba heads.

[Floodkiller]

Oka and weapon cross training now no longer increase skills past the max cap. Aptitudes were just changed to +3 (+5 had you maxing weapon skill by D:4, which was waaaay too fast in my opinion). Attributes were changed to 7 Str/7 Int/10 Dex to differentiate from humans and have a slight bonus for Dodging in the early game. Finally, the skill cap was raised to 14 because 12 was very awkward for a lot of skills (like weapons) and also because 14 is closer to being half of 27; I may change it back if 14 proves to be too easy.

Also, the plural is now Cynos instead of Cynii.

[dracos369]

Can someone give me a quick tutorial on how to compile this? I'd love to give this species a run, but I find myself a bit overwhelmed by so many files. Assume basic programing knowledge.

[Floodkiller]

Can someone give me a quick tutorial on how to compile this? I'd love to give this species a run, but I find myself a bit overwhelmed by so many files. Assume basic programing knowledge.

You should be able to download the directory manually from the git page, then all you have to do is follow these instructions:
http://s-z.org/neil/git/?p=crawl.git;a= ... a6;hb=HEAD
I'm on Windows, and preferred the MSYS2 instructions myself. Where it says you need to clone the directory from git, you can just skip that step with the manual download and cd (change directory) to the source folder. Alternatively, if you would prefer to clone it from git, run these commands in MSYS2 (or whatever you are using to connect to git):
1) git clone https://github.com/Floodkiller/crawl.git (grabs a copy of the full git directory)
2) cd crawl (go to the newly created crawl folder)
3) git submodule update --init (builds contributing libraries, just in case you are missing them)
4) git checkout cynocephalus (switches branch to dog species, try 'origin cynocephalus' if this doesn't work)
5) git pull (grabs any recent changes)
6) cd crawl-ref/source (change directory to the source folder)
7) make (or 'make TILES=Y' if you want tiles instead of console)

Then either run crawl.exe from the command prompt in crawl/crawl-ref/source or navigate to the folder where you installed MSYS2/git tools, search for the crawl folder, and navigate down to the crawl.exe in the source folder yourself.

[Implojin]

Can someone give me a quick tutorial on how to compile this? I'd love to give this species a run, but I find myself a bit overwhelmed by so many files. Assume basic programing knowledge.

  Code:

git clone https://github.com/crawl/crawl.git crawl
cd crawl
git submodule update --init
git remote add Floodkiller https://github.com/Floodkiller/crawl.git
git fetch Floodkiller
git checkout cynocephalus
cd crawl-ref
cd source
make clean
make TILES=y PCH=y -j8
./crawl


Alternately, use this for reference but point it at Floodkiller's repo instead. https://github.com/crawl/crawl/blob/mas ... NSTALL.txt

[Rast]

Limited potential (each skill can be raised to a maximum of 12, instead of 27).
All aptitudes are +5 across the board.

On the face of it, this is busted. I don't buy into the tavern meme of D1:-9 being the only hard part of crawl, but a massive boost early is not made up for by a slightly harder midgame and harder endgame.

But I'll try it out if it makes it to a public server.

[Implojin]

Oka and weapon cross training now no longer increase skills past the max cap. Aptitudes were just changed to +3 (+5 had you maxing weapon skill by D:4, which was waaaay too fast in my opinion). Attributes were changed to 7 Str/7 Int/10 Dex to differentiate from humans and have a slight bonus for Dodging in the early game. Finally, the skill cap was raised to 14 because 12 was very awkward for a lot of skills (like weapons) and also because 14 is closer to being half of 27; I may change it back if 14 proves to be too easy.

Also, the plural is now Cynos instead of Cynii.

Thought I'd try this out, so I finished a 3 rune CyFi^Hep. Morgue attached to this post.

The concept sounds interesting, but the current implementation doesn't play very well.


Gameplay thoughts:

Spoiler: show

Earlygame is a joke. +3 aptitudes across the board mean you're maxing skills accidentally before Lair. Okay, this difficulty is intended, but it also means that you end up collecting escape resources throughout earlygame that you have no reason to use. This becomes relevant later.

Midgame is still a joke. There's enough XP in Lair with +3 aptitudes to reach 14 weaponskill, Fighting, Armour, Shields, Dodging, and start working on a couple of support schools. High skills across the board like this make lairbranches trivial to wade through. *Still collecting resources.* Swamp was nothing, I had to read one or two scrolls in Snake.

Late midgame started to get interesting, I hit some speed bumps in Vaults because Cynowhatever seem to have very low MR growth. Had to use some escape scrolls to clear V1-2, somewhere around V:2 I swapped some gear around and didn't have further trouble.

Depths was kind of a joke again, by this time I had picked up a sacred scourge from mennas with a few ?ew scrolls and had things like invis and dmsl online. The character seemed a little weak to pickup rune #3 so I started to do Crypt:1, then I said fuck it and just turned around to do V:5 instead. V:5 might have been problematic if no MR gear had spawned in the game, but by that time I had some items with MR and everything was okay.

Surprisingly, Zot wasn't really different than usual. The low skillcap here didn't have nearly as much effect as I'd have expected before playing this branch. I had to play a bit more cautiously than a typical XL:26 character, but nothing particularly out of the ordinary.

I really like the idea of making a race that reverses the usual difficulty curve, but the implementation needs a lot of work if it's going to approach that concept.

Part of the problem with this race's design is that cumulative item generation pushes back against any attempt it makes at ramping up lategame difficulty.

If Cyno had an innate reduction to dungeon item generation rates, it might be able to start approaching the concept of "difficult lategame". As it is, it's more like "trivial until midgame, then you start reading a bit to get through late-midgame but it doesn't matter because you're sitting on all the resources there were no reason to spend earlier, then eventually you get enough floordrops that everything is back to normal by lategame".

(I feel dirty saying this, but Cyno might play better with some form of consumable destruction turned back on. If not that, then they need something like a passive reduction to consumable spawnrates, else you just end up collecting enough escapes from early+midgame to nullify later threats.)

I also think that low skillcaps will limit character differentiation pretty severely. With the existing aptitudes and caps, you run out of useful skills around XL:18, and you cap all of them before your 3 rune game is finished. The ripple effects from lowered skillcaps affect spell success rates, spellpower, weapon delay, relative utility of weapon basedelay, opportunity cost of SH, relative utility of SH, maximum available EV, relative utility of Dodging, relative utility of heavy armour, relative utility of summons, and relative utility of spells minimally affected by spellpower, among other things. I don't want to stress this point too much before seeing how other people play this race, but if this design were to be used as-is I expect that most Cyno characters would look very similar as a result of knockon effects from above, (1h + shield, level 6-7 spells mostly buff/summon, midheavy to heavy armour), which feels like a pretty glaring problem.


The idea is interesting, though, and I'd like to play a later revision of this race, if the implementation kinks can be worked out.

[Floodkiller]

Thought I'd try this out, so I finished a 3 rune CyFi^Hep. Morgue attached to this post.

Thanks for the feedback! This looks like it's reflecting my playthroughs as well, which is a shame. I made the following changes on my local copy for testing (I'll hold off on committing to git until I test them to ensure I didn't make the species too hard/impossible):

-Aptitudes down to +2: This was suggested in #crawl-dev, and I think it's probably going to be committed regardless of if I keep the other changes. This should make it so that you max out less skills before the end, as well as provide a bit more opportunity costs for skills early on (unless you are a CyBe, I guess). A +2 flat aptitude is still incredibly flexible as well, so I'm not too concerned about doing this.

-Skills down to 5/5/5: This is pretty drastic (lower total stats than ghouls!), but this should help improve the variability of end-game builds by forcing players to use their stat choices during level ups as a way to determine what type of end-game equipment they want to support (higher level spells, heavier armor, or better dodging) instead of getting all three at a decent level. Not much I can do about the 1-hand+shield loadout though, because it naturally tends towards that due to the skill limits, and the only way I could fix that would be to remove the ability to use shields.

-XP Modifier to -1: This should slow down the rate at which you get stat ups, more HP, etc. without affecting the aptitudes. This should force slightly more consumable use through having less leeway early on.

Edit: Pushed these changes. I might push XP down to -2 later, but I'll give it some more playtests to see if I'm just overreacting.

[Steel Neuron]

How about limiting your max skills to your XL, and having an extremely slow leveling?

• Aptitudes: +3 across the board. -4 XP aptitude.
• Lorekeeper: Your skills may not exceed your XL. (tweak this relationship to tune the late game difficulty, it could be that they can't exceed XL or 15)

This should force you to be a jack of all trades from the get go, because with the severely reduced leveling any particular skill would be behind the curve.

[milski]

How about limiting your max skills to your XL, and having an extremely slow leveling?

• Aptitudes: +3 across the board. -4 XP aptitude.
• Lorekeeper: Your skills may not exceed your XL. (tweak this relationship to tune the late game difficulty, it could be that they can't exceed XL or 15)

This should force you to be a jack of all trades from the get go, because with the severely reduced leveling any particular skill would be behind the curve.

The issue with this suggestion is that you are giving them zero advantages. With skills tied to level and an incredibly bad XP aptitude, you are going to be behind a typical character from the start (since you'd start with all class apts capped) and stay that way all game, except *maybe* having better skills than a midgame fighter that mindlessly trains fighting, weapon, dodging, shield, and armour. So I guess you'd go light armour and pick up a spellbook as well, but that limits you to always going for a book start to make sure you have spells when you're stuck with awful skillcaps.

[Floodkiller]

How about limiting your max skills to your XL, and having an extremely slow leveling?

• Aptitudes: +3 across the board. -4 XP aptitude.
• Lorekeeper: Your skills may not exceed your XL. (tweak this relationship to tune the late game difficulty, it could be that they can't exceed XL or 15)

This should force you to be a jack of all trades from the get go, because with the severely reduced leveling any particular skill would be behind the curve.

I may experiment with having the XP aptitude go that low (and maybe tweak the other parts back to make up for it), but I definitely don't want to put a skill cap based on XL level. Raising the cap slowly as you level leaves you with an incredibly tough early game unless you have a very flexible start to spread your stats around, like Wizard. Lowering the cap as you level makes the player feel frustrated when they reach a point where they can't kill monsters if they don't want to grow weaker. I do think the species is best off (even if it ends up being too flawed to be implemented in the end) to have a static skill cap.

[Implojin]

A passing thought WRT skillcaps and character differentiation:

Maybe this race could be written to have dynamic max skill caps, with a sort of pyramidal structure to its available max skill levels. Players could raise one skill to, say, 17, with one more skill at 16, one at 15, one at 14, etc., all the way down the line, with the lowest skills being capped to something like 5 or 7; just enough for slight hybridization but not enough to make them a playstyle focus.

In this way, Cyno characters would become differentiated by the order in which their players choose to raise their skills, and this would allow for things like higher-damage offensive spellcaster Cynos, or higher-EV Dodging focused Cynos, or larger weapon 2h focused Cynos, while still limiting the overall available skill levels.

It's true that this approach would make any individual Cyno less able to swap freely between weapons and spell schools, but by carefully adjusting the rate at which the max available school level decreases, one can assume that a fair balance would be able to be found between enforcing differentiation and enabling effective-enough low-skill hybridization.

edit: There are probably some usability concerns here wrt accidentally raising skills by large amounts through unique kills or potions of experience, this would obviously have to be addressed.

[Floodkiller]

A passing thought WRT skillcaps and character differentiation:

Maybe this race could be written to have dynamic max skill caps, with a sort of pyramidal structure to its available max skill levels. Players could raise one skill to, say, 17, with one more skill at 16, one at 15, one at 14, etc., all the way down the line, with the lowest skills being capped to something like 5 or 7; just enough for slight hybridization but not enough to make them a playstyle focus.

In this way, Cyno characters would become differentiated by the order in which their players choose to raise their skills, and this would allow for things like higher-damage offensive spellcaster Cynos, or higher-EV Dodging focused Cynos, or larger weapon 2h focused Cynos, while still limiting the overall available skill levels.

It's true that this approach would make any individual Cyno less able to swap freely between weapons and spell schools, but by carefully adjusting the rate at which the max available school level decreases, one can assume that a fair balance would be able to be found between enforcing differentiation and enabling effective-enough low-skill hybridization.

edit: There are probably some usability concerns here wrt accidentally raising skills by large amounts through unique kills or potions of experience, this would obviously have to be addressed.

I can already see the level of micromanagement that would follow, with players gaming the pyramid threshold levels to get a higher level of increased skills, making a hard 'soft' cap. For example, using your first recommended stopping point of 17, I would see bringing all skills to 15.9 to avoid triggering any flags in the code (say, first skill that breaks 16 is identified to be going to 17), thus getting a higher overall skill level for all of my skills and avoiding the pyramid system. If you set a lower threshold (say, 17 is the cap but the first skill to reach 10 is the only one that can do so) to stop that power gaming, you now have micromanagement of your skills after each kill to ensure that you are triggering you skills in the right order across the threshold, instead of only when you actually want to train those skills.

I could maybe see this working if you chose skills on an XL gain basis and maxed out the choice to the pyramid cap, but then you have an incredibly erratic difficulty curve that starts hard, bottoms out quickly (first level up choice, weapon skill goes from 0 to 17), and only starts to rise again after skill choices on level up become near worthless. It can't start from the bottom of the pyramid, either, because I think mummies would have a stronger all around game at that point.

[Implojin]

For example, using your first recommended stopping point of 17, I would see bringing all skills to 15.9 to avoid triggering any flags in the code (say, first skill that breaks 16 is identified to be going to 17), thus getting a higher overall skill level for all of my skills and avoiding the pyramid system. If you set a lower threshold (say, 17 is the cap but the first skill to reach 10 is the only one that can do so) to stop that power gaming, you now have micromanagement of your skills after each kill to ensure that you are triggering you skills in the right order across the threshold, instead of only when you actually want to train those skills.

The technical angle is easy to address, one example would be popping from an availability stack for each tier before allowing advancement to the next tier, but you could really handle it however you wanted to.

I agree that the usability angle is sketchy, though, especially with Crawl's soft-to-nonexistent ways to prevent skill level overruns.

[Floodkiller]

I messed around with the XP attribute using the new aptitudes and stat line, getting to the first rune with a variety of backgrounds (Fi/Be/IE/Sk) and further as I felt like it. -1 XP feels pretty good with the new stat line putting in a bit more restrictions, forcing more reliance on the good aptitudes early on, but it can still be relatively easy if you get above average floor loot. -2 XP feels like the sweet spot if I want to threaten more consumable use if the stockpiles are still too high with these changes, as I had to be a lot more liberal with my consumables due to the amount of OODs I was encountering, but it was still quite possible to make it to Lair the majority of the time. -3 and -4 XP definitely need some compensation if they are to be considered, probably with raising the stat line up again; -3 was incredibly demanding and I barely made it to Lair (all of my characters died before the first rune), and -4 felt way too weak to attempt without playing to the pinnacle of optimal play (which I definitely am not good at, so only the Be made it to Lair to die on Snake:2). I think I'll leave it at -1 for now, just in case I have too low a sample size to have a good estimate of their strength, and lower it to -2 if feedback from others says it is still too easy.

As a concerning sidenote, although I feel the low stats may have done the job for improving diversity of late game builds (armor choice was much harder to make as my characters reached Vaults, spell diversity was slightly better with the Int difference from background starts determining my late game spell choices), it might have highlighted that this species pairs maybe too well with Chei. Although slow is still a harsh downside, the high aptitudes combined with vastly improved stats brought most of the characters back to the strength I experienced with the +3 apt, 7/7/10 version of the species. I'm probably giving Chei too much credit and the slow would have killed me later on, but it is still something I should watch.

[Speleothing]

For example, using your first recommended stopping point of 17, I would see bringing all skills to 15.9 to avoid triggering any flags in the code (say, first skill that breaks 16 is identified to be going to 17), thus getting a higher overall skill level for all of my skills and avoiding the pyramid system. If you set a lower threshold (say, 17 is the cap but the first skill to reach 10 is the only one that can do so) to stop that power gaming, you now have micromanagement of your skills after each kill to ensure that you are triggering you skills in the right order across the threshold, instead of only when you actually want to train those skills.

The technical angle is easy to address, one example would be popping from an availability stack for each tier before allowing advancement to the next tier, but you could really handle it however you wanted to.

I agree that the usability angle is sketchy, though, especially with Crawl's soft-to-nonexistent ways to prevent skill level overruns.

Make it come up at set levels where the player chooses one skill to raise his/her cap on? Like choosing Str/Int/Dex on the 3s, except on the 5s, 6s, or 7s (each skill could only have its cap raised once, ofc)

You could also have the game randomly pick a skill at set intervals, to further encourage adaptation.