Unarmed troll: how to survive without luck?


Ask fellow adventurers how to stay alive in the deep, dark, dangerous dungeon below, or share your own accumulated wisdom.

Snake Sneak

Posts: 119

Joined: Tuesday, 10th September 2013, 14:55

Post Wednesday, 5th July 2017, 10:06

Unarmed troll: how to survive without luck?

I do not understand.
Trolls are good with decent armour.

But I find nearly impossible to go with a melee troll if you do not find a good dragon scale armour.
The few times I managed to start collecting runes were:
1) found +6 protection on D:2 and in the game of mutations, got +2AC and +1AC (and even in this case, was just not enough)
2) found GDA on D:1
3) found SwampDA on early Lair
The *only* alternative seems to find statue form and train for it. But... this leaves you for a long period with skill wasted on something you will need later. You need at least 7 earth and 7/8 transmutation, with good +Int ring and Wiz, to cast it. So, also here, you need a fair amount of luck.

Tried TrFi, TrGl, TrTm. The best combo is TrNe^Kiku, but anyway this does not help with armour, in the end you need some form of protection, for example in Shoal, or in Vaults.

Another partial alternative is going with Trog, hoping for a +9 ranged weapon of anything, and blasting everything from afar. But Trog cuts you away from the extended, and anyway in the end (Zot/Vaults...) you need something more than a robe to protect you.

Obviously this is true if you manage to have a shield. If you do not find even a shield, you are f***... ehm, dead.

So. I have some solutions in mind (higher troll armour base AC, leaving helmet slot usable by trolls, giving them some mutation that increases armour with levels).
But, apart from that, I'd like to know what do you think. Am I so bad at playing crawl? How do you survive when you have to kite every single death yak / hydra / catoblepa / high-AC orc?
Won: SpHu Slinger of Okawaru (3 runes) - DsGl Unarmed of Makhleb (3 runes) - TrFi Unarmed of Nem (3 runes) - GnFE of Ash (3 runes) - DsGl Unarmed (with Statue and Shatter) of Chei (15 runes!!!).
Demonspawn rocks. Unarmed also.

Ziggurat Zagger

Posts: 4432

Joined: Friday, 8th May 2015, 17:51

Post Wednesday, 5th July 2017, 10:17

Re: Unarmed troll: how to survive without luck?

d3k0 wrote:How do you survive when you have to kite every single death yak / hydra / catoblepa / high-AC orc?

By kiting them, not many characters have easy time killing them.
As I wrote in a similar thread, I had problems with Tr in Lair branches and Vaults too (late game is not problematic because you usually have some dragon armour by then).

Edit. I have just analyzed a few best TrXX scores from last tournament and found that most started using different dragon armour at XL 13-15: storm, fire, shadow, steam, even Troll leather armour. Basically wear what you found first.
Underestimated: cleaving, Deep Elf, Formicid, Vehumet, EV
Overestimated: AC, GDS
Twin account of Sandman25

Snake Sneak

Posts: 119

Joined: Tuesday, 10th September 2013, 14:55

Post Wednesday, 5th July 2017, 10:28

Re: Unarmed troll: how to survive without luck?

VeryAngryFelid wrote:(late game is not problematic because you usually have some dragon armour by then).


One of the key problems here is that *usually* is not *always*. Sometimes i drop more than 10 dragon armours. Sometimes not even one.
It depends not only on how luck you have with the single drop, but (much more) on how many dragons are generated (I mean before Zot)...
Won: SpHu Slinger of Okawaru (3 runes) - DsGl Unarmed of Makhleb (3 runes) - TrFi Unarmed of Nem (3 runes) - GnFE of Ash (3 runes) - DsGl Unarmed (with Statue and Shatter) of Chei (15 runes!!!).
Demonspawn rocks. Unarmed also.

bel

Cocytus Succeeder

Posts: 2184

Joined: Tuesday, 3rd February 2015, 22:05

Post Wednesday, 5th July 2017, 10:29

Re: Unarmed troll: how to survive without luck?

You can start as a Fi to get a shield. You have huge HP, good melee attack, good regeneration and large rocks. Use them.

You can simply run away from hydras in Lair. Or use large rocks.

In Swamp, you can use large rocks on hydras. Swamp dragon armour is fine as an intermediate armour. Before that: acid dragon armour and steam dragon armour are available.

By the time you get to Vaults, you will likely have some kind of medium dragon armour. You can usually find some better dragon armour in Vaults.
User avatar

Swamp Slogger

Posts: 140

Joined: Tuesday, 15th October 2013, 06:22

Post Wednesday, 5th July 2017, 10:30

Re: Unarmed troll: how to survive without luck?

Well you don't need to finish the game in a robe, you get dragon armours usually at latest by vaults. Sometimes earlier via shops/acq/swamps...
You have innate regen so you can kite normal speed monsters quite good.
You should never train a ranged weapon on a troll, instead use throwing which is really effective.
If you find lair etc. too hard on a troll you can also start with Lucy, as she gives you great tools to deal with many monsters and gives you a good escape option should you get hit too hard.

Cocytus Succeeder

Posts: 2229

Joined: Sunday, 18th December 2011, 13:31

Post Wednesday, 5th July 2017, 10:32

Re: Unarmed troll: how to survive without luck?

d3k0 wrote:But, apart from that, I'd like to know what do you think. Am I so bad at playing crawl? How do you survive when you have to kite every single death yak / hydra / catoblepa / high-AC orc?


You kite and you kite. By Lair trolls have some defence problem, until at least they get a dragon armor and\or a shield or statue form.

But their extremely high melee damage and high hp let them rip pretty much anything from D:1 to Z:5 when fighting 1vs1 easily, no matter what AC\EV they have.

So...kite, fight 1 enemy at time, don't let things hit you at range.

As a bonus, use large rocks to see things explode at range.
screw it I hate this character I'm gonna go melee Gastronok

Snake Sneak

Posts: 119

Joined: Tuesday, 10th September 2013, 14:55

Post Wednesday, 5th July 2017, 10:37

Re: Unarmed troll: how to survive without luck?

VeryAngryFelid wrote:most started using different dragon armour at XL 13-15: storm, fire, shadow, steam, even Troll leather armour. Basically wear what you found first.


I am not able to survive in just troll leather. Maybe troll leather, with 4 enchant armour, and any ring of protection.
Steam/shadow are fine (anyway i dropped shadow just once, in Zot). But if you do not find anything?
I mean, I know that "guaranteed drop" and "crawl" can stay in the same sentence only if there are also "never"/"cannot" and similar ;)
But what I want to say is that for the average player (not the best ones) the game become nearly impossible. Not difficult, just impossible.
After you clear Lair, Orc and Dungeon, what can you do? Tried Shoals and had to stop at Shoals:3. Tried Pit, and had to stop at Pit:3. Maybe Pit can be finished with a lot of care, but those nagas can "be teleported" (I don't know the spell they use) around you, so kiting is not easy. In the end, tried Depths (but a bunch of Iron Troll changed my mind) and Slime.
Got until Slime:3, kiting everything by the stair. Then I made a little error and an Azure thing killed me (that's fine, errors kill).
Won: SpHu Slinger of Okawaru (3 runes) - DsGl Unarmed of Makhleb (3 runes) - TrFi Unarmed of Nem (3 runes) - GnFE of Ash (3 runes) - DsGl Unarmed (with Statue and Shatter) of Chei (15 runes!!!).
Demonspawn rocks. Unarmed also.

bel

Cocytus Succeeder

Posts: 2184

Joined: Tuesday, 3rd February 2015, 22:05

Post Wednesday, 5th July 2017, 10:43

Re: Unarmed troll: how to survive without luck?

The phenomenon of Nagas "teleporting around you" refers to the special ability of Guardian Serpents: "Blink Allies Encircling". You should stay out of sight of them if in a crowded area (use fog, blink or teleport if required) and kill them quickly if you find them. The Nagas themselves don't have any special ability to encircle you, so you can kite them easily.

The Nagas are easy to separate from the other snakes because snakes are fast while Nagas are slow.

Ziggurat Zagger

Posts: 4432

Joined: Friday, 8th May 2015, 17:51

Post Wednesday, 5th July 2017, 10:53

Re: Unarmed troll: how to survive without luck?

d3k0 wrote:But what I want to say is that for the average player (not the best ones) the game become nearly impossible. Not difficult, just impossible.
After you clear Lair, Orc and Dungeon, what can you do? Tried Shoals and had to stop at Shoals:3. Tried Pit, and had to stop at Pit:3. Maybe Pit can be finished with a lot of care, but those nagas can "be teleported"


Seems very similar to my Tr game where I had problems, it was a long time ago so I am sure I was below average player threshold.
I hated those nagas, they made me retreat upstairs about a dozen times from the same floor (I didn't have rPois so I was hugely wounded after every fight). Patience is the key IMHO. If you feel you are tired of kiting, luring and resting upstairs, make a break. Eventually you will get some nice body armour and it will be fun game again.

Edit. I love Trog with Tr as it gives MR++ on demand which is great to prevent unwanted teleports on Snake:4.
Underestimated: cleaving, Deep Elf, Formicid, Vehumet, EV
Overestimated: AC, GDS
Twin account of Sandman25
User avatar

Tomb Titivator

Posts: 838

Joined: Friday, 2nd October 2015, 04:47

Post Wednesday, 5th July 2017, 12:22

Re: Unarmed troll: how to survive without luck?

The armor is a significant factor around Lair, but if you practice and try different things, it shouldn't be the be all-end all of all trolls.

Some has been mentioned, but all of these can help:

1. Throw stuff (get a few points of Throwing skill too) - doesn't always have to be rocks, javelins are great and even slings count if you get enough shots off.
2. Evocations in a pinch for wands - acid is really hot generally, clouds and scattershot are amazing around Lair (although you might want a few points in Evoc to get more than mephitic cloud).
3. Boost the heck out of yourself - play Oka or TSO, maybe Dith or Usk for reliable melee bonuses.
4. Wear lighter armors, they're really not bad at all. It is often possible to get through Lair in troll leather armor if your tactics are up to it.
5. Solid skills - I hate feeling "tracked" into the same skills over and over, but with a melee Troll there really isn't too much else to work with. 12-15 Fighting, 13-15 Unarmed is the kind of stuff I'm wishing for around Lair. I don't always get quite there, but it certainly helps with death yaks to aim for it. I'll usually skimp a bit on others - just a few points of Dodging and 2-4 points of Evoc. Invocations with Oka is surprisingly worth it though (heroism will fade pretty fast if you get less than 4-5).
6. Animate Skeleton (just a few points of Necromancy) on corpses of tougher mobs.
7. Don't start Lair when you first see it. Go a couple levels deeper in the Dungeon first and get your skills up/gather more drops. I can't emphasize this enough. If you don't feel up to Lair, sometimes it's because you aren't yet.

It's true that sometimes the order and timing of drops really messes you up (like if you occasionally get almost NO melee boosts or armor by Lair)... But trolls are extremely tough in the early game and still pretty good around Lair/Lair branches (which for me is kinda their weak point) with enough training. There's also usually some good stuff on the floor inside Lair, if you get an open enough layout or can find it/slip upstairs and go around before you draw in too many huge mobs.
Online game stats & morgues
More runes! GnWn (11, 0.21), GhMo^Makh (15, 0.17)
And a Yiuf: (1.4.6, 0.20): ImpGl^Oka (3)

For this message the author stoneychips has received thanks:
d3k0

Slime Squisher

Posts: 352

Joined: Monday, 14th December 2015, 00:43

Post Wednesday, 5th July 2017, 13:01

Re: Unarmed troll: how to survive without luck?

Trolls are very easy to play, they just aren't as beginner friendly as minotaurs. You're heavily underestimating the power of UC + Claws 3. Just lure opponents one by one and massacre them in melee. Throw things at them from distance if you need to soften them up or it's a hydra. Animate Skeleton and Confusing Touch are two good spells. For a god choose Trog, Okawaru or Hep.

I started this character literally to have an example of how to play in TLA: https://pastebin.com/aDzmUPmr
2 runes at 21k turns, I'm not sure if I'll win only because I'm playing extremely careless; a very good player will win 10 trolls out of 10 if he concentrates, no matter the background, unless he finds a gnoll pack right next to the only exit from the spawn or gets shafted twice from D:1 or something.

For this message the author Leszczynek has received thanks:
d3k0

Cocytus Succeeder

Posts: 2173

Joined: Saturday, 2nd February 2013, 09:52

Post Wednesday, 5th July 2017, 13:27

Re: Unarmed troll: how to survive without luck?

Leszczynek wrote:Trolls are very easy to play, they just aren't as beginner friendly as minotaurs.

In this case I count myself as a beginner too, despite my 92 wins. Trolls are not bad if you are religiously luring enemies and try to avoid getting into trouble at all cost. However, they are nowhere near the power level of minotaurs.

For this message the author Magipi has received thanks:
VeryAngryFelid

Ziggurat Zagger

Posts: 4432

Joined: Friday, 8th May 2015, 17:51

Post Wednesday, 5th July 2017, 13:29

Re: Unarmed troll: how to survive without luck?

  Code:
- Level 12.0 Fighting
 * Level 13.2 Throwing
 - Level 10.1 Dodging
 - Level 0.3 Stealth
 + Level 13.1 Shields
 + Level 16.7 Unarmed Combat
 - Level 7.3 Invocations

This is not melee Tr, it has too high Throwing (javelins are already at min delay with Heroism). Also maybe it is not weak because of +6 ring of evasion, pure melee and AC 15/EV 16 would made the character feel quite different IMHO
Underestimated: cleaving, Deep Elf, Formicid, Vehumet, EV
Overestimated: AC, GDS
Twin account of Sandman25

Snake Sneak

Posts: 119

Joined: Tuesday, 10th September 2013, 14:55

Post Wednesday, 5th July 2017, 14:00

Re: Unarmed troll: how to survive without luck?

Obviously a +6EV (or +6AC) ring helps, anyway quite impressed you got to Vaults:1.
Bringing Throwing to high levels is ok, the main topic here is "how to survive without luck" and if the answer is "throw anything you find on the ground to anything that moves at the edge of LOS" it's ok too.

@stoneychips: One thing that I notice here is that you say Fighting to 12-15, but I have it usually lower. 5-6 if I trained (great) shields, and around 10 if not (but I always train shield with troll...).

About Evo, now that the rods has been nerfed from "really useful and cool tool" to "almost useless/situational", I find myself with... ehm... lack of space in inventory. Let me explain this: a bunch of utility scrools, a bunch of lifesaving potions and utility potions, a long list of amulets and rings for swapping resistances with AC/EV (and Wiz, if you are not with Trog), food, throwing things (see note *)... there is still space for some powerful evocables (wand of clouds, disintegration, sack) but not for "everyday evocables"...

(*) if you rely on throwing, since throwing things are not-so-common, you end up with: large rocks, javelins, tomahawks, and their cousins "of returning", "of penetration", "of dispersal", "silver"... maybe you can drop the poison one, but the list is quite long anyway. And... if you are throwing anyway, would you renounce to a blowgun? And so: needles of sleeping, of confusion, of curare....
Won: SpHu Slinger of Okawaru (3 runes) - DsGl Unarmed of Makhleb (3 runes) - TrFi Unarmed of Nem (3 runes) - GnFE of Ash (3 runes) - DsGl Unarmed (with Statue and Shatter) of Chei (15 runes!!!).
Demonspawn rocks. Unarmed also.

Slime Squisher

Posts: 352

Joined: Monday, 14th December 2015, 00:43

Post Wednesday, 5th July 2017, 15:20

Re: Unarmed troll: how to survive without luck?

VeryAngryFelid wrote:This is not melee Tr, it has too high Throwing (javelins are already at min delay with Heroism). Also maybe it is not weak because of +6 ring of evasion, pure melee and AC 15/EV 16 would made the character feel quite different IMHO

If you play a true melee only character (no throwing, no blowguns, no untrained slings/hand xbows, no wands), you're actively handicapping yourself. This is true for absolutely every race and background in the game. The large rocks were also all a gift from the god that you can pick up somewhere on the way to Lair on your own on every character.
Am I supposed to run another character and actively avoid AC/EV rings just to take away your excuses? Trolls are easy, easier than old ogres were.

Ziggurat Zagger

Posts: 4432

Joined: Friday, 8th May 2015, 17:51

Post Wednesday, 5th July 2017, 16:30

Re: Unarmed troll: how to survive without luck?

I meant your Tr has better ranged attack than melee, you don't really need to lure anything or rest a lot because you kill them with missiles. Hence it's not melee Tr
Underestimated: cleaving, Deep Elf, Formicid, Vehumet, EV
Overestimated: AC, GDS
Twin account of Sandman25

Shoals Surfer

Posts: 276

Joined: Sunday, 6th November 2016, 19:19

Post Wednesday, 5th July 2017, 16:59

Re: Unarmed troll: how to survive without luck?

If you have TLA + Shield + Aux armour that's all the AC/EV/SH you'll need for 2 runes on a troll, you don't even need a god, that's how good Trolls are. They require a little more careful play than a heavy melee axe user, but not to the extent that something like octopode requires. Be more liberal with your consumable use and you'll be fine.

I would recommend playing Makhleb, so you don't get too worn down in big,pitched battles. Once you have him, train invo to 4, and then never train it again, it's a waste of exp. Use minor demons to kill hydras, but that's about the only time you should use his active abilities since heal-on-kills is just an incredibly good passive.
16/26 on the way to GreaterOctopode (Win all backgrounds as an Octopode)

Progress so far : OpFi, OpGl, OpWn, OpAr, OpCK, OpMo, OpBe, OpHu, OpVM, OpAM, OpWr, OpFE, OpEE, OpNe, OpTm, OpSk

For this message the author Vajrapani has received thanks:
Majang
User avatar

Tomb Titivator

Posts: 838

Joined: Friday, 2nd October 2015, 04:47

Post Wednesday, 5th July 2017, 19:22

Re: Unarmed troll: how to survive without luck?

d3k0 wrote:
@stoneychips: One thing that I notice here is that you say Fighting to 12-15, but I have it usually lower. 5-6 if I trained (great) shields, and around 10 if not (but I always train shield with troll...).
.

I don't know if I've played around with shields enough to answer this, at least not recently. They are good with unarmed; I just can't put specific numbers to em right now. I don't think you need to hit quite the recommended level in shields to get by with an early troll, because shields are quite good early on and trolls are just so tough/high damaging things.

Anyway, it's probably quite possible to have lower levels of Fighting and find a combo that gets by. I'm mostly thinking of melee with elephants and death yaks, and moving on to orc warlords etc. in a while after. Maybe 15 is a bit much. (All my faster-training ogres are probably confusing me; yeah ok, 15 seems a touch high that early for trolls.) I do hate how often people say, "Train more fighting" as the solution for everything. It might be somewhat less important also if you have say, a good collection of wands or cool missile weapons.
Online game stats & morgues
More runes! GnWn (11, 0.21), GhMo^Makh (15, 0.17)
And a Yiuf: (1.4.6, 0.20): ImpGl^Oka (3)

Tomb Titivator

Posts: 771

Joined: Tuesday, 25th November 2014, 02:47

Post Wednesday, 5th July 2017, 21:42

Re: Unarmed troll: how to survive without luck?

Vajrapani wrote:I would recommend playing Makhleb, so you don't get too worn down in big,pitched battles. Once you have him, train invo to 4, and then never train it again, it's a waste of exp. Use minor demons to kill hydras, but that's about the only time you should use his active abilities since heal-on-kills is just an incredibly good passive.


Makhleb + Tr is very good, however you are giving poor advice regarding his actives. Minor destruction gives you a functional full LoS ranged attack early and can be spammed for no piety. It is even better on trolls: it gives you a ranged attack at a cost of HP, something that Tr has plenty of to use as a resource. It allows you to play ranged troll much earlier than throw troll does. It kills hydras cleanly, and will frequently spam down the first few yaks in a pack. It takes no ammo. It scales with invocations and builds into greater demon. Greater demon is incredibly powerful but situational. Until late game have to use it carefully or it will get you killed. It is not difficult to use carefully. The only questionable ability is major destruction, which is too situational and generally inferior to wands or other options.

Makhleb is a decent god because of his passive. Makhleb is a great god because of his actives.

Makhleb invocations on troll gives you more tools in your tool kit than points put anywhere else in your tree. It is better than either throwing or evocations, its main two competitors. A low investment in evocations is still worth it. If you want to pick up throwing in the mid-late game you can do so with mid-lategame XP.

For this message the author edgefigaro has received thanks: 2
Gigaslurp, Vajrapani

Shoals Surfer

Posts: 276

Joined: Sunday, 6th November 2016, 19:19

Post Wednesday, 5th July 2017, 22:04

Re: Unarmed troll: how to survive without luck?

Admittedly, I always forget about,and underuse, minor destruction. The only 2 active abilities that tend to stand out to me are minor and greater demon, the latter requires a big exp investment to get safely castable due to troll's all-around crappy aptitudes, so I tend to just stay at 4 invo forever.

I probably should use it more, since it is essentially free to cast.
16/26 on the way to GreaterOctopode (Win all backgrounds as an Octopode)

Progress so far : OpFi, OpGl, OpWn, OpAr, OpCK, OpMo, OpBe, OpHu, OpVM, OpAM, OpWr, OpFE, OpEE, OpNe, OpTm, OpSk

cjo

Spider Stomper

Posts: 217

Joined: Tuesday, 20th September 2011, 02:03

Post Thursday, 6th July 2017, 00:27

Re: Unarmed troll: how to survive without luck?

I'm so psyched - I just won my first Troll playing as a monk of Hepliaklqana, with a hexing ancestor :) I had great luck with gear and consumables, so I can't really weigh in on how to play a troll sans luck. However, I can say that Hep rocked from the moment I converted right through Zot 5, where my ancestor's anti-magic quickblade helped me take out orbs of fire. My ancestor provided both a ton of extra damage, helping take out mobs quickly, and a ton of extra defense by virtue of confusing enemies. I would definitely recommend Hep as one of the good gods for Trolls.

Spider Stomper

Posts: 212

Joined: Monday, 3rd April 2017, 11:44

Post Thursday, 6th July 2017, 16:49

Re: Unarmed troll: how to survive without luck?

I would consider troll leather armour to be "good enough" shieldless defence to collect a rune. Perhaps not good enough for Shoal but Shoal is generally deadly. You are pretty much guaranteed to find troll leather armour or better by going into Orc or going down to D:14 before going to a rune branch. For whatever reason, I usually find ogres more vulnerable. Your criteria to collect runes seem a little high. For whatever reason I always feel old ogres are more vulnerable than trolls before troll leather armour.
User avatar

Abyss Ambulator

Posts: 1194

Joined: Friday, 18th April 2014, 01:41

Post Thursday, 6th July 2017, 22:12

Re: Unarmed troll: how to survive without luck?

use okawaru so you can get gifts and +5 to UC
remove food

For this message the author tabstorm has received thanks:
cerebovssquire

Slime Squisher

Posts: 352

Joined: Monday, 14th December 2015, 00:43

Post Thursday, 6th July 2017, 23:02

Re: Unarmed troll: how to survive without luck?

Plantissue wrote:I would consider troll leather armour to be "good enough" shieldless defence to collect a rune. Perhaps not good enough for Shoal but Shoal is generally deadly.

I've done Shoals in robes with old ogres, so it's not impossible.

Halls Hopper

Posts: 55

Joined: Monday, 6th February 2017, 00:41

Post Friday, 7th July 2017, 00:15

Re: Unarmed troll: how to survive without luck?

I just won my first troll as well, as a hunter of nemelex. I got lucky with with getting Bai Suzhen in Nest4, but yeah the armour curve is definitely their trickiest part. Nemelex is good since I think his powers "peak" in usefullness around the lair branches/early vaults, which is where trolls are often feeling the lack of heavy armour the most. Also you carry very little as a troll so you've got space for lots of decks.
Wins:
(3 Runes) MiBe^Trog, DEFE^Veh, HOMo^Beo, SpEn^Goz, DDNe^Mak, MfGl^Fed, TrHu^Nem, DrSu^Ash, CeAK^Luc, GrEE^Kiku, {GhAr^WJC, HoGl^Beo}, GnWz^Oka, OgAM^Ydr
(15 runes) GrEE^Veh
User avatar

Tomb Titivator

Posts: 838

Joined: Friday, 2nd October 2015, 04:47

Post Friday, 7th July 2017, 12:29

Re: Unarmed troll: how to survive without luck?

stoneychips wrote:
d3k0 wrote:
@stoneychips: One thing that I notice here is that you say Fighting to 12-15, but I have it usually lower. 5-6 if I trained (great) shields, and around 10 if not (but I always train shield with troll...).
.

Looking at another character with lower Fighting aptitude recently, yeah 15 is probably high. Maybe 8-10? I still feel like 5 might be lowish, but maybe with a shield etc. it's not impossible.
Online game stats & morgues
More runes! GnWn (11, 0.21), GhMo^Makh (15, 0.17)
And a Yiuf: (1.4.6, 0.20): ImpGl^Oka (3)

Spider Stomper

Posts: 212

Joined: Monday, 3rd April 2017, 11:44

Post Friday, 7th July 2017, 13:26

Re: Unarmed troll: how to survive without luck?

I got inspired from reading this thread so I decided to hash up a troll as well using the Makhleb advice. I have 4 wins so my style wouldn't be as optimised as others. https://pastebin.com/19EK9hWp
Mostly in robes and a robe of resistance. There's something funny about seeing a small swarm of killer bees at D:3 and surviving by not having to avoid the level entirely. Had a scary moment in D:11 when I got shafted through 3 floors. I basically did Orc first because I wasn't confident in facing hydras, but it turns out that Makhleb minor destruction and demon summons works well enough against them. I would guess trog summons or just plain throwing would do well enough as well.
Ignore the Robe of Vines. I only got it in Lair 6. I got troll leather armour at D:9. Acid Dragon armour at D:11 and a +2 robe of resistance at D:6. +2 Shield was only equipped in Spider. I would say there is a pretty good chance of getting tla before lair simply by going orc or by going down to D:12.
I left skills on manual equally distributed till I equipped the shield. Did I get good luck with rings? Perhaps a little I guess.

Tomb Titivator

Posts: 771

Joined: Tuesday, 25th November 2014, 02:47

Post Friday, 7th July 2017, 23:48

Re: Unarmed troll: how to survive without luck?

So when i'm talking about spamming Minor Destruction like a ranged attack, this is what I'm talking about:

Invok: Minor Destruction | | | 172 | 426 | 690 | 494 | 176 | 166 | 89 || 2213
http://crawl.berotato.org/crawl/morgue/ ... 221912.txt

or this:

Invok: Minor Destruction | | | 71 | 509 | 707 | 489 | 406 | 66 | 43 || 2291
http://crawl.berotato.org/crawl/morgue/ ... 054840.txt

Its really really good. Use it all the time against all of the things.

This character used it signifcantly less due to becoming strong enough to hold down the buttons o and tab.

Invok: Minor Destruction | | | 5 | 123 | 188 | 63 | 55 | | || 434
http://crawl.berotato.org/crawl/morgue/ ... 011403.txt

For this message the author edgefigaro has received thanks:
Gigaslurp
User avatar

Crypt Cleanser

Posts: 732

Joined: Monday, 24th April 2017, 11:46

Post Saturday, 8th July 2017, 07:29

Re: Unarmed troll: how to survive without luck?

edgefigaro wrote:So when i'm talking about spamming Minor Destruction like a ranged attack, this is what I'm talking about:

I belonged to those who discarded minor destruction before reading your post, but inspired by you I was playing a Makhleb fighter and using MD heavily, and boy were you right! This ability really is amazing. Most single attackers in Lair didn't even make it into melee range to get whacked over their head. I lost that character to my own stupidity later, but MD is much cooler than I thought. What I ask myself now: does it scale up in its power all the way up with INV 27, or does it get a cap somewhere on the way?
Maɟaŋ

Spider Stomper

Posts: 212

Joined: Monday, 3rd April 2017, 11:44

Post Saturday, 8th July 2017, 23:09

Re: Unarmed troll: how to survive without luck?

I lost focus the time I got to lair and started O tabbing.

Tomb Titivator

Posts: 771

Joined: Tuesday, 25th November 2014, 02:47

Post Sunday, 9th July 2017, 23:40

Re: Unarmed troll: how to survive without luck?

I may be reading the code wrong but:Minor Destruction is affected by one minor stepdown at 8 and one major stepdown at 16.

Edit: H/T Siegurt for correction

  Code:
case ABIL_MAKHLEB_MINOR_DESTRUCTION:
    {
        beam.range = LOS_RADIUS;

        if (!spell_direction(spd, beam))
            return SPRET_ABORT;

        int power = you.skill(SK_INVOCATIONS, 1)
                    + random2(1 + you.skill(SK_INVOCATIONS, 1))
                    + random2(1 + you.skill(SK_INVOCATIONS, 1));

        // Since the actual beam is random, check with BEAM_MMISSILE and the
        // highest range possible.
        if (!player_tracer(ZAP_DEBUGGING_RAY, power, beam, LOS_RADIUS))
            return SPRET_ABORT;

        fail_check();
        beam.origin_spell = SPELL_NO_SPELL; // let zapping reset this

        switch (random2(5))
        {
[i]        case 0: zapping(ZAP_THROW_FLAME, power, beam); break;
        case 1: zapping(ZAP_PAIN,  power, beam); break;
        case 2: zapping(ZAP_STONE_ARROW, power, beam); break;
        case 3: zapping(ZAP_SHOCK, power, beam); break;
        case 4: zapping(ZAP_BREATHE_ACID, power/2, beam); break;[/i]
        }
        break;
    }


I understand the code to read as Power = Invocations skill + 2d(invocations+1). Thus, an invocations of 5 = Spellpower minimum of 7 (5+1+1) and a Spellpower max of 17(5+6+6). The spells that Minor destruction calls have different spellpower caps.

Shock has the lowest spellpower cap of 25, so it will start to stepdown at invocations 8 and always roll max spellpower at invocations 23. This is minor compared to the spellpower 50, where Throw Flame and Stone Arrow cap. This begins at Invocations 16 (16+17+17).

Pain caps spellpower at 100, so it will never stop scaling as such. Breathe Acid Caps at 50, but spellpower is reduced by half, so it will similarly never be affected.
Last edited by edgefigaro on Monday, 10th July 2017, 07:21, edited 4 times in total.

Ziggurat Zagger

Posts: 6454

Joined: Tuesday, 30th October 2012, 19:06

Post Monday, 10th July 2017, 00:13

Re: Unarmed troll: how to survive without luck?

edgefigaro wrote:I may be reading the code wrong but: Minor Destruction is affected by one minor stepdown at 8 and one major stepdown at 16.

  Code:
case ABIL_MAKHLEB_MINOR_DESTRUCTION:
    {
        beam.range = LOS_RADIUS;

        if (!spell_direction(spd, beam))
            return SPRET_ABORT;

[b]        int power = you.skill(SK_INVOCATIONS, 1)
                    + random2(1 + you.skill(SK_INVOCATIONS, 1))
                    + random2(1 + you.skill(SK_INVOCATIONS, 1));[/b]

        // Since the actual beam is random, check with BEAM_MMISSILE and the
        // highest range possible.
        if (!player_tracer(ZAP_DEBUGGING_RAY, power, beam, LOS_RADIUS))
            return SPRET_ABORT;

        fail_check();
        beam.origin_spell = SPELL_NO_SPELL; // let zapping reset this

        switch (random2(5))
        {
[i]        case 0: zapping(ZAP_THROW_FLAME, power, beam); break;
        case 1: zapping(ZAP_PAIN,  power, beam); break;
        case 2: zapping(ZAP_STONE_ARROW, power, beam); break;
        case 3: zapping(ZAP_SHOCK, power, beam); break;
        case 4: zapping(ZAP_BREATHE_ACID, power/2, beam); break;[/i]
        }
        break;
    }


I understand the code to read as Power = Invocations skill + 2d(invocations+1). Thus, an invocations of 5 = Spellpower minimum of 7 (5+1+1) and a Spellpower max of 17(5+6+6). The spells that Minor destruction calls have different spellpower caps.

Shock has the lowest spellpower cap of 25, so it will start to stepdown at invocations 8 and always roll max spellpower at invocations 23. This is minor compared to the spellpower 50, where Throw Flame and Stone Arrow cap. This begins at Invocations 16 (16+17+17).

Pain caps spellpower at 100, so it will never stop scaling as such. Breathe Acid Caps at 50, but spellpower is reduced by half, so it will similarly never be affected.


Spellpower caps only apply to casting spells. Zappping is normally invoked by casting a spell after spellpower stepdowns and caps have been applied (by calc_spell_power in spl-cast.cc) when casting a spell, the spellpower cap of actually casting a spell isn't relevant when using mak's invokations.
Spoiler: show
This high quality signature has been hidden for your protection. To unlock it's secret, send 3 easy payments of $9.99 to me, by way of your nearest theta band or ley line. Complete your transmission by midnight tonight for a special free gift!

For this message the author Siegurt has received thanks: 2
edgefigaro, Majang

Tomb Titivator

Posts: 771

Joined: Tuesday, 25th November 2014, 02:47

Post Monday, 10th July 2017, 00:47

Re: Unarmed troll: how to survive without luck?

Power caps are in both zap-data.h and spell-data.h (I'm not saying you are wrong, just saying I'm confused, and this is the source of my confusion)

For Instance, in spl-data.h SPELL_STONE_ARROW, power_cap is listed as 50.

  Code:

/*
struct spell_desc
{
    enum, spell name,
    spell schools,
    flags,
    level,
    power_cap,
    min_range, max_range, (-1 if not applicable)
    noise, effect_noise
    tile
}
*/

{
    SPELL_STONE_ARROW, "Stone Arrow",
    SPTYP_CONJURATION | SPTYP_EARTH,
    SPFLAG_DIR_OR_TARGET | SPFLAG_NEEDS_TRACER,
    3,
    50,
    4, 4,
    3, 0,
    TILEG_STONE_ARROW,
},


That 50 is the spellpower cap.

Then in zap-data, ZAP_STONE_ARROW player_power_cap is listed as 50.

  Code:
struct zap_info
{
    zap_type ztype;
    const char* name;
    int player_power_cap;
    dam_deducer* player_damage;
    tohit_deducer* player_tohit;       // Enchantments have power modifier here
    dam_deducer* monster_damage;
    tohit_deducer* monster_tohit;      // Enchantments have power modifier here
    colour_t colour;
    bool is_enchantment;
    beam_type flavour;
    dungeon_char_type glyph;
    bool always_obvious;
    bool can_beam;
    bool is_explosion;
    int hit_loudness;
}
*/

{
    ZAP_STONE_ARROW,
    "stone arrow",
    50,
    new dicedef_calculator<3, 5, 1, 8>,
    new tohit_calculator<8, 1, 10>,
    new dicedef_calculator<3, 5, 1, 10>,
    new tohit_calculator<14, 1, 35>,
    LIGHTGREY,
    false,
    BEAM_MMISSILE,
    DCHAR_FIRED_MISSILE,
    true,
    false,
    false,
    3
},



Edit: zap is constructed in beam.cc and player_power_cap is set at the given max in zap-data using zap_ench_power.

  Code:
struct zap_info
{
    zap_type ztype;
    const char* name;           // nullptr means handled specially
    int player_power_cap;
    dam_deducer* player_damage;
    tohit_deducer* player_tohit;    // Enchantments have power modifier here
    dam_deducer* monster_damage;
    tohit_deducer* monster_tohit;
    colour_t colour;
    bool is_enchantment;
    beam_type flavour;
    dungeon_char_type glyph;
    bool always_obvious;
    bool can_beam;
    bool is_explosion;
    int hit_loudness;
};

int zap_power_cap(zap_type z_type)
{
    const zap_info* zinfo = _seek_zap(z_type);

    return zinfo ? zinfo->player_power_cap : 0;
}

int zap_ench_power(zap_type z_type, int pow, bool is_monster)
{
    const zap_info* zinfo = _seek_zap(z_type);
    if (!zinfo)
        return pow;

    if (zinfo->player_power_cap > 0 && !is_monster)
        pow = min(zinfo->player_power_cap, pow);

    tohit_deducer* ench_calc = is_monster ? zinfo->monster_tohit
                                          : zinfo->player_tohit;
    if (zinfo->is_enchantment && ench_calc)
        return (*ench_calc)(pow);
    else
        return pow;
}



My understanding is now spellpower cap and zap power cap are different, but player values for zap power is capped in zap-data while monster zap power is not. The mechanism for this cap are in beam.cc under zap_ench_power, which would imply that my previous statements about stepdowns are accurate. (it was right but for the wrong reasons at the time)

For this message the author edgefigaro has received thanks:
Majang

Ziggurat Zagger

Posts: 6454

Joined: Tuesday, 30th October 2012, 19:06

Post Monday, 10th July 2017, 07:39

Re: Unarmed troll: how to survive without luck?

edgefigaro wrote:Power caps are in both zap-data.h and spell-data.h (I'm not saying you are wrong, just saying I'm confused, and this is the source of my confusion)

For Instance, in spl-data.h SPELL_STONE_ARROW, power_cap is listed as 50.

  Code:

/*
struct spell_desc
{
    enum, spell name,
    spell schools,
    flags,
    level,
    power_cap,
    min_range, max_range, (-1 if not applicable)
    noise, effect_noise
    tile
}
*/

{
    SPELL_STONE_ARROW, "Stone Arrow",
    SPTYP_CONJURATION | SPTYP_EARTH,
    SPFLAG_DIR_OR_TARGET | SPFLAG_NEEDS_TRACER,
    3,
    50,
    4, 4,
    3, 0,
    TILEG_STONE_ARROW,
},


That 50 is the spellpower cap.

Then in zap-data, ZAP_STONE_ARROW player_power_cap is listed as 50.

  Code:
struct zap_info
{
    zap_type ztype;
    const char* name;
    int player_power_cap;
    dam_deducer* player_damage;
    tohit_deducer* player_tohit;       // Enchantments have power modifier here
    dam_deducer* monster_damage;
    tohit_deducer* monster_tohit;      // Enchantments have power modifier here
    colour_t colour;
    bool is_enchantment;
    beam_type flavour;
    dungeon_char_type glyph;
    bool always_obvious;
    bool can_beam;
    bool is_explosion;
    int hit_loudness;
}
*/

{
    ZAP_STONE_ARROW,
    "stone arrow",
    50,
    new dicedef_calculator<3, 5, 1, 8>,
    new tohit_calculator<8, 1, 10>,
    new dicedef_calculator<3, 5, 1, 10>,
    new tohit_calculator<14, 1, 35>,
    LIGHTGREY,
    false,
    BEAM_MMISSILE,
    DCHAR_FIRED_MISSILE,
    true,
    false,
    false,
    3
},



Edit: zap is constructed in beam.cc and player_power_cap is set at the given max in zap-data using zap_ench_power.

  Code:
struct zap_info
{
    zap_type ztype;
    const char* name;           // nullptr means handled specially
    int player_power_cap;
    dam_deducer* player_damage;
    tohit_deducer* player_tohit;    // Enchantments have power modifier here
    dam_deducer* monster_damage;
    tohit_deducer* monster_tohit;
    colour_t colour;
    bool is_enchantment;
    beam_type flavour;
    dungeon_char_type glyph;
    bool always_obvious;
    bool can_beam;
    bool is_explosion;
    int hit_loudness;
};

int zap_power_cap(zap_type z_type)
{
    const zap_info* zinfo = _seek_zap(z_type);

    return zinfo ? zinfo->player_power_cap : 0;
}

int zap_ench_power(zap_type z_type, int pow, bool is_monster)
{
    const zap_info* zinfo = _seek_zap(z_type);
    if (!zinfo)
        return pow;

    if (zinfo->player_power_cap > 0 && !is_monster)
        pow = min(zinfo->player_power_cap, pow);

    tohit_deducer* ench_calc = is_monster ? zinfo->monster_tohit
                                          : zinfo->player_tohit;
    if (zinfo->is_enchantment && ench_calc)
        return (*ench_calc)(pow);
    else
        return pow;
}



My understanding is now spellpower cap and zap power cap are different, but player values for zap power is capped in zap-data while monster zap power is not. The mechanism for this cap are in beam.cc under zap_ench_power, which would imply that my previous statements about stepdowns are accurate. (it was right but for the wrong reasons at the time)

The places in the code are in fact where that spellpower cap is *stored* however that's not where it's *used*

zap_ench_power is the *enchantment power* for hex-type things, which frequently get some kind of modification to raw spellpower.

If you look a little further past the
  Code:
pbolt.ench_power = zap_ench_power(z_type, power, is_monster);

line you'll see where it calculates the damage done:
  Code:
    dam_deducer* dam_calc = is_monster ? zinfo->monster_damage
                                       : zinfo->player_damage;
    if (dam_calc)
        pbolt.damage = (*dam_calc)(power);


You'll note it uses "power" here (that is to say, the spellpower as passed into the zappy function, not the one set in "pbolt.ench_power" by zap_ench_power) Which for makhleb invocations is directly the "power" calculated in the ABIL_MAKHLEB_MINOR_DESTRUCTION case you quoted above (invocations+2d(invocations)) passed through the zappy function directly to the damage-rolling function for that spell.

If you poke around a little further you'll see that pbolt.ench_power is used when deciding if a spell beats MR, and has nothing to do with damage.

Mak's invocations skip the spellpower cap of the related spell entirely, it doesn't actually *cast* the spells in question, it produces an invocation that creates the same type of bolt that those spells produce.
Spoiler: show
This high quality signature has been hidden for your protection. To unlock it's secret, send 3 easy payments of $9.99 to me, by way of your nearest theta band or ley line. Complete your transmission by midnight tonight for a special free gift!

Ziggurat Zagger

Posts: 4432

Joined: Friday, 8th May 2015, 17:51

Post Monday, 10th July 2017, 07:42

Re: Unarmed troll: how to survive without luck?

Guys, could you please write tl/dr version when you are done with investigations? I know what you are talking about but it's rather hard to follow even for those who are familiar with code IMHO. Thank you ;)
Underestimated: cleaving, Deep Elf, Formicid, Vehumet, EV
Overestimated: AC, GDS
Twin account of Sandman25
User avatar

Crypt Cleanser

Posts: 732

Joined: Monday, 24th April 2017, 11:46

Post Monday, 10th July 2017, 07:50

Re: Unarmed troll: how to survive without luck?

VeryAngryFelid wrote:Guys, could you please write tl/dr version when you are done with investigations? I know what you are talking about but it's rather hard to follow even for those who are familiar with code IMHO. Thank you ;)

Let me try to summarize what I understood so far: Indeed the power of Minor Destruction scales all the way up to Invo 27, without cap. Yippee!
Maɟaŋ

For this message the author Majang has received thanks:
VeryAngryFelid

Tomb Titivator

Posts: 771

Joined: Tuesday, 25th November 2014, 02:47

Post Monday, 10th July 2017, 08:38

Re: Unarmed troll: how to survive without luck?

There are the following lines in zappy:

if (zinfo->player_power_cap > 0 && !is_monster)
power = min(zinfo->player_power_cap, power);

Later, your aforementioned damage calculation is called.

if (dam_calc)
pbolt.damage = (*dam_calc)(power);

Does the pbolt.damage not use the power set from the first two lines?

Also, I'm missing a link between zappy() and zapping(). zappy is in beam.cc. zapping is called from ABIL_MAKHLEB_MINOR_DESTRUCTION:
case 2: zapping(ZAP_STONE_ARROW, power, beam); break;

I'm going to look for zapping() (its not in beam.cc, havent looked elsewhere) tomorrow and see if I understand whats going on better from there.

Other: You mention power as being Invocations+2d(invocations) where I offered it as 2d(invocations + 1). This is due to me not understanding how random2 works? Relevant code.
random2(1 + you.skill(SK_INVOCATIONS, 1)) - This returns an integer from 0 to Invocations rather than 1 to Invocations+1. Thus Minor Destruction cast at invocations 5 could return Spellpower from 5 to 15 rather than 7 to 17.

Its late, I'm tired. Thank you for your help Siegurt.

  Code:
void zappy(zap_type z_type, int power, bool is_monster, bolt &pbolt)
{
    const zap_info* zinfo = _seek_zap(z_type);

    // None found?
    if (zinfo == nullptr)
    {
        dprf("Couldn't find zap type %d", z_type);
        return;
    }

    // Fill
    pbolt.name           = zinfo->name;
    pbolt.flavour        = zinfo->flavour;
    pbolt.real_flavour   = zinfo->flavour;
    pbolt.colour         = zinfo->colour;
    pbolt.glyph          = dchar_glyph(zinfo->glyph);
    pbolt.obvious_effect = zinfo->always_obvious;
    pbolt.pierce         = zinfo->can_beam;
    pbolt.is_explosion   = zinfo->is_explosion;

    if (zinfo->player_power_cap > 0 && !is_monster)
        power = min(zinfo->player_power_cap, power);

    ASSERT(zinfo->is_enchantment == pbolt.is_enchantment());

    pbolt.ench_power = zap_ench_power(z_type, power, is_monster);

    if (zinfo->is_enchantment)
        pbolt.hit = AUTOMATIC_HIT;
    else
    {
        tohit_deducer* hit_calc = is_monster ? zinfo->monster_tohit
                                             : zinfo->player_tohit;
        ASSERT(hit_calc);
        pbolt.hit = (*hit_calc)(power);
        if (pbolt.hit != AUTOMATIC_HIT && !is_monster)
            pbolt.hit = max(0, pbolt.hit - 5 * you.inaccuracy());
    }

    dam_deducer* dam_calc = is_monster ? zinfo->monster_damage
                                       : zinfo->player_damage;
    if (dam_calc)
        pbolt.damage = (*dam_calc)(power);

    pbolt.origin_spell = zap_to_spell(z_type);

    if (z_type == ZAP_BREATHE_FIRE && you.species == SP_RED_DRACONIAN
        && !is_monster)
    {
        pbolt.origin_spell = SPELL_SEARING_BREATH;
    }

    if (pbolt.loudness == 0)
        pbolt.loudness = zinfo->hit_loudness;
}

Ziggurat Zagger

Posts: 6454

Joined: Tuesday, 30th October 2012, 19:06

Post Monday, 10th July 2017, 15:10

Re: Unarmed troll: how to survive without luck?

edgefigaro wrote:There are the following lines in zappy:

if (zinfo->player_power_cap > 0 && !is_monster)
power = min(zinfo->player_power_cap, power);


Yep, you're right, that is limiting the spell-power to the spell's power cap, I just missed that when scanning for the line you mentioned earlier with zap_ench_power, which has nothing whatsoever to do with this :)

Where I got Invo+2d(invo) was from this:
  Code:
        int power = you.skill(SK_INVOCATIONS, 1)
                    + random2(1 + you.skill(SK_INVOCATIONS, 1))
                    + random2(1 + you.skill(SK_INVOCATIONS, 1));

in the code you quoted from ABIL_MAKHLEB_MINOR_DESTRUCTION.

random2(x) returns a number between 0 and x-1 So random2(5) returns one of [0,1,2,3,4] (That is to say "return one of 5 different results, starting from 0")

Most people's notion of d(x) is "a number between 1 and x" so really random2(1+x) isn't "d(x)", but it's not "d(x+1)" either, either one is an approximation that's probably good enough for most people to use for estimating. As shorthand, they're ok, but both are off by about the same amount in the opposite direction, and saying d(invo) is shorter than saying d(info+1), so if I'm going to pick an incorrect approximated shorthand, I'm going to pick the shorter shorthand :)

tl;dr: Mak's invocations are spellpower-limited, I hadn't read carefully enough, and invo+2d(invo) is an approximation which is off by the same amount as invo+2d(invo+1) which is also an approximation, both are pretty close.
Spoiler: show
This high quality signature has been hidden for your protection. To unlock it's secret, send 3 easy payments of $9.99 to me, by way of your nearest theta band or ley line. Complete your transmission by midnight tonight for a special free gift!

For this message the author Siegurt has received thanks: 2
Majang, VeryAngryFelid
User avatar

Crypt Cleanser

Posts: 732

Joined: Monday, 24th April 2017, 11:46

Post Monday, 10th July 2017, 17:26

Re: Unarmed troll: how to survive without luck?

Thanks for clarifying this. Zu früh gefreut!
Maɟaŋ
User avatar

Tartarus Sorceror

Posts: 1698

Joined: Saturday, 18th June 2016, 13:57

Post Tuesday, 11th July 2017, 16:58

Re: Unarmed troll: how to survive without luck?

Thinking about my recent experience:

I started out with a troll artificer. Dropped the short sword immediately. Then got stuck with a -5 ring of slaying. Collected large rocks. Used the wand of domination when I had to run downstairs because of uniques I couldn't take on, and got attacked by ants. It's very good against insects&co. Fire wand is very good against ice beasts. I hated randomness, however -- but it's clear that it's dangerous, although it can help you take out early ogres and double ogres.
Went to Oka to get some advantages, kept training UC, throwing, and evo. However, in the beginning, when I was nerfed, I had to use rings of stealth and train a couple points of stealth. As I result, I later had ++ stealth, which isn't too bad (I actually was surprised about that).
Oka throws a lot of stuff at you. You can expect him to gift you armour and shield -- sooner or later. How soon, that's the problem. You can expect to have enough piety by the end of Lair. In the meantime, he'll have been feeding you throwables.
You probably want to be around lvl 11 before you enter L.
Now, there are two main problems in Lair if you are an unarmed troll: lack of corridors and choppable hydras. Corridors can be created if you have digging, otherwise you want to carefully explore around and look for chokepoints to backtrack to. Remember that elephants can push you around, so try to have some escape or repositioning plan. In case things go bad, you can always run away. About hydras, that's a larger problem. You can use large rocks, evocables, or, if you are with Makleb, summons (and minor destruction). Maybe you trained hexes a bit and can confuse them before they come at you. If you are with Lugonu, you can banish them.
Invisibility is your greatest friend. Most lair monsters don't see invisible creatures. This is amazing against hydras. Remember, however, that standing in water, being on fire, corona, contamination, and likely something else will make you visible.
If you want a shield, there are two-three places where it's more likely to find one. These are the snake pit, the vaults, and the lower levels of the dungeon, where tengus sometimes use them. They are dangerous enemies, however.
Poison resistance isn't as important as for other species because of your fast regeneration, but it still helps. "See invisibility" is important, because you are lightly armoured, and so unseen horrors hit harder.
Cloud immunity is very useful against catoblepas, who can see invisible.
Amulets of reflection are good against projectiles you would have little chance to dodge. They also raise your SH in general.
The last thing I can think of right now is: know when to teleport, and don't take excessive risks. If things look harder than usual, use god powers or buffs, or take a chance to run away. Stairdancing is your friend.
You can also simply try to avoid hydras, but it's not easy and it depends on how many they are and if you are already running away from something.
I Feel the Need--the Need for Beer
Spoiler: show
3DSBeTr 15DSFiRu 3DSMoNe 3FoHuGo 3TrArOk 3HOFEVe 3MfGlOk 4GrEEVe 3BaIEChei 3HuMoOka 3MiWnQaz 3VSFiAsh 3DrTmMakh 3DSCKXom 3OgMoOka 3NaFiOka 3FoFiOka 3MuFEVeh 3CeHuOka 3TrMoTSO 3DEFESif 3DSMoOka 3DSFiOka

Tomb Titivator

Posts: 771

Joined: Tuesday, 25th November 2014, 02:47

Post Wednesday, 12th July 2017, 01:18

Re: Unarmed troll: how to survive without luck?

zapping is in fact in beam.cc and calls zappy about halfway through. (code In the spoiler at the bottom, it is rather irrelevant)

Siegurt's tr;dr is accurate: Mak's invocations are zap power-limited. See Zap Power Approx for more details below on Invo+2dInvo.

Edge's TLDR for this post is as follows:
Edge what do these numbers mean?
Going from Invocations 0 to 10 increases the damage of minor destruction by ~57%.
Over 50% of the damage at all stages comes from stone arrow and breathe Acid, both of which scale well compared to the other three spells and are never or rarely resisted. This is significant at 0 invocations and they become even more dominant at higher invocations.
It doesn't really matter that shock is spellpower capped, shock represents <10% of Minor Destruction's Damage output. This is under an assumption that double zaps aren't a thing.
I never factored accuracy of spells into consideration.


The rest of this post is to give some clarification on how the power caps affect Minor Destruction.

The Shock Diminishing Returns
At Invocations 1-8, both minimum zap power and maximum zap power increase. Minimum = Invocations and Maximum = invocations x 3.
At Invocations 9-25, Minimum zap power is increased and maximum zap power gets rolled increasing frequency. Whenever invocations + 2d(invocations) - 2 >25, zap power is set to 25.
At Invocations >= 24, Zap always rolls max power.

Please note, roughly half of the gains average in zap power from 9 to 25 occur in the first 4 levels. That is the average zap power gain from invocations 9-13 is the same as the gain from 14-25. The difference in power gain going from 7 to 8 and going from 8 and 9 are vanishingly small (that is going from 8 to 9 is worth very slightly less than 7-8 but is barely noticable), and similarly the gain from 24-25 is also trivial, making going from invocations 24-25 to negligible practical purposes in the case of shock.

Throw Flame and Stone Arrow
At Invocations 1-16, both minimum zap power and maximum zap power increase.
At Invocations 17-50, minumum zap power increases and maximum zap power gets rolled with increasing frequency. Whenever invocations + 2d(invocations) - 2 >50, zap power is set to 50. Note half of the gains in average zap power from 17-50 would be clustered in levels 17-25.
At invocations 50+, Minor destruction will always roll max power on throw flame and stone arrow.

Breathe Acid and Pain
Breath Acid and Pain will increase minimum zap power and maximum zap power for all invocations 1-27.

From Power To Damage


Zap Damage Formulas.

This damage formula: return dice_def(numdice, adder + pow * mult_num / mult_denom); (code below in a spoiler)

Shock
new dicedef_calculator<1, 3, 1, 4>,
numdice=1, adder =3, pow = power, mult_num=1, and_mult denom=4
Thus, shock has a damage of 1d(3+pow*1/4) Max 1d9 @ Power 24
This is expected and matches the values on learnDB and the wiki.

Stone Arrow
new dicedef_calculator<3, 5, 1, 8>,
numdice=3, adder = 5, pow = power, mult_num=1, mult_denom=8
Thus, stone arrow has damage of 3d(5+pow*1/8) Max 3d11 @ power 48.

Throw Flame
new dicedef_calculator<2, 4, 1, 10>,
numdice=2, adder = 4, pow = power, mult_num=1, mult_denom=10
2d(4+pow*1/10). Max 2d9 @ power 50.

Breathe Acid
new dicedef_calculator<3, 4, 1, 3>,
numdice=3, adder = 4, pow = power, mult_num=1, mult_denom=3
3d(4+pow*1/3) Max 3d20 @ power 50, (note, pow here is pow/2 from Minor Destruction, for our purposes pow cannot be greater than 40.5 @ invocations 27, and caps at 3d17)

Pain
new dicedef_calculator<1, 4, 1, 5>,
numdice=1, adder = 4, pow = power, mult_num=1, mult_denom=5
1d(4+pow*1/5), max 1d(24) at power 100, (1d19 at invocations 27, also zap_pain is capped at power 100 while spell pain is capped at power 25)

Minor Destruction average damage at Various Invocation Levels

Invocations 0
Pow = 0
Shock 1d3 Ave 2
Pain 1d4 Ave 2.5
Throw Flame 2d4 Ave 5
Stone Arrow 3d5 Ave 9
Breathe Acid 3d3 Ave 6

Average Damage Minor Destrucion= 4.9

Invocation 4
Ave Power 8, 4 for Breathe Acid
Shock 1d5 Ave 3
Pain 1d6.6 --? 1d6 Ave 3.5
Throw Flame 2d4.8 --> 2d4 Ave 5
Stone Arrow 3d6 Ave 10.5
Breathe Acid 3d4.33 --? 3d4 Ave 7.5

Average Damage Minor Destruction= 5.9

Invocation 10
Ave Power 20, 10 for Breathe Acid
Shock 1d8 Ave ~= 4.5 (Slightly less due to power cap)
Pain 1d8 Ave 4.5
Throw Flame 2d6 Ave 7
Stone Arrow 3d7.5 --? 3d7 Ave 12
Breathe Acid 3d6.33 --? 3d6 Ave 10.5

Average Damage Minor Destruction~=38.5/5=7.7

Edge what do these numbers mean?
Going from Invocations 0 to 10 increases the damage of minor destruction by ~57%.
Over 50% of the damage at all stages comes from stone arrow and breathe Acid, both of which scale well compared to the other three spells and are never or rarely resisted. This is significant at 0 invocations and they become even more dominant at higher invocations.
It doesn't really matter that shock is spellpower capped, shock represents <10% of Minor Destruction's Damage output. This is under an assumption that double zaps aren't a thing.
I never factored accuracy of spells into consideration.
I don't know if 3d7.5 yields 3d7 or 3d8 or some combination of the two.

Junk Math
Average damage per point in invocations (I did these calculations before I did average damage at given invocation levels, which I could draw better conclusions from)
Spoiler: show
Max Damage
Shock gains 1/4 damage per zap power
Throw Flame gains 2/10=1/5 damage per zap power
Stone Arrow gains 3/8 damage per zap power
Pain gains 1/5 damage per zap power
Breath Acid gains 3/3=1 zap per power gains 3/6=1/2 zap per power/2

Max Damage per power= (1/4+1/5+3/8+1/5+1/2)/5. = .305
Average Damage per power= .1525

Average Power Per Level in invocations from 1-->8=2

Average damage increase per point in Invocations from 1-->8 .305 average damage
Average damage from 12-->13 ~= .28

Efficency loss = 8.2%



Other Stuff (Less relevant)
Spell Damage Info
Code for damage calculation and zap_info template, and zap data for Shock, Pain, Throw Flame, Stone Arrow, Breathe Acid, as well as Fireball and Iron Shot are under the spoiler. Something is going on with the damage for Iron Shot and Fireball and I don't understand what it is. However, it makes me question the above findings.
Spoiler: show
  Code:
class power_deducer
{
public:
    virtual T operator()(int pow) const = 0;
    virtual ~power_deducer() {}
};

typedef power_deducer<int> tohit_deducer;

template<int adder, int mult_num = 0, int mult_denom = 1>
class tohit_calculator : public tohit_deducer
{
public:
    int operator()(int pow) const override
    {
        return adder + pow * mult_num / mult_denom;
    }
};

typedef power_deducer<dice_def> dam_deducer;

template<int numdice, int adder, int mult_num, int mult_denom>
class dicedef_calculator : public dam_deducer
{
public:
    dice_def operator()(int pow) const override
    {
        return dice_def(numdice, adder + pow * mult_num / mult_denom);
    }
};

/*
struct zap_info
{
    zap_type ztype;
    const char* name;
    int player_power_cap;
    dam_deducer* player_damage;
    tohit_deducer* player_tohit;       // Enchantments have power modifier here
    dam_deducer* monster_damage;
    tohit_deducer* monster_tohit;      // Enchantments have power modifier here
    colour_t colour;
    bool is_enchantment;
    beam_type flavour;
    dungeon_char_type glyph;
    bool always_obvious;
    bool can_beam;
    bool is_explosion;
    int hit_loudness;
}
*/

{
    ZAP_SHOCK,
    "zap",
    25,
    new dicedef_calculator<1, 3, 1, 4>,
    new tohit_calculator<8, 1, 7>,
    new dicedef_calculator<1, 8, 1, 20>,
    new tohit_calculator<17, 1, 20>,
    LIGHTCYAN,
    false,
    BEAM_ELECTRICITY,             // beams & reflects
    DCHAR_FIRED_ZAP,
    true,
    true,
    false,
    1 // XXX: maybe electricity should be louder?
},

{
    ZAP_THROW_FLAME,
    "puff of flame",
    50,
    new dicedef_calculator<2, 4, 1, 10>,
    new tohit_calculator<8, 1, 10>,
    new dicedef_calculator<3, 5, 1, 40>,
    new tohit_calculator<25, 1, 40>,
    RED,
    false,
    BEAM_FIRE,
    DCHAR_FIRED_ZAP,
    true,
    false,
    false,
    2
},

{
    ZAP_STONE_ARROW,
    "stone arrow",
    50,
    new dicedef_calculator<3, 5, 1, 8>,
    new tohit_calculator<8, 1, 10>,
    new dicedef_calculator<3, 5, 1, 10>,
    new tohit_calculator<14, 1, 35>,
    LIGHTGREY,
    false,
    BEAM_MMISSILE,
    DCHAR_FIRED_MISSILE,
    true,
    false,
    false,
    3
},

{
    ZAP_BREATHE_ACID,
    "glob of acid",
    50,
    new dicedef_calculator<3, 4, 1, 3>,
    new tohit_calculator<7, 1, 6>,
    nullptr,
    nullptr,
    YELLOW,
    false,
    BEAM_ACID,
    DCHAR_FIRED_ZAP,
    true,
    false,
    false,
    6
},

{
    ZAP_PAIN,
    "",
    100,
    new dicedef_calculator<1, 4, 1, 5>,
    new tohit_calculator<0, 7, 2>,
    new dicedef_calculator<1, 7, 1, 20>,
    new tohit_calculator<0, 1, 3>,
    BLACK,
    true,
    BEAM_PAIN,
    NUM_DCHAR_TYPES,
    false,
    false,
    false,
    1 // XXX: Should this be soundless?
},

{
    ZAP_FIREBALL,
    "fireball",
    200,
    new calcdice_calculator<3, 10, 1, 2>,
    new tohit_calculator<40>,
    new dicedef_calculator<3, 7, 1, 10>,
    new tohit_calculator<40>,
    RED,
    false,
    BEAM_FIRE,
    DCHAR_FIRED_ZAP,
    false,
    false,
    true,
    0 // Noise comes from explosion
},

{
    ZAP_IRON_SHOT,
    "iron shot",
    200,
    new calcdice_calculator<9, 15, 3, 4>,
    new tohit_calculator<7, 1, 15>,
    new dicedef_calculator<3, 8, 1, 9>,
    new tohit_calculator<20, 1, 25>,
    LIGHTCYAN,
    false,
    BEAM_MMISSILE,
    DCHAR_FIRED_MISSILE,
    true,
    false,
    false,
    6
},




Something unexpected is happening with other zaps:

Fireball
new calcdice_calculator<3, 10, 1, 2>,
numdice=3, adder = 10, pow = power, mult_num=1, mult_denom=2
3d(10+pow*1/2) --> up to 3d110 at max spellpower... wait this is wrong I expected it to be 3d(3.33+pow/6) lets check something else...


Iron Shot
new calcdice_calculator<9, 15, 3, 4>,
numdice=9, adder = 15, pow = power, mult_num=3, mult_denom=4
Thus, Iron Shot has a damage of 9d(15+power*3/4) --> up do 9d165 at max spellpower...
This is unexpected, seems wrong, and does not match the wiki of 9d(1.66+power/12) capping at 9d19. At some point in time in iron shot, 15+power*3/4 gets divided by 9 to give 9d(1.55+power/12). A similar thing happens with fireball I'm clearly missing something here. This thing does not appear to be happening with stone arrow or throw flame.


Zap Power Approx
Spoiler: show
Zap Power Formula For Minor Destruction
Case A: Invocations+2d(Invocations+1)-2 is an accurate representation of zap power.
Case B: Invocations + 2d(Invocations) is a slightly better and easier to understand metric than Case C.
Case C: Invocations + 2d(Invocations+1) This is the least accurate, and leads you to believe diminishing returns begin at a invocation level 1 less than they actually are.

Given Zap power 5
Case A: 5+2d(6)-2 Will give a min of 5 + 0= 5 and a Max of 5+10=15: This is accurate.
Case B: 5+2d(5) Will give a min of 7 and a max of 15, This is 2 high for min but accurate for max
Case C: 5+2d(6) Will give a min of 5 7 and a max of 17, both of which are 2 off: This is 2 high in both min and max.

Case A is accurate, case B is conceptually the most straightforward and valuable and accurate enough.


code for zapping()
Spoiler: show
  Code:
spret_type zapping(zap_type ztype, int power, bolt &pbolt,
                   bool needs_tracer, const char* msg, bool fail)
{
    dprf(DIAG_BEAM, "zapping: power=%d", power);

    pbolt.thrower = KILL_YOU_MISSILE;

    // Check whether tracer goes through friendlies.
    // NOTE: Whenever zapping() is called with a randomised value for power
    // (or effect), player_tracer should be called directly with the highest
    // power possible respecting current skill, experience level, etc.
    if (needs_tracer && !player_tracer(ztype, power, pbolt))
        return SPRET_ABORT;

    fail_check();
    // Fill in the bolt structure.
    zappy(ztype, power, false, pbolt);

    if (msg)
        mpr(msg);

    if (ztype == ZAP_LIGHTNING_BOLT)
    {
        noisy(spell_effect_noise(SPELL_LIGHTNING_BOLT),
               you.pos(), "You hear a mighty clap of thunder!");
        pbolt.heard = true;
    }

    if (ztype == ZAP_DIG)
        pbolt.aimed_at_spot = false;

    pbolt.fire();

    return SPRET_SUCCESS;
}

For this message the author edgefigaro has received thanks: 3
Majang, pedritolo, VeryAngryFelid

Shoals Surfer

Posts: 325

Joined: Tuesday, 13th October 2015, 06:02

Post Wednesday, 16th August 2017, 18:24

Re: Unarmed troll: how to survive without luck?

Being a successful melee troll mostly means to focus on UC or your weapon choice, while putting all stat points into STR. If you have Trog it gets easier, and using Berserk often is quite efficient. Just remember to only berserk when monsters are adjacent.

If you find a wand that can disable monsters from attacking, it's worth getting enough skill points in evocables and to use a id scroll on the wand.

Thowing is also nice. As a secondary attack you want skill points to help with accuracy.
User avatar

Pandemonium Purger

Posts: 1386

Joined: Sunday, 5th April 2015, 22:37

Post Wednesday, 16th August 2017, 22:35

Re: Unarmed troll: how to survive without luck?

I think you are supposed to put points into dex with a melee troll.
http://crawl.akrasiac.org/scoring/playe ... speon.html. I started playing in 0.16.1
I achieved greatplayer in less than a year.
Remove food

Tomb Titivator

Posts: 771

Joined: Tuesday, 25th November 2014, 02:47

Post Wednesday, 16th August 2017, 22:48

Re: Unarmed troll: how to survive without luck?

I don't know if str vs dex matters. They are equally weighted for normal shields and str is better for large shields. Trolls are large and have a very hard time getting points of EV, and eventually you are going to want to be in heavy armor anyway.

Dex gives accuracy. Str gives damage. I've always been under the impression that the latter is more important. Maybe troll dex is so low and str is so high that dex wins out, but accuracy comes naturally from XL.

Fighting early, shields as soon as it is relevant, and armor later are the skills I like to train for troll defenses.

Also holy crap thread necro wtf.
User avatar

Pandemonium Purger

Posts: 1386

Joined: Sunday, 5th April 2015, 22:37

Post Thursday, 17th August 2017, 00:53

Re: Unarmed troll: how to survive without luck?

You can still get reasonable EV with a troll, and you are likely going to be in a robe or TLA or light DS like acid for a while. You also have so much Str you probably don't need more.
http://crawl.akrasiac.org/scoring/playe ... speon.html. I started playing in 0.16.1
I achieved greatplayer in less than a year.
Remove food

For this message the author WingedEspeon has received thanks:
ichbins

Ziggurat Zagger

Posts: 8786

Joined: Sunday, 5th May 2013, 08:25

Post Thursday, 17th August 2017, 01:09

Re: Unarmed troll: how to survive without luck?

XL doesn't affect accuracy unless you are in certain forms

Tomb Titivator

Posts: 771

Joined: Tuesday, 25th November 2014, 02:47

Post Thursday, 17th August 2017, 01:31

Re: Unarmed troll: how to survive without luck?

I was never able to get reasonable returns on EV as a troll going dex + dodge, even when doing 2 runes + vaults 1-4 in a +2 robe. That game a few versions ago.

Dex for accuracy I can understand, and I (clearly) don't understand the accuracy mechanics and how it relates to damage well enough to comment. It will be a rare game where I think " a single xp in dodge for this troll is better than more points in fighting or armour or shields"

Ziggurat Zagger

Posts: 8786

Joined: Sunday, 5th May 2013, 08:25

Post Thursday, 17th August 2017, 03:45

Re: Unarmed troll: how to survive without luck?

Dex's effect on accuracy is completely insignificant

Abyss Ambulator

Posts: 1233

Joined: Wednesday, 23rd April 2014, 21:57

Post Thursday, 17th August 2017, 07:54

Re: Unarmed troll: how to survive without luck?

edgefigaro wrote:I was never able to get reasonable returns on EV as a troll going dex + dodge, even when doing 2 runes + vaults 1-4 in a +2 robe. That game a few versions ago.

I guess it partly depends on what you consider "reasonable". The difference even between 10 and 15 EV is definitely worth having.

Tr already get so much Str, that putting more points in usually feels wasted. I'd say Int or Dex, to heop with getting some (very) low level spells and/or an extra few points of EV.

For this message the author 4Hooves2Appendages has received thanks:
ichbins

Ziggurat Zagger

Posts: 8786

Joined: Sunday, 5th May 2013, 08:25

Post Thursday, 17th August 2017, 08:12

Re: Unarmed troll: how to survive without luck?

If your reason for raising Str is more damage, then "Tr already gets so much Str" doesn't really apply. Strength after 10 provides a linear increase in damage.
I personally think raising str is much better than raising dex on trolls, but I value damage more and EV less than most players.

For this message the author duvessa has received thanks: 2
Gigaslurp, VeryAngryFelid
Next

Return to Dungeon Crawling Advice

Who is online

Users browsing this forum: No registered users and 12 guests

cron
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group.
Designed by ST Software for PTF.