Leo Littlebook's thread


If it doesn't fit anywhere else, it belongs here. Also, come here if you just need to get hammered.

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Wednesday, 24th April 2019, 14:21

Leo Littlebook's thread

Here seems like a reasonable place to chronicle my progress. Feedback is useful, after all!

I have been urged to join the coming CSDC. Very unlikely, barring severe illness.

My goal in playing DCSS is to crack the meta, then demo that during the launch of my PIM.

I chose DCSS because I needed a virtual life to fully expose, so I could demonstrate how to use a PIM for oneself. DCSS is totally like real life.

In the demo, I'll be showing how to use a PIM to be a better player. And also how to put together and publish my thoughts on the game.

I played through the early game a bunch of times to figure out best combos. Now I'm taking my first trip through the full game with a HOFi Beogh -> good gods named "Hiewishru".

I'm not using my PIM to improve my play right now. Just playing during downtime, and taking notes on all the errors I make.

I didn't realize when I started what a high win-rate is achievable for elite DCSS players. Although the game is notoriously difficult and forbidding for beginniners, in that sense the game is easy.

I want to demo something intricate, demanding and cerebral, so I'll probably play much more slowly than is optimal for collecting achievements of the kind that elites currently use to distinguish themselves.

However, I'm comfortable with a slower approach anyway. Humans are naturally bad at probability estimation and decisionmaking thereon. DCSS is a good way to train that, but not if you treat your virtual life cavalierly for the sake of grinding wins.

In short, I like to play immersively and treat the avatar as a real person.

A character whose approach I admire is Prudence in
Deadly Assessments
by Drew Hayes

There is a time to take a risk, in real and virtual life, and I would like my virtual lives to train me for the real thing - neither cowardly nor reckless, but courageous.

DCSS is not exactly what I expected, but I think there will be plenty of material to make an interesting demo.

I'm working out some international shipping kinks and will start recording once I have my bulk backup system in place, maybe sooner if I jury-rig things.

While I'm confident in my testing results on optimal combos, I still have plenty to learn about the game, and will be going about that methodically. Learning management is a big part of the demo, so I won't get too heavy into that yet. This is the semi-passive exposure phase.

I have not yet played online. I seem to recall some UI issues with the browser version, and certainly lag is obnoxious. Plus, it is advantageous to collect lots of scoffing to prove later that my concepts were indeed heretical at the time.

Anyhow, on with Hiewishru. He's entered Depths:1, which I don't recall being in before. Exciting times!

Hiewishru:
https://pastebin.com/7EP0Qfbd
Last edited by Leo_Littlebook on Wednesday, 24th April 2019, 16:04, edited 1 time in total.
Fork you.

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Wednesday, 24th April 2019, 16:03

Re: Leo Littlebook's thread

Enough about me, let's chum the waters.

Best troll: TrCK -> Jiyva, maybe good gods for extended.

Why is TrCK > TrBe? Because Trolls are squishy and regen kite. Xom likes high-tension chases; Trogg likes toe-to-toe deathmatches.
Fork you.

Ziggurat Zagger

Posts: 6454

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

Post Wednesday, 24th April 2019, 17:25

Re: Leo Littlebook's thread

Leo_Littlebook wrote:Enough about me, let's chum the waters.

Best troll: TrCK -> Jiyva, maybe good gods for extended.

Why is TrCK > TrBe? Because Trolls are squishy and regen kite. Xom likes high-tension chases; Trogg likes toe-to-toe deathmatches.

I find TrFi, and TrHu better than either TrCK and TrBe, even when you do stuff Xom likes, he can still kill you, and either large rocks or a shield is a more useful starting tool than "some dude who might help you and might kill you, at some random intervals"

TrAM and TrAs aren't bad either, but you don't get as much right off the bat that's useful as you do with TrHu, the spells from TrAM won't be useful for a while, and while needles don't suck and throwing is troll's best ranged option, a little dex+needles isn't superior to either large rocks or a shield. TrBe are actually pretty close with TrFi/TrHu, Trog is super powerful, but trolls don't really *need* any extra power at the start. I think starting as a TrFi/TrHu and then worshipping trog at the first altar is slightly better than starting as a TrBe.

Really the reason to go with CK as a troll is "he can probably survive the bad crap Xom does, because he's a troll, and trolls are super powerful, and CK are fun, if suboptimal"

There's no real good reason to go with Jiyva unless you encounter an early altar, you'll get just as much benefit from any number of other gods, and get them way earlier, again Jivya is a "that's a fun and different thing to do, which trolls can probably manage because they're so naturally powerful" not a "that's the optimal thing to do" Also Going CK->Jivya means enduring Xom wrath (Which is only slightly more unpleasant than actual Xom worship, meaning again, that "Trolls can probably survive it")

If your definition of "Best" is "This is what I find the most fun" rather than "This gives you the best chances of winning" then your suggestion of TrCK->Jivya is quite possibly true, but also is subjective :)
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:
Leo_Littlebook

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Wednesday, 24th April 2019, 17:37

Re: Leo Littlebook's thread

Good comment, glad I delayed posting here as long as possible.

I'm agnostic as to whether DCSS .23 is balanced the way it should be. You could be right. However, my operating hypothesis will be that the current meta manages Xom poorly, undervaluing his benefit for a Troll, and I will endeavor to prove it via video. The trick will be manipulating tension and ?other Xom mechanics, I forget.

The problem might well be that the very high elite win-rate just punishes Xom's randomness, with the benefits being surplus to survival to mid-game, when options open up and deaths become more skill-related.
Fork you.

Ziggurat Zagger

Posts: 6454

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

Post Wednesday, 24th April 2019, 17:43

Re: Leo Littlebook's thread

Leo_Littlebook wrote:Good comment, glad I delayed posting here as long as possible.

I'm agnostic as to whether DCSS .23 is balanced the way it should be. You could be right. However, my operating hypothesis will be that the current meta manages Xom poorly, undervaluing his benefit for a Troll, and I will endeavor to prove it via video. The trick will be manipulating tension and ?other Xom mechanics, I forget.

The problem might well be that the very high elite win-rate just punishes Xom's randomness, with the benefits being surplus to survival to mid-game, when options open up and deaths become more skill-related.

Well if your theory is "Xom is not as bad as people make him out to be" rather than "Xom is strictly optimal" then I would certainly agree, with any fairly robust race, his bad stuff is fairly survivable, and his good stuff outweighs the bad. I had interpreted your position as "TrCK is strictly better than any other options" rather than "TrCK is pretty decent"
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!

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Wednesday, 24th April 2019, 17:53

Re: Leo Littlebook's thread

Nope I'm saying TrCK is optimal, but all hypotheses are subject to proof!
Fork you.

Ziggurat Zagger

Posts: 5382

Joined: Friday, 25th November 2011, 07:36

Post Wednesday, 24th April 2019, 18:24

Re: Leo Littlebook's thread

Leo_Littlebook wrote:Nope I'm saying TrCK is optimal, but all hypotheses are subject to proof!

Your statement here is still too vague in order to actually be testable. By optimal, do you mean in terms of winrate? Winrate for new players, or experienced players? New players can boost their winrate by playing characters with high "power ceilings", ie, characters that have a chance of becoming very, very strong. Experienced players boost their winrates by playing characters with high "power floors", ie, that will always achieve at least some reasonable level of power.

Xom has a high power ceiling but his power floor is actually negative; he can act against you. Once you're experienced enough to have a good winrate with "average" characters, Xom is no longer going to help you.

That said, I am not a Xom expert and actively avoid playing it, so this is based off a more general crawl understanding.

The same framework also relates to Cheibriados, my favorite god. Very high power ceiling, but has drawbacks early on which results in a very dangerous period with a low power floor.

Also you use PIM without ever defining it; what does that mean?

For this message the author tasonir has received thanks:
Leo_Littlebook

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Wednesday, 24th April 2019, 18:46

Re: Leo Littlebook's thread

> Your statement here is still too vague in order to actually be testable.

You make some great points. However they are somewhat orthogonal to my methodology. I looked for "best fits" assuming that all species, gods and backgrounds would be used. Hence "true Troll" is TrCK -> Jiyva.

Whether DCSS .23 is balanced to achieve that, and for which game modes and player types, are potentially difficult questions to conclusively answer.

By optimal, I always refer to win-rate, given my philosophy.

I'm not sure I care much about what's optimal for either new players or current elite players. I'm more interested in what's optimal for me, and what's optimal for a theoretically perfect player. Because of the incentives for fast play, I doubt even elite players play close to theoretical perfection. This systematically influences the combos they prefer.

> The same framework also relates to Cheibriados, my favorite god. Very high power ceiling, but has drawbacks early on which results in a very dangerous period with a low power floor.

NaVM Chei is the answer!

> Also you use PIM without ever defining it; what does that mean?

In this case, Personal Information Manager.
Fork you.

Ziggurat Zagger

Posts: 8786

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

Post Wednesday, 24th April 2019, 19:07

Re: Leo Littlebook's thread

As much as I admire the dedication to trolling here, you really need to get some new material.

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Wednesday, 24th April 2019, 19:18

Re: Leo Littlebook's thread

I reckon Troll can't go good gods in extended due to hunger. According to wiki, Slimify only works on natural holiness. This means he's locked out of non-undead extended.

If the only Jiyva altar is on Slime 5, TrCK will have a hard time reaching it. Hafta save up rations for a dive, or else continue as Xom.

That just makes the combo more Trollish. Win-rate isn't everything. There's also /Aesthetic/.
Fork you.

Ziggurat Zagger

Posts: 6454

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

Post Wednesday, 24th April 2019, 20:06

Re: Leo Littlebook's thread

Leo_Littlebook wrote:I looked for "best fits" assuming that all species, gods and backgrounds would be used. Hence "true Troll" is TrCK -> Jiyva.

Leo_Littlebook wrote:By optimal, I always refer to win-rate, given my philosophy.


These two statements seem to be in conflict, if optimal for winrate doesn't exhaustively use all species/gods/backgrounds. For example Troll might be optimal for CK, but CK might not be optimal for Troll, or any other race.

If there is no race for which CK is the optimal background, then how do you reconcile these (Similarly, there may be no background for which Mu is optimal, but there's certainly a background which is optimal for Mu)

It may be that your actually optimizing for "least number of combos which produce a win in all species/gods/backgrounds" if that's the case, that's a more complex and broader conversation than talking about Trolls, Chaos Knights, or the TrCK combo specifically.
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!

Tartarus Sorceror

Posts: 1739

Joined: Tuesday, 13th March 2012, 02:48

Post Wednesday, 24th April 2019, 20:46

Re: Leo Littlebook's thread

duvessa wrote:As much as I admire the dedication to trolling here, you really need to get some new material.

Huh, I assumed this guy was your alt.
User avatar

Shoals Surfer

Posts: 287

Joined: Friday, 19th August 2016, 21:21

Post Wednesday, 24th April 2019, 21:48

Re: Leo Littlebook's thread

duvessa wrote:As much as I admire the dedication to trolling here, you really need to get some new material.


At 1st I actually thought it was some sort of Markov text generator prank
make food great again

For this message the author pedritolo has received thanks:
byrel

Dungeon Dilettante

Posts: 2

Joined: Thursday, 25th April 2019, 00:51

Post Thursday, 25th April 2019, 00:53

Re: Leo Littlebook's thread

Hi leo,

Do you have a complete list of all of your optimal builds?

Thanks

For this message the author trogdor has received thanks: 2
duvessa, Rast

Lair Larrikin

Posts: 21

Joined: Monday, 15th October 2018, 01:17

Post Thursday, 25th April 2019, 01:08

Re: Leo Littlebook's thread

duvessa wrote:As much as I admire the dedication to trolling here, you really need to get some new material.


I think it's extremely rude and in very poor taste to accuse this person of trolling, perhaps even stretching the boundaries of the rules of this forum (not that I expect those boundaries to ever be enforced on you in particular). It's pretty clear - at least to me - that Leo is genuine in his ideas and beliefs about the game even if I disagree with most of them. Frankly I think that it is shameful to try to exclude/ostracize/bully people in this way and I wish you'd stop.

For this message the author MalcolmRose has received thanks:
Leo_Littlebook

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Thursday, 25th April 2019, 03:05

Re: Leo Littlebook's thread

> These two statements seem to be in conflict, if optimal for winrate doesn't exhaustively use all species/gods/backgrounds. For example Troll might be optimal for CK, but CK might not be optimal for Troll, or any other race.

Sure. TrCK is one of the combos for which I'm least confident that the "true combo" is the same as the optimal combo. Of course, "optimal" itself is ambiguous and I could try to weasel on those grounds, but why bother? To rectify the names, call them "true combos". They were selected with a strong regard for viability, but that was not the exclusive consideration.

I am always talking about win-rate when I say "optimal" because I don't care about other game types such as turncount, realtime, etc. Although I have a higher interest in Extended than is perhaps optimal for win-rate, since I seek additional difficulty.

> It may be that your actually optimizing for "least number of combos which produce a win in all species/gods/backgrounds"

Not exactly. I didn't repeat any species. I was looking for the essence of each species. Hence the "true combo".

> Do you have a complete list of all of your optimal builds?

There may be some minor flaws here because I haven't compiled my notes. Question marks indicate that I haven't tested Extended at all.

HOFi Beogh -> good gods from Abyss
MiBe Trogg -> good gods extended
MfIE Fedhas -> good gods
GrEE Ely, good gods
DrSu Hep -> ?good gods
HaHu Zin, good gods
TrCK Xom -> Jiyva
GhFi Yred -> ?kiku
HuGl TSO, good gods
KoCj Sif
DsNe Nemelex
CeMo Qaz -> good gods
SpAM Goz
TeAE good gods -> Veh
DEWi Sif
OgAK
GnWa Oka -> good gods
VSTm Wu Jian -> ?good gods
VpEn Dith -> Makleb
DgSk
FoEE good gods -> Usk -> ?good gods
NaVM Chei
OpVM good gods -> Ash -> good gods
FeFE Sif
BaWp good gods -> Ru
MuNe Kiku -> Goz

I appreciate the support, Malcolm, but welcome the heat. Duvessa is just upset because I ate her brother. Maybe I will find some new material in that book he was carrying, the pages I haven't used for toilet paper anyway.
Fork you.

For this message the author Leo_Littlebook has received thanks:
MalcolmRose

Spider Stomper

Posts: 242

Joined: Friday, 17th April 2015, 16:22

Post Thursday, 25th April 2019, 03:52

Re: Leo Littlebook's thread

So looking at this, it seems like you're less interested in finding optimal backgrounds for a given race (and vice-versa) and more in creating a mapping that includes every race and background with as few duplications as possible?

There's a lot of combos that, if one were to start with just the race or the background, are very poor as first picks; CK on anything, Deep Elf Wizard, Naga Venom Mage, and Demigod Skald to say a few.

I notice Deep Dwarf is missing from the list. What are your thoughts on it?

For this message the author Doesnt has received thanks:
MalcolmRose

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Thursday, 25th April 2019, 03:57

Re: Leo Littlebook's thread

This iteration of the list was initially composed as a list of potential heresies. I forgot to re-include DD:

DDAr Makleb

> There's a lot of combos that, if one were to start with just the race or the background, are very poor as first picks; CK on anything, Deep Elf Wizard, Naga Venom Mage, and Demigod Skald to say a few.

I disagree that they are poor, much less very poor, and will endeavor to demonstrate why in my vids. For one thing, I strongly dissent from the killdudes early game doctrine. This changes which combos are best.
Fork you.

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Thursday, 25th April 2019, 04:35

Re: Leo Littlebook's thread

Another way of looking at my list:

I have described it as "the essence of each species." It is also the essence of DCSS. Due to the "use every species/god/background" rule, this is a list of maximally differentiated playstyles. The current fast-play meta is biased towards an under-differentiated streamlined playstyle, which narrows one's game experience and hence enjoyment. This is due to some combination of two factors:

1. early game is too easy for elites, allowing suboptimal play
2. game balance doesn't sufficiently incentivize playstyle differentiation; some styles dominate
Fork you.

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Thursday, 25th April 2019, 05:22

Re: Leo Littlebook's thread

The Orb is a semi-sentient largely-instinctual non-human dungeon core that affects the ambient state of the Dungeon and the branches under its influence.

Regen and appetite are accelerated within its influence, due to ambient mana. Beings without immunity to magical enslavement are rapidly brought under its control, although creatures of great will are merely influenced. Slain monsters often drop no chunks because the dungeon reabsorbs its thralls. Souls of the slain within the Orb's domain are captured, converted into essence or XP, and absorbed. Shades of strong-willed adventurers, who typically receive enchantments to resist the Orb's influence, are confined in vaults and gradually digested.

A wizard built a tower near the Dungeon's entrance and began experimenting on his slave pen to develop a method of creating immunity to dungeon core influence. The protagonist was the successful result of such an experiment. As a result, he gained immunity to all magical enslavement, which enabled him to escape, thus avoiding astral dissection.

During his flight, he stole the starting equipment, including a 52-slot bag of holding. The wizard's wrath is aroused, and exiting the Dungeon without the Orb means almost certain death. With the Orb, the protagonist can bend the wizard to his will, once the Orb is removed from its controlled territory.

Due to the the protagonist's disrupting aura, the Orb loses the ability to directly alter floors he has visited. During the Orb run, the Orb begins to summon allies to its current floor, as a costly act of desperation.

The wizard wants revenge, but he also wants the protagonist's soul for dissection. Only a bounty hunter with proper anti-Orb psychic enchantment and prepared soul trap can extract the latter if the protagonist dies within the Orb's influence. Such enchantments are inherently unreliable, as the converted uniques wandering the Dungeon demonstrate. The wizard's best chance at recovering his experiment is to set an ambush at the Dungeon entrance.

In its current state, the Orb's psychic effects are diffuse, dispersed over the vast territory it controls. However, once removed, it can be bent to the will of the wielder, directing its concentrated power irresistibly over a small area. The more runes collected, the more godlike this power becomes.
Fork you.

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Thursday, 25th April 2019, 07:34

Re: Leo Littlebook's thread

Re the question: Am I trolling?

I won't deny indulging in a bit of cheekiness here and there. For example, "Zin before Lair is optimal for 1/3 of species." Before writing this, I did a quick count and came up with over 1/3, counting a few maybes as yeses. But "about 1/3" sounds weaselly. So I went with "over". A stretch, perhaps, but can you argue with art?

https://www.reddit.com/r/dcsscirclejerk ... ihf/truth/
Fork you.

For this message the author Leo_Littlebook has received thanks: 2
duvessa, ManMan

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Thursday, 25th April 2019, 08:35

Re: Leo Littlebook's thread

commentary on
GotM 2 - Gargoyle Earth Elementalist Part 2 - Dungeon Crawl Stone Soup
uv4
https://www.youtube.com/watch?v=2-7BtjE ... dex=6&t=0s

untrained +3 flail venom
worshipping oka
D5
Int 18, Str 11
XL 7.3

Should be training Maces & Flails equally with magic skills, and leveling Str as well as Int. Especially with Oka. Heroism becomes more valuable the more is invested in physical skills. Instead of Heroism being the equivalent of a small XP investment in physical skills, it becomes equivalent to a large investment.

Looks like all physical skills are at 0. Even armor? Crazy, that's Gr's specialty. 0 sling leaves no way to engage high EV casters at range, the Gr's weakness.

Skilling appears to be pure caster. Gr is gish.

Fear of orc priests is why Gr needs Zin pre-Lair.

Wand of acid doing work, but it's at 0 Evo, of course.

@12:56 Yep 0 armor skill! Not even looking at marginal costs.

Seems like these drops are pretty good for D5.

@17:50 Shouldn't drop the +3 vamp short sword. Gr Stealth + amulet of gourmand + Petrify = HP regain on a small pool. Very good for a quick topoff in corridor fights.

Clearly killdudes is in effect.

@19:10 decided to train some Armour in a bit finally. Wearing ring mail OP drop.

@19:43 Has 3 pots mutation, won't quaff, on D6. Great value to quaff now. Lethal risk quite low.

+1 sling of freezing, still 0 sling training. Nuts.

@23:00 Missed a great chance to stairdance Prince Ribbit up, then venom flail and sandblast him down. He's much more dangerous prowling the level doing hit and run than hunting him down on a cleared floor. Streaking conditions the current meta for maladaptive caution. It's higher percentage to take an advantageous fight now than allow a possibly disastrous one later.

@24:10 Met Harold D7. He's exponentially more dangerous with Ribbit on the floor. Hence what I was saying.

@24:49 Netted and marked, itz happening! Ooh now adjacent and antimagic axed.

The solution here is trained evocations and melee, but uv4's trained neither. Deep trouble. Probably burn Blinks. Dunno what else he's got.

A Recite upon spotting Harold at distance might've enabled a successful escape up the stairs.

Hm is uv4 recording offline? I think so. Looks like he also prefers the offline UI for recording.

@29:20 not worth wasting an acid charge at 1/2 HP.

Using all acid charges, like 3x. Should just use enough to corrode, then venom flail. Didn't even swing once to get poison ticks for whole fight.

Already spent Agi and Heroism, hence why trying to stay in the fight.

Now within 1-shot range. Hafta Blink anyway.

Gonna burn a curing as well. And iceblast. Very expensive kill.

Harold is a dallyer from fullscreen, with his Marking. Could've Agi, poisoned to max, Blinked, kited, then finished with Acid or whatever. Pure caster is resource inefficient.

Got book of Debilitation. Not gonna memorize any. Nuts. Should train Hexes to primary, use it against low MR mobs. Rounds out offense. Maces & Flails needs Corona to hit. Slow cuts threat of mob by over half.

@36:40 Met Joseph. Still not starting hard fights with venom flail.
Now hitting with venom flail when OOM. Backwards, should do it at start.
Took 4 Iceblasts to kill. Terrible value.

If hadn't wasted acid charges earlier, could have corroded and then sandblasted him easily.

At half HP, got adds with Ribbit on floor. Not punished.

Had a flail of protection, wasn't using it! Should use venom, then switch to protection for GDR maximization. Then cast while AC bonus is active, swinging only to refresh. That's gish efficiency.

XL10, still no Maces skill. Killdudes in effect.

Found Acrobat amulet. Synergizes with Zin's Recite.

Can just kite Harold at fullscreen while he Marks on an empty floor and kill him with a sling.
Fork you.

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Thursday, 25th April 2019, 09:33

Re: Leo Littlebook's thread

Correction: After checking my morgue directory for this computer, which contains only two games, I noticed that this is my second attempt to beat the game. I started another HOFi Beogh "Giops" who died after polymorphing Sigmund into a Steam Dragon. YASD. It's hard to tell from the documentation what any given polymorph might result in. Anyway I was playing poorly at the time, and overconfident, and partly just wanted to see what would happen. Then I wanted to see if I could fight the steam dragon instead of running. Steam hits surprisingly hard at low XLs. This was on 20190404.

Anyway, after that I was suitably chastened and got serious, as serious as I can be during downtime.
Fork you.

Lair Larrikin

Posts: 27

Joined: Saturday, 15th September 2018, 16:54

Post Thursday, 25th April 2019, 11:10

Re: Leo Littlebook's thread

Leo_Littlebook wrote:Hm is uv4 recording offline? I think so. Looks like he also prefers the offline UI for recording.


GOTM is seeded. It's not possible to play seeded online.

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Thursday, 25th April 2019, 18:50

Re: Leo Littlebook's thread

GotM 2 - Gargoyle Earth Elementalist Part 3 - Dungeon Crawl Stone Soup
uv4
https://www.youtube.com/watch?v=QvSPAaN ... gs=pl%2Cwn

Ok, he's playing offline due to server unavailability.

@4:18 Orc High Priest and Sorcerer. I can't picture this going well.

LRD no damage on Sorcerer cuz ring mail. 2nd time 1/2 HP.

5 MP left and the retreat starts, Sorcerer still up. I think this late retreat risks death.

Advancing again, blind. Not a good idea.

Buncha Iceblast charges worked. Bet Evo is still 0.

Nothing too dangerous poured out of the FoW.

Recite would've been highly effective here. Oka didn't do much.

I take severe consumable attrition like this to indicate death spiral.

Recent play indicates the folly of going pure caster on a gish species. MP increase by HD. There is a diminishing return for casting high-level spells prematurely. Cheaper and more sustainable to develop the melee side as intended.

@09:14 There's the orc knight out of FoW I was worried about.

There goes the last of the Iceblast. Very expensive crossbow. At least movement is pretty safe with Acrobat.

More Disintegration charges spent.

Shouldn't keep training Spellcasting at primary if not using the spell levels. But should be memorizing Hexes. Great counter to orc knights and warriors.

The discarded vamp short sword in Bailey would allow regaining HP off weak mobs to extend endurance. Plenty of chunks, no time to regen. Gourmand OP.

OP hand crossbow and scroll of acquirement. Crawl too easy.

Shouldn't discard +1 sling of freezing. Good for conserving bolts vs popcorn.

No bolts for crossbow. First act on D:1 should be turning on all relevant ammo pickups. Silly to die for want of a nail.

Gloves of archery. OP drops. Pretty hard to die due to evocations depletion now.

Not learning Animate Skeleton cuz anti-synergy with archery LoS. Backwards. Fire first shot, skeletons aggro, retreat to fullscreen, let skeletons harass while you fire. Unlike summons, skeletons cost no in-combat moves. Great for 10-speed archers.

Even better in corridors. Fire a shot, step behind, LRD the queue.

@26:40 Armor skill way low, SL5 at XL 11. It's a primary skill on Gr.
Training Dodge to equal SL as Armor. Crazy. Gr terrible at Dodge, great at Armor.

@28:07 OP shield in shop, affordable. Game won. Ridiculous. Can just Acrobat regen kite on a small HP pool if hurt.

Oka certainly worked out thanks to the hand crossbow.

Coulda had a safe Bailey with Zin, then switched to Oka after finding crossbow. Good gods OP.

Stopped training Dodge, good.

Gloves of archery mean should pick up that blowgun and use it. Good for kiting scary monsters like double-headed ogres.

Should be walking around with Acrobat, not Reflection. Use Reflection when ranged monsters appear. Can't switch to Acrobat when hurt.
Fork you.

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Friday, 26th April 2019, 05:49

Re: Leo Littlebook's thread

Testing "Two State Vampires" [Dungeon Crawl Stone Soup]
Malcolm Rose
Published on Apr 23, 2019
https://www.youtube.com/watch?v=uPRRJAQ ... dex=3&t=0s

I hate this change, it ruins Vp. Maybe I joined DCSS just at the moment the zeal for pruning went from adaptive to pathological. I won't update to .24 if this change is included. I hear they're going to do terrible things to Fedhas as well.
Fork you.
User avatar

Zot Zealot

Posts: 982

Joined: Monday, 29th September 2014, 09:04

Post Friday, 26th April 2019, 06:37

Re: Leo Littlebook's thread

A forum thread isn't a blog. If you want to comment on videos, you can comment on the videos. Or start your own blog.

For this message the author chequers has received thanks:
ichbins

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Friday, 26th April 2019, 06:50

Re: Leo Littlebook's thread

Commenting UI on YouTube is annoying. I've written enough to establish a record of my ideas and their reception. Don't see the point in starting a blog yet. Will just go back to not publishing my notes until demo. Less work.
Fork you.
User avatar

Slime Squisher

Posts: 405

Joined: Sunday, 27th January 2019, 13:50

Post Friday, 26th April 2019, 10:35

Re: Leo Littlebook's thread

Leo_Littlebook wrote:Testing "Two State Vampires" [Dungeon Crawl Stone Soup]
Malcolm Rose
Published on Apr 23, 2019
https://www.youtube.com/watch?v=uPRRJAQ ... dex=3&t=0s

I hate this change, it ruins Vp. Maybe I joined DCSS just at the moment the zeal for pruning went from adaptive to pathological. I won't update to .24 if this change is included. I hear they're going to do terrible things to Fedhas as well.


I only started playing DCSS in 19.5. tiles, and from what little I've played Vp chars they were awful from an engineering perspective. The state juggles were horrendous and I can't see a lot of players (or devs, for that matter) saying that particular game aspect will be missed or even fondly remembered. While a lot of people (both on and off Tavern) are quite vocal in the support of feature removal, changing Vp or even scrapping the concept entirely would be an improvement in my eyes.
As for "ruining" Vp, you have me at a loss here. What aspect is ruined by the overhaul: the technical possibilities of the species, tacticals, flavour? I'm quite open to having your opinion on this.

As for Fedhas Madash, the gameplay has very real issues with pacing and requires very limited resources to make for solid utility. Sure, having the ability to modify terrain and get to loot that would otherwise become inaccessible by default is great. Being able to produce a TD-esque clump of greenhouse vitriolic death is also fairly neat. But almost everything good that Fedhas offers is badly represented because the core mechanics and their application to crawling are so wildly divergent from both the usual gameplay required, and the guidelines of what DCSS devs strive for with the game at large.
There is always something new to learn.

Lair Larrikin

Posts: 27

Joined: Saturday, 15th September 2018, 16:54

Post Friday, 26th April 2019, 11:06

Re: Leo Littlebook's thread

The Fedhas rework that Demise played in one of his videos in some fork looked really neat, except it was totally ridiculously unbalanced... on demand batteries of oklobs machinegunning everything in sight. But with some rebalancing that would be the best way to go.

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Friday, 26th April 2019, 13:34

Re: Leo Littlebook's thread

Hi Sorcerous. If you're talking about bloodless kicking VpTm out of a Form, then that's a bug. Ditto Berserk but IIRC that's fixed.

Otherwise, the state changes are the point of the species. The desire of players to tab-macro spam through the game regardless of species should be taken as a symptom of broken competitive incentives, not indulged.

VpEn Dith is great. You're a nightlord, trying to rack up kills silently like a shade, then gorging on blood when hurt and fighting like a troll. The multiple states give a sense of fading in and out gradually of the parched and gorged forms. It's immersive and OP.

The 2-state change ruins the immersion and makes the species trivial to play. Making the game easier is a move in the wrong direction. Now VpEn isn't forced to go Makleb in Extended. Why suffer through Mummy if you can just cast with a Vampire instead? It's dumb.

The undead species are balanced in .23 - knight, rogue and mage. The change just makes Vampire a flavorless Gary Stu who obsoletes the other two. Undead is about the tradeoffs of unlife. Each one is severely crippled in some way. Death isn't supposed to be easy.

> As for Fedhas Madash, the gameplay has very real issues with pacing and requires very limited resources to make for solid utility.

Fedhas is fine. MfIE Fedhas is an OP combo, then switch to Zin for Abyss. Fedhas is great for Ice Elementalists. Not a bad second choice for Nagas either.

The early incentive to rot corpses for piety instead of eating their chunks is a much-needed reason to play faster than the hunger clock requires, and a fair tradeoff for the power of mushrooms.

Not every god needs to be a broadly useful powerhouse like Trogg or good gods. Playstyle differentiation is more important.

You'll have to use more specifics. I don't read Tavern so I've no idea what your generalizations are about.
Fork you.
User avatar

Slime Squisher

Posts: 405

Joined: Sunday, 27th January 2019, 13:50

Post Friday, 26th April 2019, 18:31

Re: Leo Littlebook's thread

Hello Leo.

Leo_Littlebook wrote:VpEn Dith is great. You're a nightlord, trying to rack up kills silently like a shade, then gorging on blood when hurt and fighting like a troll. The multiple states give a sense of fading in and out gradually of the parched and gorged forms. It's immersive and OP.


There are multiple ways to describe repeatedly drinking or 5ing to change states, "gradually" and "immersive" are not among them.

Leo_Littlebook wrote:Now VpEn isn't forced to go Makhleb in Extended.


Rolling through extended is about far more than deity choice. DCSS forces nothing, players choose what they choose.

Leo_Littlebook wrote:Fedhas is fine. MfIE Fedhas is an OP combo, then switch to Zin for Abyss.


This combo is OP -> switch away from it. Brilliant.

Leo_Littlebook wrote:Not every god needs to be a broadly useful powerhouse like Trog or good gods. Playstyle differentiation is more important.


Your posted list of "best fit" builds above shows otherwise, because a lot of them default to the good trio. You also listed switches from Invocations-dependent gods to the opposite (GhFi Yred -> ?kiku,TeAE good gods -> Veh, BaWp good gods -> Ru ), while claiming "optimal" throughout this thread. I fail to see what is optimal about that.

Leo_Littlebook wrote:I don't read Tavern so I've no idea what your generalizations are about.


In that case, you may want to delay posting until you get to reading. Some of the posters here used to make this game, or still do.
There is always something new to learn.

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Friday, 26th April 2019, 20:12

Re: Leo Littlebook's thread

> There are multiple ways to describe repeatedly drinking or 5ing to change states, "gradually" and "immersive" are not among them.

vampire
/ˈvampʌɪə/
noun
1.
(in European folklore) a corpse supposed to leave its grave at night to drink the blood of the living by biting their necks with long pointed canine teeth.

You are playing VpEn Dith wrong. Supposed to stay bloodless to maximize stealth until hurt, then drink. Don't wait to exsanguinate, that wastes piety.

If you're complaining about how some other combo plays, that's because it's a suboptimal combo.

> This combo is OP -> switch away from it. Brilliant.

Thanks. It hardly needs mentioning that the effectiveness of a god changes depending on the branch.

>> Not every god needs to be a broadly useful powerhouse like Trog or good gods. Playstyle differentiation is more important.
> Your posted list of "best fit" builds above shows otherwise, because a lot of them default to the good trio.

Pre-Lair, Zin is optimal for about 1/3 of species, yes. The goal is to survive until the species' adolescent form is viable. Pick one species to complain about and we can talk specifics.

> You also listed switches from Invocations-dependent gods to the opposite (GhFi Yred -> ?kiku,TeAE good gods -> Veh, BaWp good gods -> Ru ), while claiming "optimal" throughout this thread. I fail to see what is optimal about that.

I don't see how you can complain about GhFi Yred -> Kiku extended. Yred is garbage in Extended, and Kiku solves the no-corpse problem. Who cares about some wasted XP, Extended has plenty.

As for the two others, survival is totally worth training Invo to primary and then ditching it. One shouldn't switch gods until ready to begin adolescence, which implies that the character is strong.

For the species that switch away from Good Gods, the GG start is to ensure power floor. The adolescent transition is to maximize power ceiling.

> In that case, you may want to delay posting until you get to reading. Some of the posters here used to make this game, or still do.

I am delaying posting in the entire forum that isn't this thread.
Fork you.
User avatar

Shoals Surfer

Posts: 287

Joined: Friday, 19th August 2016, 21:21

Post Friday, 26th April 2019, 22:26

Re: Leo Littlebook's thread

Sorry if I came across as a bit unpleasant on my last post, Leo, I thought you were a troll since your 1st post was pretty confusing.

Leo_Littlebook wrote:In this case, Personal Information Manager.


Could you explain what you mean by PIM (https://en.wikipedia.org/wiki/Personal_information_manager) is in this context, and how that'll help you to crack dcss's meta, as you said?
make food great again

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Saturday, 27th April 2019, 03:13

Re: Leo Littlebook's thread

No problem pedritolo. I think it's a legitimately confusing post for someone encountering me for the first time.

> Could you explain what you mean by PIM (https://en.wikipedia.org/wiki/Personal_ ... on_manager) is in this context, and how that'll help you to crack dcss's meta, as you said?

Well, it manages my text and binaries, so far. I process roughly 10k words per day through it.

Right now I just file all my DCSS stuff into a directory and never look at it again, so I can't say the PIM helped to crack the meta. It will help me correct, organize and present those thoughts, and help me to play DCSS better. I plan to create tactical checklists, branch guides, build guides etc for myself to follow when playing my true combos list. My PIM handles my entire daily and long-term executive function, so applying it to a virtual character is overkill.

I won't use the PIM on DCSS until I'm recording, so that I can show the whole process from scratch.
Fork you.

Lair Larrikin

Posts: 21

Joined: Monday, 15th October 2018, 01:17

Post Saturday, 27th April 2019, 06:22

Re: Leo Littlebook's thread

chequers wrote:A forum thread isn't a blog. If you want to comment on videos, you can comment on the videos. Or start your own blog.


Why try to send him away from the main forum for the game, though? Is it really hurting anything for him to have a thread where he voices a bunch of opinions? It's not like it's against the forum rules, as far as I can tell. I haven't really read too much of the thread for obvious reasons but it does appear that he's replying to people rather than just using this as a one-way soap box..

Spider Stomper

Posts: 242

Joined: Friday, 17th April 2015, 16:22

Post Saturday, 27th April 2019, 07:43

Re: Leo Littlebook's thread

If you're factoring in stuff like the "essence of the game" and maximal differentiation, it's kinda weird to call your list optimal since it's...really not going to be win-rate optimal. Anyway some of my thoughts on the list from a more standard definition of optimal:


HOFi of Beogh that converts for extended is bog standard.
MiBe of Trog is the staple.
MfIE of Fedhas is standard for all three.
GrEE of Elyvilon is functional.
DrSu of Hep->Good god is bizarre as heck. The DrSu part is fine, Hep is weird to beeline but strong enough that I'd just attribute it to taste, but ditching Hep seems to just be over-complicating things.
HaHu of Zin is curious god choice but straightforward enough.
TrCK of Xom->Jiyva is not the best Troll, but if you have to have CK somewhere might as well put it here. Same with Jiyva really.
GhFi of Yred->Kiku is fine except for the part where you go to Kiku in extended. Ditching Yred is fine, but Kiku is a really bizarre first pick for extended.
HuGl of TSO->Good god is....not great? TSO is kinda hot garbage outside extended.
KoCj of Sif is the sort of thing I'd play for fun tbh, have a hard time saying Sif is optimal though with how weak the god is prior to books.
DsNe of Nemelex is functional enough I guess, would rather have a bunch of gods over Nemelex here personally though.
CeMo of Qaz->Good god is bizarre as heck. Centaur lives and dies on its kiting ability; Qaz means you're gonna have a really rough time doing that without getting run over by a giant mob. Going on to ditch Qaz is also really not smart for surviving.
SpAM of Gozag is strong
TeAE of Good god->Vehumet is insane?? Why are you taking Zin/Ely early on a race with negative Invo aptitude? Why are you going on to ditch them for a god as weak as Vehumet? Just start with Veh if you must use him.
DEWz of Sif is a trap! Wizard wants to either be on a race that can muscle up if needed or to take a god that gives it raw immediate power. DEWz^Sif is neither; if you don't have good book luck by Lair you're gonna have a bad time.
OgAK of Lugonu is...well it's a combo, not a particularly impressive one though.
GnWn of Oka->Good god is whatever, it's really hard to roll a bad gnoll
VSTm of Wu is baffling. Even if this gets through its harrowing XL1, Wu is not an impactful enough god to save this combo IMO.
VpEn of Dith/Makh is standard, just pick one god instead of doing such a ridiculous swap though.
DgSk is not the optimal Demigod or Skald but no race really likes Skald and Demigod does them well enough
FoEE of Good god->Uskayaw->Good god is the most baffling thing on this list, why in god's name are you dipping Uskayaw
NaVM of Chei is another trap! VM sees some sick spellpower with Naga, but it has a severe issue where its spells are best if you have functioning legs. Naga does not have these. It also gives you a double hate of poison monsters, to say nothing of the actual anti-synergy of Naga and Chei; your glacial move delay can get you killed later on by threats you couldn't see when you started your move.
OpVM of Good god->Ash->Good god is another pointless god swap (Ash is bad on octo due to slot issues and also actively terrible as a dip), also a kinda dicey combo I feel like if you have VM already represented.
FeFE of Sif wants a more proactive god to baby the cat but it's basic enough.
BaWr of Good god->Ru baffles me, if you go good gods just stay there. If you want Ru just start Ru and stick with Ru. Good gods aren't doing much without Invo investment that is wasted when you go to Ru. You're also wasting a lot of time that could be spent getting your Ru sacrifices earlier.
MuNe of Kiku->Gozag is another dubious (if understandable this time) god swap on a pretty whatever combo. Mummy's apts are so bad that I would suggest something with more focused skills than Necromancer though, like Fi or FE.
DDAr of Makhleb is the staple, no disputes here.

For this message the author Doesnt has received thanks:
MalcolmRose

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Saturday, 27th April 2019, 09:07

Re: Leo Littlebook's thread

> If you're factoring in stuff like the "essence of the game" and maximal differentiation, it's kinda weird to call your list optimal since it's...really not going to be win-rate optimal.

Rough drafts have mistakes. We've already established it's better to call them "true combos". And I will attempt to prove the true combos are optimal, which I expect to shock the current meta but not totally succeed on e.g. TrCK. But I think I might succeed even on TrCK -> Jiyva.

> DrSu of Hep->Good god is bizarre as heck. The DrSu part is fine, Hep is weird to beeline but strong enough that I'd just attribute it to taste, but ditching Hep seems to just be over-complicating things.

Start Hep because Su's Summon Imp conflicts with good gods conduct. Summoner synergizes with Hep, keeping the useful ancestor alive in the press of bodies.

Don't worry too much about my strange extended choices, they're untested.

> HaHu of Zin is curious god choice but straightforward enough.

Any good gods pick will probably involve early switching. Zin for Dungeon and Orc, Ely for Lair Swamp Spider. I specify one of the good gods to show which gives the strongest overall synergy, but branch synergy often outweighs that.

> GhFi of Yred->Kiku is fine except for the part where you go to Kiku in extended. Ditching Yred is fine, but Kiku is a really bizarre first pick for extended.

Why? Gh needs to eat. Kiku gives bodies. Gh can cast Necromancy ok.

> HuGl of TSO->Good god is....not great? TSO is kinda hot garbage outside extended.

That's not what I wrote. HuGl would start Zin, then Ely Lair probably. Affinity for TSO means e.g. maybe doing Crypt earlier.

> KoCj of Sif is the sort of thing I'd play for fun tbh, have a hard time saying Sif is optimal though with how weak the god is prior to books.

One could start good gods and delay Sif, but I don't think that's better. Sif synergizes with Ko's gourmand. Use +Cast to kite cast expensive Conjurations. Conjurations' weakness is DPS/mana ratio, but with +Cast and gourmand this doesn't matter. In particular, Dazzling Spray then stab. Blowguns aptitude also synergizes with kiting.

> DsNe of Nemelex is functional enough I guess, would rather have a bunch of gods over Nemelex here personally though.

Ds are the spawn of Demons, who grow random mutations. Nemelex gifts decks of random cards written in the ichor of demons. Whatever mutations develop, Nemelex will complement the build. DsNe is strong enough to handle the cheeky Nem randomness.

> CeMo of Qaz->Good god is bizarre as heck. Centaur lives and dies on its kiting ability; Qaz means you're gonna have a really rough time doing that without getting run over by a giant mob. Going on to ditch Qaz is also really not smart for surviving.

Qaz improves Ce kiting by forcing monsters to path around clouds or take damage. In practice, a crowd winds up taking damage due to pathing failure. Qaz is bad in extended.

Ce is a mediocre kiter without Qaz, because his speed is not great enough to gain a step to fire on a short runway. With Qaz his kiting becomes much deadlier.

Obviously the danger of Qaz is getting swarmed. CeMo permits jumping straight to strong cloud cover so that his cloud-kiting is immediately effective. He must still always be conscious of his noise output. Corridors are the primary danger, because they negate clouds and permit getting trapped.

> TeAE of Good god->Vehumet is insane?? Why are you taking Zin/Ely early on a race with negative Invo aptitude? Why are you going on to ditch them for a god as weak as Vehumet? Just start with Veh if you must use him.

TeAE Veh is a bad start. An extra lvl 1 conjuration will have no impact on his survival odds in early Dungeon. It just encourages him to unwisely split his XP investments, which are already spread thin to his gish nature.

TeAE Zin for Dungeon is good. The invo is 20% underpowered at -1 Apt, but this is more than balanced by Te's kiting ability.

> DEWz of Sif is a trap! Wizard wants to either be on a race that can muscle up if needed or to take a god that gives it raw immediate power. DEWz^Sif is neither; if you don't have good book luck by Lair you're gonna have a bad time.

Current meta skills Wizard poorly, reducing his viability.

+Cast is great for DEWz, because he's such an OP caster. Wizard book isn't great for Lair, although Frost Imps help. To compensate, Conjure Flame is strong in Dungeon's tunnels, and Mephitic is strong in Orc.

One could always go good gods with +1 Invo until finding another spell book to complement Wizard, then switch to Sif. This would sacifice Summon Imp for a while, possibly until it is no longer useful.

I can't recall the results of my DEWi testing offhand, so I can't comment further.

> OgAK of Lugonu is...well it's a combo, not a particularly impressive one though.

It's great. Bend Space fixes the Ogre's low AC, keeping him at range. Lugonu abilities cost HP, which Ogre has plenty of, particularly in Troll leather armor. Wield a polearm and throw stuff, then learn spells.

> GnWn of Oka->Good god is whatever, it's really hard to roll a bad gnoll

It's great. Gnoll negates the random training disadvantage of Wn, allowing the generosity of Wn to shine. Oka's random gifts and broad skill buff both suit Gnoll eclecticism.

> VSTm of Wu is baffling. Even if this gets through its harrowing XL1, Wu is not an impactful enough god to save this combo IMO.

Wu's perfect for Vs. More chances for antimagic bite to recover MP. I don't see the problem with D1 either.

> VpEn of Dith/Makh is standard, just pick one god instead of doing such a ridiculous swap though.

Makhleb is obviously superior to Dith for Extended because there is no blood to regen in Extended. Dith is better when there is blood, for rogue playstyle.

> DgSk is not the optimal Demigod or Skald but no race really likes Skald and Demigod does them well enough

Disagree. Skald is about endurance. Dg is a gish with a long MP bar, perfect for Infusion.

> FoEE of Good god->Uskayaw->Good god is the most baffling thing on this list, why in god's name are you dipping Uskayaw

Fo wants shield and 2h sword before he converts to Usk. Then he can tank in a killhole to build piety, and stab paralyzed. Until then he's squishy and good gods help. Extended is speculative.

> NaVM of Chei is another trap! VM sees some sick spellpower with Naga, but it has a severe issue where its spells are best if you have functioning legs. Naga does not have these. It also gives you a double hate of poison monsters, to say nothing of the actual anti-synergy of Naga and Chei; your glacial move delay can get you killed later on by threats you couldn't see when you started your move.

Current meta is bad at playing the tunnel snake that is NaVM Chei. If a move might kill you, consider yell-pulling instead. Chei abilities compensate for loss of VM kiting. Pull with Poison Spit.

> OpVM of Good god->Ash->Good god is another pointless god swap (Ash is bad on octo due to slot issues and also actively terrible as a dip), also a kinda dicey combo I feel like if you have VM already represented.

Go good gods until can fill slots, then Ash. Ash value diminishes when reaching skill caps in extended. Burn off Ash wrath somewhere safe.

Ash and Op are perfect match. Op can curse all his rings, hat and weapon without losing much, once he has some spells. Other species limited to 2 ring slots can't.

> FeFE of Sif wants a more proactive god to baby the cat but it's basic enough.

Nah. Kiting Flame Tongue with +Cast is OP.

> BaWr of Good god->Ru baffles me, if you go good gods just stay there. If you want Ru just start Ru and stick with Ru. Good gods aren't doing much without Invo investment that is wasted when you go to Ru. You're also wasting a lot of time that could be spent getting your Ru sacrifices earlier.

Negative Invo means BaWr shouldn't stay good gods. +1 vision synergizes with Ru fullscreen abilities.

Starting Ru is a bad idea because one doesn't know yet which way one's build should develop.

Wasted Invo investment is fine. Stay good gods until it becomes clear which kind of sacrifices would be preferable, then switch to Ru.

Ru is a tanking god. So for example, put together a strong set of armor and shield, then switch to Ru and sacrifice magic.

> MuNe of Kiku->Gozag is another dubious (if understandable this time) god swap on a pretty whatever combo. Mummy's apts are so bad that I would suggest something with more focused skills than Necromancer though, like Fi or FE.

Current meta is bad at skilling gishes. This combo is optimal.
Fork you.
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Saturday, 27th April 2019, 12:32

Re: Leo Littlebook's thread

I mean, you can call this trolling, but I like this thread better than the other long running CYC threads people regularly post in.
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

For this message the author tealizard has received thanks: 3
Rast, tasonir, TheMeInTeam

Temple Termagant

Posts: 14

Joined: Saturday, 15th April 2017, 17:43

Post Saturday, 27th April 2019, 14:02

Re: Leo Littlebook's thread

It's been pretty entertaining yes.

Dungeon Master

Posts: 250

Joined: Thursday, 27th November 2014, 19:12

Post Saturday, 27th April 2019, 14:41

Re: Leo Littlebook's thread

While I know you want to disrupt the meta it will help if you adapt to the local vocabulary. We don’t really use gosh in the crawl world, see here for new and old vocab.

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Saturday, 27th April 2019, 16:03

Re: Leo Littlebook's thread

It's gish. If that's genuinely your local vocabulary, it doesn't look worth adopting.

For referencing the standard builds in DCSS, I prefer warrior, mage, rogue, skirmisher, gish.
Fork you.

Tartarus Sorceror

Posts: 1739

Joined: Tuesday, 13th March 2012, 02:48

Post Saturday, 27th April 2019, 18:43

Re: Leo Littlebook's thread

Sorry to disrupt the entertainment, fellas, but...

Leo, how many games of DCSS have you won? Were any of them on public servers? Please post pastebins of the morgues for your last five wins.

Thanks.

Dungeon Master

Posts: 250

Joined: Thursday, 27th November 2014, 19:12

Post Saturday, 27th April 2019, 18:44

Re: Leo Littlebook's thread

Well don’t expect to be understood if you insist on calling widepersons fish, though it’s true Op and Mf make good widegals.

For this message the author ebering has received thanks:
duvessa

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Saturday, 27th April 2019, 18:45

Re: Leo Littlebook's thread

Zero. Ebering, I'm going to mute you.
Fork you.

Spider Stomper

Posts: 242

Joined: Friday, 17th April 2015, 16:22

Post Saturday, 27th April 2019, 21:34

Re: Leo Littlebook's thread

lmao at staff complaining about the word gish

Leo_Littlebook wrote:> HuGl of TSO->Good god is....not great? TSO is kinda hot garbage outside extended.

That's not what I wrote. HuGl would start Zin, then Ely Lair probably. Affinity for TSO means e.g. maybe doing Crypt earlier.


My bad, posting at 2 AM leads to some misreading.

Leo_Littlebook wrote:> KoCj of Sif is the sort of thing I'd play for fun tbh, have a hard time saying Sif is optimal though with how weak the god is prior to books.

One could start good gods and delay Sif, but I don't think that's better. Sif synergizes with Ko's gourmand. Use +Cast to kite cast expensive Conjurations. Conjurations' weakness is DPS/mana ratio, but with +Cast and gourmand this doesn't matter. In particular, Dazzling Spray then stab. Blowguns aptitude also synergizes with kiting.


The bigger limit on conjurations is usually fail rate and skill level moreso than MP. I get the impression you're severely overrating +Cast.

Leo_Littlebook wrote:> CeMo of Qaz->Good god is bizarre as heck. Centaur lives and dies on its kiting ability; Qaz means you're gonna have a really rough time doing that without getting run over by a giant mob. Going on to ditch Qaz is also really not smart for surviving.

Qaz improves Ce kiting by forcing monsters to path around clouds or take damage. In practice, a crowd winds up taking damage due to pathing failure. Qaz is bad in extended.

Ce is a mediocre kiter without Qaz, because his speed is not great enough to gain a step to fire on a short runway. With Qaz his kiting becomes much deadlier.

Obviously the danger of Qaz is getting swarmed. CeMo permits jumping straight to strong cloud cover so that his cloud-kiting is immediately effective. He must still always be conscious of his noise output. Corridors are the primary danger, because they negate clouds and permit getting trapped.


Gonna strongly disagree with the assertion that non-Qaz Centaurs are bad at kiting. No, you don't get a gap every step, but with a decently sized pillar it's not hard to get as many gaps as you need.

> TeAE of Good god->Vehumet is insane?? Why are you taking Zin/Ely early on a race with negative Invo aptitude? Why are you going on to ditch them for a god as weak as Vehumet? Just start with Veh if you must use him.

TeAE Veh is a bad start. An extra lvl 1 conjuration will have no impact on his survival odds in early Dungeon. It just encourages him to unwisely split his XP investments, which are already spread thin to his gish nature.

TeAE Zin for Dungeon is good. The invo is 20% underpowered at -1 Apt, but this is more than balanced by Te's kiting ability.


>centaur kiting is mediocre because 0.7 movedelay isn't fast enough
>tengu kiting off 0.9 movedelay is a selling point for an invo build

Are...are you sure about that one?

> DEWz of Sif is a trap! Wizard wants to either be on a race that can muscle up if needed or to take a god that gives it raw immediate power. DEWz^Sif is neither; if you don't have good book luck by Lair you're gonna have a bad time.

Current meta skills Wizard poorly, reducing his viability.

+Cast is great for DEWz, because he's such an OP caster. Wizard book isn't great for Lair, although Frost Imps help. To compensate, Conjure Flame is strong in Dungeon's tunnels, and Mephitic is strong in Orc.

One could always go good gods with +1 Invo until finding another spell book to complement Wizard, then switch to Sif. This would sacifice Summon Imp for a while, possibly until it is no longer useful.

I can't recall the results of my DEWi testing offhand, so I can't comment further.


You could also just go Kiku, who reliably gives you useful stuff at 1* piety.

> OpVM of Good god->Ash->Good god is another pointless god swap (Ash is bad on octo due to slot issues and also actively terrible as a dip), also a kinda dicey combo I feel like if you have VM already represented.

Go good gods until can fill slots, then Ash. Ash value diminishes when reaching skill caps in extended. Burn off Ash wrath somewhere safe.

Ash and Op are perfect match. Op can curse all his rings, hat and weapon without losing much, once he has some spells. Other species limited to 2 ring slots can't.


Ashenzari doesn't give boosts per cursed item. On Octopode, having 8 cursed rings gives you the same benefits as a human with 2 cursed rings, and you need 4 cursed rings to get any ring-related benefits at all. You also need a hat to gain any defensive boosts, since that accounts for your entire armor.

Ashenzari's value does not diminish in extended; unless you're doing extreme farming, skill boosts are still relevant and Scrying is more relevant than ever while wandering Pandemonium.

> FeFE of Sif wants a more proactive god to baby the cat but it's basic enough.

Nah. Kiting Flame Tongue with +Cast is OP.


Fast Flame Tongues without +Cast are about as good as without.

> BaWr of Good god->Ru baffles me, if you go good gods just stay there. If you want Ru just start Ru and stick with Ru. Good gods aren't doing much without Invo investment that is wasted when you go to Ru. You're also wasting a lot of time that could be spent getting your Ru sacrifices earlier.

Negative Invo means BaWr shouldn't stay good gods. +1 vision synergizes with Ru fullscreen abilities.

Starting Ru is a bad idea because one doesn't know yet which way one's build should develop.

Wasted Invo investment is fine. Stay good gods until it becomes clear which kind of sacrifices would be preferable, then switch to Ru.

Ru is a tanking god. So for example, put together a strong set of armor and shield, then switch to Ru and sacrifice magic.


>Starting Ru is a bad idea because one doesn't know yet which way one's build should develop.
>Ru is a tanking god. So for example, put together a strong set of armor and shield, then switch to Ru and sacrifice magic.

Or you can just, take Ru early and sac things you aren't going to use later. Ru's hardly a weak god in Lair if you go for big sacrifices.

> MuNe of Kiku->Gozag is another dubious (if understandable this time) god swap on a pretty whatever combo. Mummy's apts are so bad that I would suggest something with more focused skills than Necromancer though, like Fi or FE.

Current meta is bad at skilling gishes. This combo is optimal.


It's not a meta thing; it's the fact that Mummy is overall so poor that if you spread yourself thin you wind up with poor raw power to work through problems with. A passable fighter or a passable dedicated caster tends to do better than a character with cruddy combat and cruddy magic in Crawl.



I'm going to echo Rast's question; how many games have you won, and can we see morgues for the last few? If there are none that's fine, I look forward to seeing how your opinions develop as you get more experienced.

For this message the author Doesnt has received thanks: 2
Leo_Littlebook, MalcolmRose
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Saturday, 27th April 2019, 23:46

Re: Leo Littlebook's thread

He answered: "Zero."

wiki -> salad shooter -> Leo Littlebook's thread.
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Sunday, 28th April 2019, 01:03

Re: Leo Littlebook's thread

> My bad, posting at 2 AM leads to some misreading.

No problem. I just wanted to emphasize the difference between , and -> in my list.

> The bigger limit on conjurations is usually fail rate and skill level moreso than MP. I get the impression you're severely overrating +Cast.

If you're trying to cast above your level, sure. The big spell for early KoCj is Dazzling Spray, which has a terrible damage:MP ratio vs single targets, and is unreliable at blinding. +Cast changes this.

I argue Sif and therefore +Cast on three species: Fe, Ko, DE. All three are fragile mages who can die if they stop casting due to OOM. It could be that all three Sif should delay Sif in favor of good gods until a second spellbook is found. But this would also upset the meta.

> Gonna strongly disagree with the assertion that non-Qaz Centaurs are bad at kiting. No, you don't get a gap every step, but with a decently sized pillar it's not hard to get as many gaps as you need.

I said they are mediocre at kiting. A pillar permits infinite kiting, assuming no adds. Runway is not always infinite.

Centaur should start off kiting with a trident. Kiting a crowd while maintaining a one-tile separation is optimal Qaz kiting.

>>centaur kiting is mediocre because 0.7 movedelay isn't fast enough
>>tengu kiting off 0.9 movedelay is a selling point for an invo build

> Are...are you sure about that one?

Yes. The difference is that for Ce I mean pre-Lair kiting with trident and bow, and for Te pre-Lair kiting with Recite. TeAE doesn't kite with his primary offense, Shock. Instead he uses his speed to obtain a double-bounce position for Shock.

Trident and bow attack at slower than 1 turn and ideally occur every other turn when kiting, whereas Recite uses 1 turn in ?10-20.

>> One could always go good gods with +1 Invo until finding another spell book to complement DEWi, then switch to Sif. This would sacifice Summon Imp for a while, possibly until it is no longer useful.
> You could also just go Kiku, who reliably gives you useful stuff at 1* piety.

Kiku precludes a good gods opener. It is implied in my list that when possible one picks up a good god before finding one's preferred god.

Paying -1 HP per cast of Pain on a small DE HP pool is painful.

Zin's Recite is a better deal in Dungeon, and Ely's value is boosted by the low HP.

> Ashenzari doesn't give boosts per cursed item. On Octopode, having 8 cursed rings gives you the same benefits as a human with 2 cursed rings, and you need 4 cursed rings to get any ring-related benefits at all. You also need a hat to gain any defensive boosts, since that accounts for your entire armor.

That is why Op delays Ash until all slots are full, rendering the slow buildup to fully cursed slots irrelevant.

> Ashenzari's value does not diminish in extended; unless you're doing extreme farming, skill boosts are still relevant and Scrying is more relevant than ever while wandering Pandemonium.

Yes. So when doing extreme extended farming, when skill boosts lose relevance, switch to good gods.

> Fast Flame Tongues without +Cast are about as good as without.

Assuming infinite runway and no monster regen. In practice, the less runway one needs, the safer. Morever, running OOM precludes Conjure Flame, a major safety measure in corridors. +Cast also allows spamming Inner Flame, a FeFE specialty.

> Or you can just, take Ru early and sac things you aren't going to use later. Ru's hardly a weak god in Lair if you go for big sacrifices.

Not as strong. Zin/Ely are OP through Swamp/Spider. Ru is weaker until character is fully tankified, due to Aura of Power. Makes sense to delay him on a gish that starts squishy. One has vastly more information on build direction by mid-game than at D1. Ru is balanced for early-game decisionmaking; one can break this balance by waiting until loot drops have altered sacrifice incentives. Warper needs mid-game XLs to wear shield and plate and still cast Shroud.

>> Current meta is bad at skilling gishes. This combo is optimal.
> It's not a meta thing; it's the fact that Mummy is overall so poor that if you spread yourself thin you wind up with poor raw power to work through problems with. A passable fighter or a passable dedicated caster tends to do better than a character with cruddy combat and cruddy magic in Crawl.

That is what I'm saying. The doctrine you expressed is due to inefficient gish skilling in the current meta.

> If there are none that's fine, I look forward to seeing how your opinions develop as you get more experienced.

I have a lot of experience playing through the early game and quitting after Lair, because early game makes or breaks combos.
Fork you.

Mines Malingerer

Posts: 38

Joined: Wednesday, 24th April 2019, 13:53

Post Sunday, 28th April 2019, 12:34

Re: Leo Littlebook's thread

Nobody has said anything positive about the idea of exclusively studying DCSS' early game in order to figure out true combos. Instead there's been only demands for proof of wins. Which leads me to believe that nobody has done what I did:

Play all species simultaneously, switching to the lowest XL one after doing a few floors. Quit sometime between Lair and Depths, or sooner if the combo+playstyle feels wrong. Whenever quitting a species, start a new game for that species. Infinitely loop until confident that have found true combos and playstyles.

It is much harder to learn a skill when one studies the whole thing instead of working on one chunk at a time. This explains how people with tens (hundreds?) of DCSS wins may never reach the conclusions I did through focused immersion in arguably the simplest part of the game. Their minds were distracted by too much information.

Starting combo has the greatest impact on the early game. When one has little to work with, one must learn to be efficient. The early game is a good place to learn the basic mechanics, before power growth permits sloppiness.

Mastering DCSS from mid-game to extended is largely a process of getting spoiled. But mastering the early game requires understanding combo selection. One may play full DCSS games indefinitely without ever advancing beyond the early-game understanding required to streak.
Fork you.

Lair Larrikin

Posts: 21

Joined: Monday, 15th October 2018, 01:17

Post Sunday, 28th April 2019, 12:49

Re: Leo Littlebook's thread

Leo_Littlebook wrote:Zero. Ebering, I'm going to mute you.


He is trolling you, which is honestly kind of a bad look for a dev of the game even if it isn't meant maliciously. Upsetting to see, tbh.
Next

Return to Crazy Yiuf's Corner

Who is online

Users browsing this forum: No registered users and 25 guests

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